OR WAIT null SECS
Validation of analytical methods can be more easily accomplished by breaking the task down into a series of planned steps.
Regulatory guidance documents are written by committees, resulting in statements that are both exact and generic. Meeting regulatory requirements involves not only interpreting these documents correctly but also addressing their omissions. This article provides practical guidance on issues that are not thoroughly covered by current guidance documents regarding validation of analytical methods for biopharmaceuticals.
Four of the existing regulatory guidance documents on methods validation state, "Methods validation is the process of demonstrating that analytical procedures are suitable for their intended use."1–4 We have all read, and likely used, this phrase many times when summarizing method-validation results. According to Muire-Sluis, development scientists often point out that "validated methods may not be valid."5 The question therefore arises, what exactly makes a validated method valid? According to the Center for Biological Evaluation and Research (CBER), "the acceptability of analytical data corresponds directly to the criteria used to validate the method."4
Table 1. Validation characteristics per ICH Q2(R1) and relevant product specifications
We can generate evidence for the validity of analytical data in the formal method-validation program where all critical parameters are extensively tested under a detailed protocol that includes scientifically justified and logical step-by-step experimental approaches. All planned data sets must fall within pre-established protocol acceptance criteria limits. These criteria should be derived from and justified in relation to historical data and product specifications. Once evidence for all critical elements is provided, the validated method will become the official, licensed procedure for that particular product and process step, and it will then support production and product release. The relationship between "valid" or "suitable and validated" is often overlooked, but there is a high price when "validated" test systems are simply inappropriate.
Incentives to replace existing licensed test procedures may come from regulatory agencies, or they could be motivated by potential cost savings, ease of use (automation), and the opportunity to generate more accurate and reliable results.
The International Conference on Harmonization (ICH)'s Q2(R1),1 should be used for basic guidance. However, following just these guidelines will not necessarily produce a "valid" method and may not provide sufficient evidence that this method is suitable for product release. The intent of USP 30 <1225> is to provide guidance only on validation requirements for test methods for inclusion into USP with the expectation that validated USP methods still require verification from users.6–7 The US Food and Drug Administration (FDA) and European Medicines Agency (EMEA) provide guidance on some of the scientific issues that are not covered by Q2(R1).
Process Map. A process map showing the recommended steps for the selection, development, validation, and potential transfer of analytical methods, illustrating all proposed functional responsibilities was developed. Frequently, larger companies have separate functional units for method development, validation, and testing. The process flow in Figure 1 describes an ideal sequence of steps for better analytical method validation (AMV).
The rigorous standards suggested here are ideal but they are not necessarily required or followed during method development. Methods can be developed without strict adherence to good manufacturing practices (GMP) regulations if adequate documentation systems are used.
Many new technologies are now available for biopharmaceutical development. These analytical advances and their appropriate application are discussed in detail in the literature.8–10 Because these technologies are constantly improving—resulting in shorter testing times and increased throughput, ease of use, sensitivity, selectivity, and precision—at some point existing methods will be replaced with better ones. Automating a procedure, resulting in long-term savings and fewer operator errors, is one reason to follow this process. A more sensitive method may increase the likelihood of observing impurities at an upstream-process stage where corrective action is less expensive.
Whether it is the initial license procedure, an update, or a replacement to a licensed procedure, qualified personnel should carefully select a new test methodology and its appropriate instrumentation. Changing biological assays (for instance, impurity testing by enzyme-linked immunoabsorbant assay [ELISA]) requires extra care. Literature should be reviewed before selecting a new method. Accuracy is a prime consideration, because any bias in results must be reflected in the release specifications.11 When replacing existing technologies with automated or more sensitive ones, alert and action levels and associated specifications must be adjusted if needed. In-process and product specifications should reflect production process consistency and analytical capabilities, unless otherwise dictated by regulatory authorities.11
Current GMP guidelines state that GMP documentation and the detail of validation activities should increase as the production process progresses.12 Testing upstream stages may actually be more critical than many final container release tests because it provides evidence of fermentation quality and the efficiency of impurity removal—although the tests are more uncertain and variable. Final container testing attests that active and inactive formulation components remain at predicted levels with little variability.
Science- and risk-based testing should carefully evaluate different product quality attributes that can impact overall product quality. Testing for in-process impurities (such as host cell proteins) should emphasize overall measurement sensitivity, selectivity, and precision. In other words, the analytical method should detect batch-to-batch variations; whether the measurements are extremely accurate (100% recovery) is not as important.
Some of the most advanced and innovative analytical technologies may be extremely informative for characterization of product, impurities, or the product matrix, but may not be appropriate for product release testing. When selecting an appropriate quality control (QC) method, the pros and cons should be carefully weighed against each other. Solid evidence that the new method will provide equivalent or better results is necessary when submitting a license change to regulatory authorities. The method's requirements should be similar to instrument requirements and based on the expected capability of the new method, as determined by a careful data review and identification of critical assay characteristics.
It is the responsibility of the analytical method development (AMD) scientist to include the test method's details in the standard operating procedure (SOP), including optimization of assay elements (such as mixing volumes, number of replicates, and statistical data reduction). If practical, all AMD data should ideally be generated in a GMP environment. In other words, we should generate all development data with qualified equipment by qualified personnel, and properly document and summarize the data in an AMD report approved by quality assurance (QA).
Often, methods are not developed from the ground up, but are optimized for a particular product and product matrix. In any case, always follow a thorough optimization process, which includes incorporating the best-fit data reduction function (for example, five-parameter, logistic parallel-line statistics for an ELISA assay).13,14
A well-planned and controlled experimental design that emphasizes QC release-testing suitability will prevent multiple, unsuccessful trial-and-error efforts. Scientific and regulatory concerns must be balanced with potential economical restrictions. There are excellent tools published by the American Society for Standards and Testing (ASTM) to establish efficient design of experiment (DOE) templates and to help establish appropriate system suitability criteria.15–19
QA approval is required at many points in method development and validation (as indicated in red in Figure 1). Ideally, the process does not continue until it has been approved. Data generated using a final, optimized method may be used to set acceptance criteria for the AMV protocol. All instruments and equipment should be qualified and all relevant software should be validated, ensuring that all AMD data and results (summarized later in the AMV protocol) are valid from a compliance perspective. The main tasks of analytical method development and optimization (indicated in yellow in Figure 1) are discussed below.
ICH Q2(R1) Validation. The ICH Q2(R1) validation requirements should be evaluated twice: once (at least partially) during or after method optimization and once formally during the AMV studies. We need to know before writing the AMV protocol whether the method is suitable to support a desired process capability in relevant in-process or product (target) specifications. Table 1 lists all validation characteristics that typically must be evaluated for each analytical procedure. The corresponding anticipated in-process and product specifications for the new method form a checklist. An alternate, graphical presentation of the product specifications can be found in Figure 2. This is a typical example, not a universal table.
Results in and outside the product specifications must be reliable. If the boundaries are fuzzy, it is not possible to clearly differentiate between acceptable and unacceptable (out-of-specification) results, and material may be improperly accepted or rejected. The analytical method and instrumentation must be capable of bracketing the assay ranges required by ICH Q2(R1) (illustrated with red arrows in Figure 2) to ensure that these requirements can be met during AMD, AMV, and routine testing. This is why it is critical that instrument and method requirements (design qualifications) are thoroughly considered during selection of the new method.
Figure 2. All routine release specifications are shown that will generate numerical results. Pass-fail (or present-absent) is not shown, as it cannot be represented graphically. The green (bold) arrows represent examples of product specifications. The valid assay range of the new method must be capable of bracketing the product specifications. This is reflected in the ICH Q2(R1) requirements (yellow arrows) for assay range.2 The instrument (and method) design requirements must bracket the ICH Q2(R1) requirements. The red (dotted) arrows represent examples of the instrument (and method) design requirements for assay range. A, B, C, and D correlate to ICH Q2(R1) categories 2, 3, 4, and 5 as presented in Table 1.
Assay Bias and Analyte Response Factors. All analytical procedures are associated with a degree of bias, particularly biological assays that test for the purity, potency, and molecular interactions of biopharmaceuticals.5 Also, appropriate reference standards may not be readily available, because the product may be one of a kind. The evaluation of the assay's accuracy and bias can be the most difficult part of the development and validation process. Comparing the results of the new method to those of the old method is often meaningful only when accounting for assay bias. As long as we can compensate for the bias by modifying release specifications, we should be able to properly assess the quality of the production process and the product and remain in compliance.
Accuracy can be estimated by measuring the recovery of various spiked levels of particular analytes. Many critical assays of product purification efficiency and product quality determine product purity and impurities simultaneously (for example, protein composition by capillary zone electrophoresis [CZE] or high performance size exclusion chromatography [HP-SEC]). Whenever relative percentages of various analytes are estimated using a single assay, response factors must be established and integrated (normalized) in the calculations in order to accurately report purity and impurity levels. Using different detectors to measure analyte signals (for example, HP-SEC with ultraviolet detection to measure protein aggregation versus laser-light scattering or refractive-index detection) affects these relative percentages and should be thoroughly evaluated during AMD. A simple way to directly compare response factors from various detectors during AMD is to connect all detectors in-parallel (or inline).10
Stability. Samples, standards (secondary, in-house, or working), controls, and critical reagents should be evaluated for degradation during storage and potential freeze-thaw cycles. For final container testing, the analyte and matrix of samples, in-house standards, and controls may be similar since they all could come from the same production process. In any case, the negative effects of time on the bench during actual testing (room temperature), repetitive freeze-thaw cycles, and long-term storage of all materials used to generate test results should be evaluated and expiration times should be established. Reviewing and integrating historical data from the previous development and validation of the current method or the stability program may save time. Reagent expiration times should be evaluated carefully—any degradation could negatively impact test result quality. In general, vendor certificates of analysis can be used as evidence of reagent stability unless reagents are diluted or otherwise changed before storage and use.
System Suitability. The test system must be properly controlled to ensure reliable release-testing results. The system suitability criteria should be established during the AMD and optimization phase. This is usually accomplished by running a set of control points. For each test, system suitability will be satisfied (valid test results generated) if all control points are in established limits. A test system must be able to reproduce measurable results of a homogeneous sample (control) to allow examination of differences between batches. Small differences in batches are normal and acceptable, but the sources of variation should be identified. Ultimately, we will have more certainty when we can separate differences in production batches from assay variability.20
Sample Suitability. Technically, sample suitability is part of system suitability so these parameters can be evaluated together. Sample suitability should be established during AMD and should ideally ensure that samples, controls, and standards are prepared identically and run simultaneously. In addition, sample suitability should include a statistical analysis of the number of replicates needed to generate significant release results. Single measurements may be acceptable if the production-process sampling can deliver truly batch-representative samples and the precision of assay repeatability is high compared to the product specifications—and therefore high compared to the batch-to-batch variation on which these specifications are based. Often, however, assay precision is relatively low, and multiple measurements will substantially increase the level of certainty in the corresponding test results. Several detailed standard practices are published by ASTM.15–19
Also, we should evaluate batch sampling to ensure capturing and accounting for apparent variability. For example, product potency in final containers may vary among samples taken from the beginning, middle, and end of fill due to protein adsorption or protein aggregation during fill. In this case, samples should be taken from each of the three fill stages to ensure a more accurate representation of the batch's average potency.20
Statistical Data Reduction. Technically, statistical data analysis also is part of overall system suitability. Use multiple statistical values (for instance, regression line correlation coefficients or p-values for the testing of standard-to-sample line parallelism) to verify that test system performance is acceptable.
Sometimes, data transformation (for example, logarithmic conversion) may lead to improved linearity. However, most biological assay response curves are not linear, even after mathematical transformation.20 These are particularly difficult to deal with and should only be handled by experienced statisticians. Nonlinear models, such as four or five-parameter logistic functions with weighted factors and tests for parallelism, may be the best approach.20 Just as different test methodologies have different biases, changing statistical models may significantly change the final results. Some models may simply be inappropriate or may not provide acceptable results over the entire assay range.20
Robustness. Robustness, defined as the lack of a significant effect when small changes are deliberately introduced into the test system, should ideally be addressed during the method optimization phase and not as part of AMV. We should know the degree of robustness of a method before starting the formal AMV phase. Critical test system parameters (for example, the acceptable range of diluting the test sample) must be identified and controlled with appropriate operational limits. These limits should be described in the AMD report and documented in the method SOP. The SOP will then contain operational limits which are in the context of the overall system suitability criteria and which are adhered to during the validation phase. In addition, robustness should be tested in the AMD phase during or after method optimization because significant differences in the AMV results (from challenging the critical operational limits) may be difficult to explain in the AMV report.20
We must remember that AMV is the formal evidence that this method is suitable to be used under strictly controlled QC testing conditions. The AMV protocol should be set up to deliver this evidence through appropriate acceptance criteria by varying sample batches and concentrations, operators, instruments, days, and other factors that are expected to vary during routine testing—in established sample and system suitability conditions and operational limits.20
1. International Conference on Harmonization (ICH). Q2(R1), Validation of analytical procedures. Current Step 4 Version. Geneva, Switzerland; 2005.
2. Eurachem Guide. Fitness for Purpose of Analytical Methods. Teddington, UK; 1998.
3. Center for Drug Evaluation and Research. Guidance for Industry. Bioanalytical Method Validation. Bethesda MD; 2001.
4. Center for Biologics Evaluation and Research. Draft Guidance for Industry. Analytical procedures and methods validation. Bethesda, MD; 2000.
5. Muire-Sluis A. Presented at Biological Assay Development and Validation; 2004 Apr 26–28; San Diego, CA.
6. United States Pharmacopoeia. USP 30 <1225>. Validation of Compendial Methods. 2007.
7. Pharmacopoeial Forum 32 <1226> Verification of Comendial Test Procedures. PF, Vol 32 (4), July August 2006.
8. Schenerman MA, et al. CMC strategy forum report: analysis and structure characterization of monoclonal antibodies. BioProcess Int 2004;(2):42–53.
9. Lucy PK, Beri RG. Key considerations in process transfer. BioProcess International 2003;(8):36–43.
10. Nguyen LT, et al. Characterization methods for the physical stability of biopharmaceuticals. PDA J Pharm Sci Technol. 2003;57(6):429–45.
11. ICH. Q6B, Specifications: test procedures and acceptance criteria for biotechnological/biological products. ICH Harmonized Tripartite Guideline. Geneva, Switzerland; 1999.
12. ICH. Q7A, Good manufacturing practice guide for active pharmaceutical ingredients. Draft Consensus Guideline. Geneva, Switzerland; 2000.
13. Capen R. Revising USP <111> Design and analysis of bio-assays: current status and future plans. Presented at Biological Assay Development and Validation; 2004 Apr 26–28; San Diego, CA.
14. Callahan J. New USP guidelines for parallelism. Presented at Biological Assay Development and Validation; 2004 Apr 26–28; San Diego, CA.
15. American Society for Testing and Materials (ASTM). ASTM E 1488-02, Standard guide for statistical procedures to use in developing and applying test methods. West Conshohocken, PA; 2002.
16. ASTM. ASTM E 1169-02, Standard guide for conducting ruggedness tests. West Conshohocken, PA; 2002.
17. ASTM. ASTM D 4853-97, Standard guide for reducing test variability. West Conshohocken, PA; 1997.
18. ASTM. ASTM D 4854-95, Standard guide for estimating the magnitude of variability from expected sources in sampling plans. West Conshohocken, PA; 1995.
19. ASTM. ASTM D 4855–97, Standard practice for comparing test methods. West Conshohocken, PA: ASTM; 1997.
20. Krause, SO, Validation of Analytical Methods for Biopharmaceuticals—A guide to risk-based validation and implementation strategies. PDA/DHI Publishing. Bethesda, MD; 2007 Apr.