FDA has issued guidances on analytical method development and validation for over 30 years. The latest, published in July 2015, is considerably less prescriptive than the previous version. While substantially shorter, it puts the burden on the user to understand its supporting documentation. The document asks for a holistic view of the development and validation of the methodology. It requests an in-depth understanding of the reasoning behind the choice of methodologies in the proper characterization of the material, alone and as a whole, along with the reasoning, specifications and results associated with the method validation procedures.

In addition, there is a very strong emphasis on lifecycle management. In this article we discuss the changes from the previous guidance, both in content and approach, and examine how it might, in the context of broader FDA and ICH initiatives, have an impact in how we think about method development and validation for drug substances and drug products.

Background
FDA has issued guidance on method development and validation back at least as far as 1987 with its guidance, “Guidelines for Submitting Samples and Analytical Data for Method Validation.” This document was updated in 2000 with the guidance, “Analytical Procedures and Method Validation,” and again in July 2015 with the current guidance, “Analytical Procedures and Methods Validation for Drugs and Biologics.”¹ The latest guidance is about half the length of the 2000 guidance, and represents a shift away from a prescription for how to do method development and validation to more of a risk-based approach emphasizing a more holistic understanding of the molecules, methods and processes that go into method development and validation. This shift is consistent with other FDA and ICH initiatives to drive in-depth understanding of the science behind the methodology to optimize the value of the methods and their validation.

Why update?
Over the last several years, FDA has, in many cases, become less prescriptive in terms of submission requirements, focusing instead on a higher-level science and risk-based approach that puts the onus on the submitter to ensure and justify that the proper tests are performed and they are sufficiently validated for their intended purpose. In addition, there has been an increased emphasis on lifecycle management—demonstration that the methodology is capable of detecting changes to the drug substance (e.g., polymorphic form), its impurities or degradation products, or drug product (e.g., degradation products resulting from excipient changes) as the material changes through the lifecycle of the product from launch to removal from the market. Thus there is an expectation that the holder of the marketing authorization is thinking about the lifecycle management approaches described in, among others, ICH Q8, Q9 and Q10 in terms of continuous improvement—including new technology, if needed—risk assessment, and updating methodology as processes or specifications change, all inside of an established quality management system.

Part of FDA’s goal is to drive the Quality by Design (QbD) approach down into method development and validation. This requires the sponsor, i.e., holder of the NDA, ANDA or BLA, to think about what tests are really needed to provide the proper characterization of the drug substance and drug product, from drug substance synthesis and testing through to the end of the drug product’s life. In addition, it requires that the sponsor keep an eye on the product throughout this evolution. Is anything changing—new API synthetic route, new suppliers, manufacturing process change? What’s the impact and how do you know? Validation gets at how you know your method does what you say it does.

A key goal is to properly evaluate results generated from new technology or the new application of existing technology that could reveal issues that were not previously observed. During the process of continuous improvement, unavoidable situations can arise. For example, a previously unobserved impurity leads to a method revision or a new related substance method is developed. If this occurs, it is critical to test old batches—consider expired stability samples—by both the old and new procedure to establish that the observed changes are due to better methodology, not a change in product purity.

How does good method development and validation fit into the big picture?
A successful filing requires all the pieces associated with the development of the drug to fit together seamlessly. From a CMC perspective, the review is ideally done in a holistic way. That is, a reviewer is responsible for telling the CMC story to the rest of the team. The better the validation—the logic, foundation, and supporting data behind the measurement—the easier the test and result are to defend. At the individual test level, validation goes beyond figures of merit—precision and accuracy—to a more fundamental level. Why that test—IR vs. Raman vs. Near IR?  What are you trying to learn with the test and why is it the best technique to do it? More broadly:

• What do you want to learn about the drug substance, blend, formulation or product?
• What tests do you need to do that?
• Why are these the best tests to accomplish your objective?
• What are the controls (acceptance criteria) needed to make the tests meaningful

At a high level, there is a need to tie all this back to your critical quality attributes and critical process parameters. Putting the story together that you understand the drug substance, formulation, process and stability well enough to keep everything in control now and through the life of the product provides assurance regarding the current and on-going quality of the drug product.

Intent of the guidance
The guidance provides recommendations on how to develop and submit analytical procedures and method validation data to support documentation of identity, strength, quality, purity and potency of drug substances and drug products. It is intended to help assemble information and present data for NDAs, BLAs, ANDAs and supplements as well as OTC monograph products in a consistent, logical way. Finally, it is intended to focus on getting the science right, not just providing a checklist against which to evaluate your methodology or the extent to which you have properly characterized the material of interest.

As discussed, this is an NDA/ANDA/BLA document. It does not address IND method validation issues. FDA is advocating for a phased validation approach for each stage of development, thus expects the amount of information obtained and submitted to increase as the drug progresses through development. These expectations are set out in other documents.^2-4

Analytical method development
This document doesn’t change the goals of method development. One needs to develop the method based on what is being measured and how well it needs to be measured—the characteristic and its acceptance criteria. The data should demonstrate a mechanistic understanding of the basic methodology. For example, if you submit a Raman spectrum rather than an IR spectrum, what is the rationale? Specificity, linearity, LOD, LOQ, range, accuracy and precision should be considered, as needed based on the requirements of the procedure.

A few things have changed. The guidance recommends that development data be submitted within the method validation section if they support the validation of the method. There is also an increased emphasis on robustness. The guidance recommends demonstrating a systematic approach to examining the effect of method parameter variation on method performance—initial risk assessment followed by multivariate Design of Experiments—and recommends that the sponsor demonstrate an understanding of the sources of variation in the method and their effects on method performance. This may require the evaluation of performance on samples obtained from different stages of the manufacturing process, from in-process to finished product.

Method development is the best time to stress the various parameters in a method to ensure the method is sufficiently robust to be validated and to withstand the test of time. Going the extra mile to assess the Relative Retention Factors based on the slope of multiple concentrations versus a single-concentration determination is one way to help reduce method variability. Assessing the need for sensitivity solutions in system suitability experiments on related-substance runs is an example of building in a control to ensure the system is performing as it needs to.

Content of analytical procedures
What is to be expected in the content of analytical procedures has not changed substantially. There is an expectation that the procedures be described in sufficient detail to allow a competent analyst to reproduce the necessary conditions and obtain results that are within the specification. An analytical procedure from an FDA-recognized source may be referenced, such as USP or AOAC-International. If the method isn’t changed from what is allowed in the publication, it can be used as-is. However, if changes are made, it is expected that they will be appropriately identified with inclusion of the appropriate validation data. Detailed procedures need to be provided if the method is from other published sources.

Broadly speaking, the goal is to describe how and why the methodology chosen is the right one, and how you know the methodology will remain in control. Thus, the Principle/Scope section describes the basic principles of the technology—separation, detection—target analyte(s),  and sample matrix type—drug substance, drug product. The Apparatus/Equipment and Operating Parameters section needs to describe all required qualified instruments and components and all qualified optimal settings critical to the analysis. With respect to Reagents and Standards, the goal again is consistency and control. Important to this are parameters such as the grade of chemical (USP, NF, ACS), source (USP, WHO, CBER, qualified in-house reference material), purity, storage conditions, directions for safe use and documented shelf life. For the preparation of samples, standards and other control solutions, detailed sample preparation steps should include storage conditions and solution stability, again to help demonstrate long-term control.

The procedure should be described step by step with sufficient detail to allow a competent analyst to reproduce it. This should include special emphasis on any steps that are critical to the method’s performance. Many of us have been in the situation where we learn in retrospect about a “trick” to one of the steps that helps improve method performance. Sufficient information needs to be provided to permit the evaluation of method performance—system suitability, assay validation criteria—and evaluation of results. System Suitability should include confirmatory tests to ensure that the system will function correctly as an integrated system at the time of use.

Calculations need to include the integration method—peak area vs. peak height—and representative calculations/equations for data analysis (standards, samples, controls) for tests based on label claim and specification (e.g., assay, impurities, relative response factor). Include any mathematical transformations and scientific justification for correction factors. Data reporting should be done in a way that is consistent with instrumental capabilities and acceptance criteria. Format for reporting (% label claim, w/w, w/v) and significant figures/rounding are important. Include units, not just numbers.
The takeaway message here is that details matter. Why is this the right methodology? How does it work? What grades of materials are important and how do you ensure they stay stable? What are the key steps in the procedure? Are they well described? Are the calculations well described and are the data reported out in a format compatible with their intended use? The more attention is paid to the logic of how all this is laid out and described, the fewer issues a reviewer will have in agreeing that indeed the content of the procedure is sufficient for its intended use.

Reference standards and materials
Selection of appropriate reference standards is critical. The guidance points to ICH Q6B “Specifications: Test Procedures and Acceptance Criteria for Biotechnological/Biological Products” and ICHQ7 “Good Manufacturing Practice Guidance for Active Pharmaceutical Ingredients” for information regarding primary and secondary standards. A key change between the current and previous guidance is the lack of specifics in this guidance and the reliance on other FDA and ICH documents to fill in the details. This results in a much shorter guidance, but puts the burden on the user to be familiar with a wide range of FDA, ICH and USP documents.

Rather than prescribe the characterization and acceptance criteria of these materials in detail, the document prescribes only high-level guidance, with the goal of demonstrating that the standards/materials are suitable for their intended use. For example, it recommends that standards be the same chemical entity as the analyte to the extent possible, that there is a comparable physical response for standards used to measure physical properties (e.g., particle size) and that recommended storage and usage conditions are carefully followed. While not describing what tests are important for sufficient characterization, there is an expectation that the sponsor will know and report what tests were done, what the results were, and that the results indicated that the material was suitable for use. Thus for all standards one needs to include certificates of analysis, qualification tests, stability protocols, reports, or relevant impurity profiles. For BLAs it is important that qualification of subsequent reference standards lots should be included in the annual report. It is also highly recommended that one use Authoritative Standards whenever possible (USP, EP, JP, WHO, NIST, CBER). CBER standards must be used for certain biological products before the product can be released in the U.S.

The topic of secondary standards, from other sources or in-house, is also briefly discussed. These reference standards should be characterized by procedures including routine, and beyond routine, release testing (see ICH Q6A and ICH Q6B). For certification of secondary standards, going beyond typical ICH Q6A test is typically an expectation. Orthogonal forms of chromatography, use of photodiode array detectors and absolute assays to assess purity are examples of steps one might take to help assess the purity of small molecules. For biologics it is important to qualify secondary standards against primary standards, not against a previous lot of secondary standard to avoid drift. It is also important to link the material to clinical trial material and the current manufacturing process to avoid drift in the quality of the biologics standard.

Analytical method validation: Non-compendial procedures
When the sponsor is confident the method is suitable for use, robust and described in sufficient detail, it is ready for validation. Validation is where one actually demonstrates that the procedure is suitable for its intended purpose. It builds off of good method development and optimization studies. Validation must be generated under a sponsor-approved protocol with substantiated, pre-determined acceptance criteria. The protocol has to include a description of methodology of each validation characteristic, contain predetermined and justified acceptance criteria, and be executed using qualified instrumentation under appropriate good manufacturing conditions. Without a protocol and predefined acceptance criteria, you may have a well characterized procedure, but not a validated one.

Validation characteristics
As with the previous sections, only general guidelines are provided regarding validation characteristics. Recommended for inclusion are specificity, linearity, accuracy, precision (repeatability, intermediate precision and reproducibility), range, quantitation limit and detection limit. It is also recommended to include system precision in the evaluation to evaluate its contribution to repeatability. This guidance has eliminated many of the details from the previous guidance—those contained in ICH Q2(R1) or USP General Chapter —and sends the user to the appropriate ICH, USP or FDA documents.

Validation characteristics for stability indicating methods are discussed in some detail with an emphasis on demonstrating specificity. Analyze samples in the presence of all actual and potential impurities and degradation products, including sample subject to stress conditions—acid, base, oxidation, heat, humidity and light. One can use samples spiked with target analytes and all known interferences, stressed samples, and actual product samples (final manufacturing process) that are aged.

Submission obligations
The holder of the submission (NDA, ANDA or BLA) is responsible for submission of the data. Submit data used to establish that the procedures are appropriately validated. Notify FDA about each change in each condition in an approved application beyond the variations already approved. A QbD approach can help here in terms of allowable design space. This includes changes to the analytical procedures and controls.

Analytical method validation: Compendial procedures
For compendial methodology, there is a need to verify that the method works under your conditions of use using actual DS, DP or excipient (see USP ).

A section new to this guidance requires the demonstration of method suitability—drug substance, drug product or excipient—be done under a verification protocol and included in the submission. Include in the Verification Protocol the compendial methodology, the predetermined acceptance criteria, and details of the methodology. The procedure and extent of verification will determine which validation tests should be in the protocol (e.g., specificity, precision, accuracy). Situations such as tighter than USP specifications may require further work.  It is noted that robustness studies are not needed if methods are followed without deviations.

Lifecycle management: FDA guidance perspective
The area of lifecycle management is one of the key differences in this guidance. One sees here the emphasis on not only how you know the procedures are working today, but how do you know they are going to continue working in the future? What is your plan to respond to changes in drug substance synthesis, manufacturing changes, reagent or excipient changes, or improvements in measurement technology?

There is the expectation that once validated/verified and implemented, the method should be followed through the life cycle of the product to assure it remains fit for purpose. When should a method be changed? Some examples include:

• When trend analysis (performed regularly) suggests the method is no longer optimal or that all or part needs to be revalidated;
• When requisite consumables become unavailable or are inherently changes by the manufacturer (e.g. chromatographic columns, filters, etc.); 
• When system suitability criteria are difficult to meet (repeated adjustments needed); 
• Availability of new product knowledge;
• Discovery of a new impurity, or a better understanding of CQAs;
• Change in product formulation or manufacturing; and
• Availability of new technology.

Regarding data trending analysis, periodic evaluation of release data is crucial to evaluation of test method performance. A conventional ± 3 standard deviation plot (Levey-Jennings) can provide an instant picture of how data is trending in relation to existing acceptance criteria. While observed variation can of course be attributed to manufacturing of the API or drug product, there is also an analytical method-based evaluation that is inherently conducted when reviewing these plots. Assuming that the manufacturing process is fully validated and well-qualified analysts are executing the test methods, trending plot review can provide an additional tool to evaluate trends that can originate with inherent test method variability. Furthermore, laboratory investigations of aberrant data, along with the corresponding quality-based CAPA system, can also uncover method-related issues that can potentially lead to revalidation. 

The guidance also stresses the need to periodically assess technologies to evaluate appropriateness of current methods and consider alternative methods as needed. 

Comparative studies (comparability reports)
Comparability studies play an important role in the context of continuous improvement and demonstration of continuous control of the methodology and materials. Situations where they are important include the development of alternative analytical procedures or the transfer of an analytical procedure. Either may require revalidation, a comparability study, or combination. It is important to have sufficient retention samples to allow for comparative studies or revalidation, including, if possible, marketed product and pivotal clinical trial material.

Revalidation
The same principles apply for revalidation as for validation. Consider revalidation when a change is made to the procedure (e.g., equipment, reagent), when there is a change in the formulation or manufacturing process, or changes such as route of synthesis or fermentation. Revalidate all or part of the method as appropriate to ensure the method maintains its critical performance characteristics—specificity, linearity, accuracy, and precision.

Analytical method comparability studies: Alternative analytical procedures
Once a method is approved by FDA, care must be taken in replacing it with an alternative procedure. For NDA or ANDA, include a description of the proposed new procedure in the application. Post marketing, addition, revision or deletion of an alternative method is submitted as directed in 21 CFR 314.70 (NDA, ANDA) or 601.12 (BLA) and documented in the next annual report or, if required, a Changes Being Effected or Pre-Approval Supplement document. Some biological products may have methods included in an FDA regulation. Change requires review and approval as per 21 CFR 610.9(a) and approval with the application or in a post approval supplement (see 21 CFR 610.9(b)).

Comparability studies
Guidelines for the use of comparability studies when providing an alternative analytical procedure include:

• Identify the use of the method (e.g., release, stability testing);
• Provide rationale for inclusion, validation data and comparative data to FDA approved method;
• The study should demonstrate that the method coupled with any controls is the equal or better than the current one relative to demonstration of identity, strength, quality, purity or potency of material, including demonstrating that the method is not more susceptible to matrix effects; and
• If the method reveals new process or product-related variants, or new impurities are discovered, test historical batches. The sponsor needs to demonstrate the newly observed impurities are the result of increased sensitivity or selectivity and not due to a change to process-related impurities. The corresponding data from the revised method may lead to a proposed change in product release and stability acceptance criteria. 

Method transfer studies
Method transfer studies need to be managed under a transfer protocol. These typically involve two or more laboratories or sites. Detail the parameters evaluated and the predetermined acceptance criteria. Use a sufficient number of representative test articles—same lots of drug substance or drug product. Evaluate precision and accuracy with attention to inter-lab variability. For stability-indicating methods, forced degradation samples or samples containing pertinent product-related impurities should be analyzed by both sites. Avoid the use of samples that are very low in related substance content—spike if necessary to obtain relevant levels of analyte. Also, in some cases it may be important to provide a time frame in which the multiple sites need to complete their analyses, so that an appropriate comparison can be conducted. USP general chapter Transfer of Analytical Procedures provides useful guidance.

Conclusions
The current FDA guidance is much less prescriptive than its predecessors, but in its own way more demanding. It puts the onus on the sponsor to demonstrate that the procedures are the right ones for the task, that they are described in sufficient detail that they can be executed by competent analysts, that they are robust enough to ensure that the material under test remains in control, and that they are routinely monitored and modified or replaced, as necessary, to ensure they are suitable for use throughout the lifecycle of the material.

Taken as a whole, this document embodies FDA’s Quality by Design principles—encompassing the principles of ICH Q8, Q9 and Q10—requiring a complete understanding of the goals and science behind the tests, and the expectation that these tests will be continuously monitored and improved as needed to remain fit for purpose throughout the material’s lifecycle.

Finally, the concepts presented in the guidance are important as they relate to the new ICH initiative Q14 “Analytical Procedure Development and Revision of Q2(R1) Analytical Validation.”⁵  It is anticipated that the Expert Working Group dedicated to this ICH initiative will be able to build on many of the concepts FDA presents in the current guidance. 

References

1. “Analytical Procedures and Methods Validation for Drugs and Biologics”, Guidance for Industry, July, 2015.   https://www.fda.gov/downloads/drugs/guidances/ucm386366.pdf.  Accessed Jan 9, 2019.
2. “Content and Format of INDs for Phase 1 Studies Including Well-Characterized, Therapeutic, Biotechnology-Derived Products”, Guidance for Industry, Nov. 1995. https://www.fda.gov/downloads/drugs/guidancecomplianceregulatoryinformation/guidances/ucm071597.pdf.  Accessed Jan 8, 2019.
3. Guidance for Industry, May, 2003. INDs for Phase 2 and Phase 3 Studies – Chemistry, Manufacturing, and Controls Information. https://www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/UCM070567.pdf.  Accessed Jan 8, 2019.
4. “IND Meetings for Human Drugs and Biologics, Chemistry, Manufacturing and Controls Information”. Guidance for Industry, May, 2001. https://www.fda.gov/downloads/drugs/guidancecomplianceregulatoryinformation/guidances/ucm070568.pdf.  Accessed Jan 8, 2019.
5. “Analytical Procedure Development and Revision of Q2(R1) Analytical Validation https://www.ich.org/products/guidelines/quality/article/quality-guidelines.html#2-1.  Available at ICH.org.  Accessed Jan 8, 2019.

---

Tony DeStefano is a member of the Editorial Advisory Board of Contract Pharma. He was Vice President and Sr. Vice President of General Chapters at USP from 2008-2013. He currently is a pharmaceutical industry consultant in the areas of CMC, bioanalysis and compendial issues.

Thomas Kester is currently the Director of Analytical Services at Recordati Rare Diseases, Inc. in Lebanon, NJ. He has over three decades of experience in the analytical sciences with-in the pharmaceutical, biological and beverage industries.