Cleaning validation in the pharmaceuticals industry:

Ferrum – Supplier of Centrifuge Technology Solutions for Solid/Liquid Separation

Category: Centrifuge Technology Solutions

Good pharmaceuticals manufacturing practice requires from pharmaceuticals companies that rooms and apparatus such as centrifuges and other devices must be cleaned according to written methods (“Good Manufacturing Practice” or GMP).

The most suitable method must be validated by the respective pharmaceuticals company on the basis of regulatory requirements [1] and their own expertise and technological advances in apparatus engineering. This takes place as part of cleaning validation; and this is precisely where the innovations of Ferrum in the area of vertical scraper centrifuges offer further alleviations.

Not without “my” risk assessment

The word cleaning validation represents a real challenge to the pharmaceutical, apparatus and plant engineering industries. This does not just simply involve complying with regulatory standards. The safety of pharmaceuticals, feasibility and efficiency are main aspects.

At the start of every cleaning validation is the validation plan, which can be divided into three phases, see also [2], [3]. The providers of centrifuge technology solutions can make an essential contribution in all three phases towards realisation and efficiency. This can only be achieved by working together and harnessing all available relevant knowledge available.

The three phases can be briefly described as follows:

1)Internal status inspection of planned production line
This concerns the question of which active and inactive ingredients are to be produced or used? The product change frequency has a considerable influence on the efficiency. One must therefore know which cleaning agent and method should or can be used.

This is where the latest innovative VBC centrifuge technology comes in; based on the expertise of the machine supplier in apparatus engineering and construction in line with the latest advances in mechatronics as well as design aspects and process sequences of solids-liquid separation and cleaning. The machine supplier is not responsible for the active ingredients however.

2)Risk assessment of products and facilities
The internal pharmaceuticals status inspection must be followed by risk assessment for all products, the aim being to identify substances that are a particular hazard. Responsible is the pharmaceuticals company, see graphic “worst-case” analysis according to Borchert [6].
Centrifuge manufacturers can make a valuable contribution with their expertise and years of experience in the assessment of design-related cross-contamination (e.g. difficult to access or absorbent surfaces, dead ends in pipes and extraction points, etc.).

3)Determination of extent of validation
On completion of the internal status inspection and joint risk assessment (machine supplier and pharmaceuticals company), the extent of validation can (must) be determined by the pharmaceuticals company, see also [4].

In this phase, Ferrum is able to offer the possibility of validating design-related critical points in its own assembly halls following assembly and so reduce by this verification item, time-consuming validation within the pharmaceuticals company. By means of a so-called riboflavin test, for example, the effectiveness and wettability within the centrifuge can be verified or also the effectiveness of CIP cleaning of inert material at “critical points”.

It is therefore in the interest of the pharmaceuticals company to complete validation quickly and if possible in the phase prior to commissioning. This is only possible in cases where existing facilities are duplicated. As maximum flexibility in the manufacture of pharmaceuticals is of the essence today, apparatus such as centrifuges must be appropriately flexible in design.

Precisely this step was taken at Ferrum with the latest VBC vertical pharmaceutical centrifuge. The processes of rinsing, washing, spraying, measuring, analysing, scraping, blowing out and even flooding have been greatly improved in the new generation of centrifuges over that which was state of the art a few years ago. The special advantage of Ferrum centrifuge technology solutions is that many of the different function modules can be integrated flexibly both in the initial design of the machine as well as retrofitting. The cleaning process can be optimally adapted to the respective production sequence in a highly flexible manner.

It is therefore extremely important that the experienced and innovative centrifuge supplier is included in process selection already in the planning phase.

Cleaning procedure validation and selection

In order to locate fouling on machine parts, specific samples are taken before and after the cleaning procedure. In the PIC document PI 006, sampling procedures using the wipe or swab test and flush or rinse test are considered suitable. [5]

One advantage of the swab test is that it provides information on where the fouling is located, e.g. in bends or branches of pipe systems.

Useful is the inclusion of global analytical methods. An example is TOC determination for organic loading, conductivity measurement for ionic residues and ph measurement for cleaning agent residue detection. These analytical methods can be included as online measurements or installed in the centrifuge. Such analytical methods can be used for multifunction systems to provide maximum flexibility during commissioning, as all possible active and inactive ingredients are often not known. Offered is a retrofitting option; this is usually possible in the majority of cases without redelivery to the manufacturer’s workshop due to the modular design of the VBC.

Modularity in use

Passive contribution towards cleanliness in scraper centrifuges

It may sound simple, but good access to the centrifuge is a precondition for its cleaning and analysis, even when fully automatic CIP systems are usually installed. The modular design of the VBC vertical pharmaceutical centrifuge takes this into account. The design of the cover opening, the position of the outlet and the basket drive can be selected in a wide range of variants and combinations. This enables the machine to be optimally adapted to local space conditions at the site of installation without additional expenditure; this is something that will be appreciated by structural engineers and plant constructors as well as those responsible for maintenance.

Until now, only so-called horizontal centrifuges where considered suitable for installation in a clean room. With the introduction of the VBC, a vertical scraper centrifuge now meets all requirements for installation in a clean room concept, as the complete drive can be arranged below the vibrating plate. This allows the technical area to be separated from the clean room area by means of a membrane in the floor/ceiling.

This method of installation complies with the wishes of many pharmaceuticals companies as the entire production flow takes place gravimetric vertical. The VBC vertical scraper centrifuge thus requires less space than a conventional horizontal centrifuge, as this additional clean room area is required for opening the horizontal housing and positioning the pipes with respect to the vertical product flow into the horizontal machine.

A further important item in the design of the VBC is the one-piece concept of the housing and base plate, avoiding numerous edges and transitions with the advantage of less fouling.

Active contribution towards cleanliness in scraper centrifuges

The function modules that contribute towards active cleanliness include CIP nozzles. The principle applies: As much as necessary, as little as possible. Especially in the case of multifunction systems, the use of an additional CIP nozzle may be necessary. In the modular concept of the VBC centrifuge, this does not present a problem as the CIP nozzles are flanged and easily retrofitted (during production) without welding. It goes without saying that these flanges are all provided with GMP compliant seals.

Those responsible in pharmaceuticals companies can face a far greater problem if the cleaning process must subsequently be changed from CIP cleaning to flood cleaning with a change of product, see “worst-case” analysis. This is another problem that can easily be solved with the design concept of the VBC, as the complete bearing and sealing system is in a modular design. The so-called bearing cartridge can be prefitted as a floodable version and exchanged for the fitted cartridge (during production); and all this without removing the complete centrifuge and sending it to the manufacturer’s works. After the conversion, the complete centrifuge can be flooded up to the cover.

A further important innovation measure was achieved by the optimal use of pneumatic drives for operating the scraper and cover, whereby movement is effected by self-locking spindle drives. All hydraulic oil is thus banned from the pharmaceuticals area and a further risk factor eliminated.

Conclusion

Innovation in the engineering industry today is based on a large proportion of modularisation of functional units. An innovative and modular design in combination with the latest control concept enables users in pharmaceuticals production to adapt machines precisely to suit their individual requirements and to bring function and costs into accord.

A technical feature does not often produce the best result with regard to reducing consumption costs. Only the sum total of all activities in one solution enables costs to be minimised within a production plant. In short, Ferrum “Centrifuge Technology Solutions.

SMOL Schneider; Dipl. Phys. ETHZ, Ferrum Marketing& Sales, Business Unit Centrifuge Technology /29.2.04

Bibliography
[1] EG-GMP Guide, Appendix 15

[2] Jörg Koppenhöfer; Efficient and cost-saving cleaning validation in the area of active ingredients and substances in multifunction systems, gempex, GmbH, Mannheim, Source: http://www.gempex.com/

[3] Dr. Bernd Köhler und Dr. Carsten Richling; Planing, implementation and documentation of cleaning validation in the pharmaceuticals industry, SWISS PHARMA 25 (2003) No. 9.

[4] FDA Guide for inspection of the validation of cleaning processes; http://www.fda.gov/ICECI/Inspections/default.htm

[5] PIC/S PI 006; Recommendations for validation master plan; GMP Consultant, GMP-Verlag, Schopfheim (2003)

[6] D. Borchert; Cleaning validation (Bd. 1, Chapter 8.B-8.K), GMP Consultant, GMP-Verlag, Schopfheim (2003)

HPLC Method Development and Validation for Pharmaceutical Analysis

HPLC Method Development and Validation for Pharmaceutical Analysis

The wide variety of equipment, columns, eluent and operational parameters involved makes high performance liquid chromatography (HPLC) method development seem complex. The process is influenced by the nature of the analytes and generally follows the following steps:

• step 1 - selection of the HPLC method and initial system
• step 2 - selection of initial conditions
• step 3 - selectivity optimization
• step 4 - system optimization
• step 5 - method validation.

Figure 1: A flow diagram of an HPLC system.

Depending on the overall requirements and nature of the sample and analytes, some of these steps will not be necessary during HPLC analysis. For example, a satisfactory separation may be found during step 2, thus steps 3 and 4 may not be required. The extent to which method validation (step 5) is investigated will depend on the use of the end analysis; for example, a method required for quality control will require more validation than one developed for a one-off analysis. The following must be considered when developing an HPLC method:

• keep it simple
• try the most common columns and stationary phases first
• thoroughly investigate binary mobile phases before going on to ternary
• think of the factors that are likely to be significant in achieving the desired resolution.

Mobile phase composition, for example, is the most powerful way of optimizing selectivity whereas temperature has a minor effect and would only achieve small selectivity changes. pH will only significantly affect the retention of weak acids and bases. A flow diagram of an HPLC system is illustrated in Figure 1.

Table I: HPLC detector comparison.

HPLC method development Step 1 - selection of the HPLC method and initial system. When developing an HPLC method, the first step is always to consult the literature to ascertain whether the separation has been previously performed and if so, under what conditions - this will save time doing unnecessary experimental work. When selecting an HPLC system, it must have a high probability of actually being able to analyse the sample; for example, if the sample includes polar analytes then reverse phase HPLC would offer both adequate retention and resolution, whereas normal phase HPLC would be much less feasible. Consideration must be given to the following:

Sample preparation. Does the sample require dissolution, filtration, extraction, preconcentration or clean up? Is chemical derivatization required to assist detection sensitivity or selectivity?

Types of chromatography. Reverse phase is the choice for the majority of samples, but if acidic or basic analytes are present then reverse phase ion suppression (for weak acids or bases) or reverse phase ion pairing (for strong acids or bases) should be used. The stationary phase should be C₁₈ bonded. For low/medium polarity analytes, normal phase HPLC is a potential candidate, particularly if the separation of isomers is required. Cyano-bonded phases are easier to work with than plain silica for normal phase separations. For inorganic anion/cation analysis, ion exchange chromatography is best. Size exclusion chromatography would normally be considered for analysing high molecular weight compounds (.2000).

Table II: The basic types of analytes used in HPLC.

Gradient HPLC. This is only a requirement for complex samples with a large number of components (.20–30) because the maximum number of peaks that can be resolved with a given resolution is much higher than in isocratic HPLC. This is a result of the constant peak width that is observed in gradient HPLC (in isocratic HPLC peak width increases in proportion to retention time). The method can also be used for samples containing analytes with a wide range of retentivities that would, under isocratic conditions, provide chromatograms with capacity factors outside of the normally acceptable range of 0.5–15.

Gradient HPLC will also give greater sensitivity, particularly for analytes with longer retention times, because of the more constant peak width (for a given peak area, peak height is inversely proportional to peak width). Reverse phase gradient HPLC is commonly used in peptide and small protein analysis using an acetonitrile–water mobile phase containing 1% trifluoroethanoic acid. Gradient HPLC is an excellent method for initial sample analysis.

Column dimensions. For most samples (unless they are very complex), short columns (10–15 cm) are recommended to reduce method development time. Such columns afford shorter retention and equilibration times. A flow rate of 1-1.5 mL/min should be used initially. Packing particle size should be 3 or 5 μm.

Detectors. Consideration must be given to the following:

• Do the analytes have chromophores to enable UV detection?
• Is more selective/sensitive detection required (Table I)?
• What detection limits are necessary?
• Will the sample require chemical derivatization to enhance detectability and/or improve the chromatography?

Fluorescence or electrochemical detectors should be used for trace analysis. For preparative HPLC, refractive index is preferred because it can handle high concentrations without overloading the detector.

UV wavelength. For the greatest sensitivity λ_max should be used, which detects all sample components that contain chromophores. UV wavelengths below 200 nm should be avoided because detector noise increases in this region. Higher wavelengths give greater selectivity.

Fluorescence wavelength. The excitation wavelength locates the excitation maximum; that is, the wavelength that gives the maximum emission intensity. The excitation is set to the maximum value then the emission is scanned to locate the emission intensity. Selection of the initial system could, therefore, be based on assessment of the nature of sample and analytes together with literature data, experience, expert system software and empirical approaches.

Table III: HPLC optimization parameters.

Step 2 - selection of initial conditions. This step determines the optimum conditions to adequately retain all analytes; that is, ensures no analyte has a capacity factor of less than 0.5 (poor retention could result in peak overlapping) and no analyte has a capacity factor greater than 10–15 (excessive retention leads to long analysis time and broad peaks with poor detectability). Selection of the following is then required.

Mobile phase solvent strength. The solvent strength is a measure of its ability to pull analytes from the column. It is generally controlled by the concentration of the solvent with the highest strength; for example, in reverse phase HPLC with aqueous mobile phases, the strong solvent would be the organic modifier; in normal phase HPLC, it would be the most polar one. The aim is to find the correct concentration of the strong solvent. With many samples, there will be a range of solvent strengths that can be used within the aforementioned capacity limits. Other factors (such as pH and the presence of ion pairing reagents) may also affect the overall retention of analytes.

Gradient HPLC. With samples containing a large number of analytes (.20–30) or with a wide range of analyte retentivities, gradient elution will be necessary to avoid excessive retention.

Determination of initial conditions. The recommended method involves performing two gradient runs differing only in the run time. A binary system based on either acetonitrile/water (or aqueous buffer) or methanol/water (or aqueous buffer) should be used.

Figure 2: The chemical structure of progesterone and Figure 3: Amount injected versus peak area of progesterone standard to demonstrate linearity.

Step 3 - selectivity optimization. The aim of this step is to achieve adequate selectivity (peak spacing). The mobile phase and stationary phase compositions need to be taken into account. To minimize the number of trial chromatograms involved, only the parameters that are likely to have a significant effect on selectivity in the optimization must be examined. To select these, the nature of the analytes must be considered. For this, it is useful to categorize analytes into a few basic types (Table II).

Once the analyte types are identified, the relevant optimization parameters may be selected (Table III). Note that the optimization of mobile phase parameters is always considered first as this is much easier and convenient than stationary phase optimization.

Selectivity optimization in gradient HPLC. Initially, gradient conditions should be optimized using a binary system based on either acetonitrile/water (or aqueous buffer) or methanol/water (or aqueous buffer). If there is a serious lack of selectivity, a different organic modifier should be considered.

Step 4 - system parameter optimization. This is used to find the desired balance between resolution and analysis time after satisfactory selectivity has been achieved. The parameters involved include column dimensions, column-packing particle size and flow rate. These parameters may be changed without affecting capacity factors or selectivity.

Table IV: Accuracy/recovery of progesterone from samples of known concentration.

Step 5 - method validation. Proper validation of analytical methods is important for pharmaceutical analysis when ensurance of the continuing efficacy and safety of each batch manufactured relies solely on the determination of quality. The ability to control this quality is dependent upon the ability of the analytical methods, as applied under well-defined conditions and at an established level of sensitivity, to give a reliable demonstration of all deviation from target criteria.

Analytical method validation is now required by regulatory authorities for marketing authorizations and guidelines have been published. It is important to isolate analytical method validation from the selection and development of the method. Method selection is the first step in establishing an analytical method and consideration must be given to what is to be measured, and with what accuracy and precision.

Method development and validation can be simultaneous, but they are two different processes, both downstream of method selection. Analytical methods used in quality control should ensure an acceptable degree of confidence that results of the analyses of raw materials, excipients, intermediates, bulk products or finished products are viable. Before a test procedure is validated, the criteria to be used must be determined.

Analytical methods should be used within good manufacturing practice (GMP) and good laboratory practice (GLP) environments, and must be developed using the protocols set out in the International Conference on Harmonization (ICH) guidelines (Q2A and Q2B).^1,2 The US Food and Drug Administration (FDA)^3,4 and US Pharmacopoeia (USP)⁵ both refer to ICH guidelines. The most widely applied validation characteristics are accuracy, precision (repeatability and intermediate precision), specificity, detection limit, quantitation limit, linearity, range, robustness and stability of analytical solutions. Method validation must have a written and approved protocol prior to use.⁶

Equation 1 and Figure 4: HPLC chromatograms of (a) progesterone reference standard; (b) separation of progesterone gel sample; (c) placebo formulation.

This article reviews and demonstrates practical approaches to analytical method validation with reference to an HPLC assay of progesterone (Figure 2) in a gel formulation. Progesterone is widely used for dysfunctional uterine bleeding or amenorrhoea,^7,8 for contraception (either alone or with, for example, oestradiol or mestranol in oral contraceptives) and in combination with oestrogens for hormone replacement therapy in postmenopausal women.^9,10

Experimental Chemicals and reagents All chemicals and reagents were of the highest purity. HPLC-grade methanol was obtained from Merck (Darmstadt, Germany). Progesterone reference standard was purchased from Sigma Chemicals (St Louis, Missouri, USA). Deionized distilled water was used throughout the experiments.

HPLC instrumentation The HPLC systems used for the validation studies consisted of Series 200 UV/Visible Detector, Series 200 LC Pump, Series 200 Autosampler and Series 200 Peltier LC Column Oven (all Perkin Elmer, Boston, Massachusetts, USA). The data were acquired via TotalChrom Workstation (Version 6.2.0) data acquisition software (Perkin Elmer), using Nelson Series 600 LINK interfaces (Perkin Elmer).

All chromatographic experiments were performed in the isocratic mode. The mobile phase was a methanol/water solution (75:25 v/v). The flow rate was 1.5 mL/min and the oven temperature was 40 ºC. The injection volume was 20 μL and the detection wavelength was set at 254 nm. The chromatographic separation was on a 25034.6 mm ID, 10 μm C₁₈ μ-Bondapak column (Waters, Milford, Massachusetts, USA).

Table V: Demonstration of the repeatability of the HPLC assay for progesterone.

Results and discussionLinearity and range The linearity of a test procedure is its ability (within a given range) to produce results that are directly proportional to the concentration of analyte in the sample. The range is the interval between the upper and lower levels of the analyte that have been determined with precision, accuracy and linearity using the method as written. ICH guidelines specify a minimum of five concentration levels, along with certain minimum specified ranges. For assay, the minimum specified range is 80–120% of the theoretical content of active. Acceptability of linearity data is often judged by examining the correlation coefficient and y-intercept of the linear regression line for the response versus concentration plot. The regression coefficient (r²) is .0.998 and is generally considered as evidence of acceptable fit of the data (Figure 3) to the regression line. The per cent relative standard deviation (RSD), intercept and slope should be calculated.

In the present study, linearity was studied in the concentration range 0.025–0.15 mg/mL (25–150% of the theoretical concentration in the test preparation, n=3) and the following regression equation was found by plotting the peak area (y) versus the progesterone concentration (x) expressed in mg/mL: y53007.2x14250.1 (r²51.000). The demonstration coefficient (r²) obtained for the regression line demonstrates the excellent relationship between peak area and concentration of progesterone. The analyte response is linear across 80-120% of the target progesterone concentration.

Accuracy A method is said to be accurate if it gives the correct numerical answer for the analyte. The method should be able to determine whether the material in question conforms to its specification (for example, it should be able to supply the exact amount of substance present). However, the exact amount present is unknown, which is why a test method is used to estimate the accuracy. Furthermore, it is rare that the results of several replicate tests all give the same answer, so the mean or average value is taken as the estimate of the accurate answer.

Some analysts adopt a more practical attitude to accuracy, which is expressed in terms of error. The absolute error is the difference between the observed and the expected concentrations of the analyte. Percentage accuracy can be defined in terms of the percentage difference between the expected and the observed concentrations (Equation 1).

Percentage accuracy tends to be lower at the lower end of the calibration curve. The term accuracy is usually applied to quantitative methods but it may also be applied to methods such as limit tests. Accuracy is usually determined by measuring a known amount of standard material under a variety of conditions but preferably in the formulation, bulk material or intermediate product to ensure that other components do not interfere with the analytical method. For assay methods, spiked samples are prepared in triplicate at three levels across a range of 50-150% of the target concentration. The per cent recovery should then be calculated. The accuracy criterion for an assay method is that the mean recovery will be 100±2% at each concentration across the range of 80-120% of the target concentration. To document accuracy, ICH guidelines regarding methodology recommend collecting data from a minimum of nine determinations across a minimum of three concentration levels covering the specified range (for example, three concentrations, three replicates each).

In the present study, the accuracy of the method was evaluated by recovery assay, adding known amounts of progesterone reference standard to a known amount of gel formulation, to obtain three different levels (50, 100 and 150%) of addition. The samples were analysed, and mean recovery and %RSDs calculated. The data presented in Table IV show that the recovery of progesterone in spiked samples met the evaluation criterion for accuracy (100±2.0% across 80–120% of target concentrations).

Specificity Developing a separation method for HPLC involves demonstrating specificity, which is the ability of the method to accurately measure the analyte response in the presence of all potential sample components. The response of the analyte in test mixtures containing the analyte and all potential sample components (placebo formulation, synthesis intermediates, excipients, degradation products and process impurities) is compared with the response of a solution containing only the analyte. Other potential sample components are generated by exposing the analyte to stress conditions sufficient to degrade it to 80–90% purity. For bulk pharmaceuticals, stress conditions such as heat (50–60 ºC), light (600 FC of UV), acid (0.1 M HCl), base (0.1 M NaOH) and oxidant (3% H₂O₂) are typical. For formulated products, heat, light and humidity (70-80% RH) are often used. The resulting mixtures are then analysed, and the analyte peak is evaluated for peak purity and resolution from the nearest eluting peak.

Once acceptable resolution is obtained for the analyte and potential sample components, the chromatographic parameters, such as column type, mobile phase composition, flow rate and detection mode, are considered set. An example of specificity criterion for an assay method is that the analyte peak will have baseline chromatographic resolution of at least 2.0 from all other sample components. In this study, a weight of sample placebo equivalent to the amount present in a sample solution preparation was injected to demonstrate the absence of interference with progesterone elution (Figure 4).

Precision Precision means that all measurements of an analyte should be very close together. All quantitative results should be of high precision - there should be no more than a ±2% variation in the assay system. A useful criterion is the relative standard deviation (RSD) or coefficient of variation (CV), which is an indication of the imprecision of the system (Equation 2).

According to the ICH,² precision should be performed at two different levels - repeatability and intermediate precision. Repeatability is an indication of how easy it is for an operator in a laboratory to obtain the same result for the same batch of material using the same method at different times using the same equipment and reagents. It should be determined from a minimum of nine determinations covering the specified range of the procedure (for example, three levels, three repetitions each) or from a minimum of six determinations at 100% of the test or target concentration.

Intermediate precision results from variations such as different days, analysts and equipment. In determining intermediate precision, experimental design should be employed so that the effects (if any) of the individual variables can be monitored. Precision criteria for an assay method are that the instrument precision and the intra-assay precision (RSD) will be ≤2%.

In this study, the precision of the method (repeatability) was investigated by performing six determinations of the same batch of product. The resulting data are provided in Table V, which show that the repeatability precision obtained by one operator in one laboratory was 0.28% RSD for progesterone peak area and, therefore, meets the evaluation criterion.

Table VII: Stability results of progesterone samples and standard solutions (n53).

The intermediate precision was demonstrated by two analysts, using two HPLC systems and who evaluated the relative per cent purity data across the two HPLC systems at three concentration levels (50%, 100%, 150%) that covered the assay method range (0.025–0.15 mg/mL). The mean and RSD across the systems and analysts were calculated from the individual relative per cent purity mean values at 50%, 100% and 150% of the test concentration. The data are presented in Table VI, and show ≤2.0% RSD, therefore, meeting the evaluation criterion.

Limits of detection and quantitation The limit of detection (LOD) is defined as the lowest concentration of an analyte in a sample that can be detected, not quantified. It is expressed as a concentration at a specified signal:noise ratio,² usually 3:1. The limit of quantitation (LOQ) is defined as the lowest concentration of an analyte in a sample that can be determined with acceptable precision and accuracy under the stated operational conditions of the method. The ICH has recommended a signal:noise ratio 10:1. LOD and LOQ may also be calculated based on the standard deviation of the response (SD) and the slope of the calibration curve(s) at levels approximating the LOD according to the formulae: LOD53.3(SD/S) and LOQ510(SD/S).

The standard deviation of the response can be determined based on the standard deviation of the blank, on the residual standard deviation of the regression line, or the standard deviation of y-intercepts of regression lines. The method used to determine LOD and LOQ should be documented and supported, and an appropriate number of samples should be analysed at the limit to validate the level. In this study, the LOD was determined to be 10 ng/mL with a signal:noise ratio of 2.9. The LOQ was 20 ng/mL with a signal:noise ratio of 10.2. The RSD for six injections of the LOQ solution was ≤2%.

Analytical solution stability Validation of sample and standard solution preparation may be divided into sections, each of which can be validated. These include extraction; recovery efficiency; dilution process when appropriate; and addition of internal standards when appropriate. Although extraction processes do not actually affect the measuring stage they are of critical importance to the analytical test method as a whole. The extraction process must be able to recover the analyte from the product; it must not lose (for example, by oxidation or hydrolysis) any of the analyte in subsequent stages, and must produce extraction replicates with high precision. For example, during analysis of an ester prodrug the extraction process involves the use of strongly alkaline or acid solutions, it may cause some of the prodrug to be hydrolysed and, therefore, give false results.

Reference substances should be prepared so that they do not lose any of their potency. Thus it is necessary to validate that the method will give reliable reference solutions that have not been deactivated by weighing so little that an error is produced; adsorption onto containers; decomposition by light; and decomposition by the solvent. If the reference is to be made up from a stock solution then it must be validated that the stock solution does not degrade during storage. Reagent preparation should be validated to ensure that the method is reliable and will not give rise to incorrect solutions, concentrations and pH values.

Samples and standards should be tested during a period of at least 24 h (depending on intended use), and component quantitation should be determined by comparison with freshly prepared standards. For the assay method, the sample solutions, standard solutions and HPLC mobile phase should be stable for 24 h under defined storage conditions. Acceptable stability is ≤2% change in standard or sample response, relative to freshly prepared standards. The mobile phase is considered to have acceptable stability if aged mobile phase produces equivalent chromatography (capacity factors, resolution or tailing factor) and the assay results are within 2% of the value obtained with fresh mobile phase.

In the present study, the stabilities of progesterone sample and standard solutions were investigated. Test solutions of progesterone were prepared and chromatographed initially and after 24 h. The stability of progesterone and the mobile phase were calculated by comparing area response and area per cent of two standards with time. Standard and sample solutions stored in a capped volumetric flask on a lab bench under normal lighting conditions for 24 h were shown to be stable with no significant change in progesterone concentration during this period (Table VII).

Robustness Robustness measures the capacity of an analytical method to remain unaffected by small but deliberate variations in method parameters. It also provides some indication of the reliability of an analytical method during normal usage. Parameters that should be investigated are per cent organic content in the mobile phase or gradient ramp; pH of the mobile phase; buffer concentration; temperature; and injection volume. These parameters may be evaluated one factor at a time or simultaneously as part of a factorial experiment. The chromatography obtained for a sample containing representative impurities when using modified parameter(s) should be compared with the chromatography obtained using the target parameters.

Conclusion Method development involves a series of sample steps; based on what is known about the sample, a column and detector are chosen; the sample is dissolved, extracted, purified and filtered as required; an eluent survey (isocratic or gradient) is run; the type of final separation (isocratic or gradient) is determined from the survey; preliminary conditions are determined for the final separation; retention efficiency and selectivity are optimized as required for the purpose of the separation (quantitative, qualitative or preparation); the method is validated using ICH guidelines. The validated method and data can then be documented.

References 1. International Conference on Harmonization, "Q2A: Text on Validation of Analytical Procedures," Federal Register 60(40), 11260–11262 (1995).

2. International Conference on Harmonization, "Q2B: Validation of Analytical Procedures: Methodology; Availability," Federal Register 62(96), 27463–27467 (1997).

3. FDA, "Analytical Procedures and Methods Validation: Chemistry, Manufacturing and Controls Documentation; Availability," Federal Register (Notices) 65(169), 52776–52777 (2000).

5. USP 25–NF 20, Validation of Compendial Methods Section (1225) (United States Pharmacopeal Convention, Rockville, Maryland, USA, 2002) p 2256.

6. G.A. Shabir, "Validation of HPLC Chromatography Methods for Pharmaceutical Analysis. Understanding the Differences and Similarities Between Validation Requirements of FDA, the US Pharmacopeia and the ICH," J. Chromatogr. A. 987(1-2), 57-66 (2003).

7. C.E. Wood, "Medicare Program; Changes to the Hospital Outpatient Prospective," Med. J. Aust. 165, 510–514 (1996).

8. A. Prentice, "Medical Management of Menorrhagia," Br. Med. J. 319, 1343–1345 (1999).

9. D.T. Baired and A.F. Glasier, "Hormonal Contraception," New Engl. J. Med. 328, 1543–1549 (1993).

10. P.E. Belchetz, "Hormonal Treatment of Postmenopausal Women," New Engl. J. Med. 330, 1062–1071(1994).

Statistically Justifiable Visible Residue Limits

The standard of visual cleanliness is commonly applied to the evaluation of surface contamination. Numerous published studies have examined the visually clean standard as a means of verifying cleaning effectiveness in pharmaceutical manufacturing, and methods for the quantitation of visible residue limits (VRLs) have been provided. Current methods for establishing VRLs are not statistically justifiable, however. The author proposes a method for estimating VRLs based on logistic regression.

Visually clean (VC), a term that refers to inspection with the naked eye, is a common cleanliness standard employed for evaluating surface contamination and cleaning in high-technology manufacturing, including that of pharmaceuticals, where surface cleaning is of utmost importance. The importance of the VC standard for pharmaceutical manufacturing is evident in the following facts:

• It is one of the acceptance criteria for establishing the limits for cleaning-validation (CV) studies (1)
• Visual examination of equipment surfaces for cleanliness immediately before use is required by good manufacturing practice (GMP) regulations (2)
• Even before the issuance of the GMP regulations, most companies used to a VC standard (3)
• A VC approach to controlling cross-contamination in processing and manufacturing operations provides a practical and effective method of risk management (4, 5)
• It is one of the means of evaluating cleaned surfaces during the development, optimization, and validation of cleaning processes
• It is the only tool available to operators for examining equipment surfaces to verify that they have been cleaned effectively
• Manufacturers employ it for routine monitoring of the cleaning process.

Table I: Advantages and disadvantages of the visually clean standard.

Many manufacturers believe that compliance with a requirement that the surface be visually clean ensures only the absence of gross amounts of contamination and may be regarded as the lowest cleanliness standard because of its subjectivity and variability. Many studies of VC as one of several criteria for evaluating surface contamination have been published. In light of its advantages and disadvantages, which are listed in Table I, visual inspection of surfaces, combined with a few simple tools, is still regarded as an effective and inexpensive primary way for evaluating surface cleanliness. Visually clean criterion for CV studies

In the pharmaceutical industry, cleaning is defined as limiting contamination to a level below practical, achievable, justifiable, and verifiable limits. CV is the documented evidence of cleanliness.

Common bases for establishing CV acceptance limits, as described in literature and in regulatory guidance documents, include the following (6, 7):

• Therapeutic daily dose
• Toxicological data
• The 10-ppm criterion
• The VC criterion.

Table II: Definitions of visually clean and visible residue limit from the literature.

The method that yields the lowest acceptance limit is selected, and the value is considered the maximum allowable carryover (MACO) limit for CV studies. The VC criterion still holds, however, and is independent from the established MACO limit. Regardless of whether the established visible residue limit (VRL) is lower or higher than the MACO values, noncompliance with the VC requirement indicates the failure of CV. The Pharmaceutical Inspection Convention and Pharmaceutical Inspection Cooperation Scheme (PIC/S) requires the VC criterion to be verified through well-documented spiking studies before it can be used for CV studies (1).

Because the VC standard is relevant to many technological areas, tremendous efforts have been devoted to defining and devising novel and efficient ways to develop justifiable and quantifiable VRLs for monitoring and validating cleaning procedures. Table II lists some definitions of the VC standard from the literature. The VC standard and VRLs are based on the following common principles:

• Particles deposited on the surface tend to reduce the reflection of light.
• The unaided human eye (with or without corrected vision) can detect particles as small as 40–50 μm under ideal conditions (8).
• The viewer's state of mind could affect his or her ability to detect the residue visually (e.g., the residue might be visible but unseen because the observer is inattentive).
• The brighter a residue in comparison with its background, the higher the probability of its detection through visual inspection.

Although the VC standard may be highly subjective, personnel have successfully quantified VRLs by establishing well-controlled experiments and programs. For industries other than pharmaceuticals, variables and parameters associated with the VC standard (e.g., viewing distance and light intensity) have been quantified and well documented (8).

The most popular method, henceforth referred to as the current method, for determining VRLs in the pharmaceutical industry involves spiking the selected material surface with known amounts of residue at concentrations of about 0–10 μg/cm². Trained inspectors then examine the surfaces under controlled viewing conditions (e.g., light, viewing angle, and viewing distance) for the presence of residue (9–11). The lowest level of residue that is detected is then considered the VRL for that particular residue. The only drawback with the method is that it is not statistically justifiable and, hence, not scientifically definable. The primary objective of this article is to establish a method for setting scientifically and statistically justifiable VRLs and to provide a meaningful definition of the VC standard.

Statistical limitations of the current method

One statistical limitation of the current method is that the VRL is determined based on observed data without describing a relationship between observations and the experimental parameters. Suppose that in a VC verification study, a residue is spiked at levels of 0, 0.5, 1.0, 2.0, 3.0, and 4.0 μg/cm². If all four inspectors detect residue at 2.0 μg/cm² and only three inspectors detect residue at 1.0 μg/cm², then the VRL would be 2.0 μg/cm², assuming that the residue levels between 1.0 and 2.0 μg/cm² would not be detected by all inspectors. The VRL is inappropriate because a residue level between 1.0 and 2.0 μg/cm² could possibly have been detected by all inspectors.

To predict the number of observers that would detect residue at levels other than those spiked (e.g., 1.5 μg/cm²), the observed data must be incorporated into a reasonable model that describes a relationship between an outcome and a set of independent variables. The results obtained from spiking studies for verifying the VC criterion are binary (i.e., only two values are possible) rather than continuous. The regulatory guidelines and the available literature do not explain how to establish a modeling procedure, based on these discrete responses, that could be used to derive VRLs, however.

Another important parameter in determining appropriate VRLs is the sample size (i.e., number of inspectors and total number of observations) for VC verification studies. Most published studies are based on relatively small sample sizes. Forsyth's studies are based on only four observers (4, 10, 11). Because the VRL depends on the proportion of detection (i.e. the number of detections of residue to the number of inspections), a small sample size increases the width of the confidence interval and the margin of error. Thus, an 0.8 proportion of detection with a sample size of five would result in a 95% exact-confidence interval of 0.2836–0.9949 and an approximately 35.57% margin of error. At the same proportion of detection, a sample size of 25 would result in a 95% exact-confidence interval of 0.5930–0.9317 and an approximately 16.94% margin of error. Because no consensus has been established about the appropriate number of observers for VC verification studies, the sample size of the study could cause over- or underestimation of VRLs.

Logistic regression

The objective of VC verification studies is to prove that the VC criterion would ensure cleanliness if implemented in the manufacturing setting. During VC verification, spiking studies are performed and inspectors state whether they can detect residue visually under controlled viewing conditions. VC verification studies thus provide a basis for the establishment of VRLs and the determination of appropriate viewing conditions. The lowest concentration of residue that is visually detected by all the observers is then used as VRL. In the current method, VRL could be defined mathematically as the lowest residue concentration for which the ratio of the number of observers able to detect the residue to the total number of observers is equal to 1. As discussed earlier, any knowledge about the outcome in future situations could not be obtained from the observed data unless the data were fitted with the most conservative model that explains the data.

One of the most common examples of modeling is the linear-regression technique. However, linear regression is not suitable for binary data. If we represent the binary responses "Yes" and "No" with values of 1 and 0, respectively, then the mean is the proportion of cases with a value of 1 and can be interpreted as proportion or probability of detection. Although the proportions and probabilities cannot exceed 1 or fall below 0, fitting the data with linear regression could give predicted values of the response variable above 1 and below 0. Clearly, linear regression is not appropriate when the data must lie between 0 and 1 because predictions from the model are not similarly constrained. Other problems that arise when fitting binary data with linear regression are that the variance of the error term is not constant and that the error term is not normally distributed.

Table III: The outcome of visually clean verification studies.

The most suitable modeling technique that could be applied to describe a relationship between explanatory variables (e.g., experimental parameters such as residue concentration, viewing distance, viewing angle, and light intensity) and the binary-response variable is logistic regression. Logistic regression allows scientists to predict probabilities of detection of residue based on experimental variables, which is an advantage compared with other prediction techniques. Logistic regression is a flexible and easily applied modeling technique that can be used effectively to model data with continuous or discrete explanatory variables. In addition, the technique accommodates response variables that are not normally distributed.

Figure 1: Plot of probability of detection against the residue concentration for the data presented in Table III.

The modeling procedure based on logistic regression is explained here using a hypothetical data set (see Table III). This data set may represent an ideal case and is certainly within the experience of those involved in conducting VC verification studies (4, 11, 12). The response variable is binary and indicates two different outcomes (i.e., "Yes" or "No") based on the detection of residue by the observers at a specific viewing condition. The continuous explanatory variable is the measure of the theoretical residue concentration spiked on the model surface. At each spiking level, five replicates indicate the total number of observers used for the visual inspection of the surface. The observed proportion and probability of detection for each spiking level is the ratio of observers that detected the residue to the total number of observers. Figure 1 shows a plot of these observed probabilities of detection. It suggests that the probability of detection increases with the spiked residue concentration. The relationship is nonlinear, however, and the probability of detection changes little at the high extreme of spiked residue. This pattern is typical because proportions and probabilities cannot lie outside the range of 0 to 1.

Although the relationship between observed probability of detection and the residue concentration is nonlinear, a generalized linear modeling technique can be applied to these data. The logistic-regression technique fits the observed data with a linear model, the parameters for which are estimated using the maximum likelihood technique. Next, logistic regression transforms this linear model into a nonlinear logistic curve also known as an S-shaped or sigmoid curve. Logistic regression can therefore be seen as the conversion of a linear model into a nonlinear model that is naturally suited to the description of a binary response variable (13). The link function, commonly known as logit (the logarithm of odds), is used for converting the linear model to nonlinear logistic model and vice versa.

The logistic regression model is represented by the following equation (13):

in which P(Y =1) represents the predicted probability of response being equal to 1 (i.e., the predicted probability of detection); e is the exponent function; β₀, β₁, β₂, ... β_k are coefficients estimated from the data (obtained using the method of maximum likelihood); x₁, x₂, ... x_k are independent variables; and k is the number of independent variables.

The term β₀ + β₁x₁ + β₂x₂ ... + β_k x_k is the logit function and is defined as a natural logarithm of odds (e.g., the probability that an observer detects the residue divided by the probability that he or she does not detect it). Odds are expressed by the following equation:

For the data presented in Table III, which involves only one independent variable (i.e., residue concentration), logit = β₀ + β₁x₁. Once a meaningful relationship is defined between spiked residue concentration and probability of detection, VRL could easily be obtained from the regression model.

Results and discussion

Table IV: Results of fitting the logistic regression model to the data presented in Table III.*

The logistic-regression model is applied to develop a relationship between the residue concentration and its probability of detection. The summary of model parameters, obtained after fitting the data with logistic regression, is provided in Table IV. With a 0.1-μg/cm² increase in the residue concentration on the model surface, the odds of detecting the residue increase by 42.710%. An increase in concentration of this magnitude is likely to increase the probability of detection by 20.150–69.506%. Thus, at the 95% two-tailed level of significance, the residue concentration does have an effect on the response variable. The statistical significance of regression coefficients is tested using the Wald X² statistic. Table IV shows that the estimated coefficient for residue concentration has a p value of less than 0.05, indicating that the coefficient is probably not zero using an α level of 0.05. Therefore, residue concentration is a significant predictor of the probability of detection. The goodness-of-fit tests, with p values ranging from 0.979 to 0.999, indicate that the model fits the data adequately. In other words, the null hypothesis of a good model fit to data is tenable.

Table V: Relationship between spiked residue concentration and predicted probability of detection based on logistic regression model.

To understand fully the predictions made from the model, all values were transformed into direct measures of probability, and confidence intervals for these probabilities were derived using the method of Fleiss et al. to describe the uncertainty associated with fitting the model (14). Table V lists the logit and predicted probabilities of detection for each residue concentration obtained after fitting the data with logistic regression. Logit and predicted probabilities, when plotted against residue concentration, gave a straight line and an S-shaped curve, respectively, as may be seen in Figure 2. Table V shows that the higher the predicted probability of detection for a residue concentration, the more likely an observer will visually detect the residue. A comparison with the data in Table III implies that, based on the current method, the VRL should be 1.80 μg/cm² because it is the lowest concentration of residue for which the observed probability of detection is equal to 1. However, based on the logistic regression model (see Table V), the predicted probability associated with a residue concentration of 1.80 μg/cm² is only 0.949 (i.e., approximately 95 out of 100 observers are predicted to detect the residue).

Figure 2: Logistic regression model of the relationship between residue concentration and (a) Logit and (b) probability of detection. CI is 95% confidence interval, O is observed probability of detection, and P is predicted probability of detection.

Table V also compares observed and predicted probabilities and lists the expected number of detections for each residue concentration calculated as the predicted probability multiplied by the number of observers in each category. In logistic regression, the standard error and confidence interval for the model-based probabilities tend to be much smaller and narrower, respectively, than the ones based on the sample proportion. Instead of using a sample of only five observations as the current method does, the logistic-regression model uses information from all the observations in estimating the probabilities of detection. The result is a more precise estimate. Using the logistic-regression model to estimate the probabilities of detection instead of simply using observed proportions is therefore justifiable

Figure 3: Point of intersection for the logistic curve and given acceptance criterion. CI is confidence interval.

Irrespective of which method is used, two aspects of considerable importance in determining VRLs are the acceptance criterion for establishing VRLs and the number of observers participating in the VC verification studies. The acceptance criterion indicates the probability of detection (observed or predicted) at which a specific residue concentration could be regarded as the VRL. For the current method, the acceptance criterion is 1, which is based on the idea that if all the observers are able to detect a specific residue concentration during the VC verification phase, then the same residue concentration could also be detected by an observer in future situations. However, the same acceptance criterion could not be used for the logistic-regression model because the probabilities that it predicts can neither be less than 0 nor greater than 1.

Table VI: Predictions (point estimates) derived from the logistic regression model.

To estimate VRLs at different acceptance criteria, the model was inverted to estimate the concentrations that yield a certain response probability. The residue concentration at which the given acceptance criterion intersects with the logistic curve (see Figure 3) was determined. Table VI shows the residue concentrations thus obtained. The notation P_x (e.g., P₅₀) denotes the residue concentration that would give a response of x% according to the model (e.g., the probability that the residue would be detected by 50% of the observers). Confidence intervals for P_x were then derived using Fieller's theorem.

These point estimates provided a framework for evaluating the reliability of the logistic model. The logistic models and associated point estimates could be considered reliable if the observed probability of detection was found to be consistent with the predicted probability of detection. If one assumes 0.999 to be approximately equal to 1, then the VRL for the given residue should be 2.921 μg/cm², which is larger than the one obtained with the current method (see Table VI). Thus, based on the logistic-regression model, the residue concentration at 2.921 μg/cm² is predicted to be detected by all the observers with a 95% confidence interval of 2.266–4.761. Because the probability of detection increases significantly with an increase in spiked-residue concentration, setting higher acceptance criteria would give larger VRLs. Table VI shows that as the acceptance criterion approaches 1, the relationship requires a larger change in the explanatory variable to have the same effect as a smaller change in the explanatory variable at the middle of the curve. For example, a change in the predicted probability of detection from 0.9 to 0.99 requires a larger change in residue concentration (i.e., a change of 0.674 μg/cm²) than does a change in the probability from 0.5 to 0.6 (i.e., a change of 0.114 μg/cm²). Similarly, the confidence interval for these VRLs would tend to be wider as the acceptance criterion increases. Logistic regression may provide a much larger VRL than the current method. However, manufacturers may achieve lower VRLs by adjusting the acceptance criterion.

Unlike continuous responses, binary responses require a large number of observations. The more trials are attempted, the more accurate the estimated probability is. For VC verification studies with a small number of observers, a large number of observations with some replicates at each spiking level is recommended. However, for an accurate estimation of sample size, one may use the formula proposed by Hsieh et al. (15).

Logistic regression, as previously described, can be generalized to incorporate more than one explanatory variable, which may be continuous or categorical. However, care should be taken when interpreting and reporting results from multiple logistic-regression models. To correctly interpret the results from a multiple logistic-regression analysis and arrive at meaningful conclusions, appropriate steps must be taken to incorporate statistical interaction or curvilinear effects properly (e.g., including additional x₁ × x₂ or polynomial terms such as x₁² in the systemic component of the model) (13). If the logistic coefficient for the product or polynomial term is not statistically significant, then the interaction or curvilinear effect is not statistically significant. One problem that may arise while modeling multiple explanatory variables is that sometimes the value of one or more independent variables may raise the probability of the dependent variable close to 1, therefore the effects of other variables cannot have much influence. In that case, such variables should be excluded from the model or individual VRLs should be determined for the most appropriate viewing conditio

Conclusion

Logistic regression was demonstrated to be a better approach than the current method for estimating accurate and statistically justifiable VRLs based on discrete responses. It has the advantage of always making biologically meaningful predictions and, in most cases, its predictions closely reflect observations. Logistic regression should be used to determine VRLs rather than current method. Because the model may provide a much larger VRL than the current method does, the quality of visual inspection, in terms of the discrete response, can be improved by properly controlling the experimental variables and defining the acceptance criterion for the estimation of VRL.

Based on the modeling procedure, VRL can be defined as a scientifically justifiable residue concentration that, when viewed with the unaided eye, as measured by a specific method, would be detected by the observers with a predefined acceptance criterion. Once established, the VRL could then be used for CV and routine monitoring purposes. It would be appropriate to define the VC criterion as the absence of all particulate and nonparticulate contaminants above the established VRL from the surface when viewed with the unaided eye under preverified viewing conditions.

M. Ovais is a senior pharmaceutical scientist at Xepa-Soul Pattinson, 1-5, Cheng Industrial Estate, 75250 Melaka, Malaysia, tel. +60 63351515, fax +60 63355829, mohammad@xepasp.com
.

References

1. PIC/S, "Recommendations on Validation Master Plan, Instal-lation and Operational Qualification, Non-Sterile Process Validation, Cleaning Validation," (PIC/S, Geneva, Aug. 2002).

2. PDA Pharmaceutical Cleaning Validation Task Force, PDA J. Pharm. Sci. Technol. 52 (6), 1–23 (1998).

3. W. Hall, J. Val. Technol.14 (1), 42–49 (2007).

4. R.J. Forsyth, J. Hartman, and V. Van Nostrand, Pharm. Technol. 30 (9), 104–114 (2006).

5. R.J. Forsyth and J. Hartman, Pharm. Eng. 28 (3), 1–10 (2008).

6. European Chemical Industry Council, Guidance on Aspects of Cleaning Validation in Active Pharmaceutical Ingredient Plants, (CEFIC, Brussels, Dec. 2000)

7. G.L. Fourman and M.V. Mullen, Pharm. Technol. 17 (4), 54–60 (1993).

8. NASA, "Space Shuttle, Contamination Control Requirements (Lyndon B. Johnson Space Center, Houston, TX)," SN-C-0005, Rev. D, 1–3 (July 20, 1998).

9. D. A. LeBlanc, Pharm. Technol. 22 (10), 136–148 (1998).

10. R.J. Forsyth, V. Van Nostrand, and G. Martin, Pharm. Technol. 28 (10), 58–72 (2004).

11. R.J. Forsyth and V. Van Nostrand, Pharm. Technol. 29 (10), 152–161 (2005).

12. R.J. Forsyth and V. Van Nostrand, Pharm. Technol. 29 (4), 134–140 (2005).

13. G. Hutcheson and N. Sofroniou, "Logistic Regression," in The Multivariate Social Scientist: Introductory Statistics Using Generalized Linear Models (Sage Publications, London, 1st ed., 1999), pp. 113–152.

14. J.L. Fleiss, B. Levin, and M.C. Paik, "Logistic Regression," in Statistical Methods for Rates and Proportions, W.A. Shewart and S.S. Wilks, Eds. (John Wiley and Sons, Hoboken, NJ, 3rd ed., 2003), pp. 284–339.

ns.

Cleaning Validation Procedures

CONTAMINATION CONTROL

By Eric Lingenfelter, Wes Atkins, and Henry Evans

It’s Clean, but Can You Prove It?

Validation and revalidation are key when establishing cleaning methods

ALL IMAGES COURTESY OF Lancaster laboratories inc.

Editor’s Note: This article is the second in a two-part series on cleaning validation methodology. Part one, “How to Improve Cleaning Processes,” appeared in our June issue.

Method limits, selection of cleaning techniques, and selection of method detection were addressed in the first part of this article. Part two will address method validation, the importance of stability for cleaning validation samples, when revalidation of a cleaning method is necessary, the use of correction factors, and how to handle failing results.

Once the basic cleaning procedure elements have been established (establishment of limits, cleaning procedures, master plan, cleaning protocols, development of analytical method), the method is ready to be validated.1 This section outlines typical components utilized to validate the analytical method. The validation components presented below are based upon International Conference on Harmonisation (ICH) and United States Pharmacopeia (USP) guidelines.

Accuracy/Precision: The accuracy of an analytical procedure expresses the closeness of agreement between the value that is accepted either as a conventional true value or an accepted reference value and the value found. The precision of an analytical procedure expresses the closeness of agreement (degree of scatter) between a series of measurements obtained from multiple sampling of the same homogeneous sample under the prescribed conditions.2

Accuracy/precision should be assessed using a minimum of three concentration levels, each prepared in triplicate. Accuracy/precision is typically performed with concentrations ranging from 80 to 120% of the final theoretical sample concentration (based upon the maximum contamination limit or MCL), although a wider range may be more appropriate in certain instances.

Swab accuracy determines a method’s ability to recover the compound of interest directly from the swab head.

There are various categories of accuracy/precision that need to be established as part of the method validation.

• Solution accuracy is the measurement of the compound of interest added directly to the extraction solution. Reference standard solution is spiked directly into the diluent to prepare the three levels of concentration. Solution accuracy is performed as a control—usually in triplicate—to prove recovery of the analyte from the extraction solution. This recovery can then be compared to both swab and surface accuracy recoveries. If rinseates are being analyzed, this test is the only one needed to prove accuracy/ precision.
• Swab accuracy determines the method’s ability to recover the compound of interest directly from the swab head. These studies are performed by directly adding standard material to the swab head and then extracting as per the analytical method. Typically, three replicate-spiked swabs are prepared at the high and low concentrations, while six replicates are prepared at the 100% level.
• Surface accuracy determines the method’s ability to recover the compound of interest directly from a defined surface. Coupons of the defined surface material are spiked with reference standard at the three concentration levels mentioned above. The area of the coupon spiked with standard is dependent on the actual cleaning procedure and the surface area typically sampled after manufacturing. Typical surface areas sampled are 25 to 100 cm2.

Acceptance criteria should be evaluated and determined during the development of the analytical methods. Many factors influence the establishment of appropriate criteria, including the surface being swabbed, MCL, type of swab, and instrumentation. Intermediate accuracy/precision should be performed by a second analyst repeating the accuracy/ precision tests listed above. If multiple surfaces are involved in the validation, the second analyst can perform accuracy/precision on select surfaces (worst case) if appropriate.

Linearity: The linearity of an analytical procedure is its ability, within a given range, to obtain test results that are directly proportional to the concentration (amount) of analyte in the sample.2

A minimum of five concentration levels are typically evaluated, with duplicate injections at each level. Concentrations ranging from the limit of quantitation to 200% of the MCL are typically evaluated during validation. Acceptance criteria are generally based upon either the correlation coefficient or the coefficient of determination of the linear plot. In addition, criteria can be established around both the slope and Y-intercept of the plot.

Table 1. Summary of Typical Validation Components (Click to Enlarge)

Specificity: Specificity is the ability to assess the analyte unequivocally in the presence of components that may be expected to be present.2

Swab type and surface type are typically evaluated to determine if interferences are present in the method. Although each of these components is typically examined during method development, they should be included in the validation process and shown, under protocol, to have little or no interference. Swabs and surfaces are prepared as per the method without the introduction of the analyte of interest. Interference from either the swab or surface should be less than 10% of the MCL. Lower limits for specificity may be appropriate depending on the method conditions.

Limits of Detection and Quantitation: The limit of detection (LOD) of an individual analytical procedure is the lowest amount of analyte in a sample that can be detected but not necessarily quantitated as an exact value. The limit of quantitation (LOQ) of an individual analytical procedure is the lowest amount of analyte in a sample that can be quantitatively determined with suitable precision and accuracy.2

Both the LOD and LOQ should be verified by a suitable number of preparations known to be prepared near the respective limit being evaluated.4 LOD and LOQ can be estimated using a signal-to-noise approach with typical values of three-to-one for LOD and 10-to-one for LOQ.

During method validation, standard solutions are prepared at the estimated LOD and LOQ (three preparations for LOD and three preparations with duplicate analyses for LOQ). Typical acceptance criteria for LOD require that the analyte be detected in each analysis. For LOQ, the percent recovery is determined for each of the six measurements and should fall between 75% and 125% recovery. The relative standard deviation is also determined for the six measurements and should be less than 25%.

Robustness–Chromatographic Conditions: The robustness of an analytical procedure is a measure of its capacity to remain unaffected by small but deliberate variations in method parameters and provides an indication of its reliability during normal usage.2

Instrument and reagent variations, for example, may be examined as part of robustness to ensure that the method provides reliable data under varying conditions. Robustness is not a critical validation component according to ICH guidelines, but should be considered on a case-by-case basis. Robustness may be determined during development of the analytical procedure, and if measurements are susceptible to variations in analytical conditions, these should be suitably controlled, or a precautionary statement should be included in the procedure.3

Stability–Stock Standard, Working Standard, Working Swab: Stability of stock standards, working standards, and working swab or rinseate samples are evaluated as part of the validation. Stability can be evaluated under various conditions such as refrigeration or protection from light, but ambient conditions are preferred. This stability period is necessary to ensure that cleaning validation samples can be collected, shipped to the testing facility, and analyzed. The length of stability is particularly important for cleaning validation sample solutions and should be at least one week old, preferably two weeks.

Table 1 lists a summary of the validation components involved in a typical cleaning validation, along with examples of typical acceptance criteria that can be set for those tests. Please note that the acceptance criteria are listed for informational examples only and that the actual acceptance criteria must be determined on a case-by-case basis, depending on validation specifics.

Cleaning validation procedures should be revalidated when the equipment train of the manufacturing process is changed.

Revalidation

Cleaning validation procedures should be revalidated when the equipment train of the manufacturing process is changed. Possible changes in the equipment train include the surface type utilized and/or surface area, which can lead to the establishment of a new MCL. A full validation can usually be avoided, and only certain elements of the cleaning validation need to be revalidated. If the new limit is within the previously established linear range, only surface recoveries (bracketing the new limit) and surface residue specificity need to be revalidated. These same two elements must be revalidated if a surface type is changed.

If the new limit is outside the previously established linear range, linearity must be extended above or below the new limit, and swab recovery, surface recovery, and surface residue specificity need to be revalidated. For a new limit below the established linear range, a new standard concentration at this level may be recommended. If the method is not linear through the new level, however, a new standard concentration is necessary. A new standard concentration requires a full revalidation.

Other possible but less likely reasons to revalidate swab recovery, surface recovery, and surface specificity include a change in the type of swab or swabbing pattern. For a change in swab type, swab specificity also needs to be revalidated. For any of the previously listed changes, elements that do not require revalidation are LOD and LOQ.

The prior revalidation discussion assumes that the validated method was for swab samples and not for rinse samples. For rinse samples, validation elements involving swabs and surfaces do not need to be conducted. Additionally, any changes in the synthesis of the drug substance, changes in the composition of the finished product, or changes in the analytical procedure require revalidation according to ICH guidance.2

There are a variety of swabs to pick from, but when a change in swab type takes place, swab specificity also needs to be revalidated.

Correction Factors

Sometimes in cleaning validation studies, it is determined that not all the residue on a surface can be fully recovered, thus producing lower recoveries. In these instances, it may be necessary to apply a recovery factor. If a recovery factor is deemed appropriate, several issues must be considered before it is set:

• Recovery factors are usually not applied if recovery results are above 70%; however, there is no standard limit.
• Recovery factors must be set under sound scientific justification.
• Recovery factors should not be used if recoveries are too low. (For example, if recoveries are consistently around 10%, a 10X factor would not be appropriate.)
• Recovery factors need to be set prior to or during validation, not during routine monitoring.
• All results used to determine the recovery factor need to be consistent and reproducible.

Recovery factors are often seen as a last resort to salvage a mediocre method. Recovery method optimization should always be explored as an alternative prior to using recovery factors.

No matter which scientific field you are in, the question of how to handle failing data during routine sample testing will arise; the world of cleaning validation is no different.

Failing Data

No matter which scientific field you are in, the question of how to handle failing data during routine sample testing will arise; the world of cleaning validation is no different. The best way to approach this issue is to address it before it becomes a problem. When developing the cleaning validation master plan or protocol, dedicate a section to appropriate handling of failing results. Here, a step-by-step investigation of the results can be laid out in advance, so that decisions won’t be made based on instance-by-instance circumstances. When reviewing data, regulatory agencies like to see that failing results were handled in a consistent and systematic manner.

All data that do not meet protocol or master plan acceptance criteria need to be treated as a deviation. They must be handled by first being verified, resolved, and approved. This may require that samples be retested. Sometimes more samples may need to be collected to verify outlying results. If results indicate that a criterion or limit is not attainable under set conditions, modifications to the method, protocol, standard operating procedure, or master plan may be entertained. Again, all of these scenarios should be investigated during the feasibility/method development/validation stage of the cleaning validation study. n

Lingenfelter, Atkins, and Evans are senior chemists in the method development and validation group at Lancaster Laboratories Inc. For more information, reach Lingenfelter at (717)656-2300 begin_of_the_skype_highlighting (717)656-2300 end_of_the_skype_highlighting, ext. 1449, or at elingenfelter@lancasterlabs.com.

References

1. Active Pharmaceutical Ingredients Committee (APIC). Cleaning validation in active pharmaceutical ingredient manufacturing plants. Washington, DC: APIC; 1999. Available at: http://apic.cefic.org/pub/4CleaningVal9909.pdf. Accessed July 10, 2009.
2. International Conference on Harmonisation (ICH). Harmonised tripartite guideline: validation of analytical procedures: text and methodology Q2(R1). Geneva, Switzerland: ICH; 2005. Available at: http://www.ich.org/LOB/media/MEDIA417.pdf. Accessed July 10, 2009.
3. United States Pharmacopeia/National Formulary. USP 32/NF 27, General Chapters: <1225>Validation of Compendial Procedures. Rockville, Md.: United States Pharmacopeial Convention; 2009.