Developing a Cleaning Process: Cleaning in Development By Birgitte Holst

INTRODUCTION The discipline of cleaning validation is as individual as process validation, but the methods, techniques, and risks are almost identical in all kinds of pharmaceutical manufacture. Process understanding is the magic term when assessing what must be done and the effort to be put into the documentation of equipment status. A smooth technology transfer and successful cleaning validation in the manufacturing plant can optimise time to market for a new product. Lack of cleaning process understanding can cause expensive surprises in the late phases of bringing products to market. It is, therefore, worthwhile to invest development time in the cleaning process. A number of crucial conditions are needed for optimising time to market. Technology transfer of the cleaning process should encompass these conditions: 1. Cleaning process impact on product residues and equipment surfaces 2. Rationale for testing residues after cleaning in the manufacturing plant 3. Test and analytical method validation for testing residues in the manufacturing plant 4. Cycle development of the cleaning process in the manufacturing plant [Make sure the process is able to clean the equipment with relevant product residues before cleaning validation begins.] 4

Institute of Validation Technology

Birgitte Holst

Technology transfer of the cleaning process does not encompass any focus on the microbial condition in the equipment after cleaning, because this parameter is not scalable. The risk this implies to the patient and the product should rule the challenge and monitoring of this parameter in the individual equipment setup - in the pilot facility as well as in the manufacturing facility. The primary scope of this article is to establish timely cleaning process understanding and a perspective for the use of this knowledge during manufacturing clinical trials. The intention is to present important general aspects for handling cleaning processes in development when preparing for technology transfer (Part I) as well as when manufacturing for clinical trials (Part II). Included with the general aspects some specific examples for performing risk assessment are stated. The examples included are from the author’s daily experience with cleaning validation in Novo Nordisk A/S, a company manufacturing biotechnology active pharmaceutical ingredients (API) as well as related parenteral finished pharmaceuticals.

PART I DEVELOPMENT OF A CLEANING PROCESS Timeliness

What is timely process understanding and timely development of a cleaning process? Understanding of a cleaning process should be documented as part of the development report. Set aside time for the development of the cleaning process as product development progresses. Cleaning process understanding can imply almost as many “preparing for Good Manufacturing Practice (GMP)” disciplines as traditional development of drug product processes or API processes require, but the time spent is a minimal fraction when invested at the right moment. When developing a cleaning process, the process should be as effective, as safe, as environmentally sound, and as cost-efficient, as possible. The technology transfer of a product to the manufacturing plant should encompass a cleaning process designed for manufacture: a robust and scalable cleaning process. The cleaning process in the manufacturing plant must support expectations from authorities as indicated in Reference 1: Food and Drug Administration (FDA) Draft Guidance from 2005, “Amended New Drug Applications (ANDA) - Impurities in Drug Products:”

Cleaning Validation

5

Birgitte Holst

“We recommend that ANDA sponsors develop robust formulations and manufacturing processes that are based on sound state-of-the-art scientific and engineering principles and knowledge. Although routine manufacturing variations are expected, significant variation in batch-to-batch impurity levels or an unusually high level of impurity may indicate that the manufacturing process of the drug substance is not adequately controlled or designed.”

Costly time in the manufacturing facility can be wasted if a mistake in design of equipment or cleaning process occurs because of a lack of understanding. Gathering know-how before the technology transfer can help to foresee some of the difficulties the manufacturing facility might face when scaling-up the cleaning processes. Gathering Know-how

What kind of know-how is needed and what effort does it require? You do not have to spend a lot of extra time and resources when gathering know-how about the cleaning process, just remember to relate the gained results from product development (e.g., regarding degradation, stability, etc.) to the production and cleaning process residues in the equipment. Know-how and process understanding of the cleaning process includes knowledge about: • Level of toxicity and potency (basis for establishment of Maximum Allowed Carry-over (MACO)) and cleanability (basis for worst-case evaluations) of : ✓ Product residues ✓ Degradation of products generated in the time lapse to cleaning ✓ Degradation of products generated in the cleaning process • Impact on stability from carry-over of product residues, cleaning agent, and degradation products (basis for establishment of MACO) • Level of toxicity of the cleaning agent and residues from the cleaning agent after the cleaning process (basis for establishment of MACO) • Rationale for the: ✓ Functionality and optimal concentration of the cleaning agent ✓ Cleaning process sequences (e.g.: pre-rinse, alkaline wash, rinse, acidic wash, rinse, drying) 6

Institute of Validation Technology

Birgitte Holst

✓ Cleaning process parameters (time, temperature, time lapse from use to cleaning) • Compatibility of the agent with different types of surface materials Know-how and process understanding in development can come from: • Experiments (preliminary studies in discovery, process development studies, process challenge studies) • Experience and data obtained from test and toxicology productions • Experience and data obtained from similar processes (e.g.: existing processes) • Scientific rationales Can we, in advance, scientifically justify application of a cleaning process when we do not have the final equipment for testing? Yes, if we base the scientific justification for applying the same cleaning process to different products on similarities from product to product, compare residues from different products in the rationale, and discuss similarities in characteristics. In addition, laboratory tests can justify similarities or differences in cleanability. When introducing a new product in an existing manufacturing plant, can we reduce the cleaning validation effort by performing laboratory tests? Yes, if former cleaning validation studies can be nominated as worst-case scenarios and cover cleaning validation studies for the new product partly or totally, when scientific justification for a better cleanability is present. If any doubt is present, verify the theory in practice by performing a confirmation study at final scale at essential, hardest to clean areas. Let the risk that the cleaning process applies to the next product rule the scope of the confirmation study. Remember that impact from differences in the process time and time lapses prior to cleaning can change the cleaning scenario and necessitate a nomination of a new worst-case scenario. This situation should lead to new validation studies despite the scientific justification for better cleanability of the different products in the small scale studies. Perform cleanability studies, e.g., as laboratory tests, but consider pros and cons when choosing the laboratory test method. Justify and document the use and limitations of the test.

Cleaning Validation

7

Birgitte Holst

Methods for cleanability comparison: • • •

Solubility in relevant solvents Simulated clean-in-place 2 (CIP) Dip test performed with stainless steel sintered tablets 3: ✓ Dip in cleaning solvent with controlled stirring ✓ Analyse tablets by weighing before and after dipping and drying • Using radioactive marked material • Other tests

Use only well characterised laboratory tests with documented reproducibility when tests are used to compare different scenarios and limit final scale studies.

PART II CLEANING PERSPECTIVE FOR CLINICAL TRIALS MANUFACTURE In this section, some of the scientific rationales related to the tasks in cleaning validation will be handled in the context of the effort needed for documenting a clean surface in a pilot facility. Once all the equipment in a pilot facility is qualified, the author has found that the majority of the effort for defining a cleaning validation should be performed at the desk, because it is important that everything we do, or choose not to do, is justified with scientific rationales. These rationales must be present prior to performing any test or study that will be used to document the status for the equipment. We seldom have identical plants with identical production schedules; therefore, the arguing and risk assessment should be documented in an individual cleaning rationale for the specific product and pilot facility in question. Most of the arguments and risk assessment possibilities discussed below can enter into a cleaning validation rationale for a manufacturing plant as well as for a pilot facility. Qualification of the Equipment – a Precondition To ensure consistency of an automated cleaning process, the state of 8

Institute of Validation Technology

Birgitte Holst

control for a qualified cleaning process should be kept by configuration control and appropriate surveillance of cleaning process parameters: Time, Action, Concentration, and Temperature (TACT). Validation in Pilot Facilities

Must a cleaning process in a pilot facility be validated? As with all other questions regarding validation, the answer is: “It depends!” It depends upon the manufacturing situation. For example, is reproducibility of the cleaning process to be used or needed in the manufacturing context? Could means other than validation contribute to the safety of the patients receiving the clinical trial product? Yes, verification or monitoring from time to time can be an option, too. The definitions are as noted in Figure 1. Figure 1 Definitions Term

Cleaning Validation

Cleaning Verification

Cleaning Monitoring

Definition Documented verification that the cleaning process, can perform effectively and reproducibly, based on the approved cleaning method and cleaning acceptance criteria. A confirmation by examination and provision of objective evidence that specified cleaning requirements have been fulfilled. Verification studies must be planned and documented in the same manner as validation studies. Documented routine examination performed as a process control based on approved cleaning acceptance criteria. Monitoring studies are usually planned in standard operating procedures (SOP) and documented as part of the batch record.

To do nothing is not an option because the cleaning process has the possibility to contribute to the API and/or the drug product with impurities not originated from the production process. These impurities are oftentimes contamination beyond product specification. Contamination beyond product specification must always be minimized to a well-defined and acceptable level whether generated by the cleaning process or by residues from the former production process. Verification and monitoring may provide the insurance needed for avoiding risk to the product and patient, but how to decide what to do will be discussed in the next sections. Cleaning Validation

9

Birgitte Holst

Why is control of residues after a cleaning process needed in preclinical and early clinical studies, when it is a fact that APIs in these studies have a high level of impurities, impurities originated from the production process? Because impurities originated from sources other than the production process or the product itself are regarded as contaminants and can blur the preclinical and clinical results.4 Taken to its logical conclusion, residues can be toxic, and must be questioned as needed. In other words, the contribution to the impurity profile for the next batch or the next product depends on the scenario for what is going on in the equipment before or while the cleaning process is performed prior to production of the next batch or the next product.

The Regulatory Frame Interpretation of regulatory requirements can answer some of the tricky questions. Do authorities expect us to perform cleaning validation or verification in a pilot facility? Yes they do. The regulatory frame for cleaning is as follows: • API: [Reference 5] - ICH Q7A, EU GMP Part II • Drug products: [Reference 6] - EU GMP Annex 13 The International Conference on Harmonization (ICH) Q7A and the European Union (EU) GMP Part II say: “During all phases of clinical development, equipment should be clean and suitable for its intended use.” EU GMP Annex 13 says:

“In clinical trials there may be added risk to participating subjects compared to patients treated with marketed products. The application of GMP to the manufacture of investigational medicinal products is intended to ensure that trial subjects are not placed at risk, and that the results of clinical trials are unaffected by inadequate safety, quality, or efficacy arising from unsatisfactory manufacture. Equally, it is intended to ensure that there is consistency between batches of the same investigational medicinal product used 10

Institute of Validation Technology

Birgitte Holst

in the same or different clinical trials, and that changes during the development of an investigational medicinal product are adequately documented and justified. “The toxicity, potency, and sensitising potential may not be fully understood for investigational medicinal products and this reinforces the need to minimise all risks of cross-contamination. The design of equipment and premises, inspection or test methods and acceptance limits to be used after cleaning should reflect the nature of these risks. Consideration should be given to campaign working where appropriate. Account should be taken of the solubility of the product in decisions about the choice of cleaning solvent.”

The requirements are interpreted this way: • A scientifically sound rationale must give an account of the risk for carry-over to both the next batch in a campaign and the next product when bringing a new product into a pilot facility. • When there is a risk to the patient or the product, we must document that the risk is minimised to an acceptable level by cleaning validation, verification, or monitoring. The context will rule what to do. A Changing Environment

Cleaning processes and documentation of the equipment’s cleaning status is a greater challenge in a pilot facility than in a manufacturing facility because new products or changed products are introduced all the time. Developing means that changes can happen from the production of one batch to the production of the next batch of the same product. Such a scenario can lead to a change in product residues before cleaning and the carry-over of different residues in a campaign as illustrated in Figure 2. In most phases, performing cleaning validation in process development will not be appropriate because changes in the production process can result in changes of residues. As partly indicated in Figure 2, it is a fact that a cleaning process generates different impurities dependent upon the product residues, the actual equipment, and the use patterns for the actual equipment. The simple illustrations in Figures 3 and 4 show examples of other implications for the changing environments in development. The initial production processes for preclinical and clinical batches might

Cleaning Validation

11

Birgitte Holst

not generate identical impurities and all relevant impurities might not have been part of the preclinical batches (no toxicology study has implied the impurities). That is one more very good reason for testing the cleaning status of the equipment with validated test and analytical methods prior to production of a clinical batch production. The test and analytical methods must be validated in relation to the content in the equipment prior to the cleaning process. The challenge is to relate existing validated processes and methods to new situations. The new situation might compromise the validated state of the processes and methods. A sound and scientific cleaning rationale for the new or changed product must account for whether or not the validated state rules or a new or modified approach must be used. Cleaning Rationale for a New Product in the Pilot Facility The scientifically sound rationale must take into account the risk of bringing a new product into a pilot facility. Three types of carry-over are in question: • Carry-over to the next batch in a campaign • Carry-over to downstream processes if equipment is shared • Carry-over to the product when product change-over occurs Accounting for the risk implies a scientific discussion about: • How risk from some residues can be regarded as mitigated • How accept limits for what is left will be established • How compliance to accept limits will be documented

Figure 2 Residue Carry-over: Changes in residues can provide changes in carry-over despite manufacturing the same product in development Residues from B

12

Institute of Validation Technology

Birgitte Holst

Figure 3 Illustration of a Scenario that Provides One Matrix of Carry-over to Product B

Figure 4 Illustration of a Scenario Providing another Matrix of Carry-over to Product B

Figure 5 Batch-to-batch Contamination Can Occur in a Campaign Residues after clean

Production mode

Equipment ready for cleaning

Cleaning process

Clean equipment

Residues after clean

Production mode

Equipment ready for cleaning

Cleaning process

Clean equipment

Production mode

Figure 6 Downstream Contamination Is Possible when Equipment Is Shared A in process A

Production mode

Residues from A

Equipment ready for cleaning

Cleaning process

Residues after clean

Clean equipment

A in process B

Production mode

Residues from B

Equipment ready for cleaning

Residues after clean

Cleaning process

Clean equipment

Cleaning Validation

A in process A

Production mode

13

Birgitte Holst

Figure 6 Downstream Contamination Is Possible when Equipment Is Shared A in process A

Production mode

Residues from A

Equipment ready for cleaning

Cleaning process

Residues after clean

Clean equipment

A in process B

Production mode

Residues from B

Equipment ready for cleaning

Residues after clean

Cleaning process

Clean equipment

A in process A

Production mode

Carry-over to the Next Batch in a Campaign A pilot facility often manufactures in campaigns, a kind of dedication. Carry-over of residues to the next batch in a campaign does not have the same perspective as carry-over to the next product and the only concern should be that the batch receives acceptable limits of: • Degradation products • Residues from the cleaning process If the product is a final dosage form, you must, in addition, look at the impact of the strength of the product. ➣ Scenarios in a campaign •

If the product is produced in a campaign with no changes in the campaign that can affect the residues as indicated in Figure 5 and Figure 6, perform appropriate verification and monitoring documenting that batch-to-batch contamination is acceptable. ✓ If the campaign is long, the verification and monitoring studies can be collected with a conclusion as though it were a validation study. ✓ Afterward, no verification need be performed because the validated state for cleaning between batches has been established.

•

If the product is produced in a campaign and changes as indicated in Figure 2 take place, perform appropriate concurrent verification and monitoring and document that batch-to-batch contamination is acceptable all through the campaign.

14

Institute of Validation Technology

Birgitte Holst

Figure 7 Next Product Contamination after Change-over Product A

Residues from A

Residues after cleaning process

Product B with residues from A

When producing as in either scenario one or two, verification should document that the equipment is clean before change-over to alternative production can take place. The scope of this study might be quite different from the study that documents no risk for batch-to-batch carry-over. Carry-over to Downstream Processes when Equipment Is Shared A batch can receive carry-over from the upstream process when two or more steps in a production process are performed in the same equipment. This scenario is illustrated in Figure 6. Carry-over to the Product when Product Change-over Occurs Contamination of next product, illustrated in Figure 7, can come from three sources: • Residues from a former production process • Degradation products from a former production process • Residues from the cleaning process ➣ Scenarios for change-over Compared to contamination in a campaign, a simpler picture can be drawn when having a product change-over, but the risk is usually higher to the patient and the product. A higher risk should elicit a greater effort when documenting the cleaning status of the equipment. If the product comes into the equipment only once, verification should document that the equipment is clean before change-over so alternative production can take place.

Cleaning Validation

15

Birgitte Holst

How Risk from Some Residues Can Be Regarded as Mitigated

Could residues from different products be compared and justification found for applying the same cleaning process to different products without applying full blown cleaning validation or cleaning verification? Yes, but be careful and scientific. Process knowledge and understanding are the key words. Use experience from former studies when preparing the rationales prior to entering a new product in the equipment and the associated facility. The approach to be used is founded on what usually is referred to as the matrix approach, product grouping or bracketing, in a manufacturing multi-purpose plant. The new product should be defined with regard to the following attributes: • Composition • Solubility • Cleanability • Treatment regime • Potency • Toxicology • Allergenicity (e.g.: mammalian host cell proteins) If the new product, compared to products already manufactured and tested in the facility, can be regarded as less challenging in each of the above mentioned groups, but the last, the rationale can state that already performed cleaning validation is representative for the new product. This cannot be done for allergenicity. When there is a risk of generating levels of concern of allergenic compounds, special attention must be taken. The rationale for evaluating that the initial, validated cleaning process still applies, must also come from the use patterns of the equipment in the production process: • Identical production processes regarding: ✓ Process conditions: temperature, time, agitation, etc. ✓ Mother liquor/content in the process stream ✓ Time lapses • Possibilities or risks for upstream or downstream cross-contamination Use a risk-based approach to evaluate the areas of risk: • Limit the cleaning verification study to these areas • Balance the scope of the study to the risk it implies • Small changes may call for a confirmation study only 16

Institute of Validation Technology

Birgitte Holst

•

Appropriate monitoring should be applied, e.g.: visual clean

Residues from the Former Production Process Cleaning process knowledge and understanding is crucial when defining product residues from a production process. Product development activities should have revealed, to some extent, the composition of the product including impurities present in the product. Impurities may already be defined with regard to potency and toxicity. Use information from already performed degradation and stability studies combined with toxicology studies. This information is important knowledge for risk assessing and for establishing limits.

➣ Almost identical processes The former production process could be almost identical to the next production process. The rationale for evaluating that the initial, validated cleaning process still applies can be established on process knowledge and understanding from: • • •

Clarification studies performed in the laboratory regarding solubility and ratio of cleanability of product residues Similarities in substances, e.g.: of broth and strain of organisms, solvents, salts, preservatives, etc. Similarities in process steps

For example, in fermentation it can be argued that the same strain of organism produces approximately the same impurities and, therefore, the challenges for the cleaning process could be identical for different products. ➣ Inactivation In addition, the rationale can be built on process knowledge and on an understanding of the ability of the cleaning process to inactivate the product residues. Again, already performed degradation studies can provide some science to this rationale. Degradation Products from the Former Production Process In the interval between the production process and the cleaning process, degradation products can be generated. The residues from the production process can undergo changes that will render a different

Cleaning Validation

17

Birgitte Holst

cleaning scenario compared to the cleaning scenario immediately after the production process: • Moisture will evaporate from the surface and leave residues more fixed to the surface • Air can promote the degradation process (oxidation) • Microbial growth can promote the degradation process The amount of degradation products can be lowered by limiting the time lapses from use to cleaning. Some knowledge regarding degradation products can already be present e.g., in API recovery or purification steps, degradation products might be similar to impurities already present in residues from the former production process. Degradation products = By-products

Residues from the Cleaning Process

“Compliance must be documented by adherence to acceptance criteria derived from established MACO. Documentation for compliance will then be found in trend reports, verification, or validation reports.”

Degradation products from the cleaning process and residues from the detergent should be considered. If the cleaning process performs satisfactorily by using time, action, and temperature only (TAT and not TACT), residues can be mitigated by not using detergent. By using agents that are identical to raw materials used in the production process (e.g., acids and alkaline solutions) limits can be established from using knowledge about the products’ sensibility to these agents (e.g.: pH challenge and justification studies). Residues from acids and alkaline solutions can easily be tested by simple methods (pH and conductivity).

Final Production Steps At early steps in product manufacture for APIs, a high level of by-products is present and later purification steps are specifically designed to take 18

Institute of Validation Technology

Birgitte Holst

care of these impurities. The rationale for not testing carry-over from the cleaning process can be process knowledge and understanding. If by-products from the production process and degradation products from the cleaning process are shown to be identical and later purification steps are designed and tested to remove the relevant level of impurities of this origin, no validation or verification should be needed. Contrary to this, special attention should be given to the final production steps where no purification steps can remove any impurities. The total amount of carry-over will contribute to the impurity profile of the final products. What if we have limitations for scientific testing? In some situations we do not have any product that can be used for the development of test methods for cleaning verification before the first production has taken place. Then, we can always reduce the risk by: • Ensuring visible cleanliness • Using the first batch after a change-over only for test and not for clinical trials • Using a blank run with the relevant mother liquor (relevant to the change-over batch), to ensure a homogeneous distribution of possible residues in the liquor and analyse a sample of the liquor for content of the previous product • Using disposable equipment where leachables are eliminated as a potential risk • Handling the equipment in dedicated cleaning cycles • Placing the used equipment in quarantine until the clean status can be documented with validated methods A “Forced Ranking” tool for systematically evaluating the risk and decision-making is provided in the Appendix.[See Reference 7 and Acknowledgement 2.]

How Maximum Allowed Carry-over will be Established How MACO will be established has been covered by many other articles, but two subjects under this heading should be discussed: How to continue when next product is unknown? MACO and the related acceptance criteria for product A in question must be calculated by using knowledge of the next product B.

Cleaning Validation

19

Birgitte Holst

Typically, that knowledge is not present when planning production of A, but by using a worst-case approach for carry-over, an approximation can be made: • Smallest batch size and lowest potency ever seen in the pilot environment • Largest area of shared surface When the next product B is known, check that the data match the worst-case approach: • If the new worst-case scenario is not present – you are home safe • If the new worst-case scenario is present – you must show a lower carry-over of the previous product A How will compliance to MACO be documented? We can document conformance to specifications for cleaning by cleaning validation, cleaning verification, or cleaning monitoring. Which to use should be discussed and scientifically justified in the rationale. Documents Compliance must be documented by adherence to acceptance criteria derived from established MACO. How to comply must be planned either in SOP’s, verification, or validation protocols. Documentation for compliance will then be found in trend reports, verification, or validation reports. If verification studies are done three or more times, the verifications studies might be presented with a conclusion as though it were a validation study. Thereafter, the validated state for cleaning prior to change-over to next product is in place and no further studies are needed. ➣ Controlling the validated state Being in the validated state must appear from planning and controlling the campaign e.g.: in SOP’s and in process control records.

20

Institute of Validation Technology

Birgitte Holst

CONCLUSION Cleaning validation might apply in a pilot facility, but if the risk assessment shows that cleaning validation studies in a campaign do not provide higher quality to the product and cleaning validation only raises the development costs, cleaning validation should not be done. Validation of a cleaning process toward a specific product in a pilot facility can be of limited value because the production process changes concurrently with the development phases, other means, such as cleaning verification and monitoring, may be sufficient. When change-over to next product occurs, means other than cleaning validation can show that risk to patient and product quality has been mitigated by a well performed cleaning process. The effort for control of a new product in a plant must be looked upon in context of former assessments and former cleaning activities. Gather know-how and reuse already gathered know-how. If a proper effort for the cleaning process is invested in development, it can speed up the time to market. In addition, an effort for gathering cleaning process understanding in development is in-line with Process Analytical Technologies (PAT) thinking. Cleaning processes will continue to be a challenge with the same velocity as the industry invents new products. ❏

REFERENCES 1. FDA draft guidance, 2005: ANDAs: Impurities in Drug Products 2. Alfredo Canhoto, Ph.D. Wyeth BioPharma. “A Novel Bench Scale Apparatus to Model and Develop Biopharmaceutical Cleaning Procedures.” Journal of Validation Technology, Vol. 11, Nov. 2004. 3. Borer Chemie are using the method as part of proactive R&D of cleaning agents to customers. http://www.borer.ch 4. ICH Q6B: Specifications: Test Procedures and Acceptance Criteria for Biotechnological / Biological Products 5. ICH Q7A identical to EU GMP Part II: Basic Requirements for Active Substances used as Starting Materials 6. EU GMP Annex 13: Manufacture of Investigational Medicinal Products 7. Rebecca Brewer, Dober Group, 1-800-323-4983, [email protected]

Cleaning Validation

21

Birgitte Holst

ACKNOWLEDGEMENTS 1. Great thanks to my patient husband and banker, Henrik Rong, former novice in cleaning, but in theory no more, he made the article readable for English-speaking readers. 2. The author extends great thanks to Rebecca Brewer for an always inspired approach to cleaning validation. [See reference number 7.]

ABOUT THE AUTHOR Birgitte Holst has worked for Novo Nordisk for more than 20 years, always with validation as a theme. Ms. Holst’s main experience has been gathered in the manufacture of aseptic products and biotech API’s. Currently, she holds the position as manager for a corporate expert team in validation that is serving as internal consultants. Ms. Holst holds a M.Sc. in Pharmacy from The Danish University of Pharmaceutical Science. She may be reached by telephone at +45 4444 8888 / + 45 3075 8019 or by email at: [email protected]. Originally published in the April 2006 issue of the Journal of GXP Compliance

22

Institute of Validation Technology