Challenges in Analytical Method Development and Validation

The manufacture of biopharmaceuticals presents some unique challenges when ensuring product quality and patient safety. Analytical testing can provide the data needed to produce a safe and effective drug product. The development and validation of analytical methods is crucial in drug development. While the biopharmaceutical industry has had success in reducing risk by incorporating analytical platform technologies for monoclonal antibody products, new molecules and drug conjugates pose additional challenges, according to Stephan Krause, principal scientist, analytical biochemistry, at MedImmune.

The extended drug development timeline of biopharmaceuticals also poses a challenge. According to Yong Guo, professor of pharmaceutical sciences at the School of Pharmacy, Fairleigh Dickinson University (FDU), analytical development and cross-functional development teams should work together to anticipate changes when creating the development timeline. Alice Krumenaker, manager, QC Stability Administration, at West-Ward Pharmaceuticals, suggests that an analytical method may need to be modified and/or updated during the drug-development process to minimize inaccurate data, and the modified method may need to be revalidated.

Technologies, such as FDA’s quality-by-design (QbD) initiative, may have a positive impact on analytical method development and validation according to Paul Smith, EMEAI laboratory compliance productivity specialist at Agilent Technologies, because methodologies are identified early in the development process.

BioPharm International spoke with Krumenaker, Krause, Guo, and Smith about the challenges involved in analytical method development and validation in biopharmaceutical manufacturing.

Biopharmaceutical ChallengesBioPharm: What are the challenges involved in analytical method development and validation for biopharmaceutical manufacturing?

Krause (MedImmune): The use of analytical platform technologies for common product types, such as monoclonal antibodies, has lowered the uncertainty/risk for the manufacturer; therefore, less development, qualification, and/or validation work is required.  More challenges typically exist for new types of molecules and/or conjugate products. For example, patient-specific cancer vaccines require conceptually a different approach in that some (routine) test methods should test for non-patient-specific manufacturing (batch-to-batch) consistency versus those that are patient-specific critical quality attributes. A direct potency test method may not exist, and instead, several surrogate test methods may need to be used for potency. Obviously, whenever new test methods are to be used throughout product development, more focus should be on the development, optimization, and validation of these new methods.

Guo (FDU): As in any drug-development program, there are, of course, many challenges involved in analytical method development and validation for biopharmaceuticals.  I would like to mention one particular challenge to analytical development. Often times, the development timelines are compressed for various reasons; however, insufficient consideration is given to analytical method development and validation. People expect the analytical methods to be ready all the time regardless of timeline or resource changes. To handle this challenge, the analytical development team needs to be closely working with the cross-functional development team to anticipate the changes and actively participate in any relevant discussion regarding the timeline. It is also important for analytical development to educate and work with the project managers to be more visible during the development process.

Biopharmaceutical Development TimelineBioPharm: How is analytical method development incorporated into the long development timeline of biopharmaceuticals? Can methods be changed mid-stream? At what point should analytical methods be validated?

Krumenaker (West-Ward): Method development and validation are typically done in the product development stage. While the method may go through various iterations during product development, as the final formulation emerges, the method will be optimized and ready for validation. The analytical method will become part of the submission package and, once a submission is approved, there is usually great resistance to submit changes to FDA. So at this point, it is important to have confidence in the method and its ability to be transferred from R&D to quality control or, in some cases, to a contract organization that will be doing the testing. A successful validation will provide this level of confidence.

As far as making changes to the method goes, if the changes are necessary, they must be made. Sometimes, changes to a process, necessary reagents no longer being manufactured, improvements in technology or other circumstances out of the lab’s control may result in a method no longer being suitable as written and/or as validated. Depending on the extent of the changes, some form of revalidation is likely to be needed. This may range from a simple verification, demonstrating that the method still performs as intended, to a full-blown validation for significant changes. In addition, this may impact the regulatory submission; however, the bottom line is if the original method isn’t accomplishing what it needs to accomplish, it needs to be modified and updated to minimize the possibility of inaccurate results. In that event, all appropriate amendments must be made to the original submission.

Krause (MedImmune): The timing and effort spent on method development correlates with the degree of novelty of the product type and manufacturing process.  The sooner methods are optimized, the lower the risk will be to overall product development.  

Methods can be changed any time during and/or after product development. The change to faster, more sensitive, accurate, and/or reliable test methods is encouraged by the regulators.  Besides providing sufficient qualification or validation results for the new method, method comparability results should also be provided. In some cases, product specifications may need to be re-evaluated and potentially adjusted to the established bias and/or increased sensitivity of the new method.  

Regulatory expectations vary a little bit depending on product type and/or quality attribute tested.  For example, for blood or blood plasma products, validation of potency and some safety test methods is expected much earlier (clinical Phase 1). For most other products, method validation is typically executed against the to-be-commercial specifications prior to process validation, which is typically initiated during the pivotal clinical phase. Method validation is usually completed one to two years prior to commercial license application(s), depending how much real-time stability data are desired following the process qualification runs (process validation stage 2).

Guo (FDU): The analytical methods need to be validated for any GMP activities. This is a GMP requirement as well as FDA expectation. The analytical methods should be properly validated even to support Phase I studies. Appropriate approaches should be considered to validate the analytical methods to support different clinical phases. The concept of ‘phase appropriate validation’ has been proposed and applied to method validation.

Regulatory ConsiderationsBioPharm: What regulatory parameters exist for analytical method development and validation for biopharmaceuticals?

Krause (MedImmune): A FDA draft guidance for development and validation was recently made available for public commenting (1). Biopharmaceuticals are in the scope of this draft guidance. A more comprehensive practical guidance, Technical Report 57, published by the Parenteral Drug Association (PDA) and specific for biopharmaceuticals, provides the best practical guidance currently available to industry and regulators (2). FDA used much of PDA TR 57’s content for their current draft guidance.

Guo (FDU): The International Conference on Harmonization (ICH) has a general guidance on method validation, ICH Q2(R1): Validation of Analytical Procedures: Text and Methodology (3). This is the guidance that the industry generally follows in performing analytical method validation. The ICH guidance, however, does not specifically address method validation for biopharmaceuticals. Interestingly, FDA issued a new draft guidance for industry on Analytical Procedures and Method Validation for Drugs and Biologics early this year (1). The new draft guidance supersedes the 2000 draft guidance for industry on Analytical Procedures and Methods Validation, and covers both small-molecule drugs and biologics. In comparison to the 2000 guidance, the new guidance specifically addresses analytical method development and suggests the parameters that should be evaluated during method development, including specificity, linearity, limits of detection (LOD) and quantitation limits (LOQ), range, accuracy, and precision. The robustness of methods is particularly discussed in the new draft guidance, and a systematic approach for method robustness study (e.g., design of experiments) should be adopted to ‘fully understand the effect of changes in method parameters on an analytical procedure’ (1). The new draft guidance refers to the ICH guidance Q2(R1) for more details on each method validation characteristics and also removes the recommended validation characteristics for various types of tests.

For the qualification of new reference standards, the new draft guidance recommends a two-tiered approach, which involves ‘a comparison of each new working reference standard with a primary reference standard so that it is linked to clinical trial material and the current manufacturing process’ (1). The new draft guidance does not address specific method validation recommendations for biological and immunochemical assays.

Quality by DesignBioPharm: How can method validation benefit from a QbD approach?

Krause (MedImmune): For a QbD approach for analytical methods, the analytical method lifecycle stages should consider the desired method performance at each product development stage. Starting with the appropriate, pre-established method performance expectations (Analytical Target Profile), the selection of the test methodology and instrumentation are the first step. Depending on the intended use of the method, typical performance criteria (accuracy, reliability, specificity, sensitivity, range, robustness/maintenance, as well as speed and throughput capacity, but possibly also costs and ease of operations) should be established early and re-evaluated as needed throughout product development.

The use of analytical platform technology (APT) methods can also greatly reduce the selection, development, and validation effort and lower the uncertainty/risk(s) for unsuitable method performance during product development. Since method performance criteria are well established, only product-specific suitability would need to be confirmed.  

Guo (FDU): The QbD approach is more applied to method development than method validation since method validation is the process of demonstrating that a well-developed analytical method is suitable for its intended purpose. The QbD concept is often narrowly interpreted in literature for analytical method development, and the method robustness study is often used as an example of the application of the QbD approach.  Well, the QbD approach is certainly applicable to the robustness studies and also consistent with the FDA expectation. The results of the robustness studies do not actually define the design space. A broader design space can be evaluated using the QbD approach during analytical method development. This would mean a significant more upfront effort in method development.

Krumenaker (West-Ward): QbD is not always employed in method development and validation because it is often considered to build time into the process. Often, labs are working under aggressive timelines to develop and validate methods. Using a pre-established set of parameters for method validation may be helpful in expediting the validation process, but it doesn’t necessarily provide relevant information about how the method will perform in ‘real world’ conditions.  When QbD is utilized in the development stages, critical product attributes are identified. Why not utilize this information in regards to analytical methods? For small molecule products, light or humidity may affect the product, so there may be potential problems related to these conditions that could arise during testing. The same is true for biopharmaceuticals.

Until the method has been stressed to the point where a problem occurs, you don’t know what the true boundaries are. Don’t just assume that by modifying a mobile phase by 2% or varying a pump speed by 0.1 mL that you have demonstrated that the method is sufficiently robust. Issues may not develop until a larger change has been made. The goal in method validation isn’t just to document that the method is suitable for use. It should be to truly confirm that it would work as expected, regardless of the analyst, the lab, and within reason, the environmental conditions. It may take longer to incorporate ‘real-world’ parameters into the process, but there will be more value in the day-to-day use after validation is complete and the method is transferred.

Smith (Agilent): The application of QbD principles to analytical method development and validation will have a significant positive impact. Fundamentally, analytical QbD requires that the analytical target profile (ATP) is identified before the analytical technology is considered. This means that fundamental requirements of the methodology are identified ‘up front.’ For HPLC methods, systems are available that integrate experimental design software with the chromatography data system (CDS) and analytical instrument, so that analytical method screening can be performed in an automated manner to identify a lead column, gradient, and mobile phase combination. This ‘Lead’ system is then subject to additional optimization experiments and final verification of the analytical design space.  This approach has the potential to significantly speed up analytical method development and result in HPLC methods that are more robust, and therefore, analytical transfer problems are reduced.

Deviations and VariationsBioPharm: What deviations or variations may occur during analytical method validation in biopharmaceutical manufacturing?

Krause (MedImmune): We should distinguish a validation deviation from a validation failure. Both can occur, but the frequency of these events should be low so that the manufacturer can maintain a state of control. In general, an unplanned deviation is typically easier to resolve and may not hold up the completion of the validation study. For example, the use of an incorrect analyte spike or a confirmed sample mix-up may require an execution repetition but may not need to be treated like a (potential) validation failure once the deviation has been confirmed.  Often, a lack of clarity in the validation protocol, poor preparation and planning, and/or insufficient analyst training can result in unplanned deviations.

A validation failure results from failing to pass the protocol acceptance criteria (or test method performance specifications).  Validation failures are always serious and difficult to deal with under current interpretation of regulatory compliance. PDA TR 57 provides good ideas and suggestions on how to systematically deal with such events and to set up an appropriate quality system (similar to dealing with out-of-specification results).

Method Transfer
BioPharm: What are the challenges in method transfer?

Krause (MedImmune): Like method validation deviations and failures (see ‘Deviations and Variations’), similar root causes may also exist for method transfers (validation extensions). The readiness of the receiving laboratory should be evaluated. Consideration should be given to the availability of required analytical and supporting equipment, software, critical reagents, standards, controls, and analysts who are skilled in the relevant analytical techniques as well as the qualification status of all materials, equipment, and analysts.  

If gaps are identified (e.g., the receiving laboratory has a similar analytical instrument), a risk assessment should be performed before execution of the formal transfer studies. Shipment and receiving procedures are needed to allow transfer of critical reagents, standards, and samples between laboratories. Incorporation of the test method procedure into the receiving laboratories quality system is also part of the transfer process.

The sending laboratory should provide hands-on training in the specific test method to analysts at the receiving laboratory, if needed. The type and amount of training needed will vary depending on the analytical method transferred and the existing experience of the receiving lab and its personnel.  The evaluation of the capability of the receiving laboratory to execute the system suitability requirements of the method successfully during the training is recommended. It is important that all responsibilities between sending and receiving laboratories are established. PDA TR 57 provides additional practical tips on what to consider for analytical method transfers.

Guo (FDU): Method transfer is a very challenging exercise and can become more challenging if the receiving laboratories are in different countries/continents. Planning is definitely the key to a successful method transfer. The timelines for method transfer need to be discussed and clearly understood by both the originating and receiving labs. The methods transfer protocols should be carefully reviewed and approved by both labs. Logistics should be well coordinated especially when shipping the reference standards and samples overseas. Proper documentation including import licenses, if necessary, should be obtained in advance. To ensure the success of method transfer, the analytical methods should be evaluated by the receiving laboratory before the initiation of method transfer. The samples for method transfer should also be carefully selected. The new FDA draft guidance recommends that forced degradation samples or samples containing pertinent product-related impurities should be analyzed at both labs for a stability indicating method. For biopharmaceuticals, it might be difficult to ship the forced degradation samples to the receiving lab. A suitable protocol may be needed to guide the receiving labs to generate similar forced degradation samples themselves.

Krumenaker (West-Ward): When transfers are being done between sites, or with contract facilities, issues may arise with different makes of instruments. While an HPLC has the same basic components, regardless of manufacturer, there may be subtle differences in performance. For example, the construction of the optics in the detector may be just different enough to present unexpected issues with linearity or sensitivity. In some cases, older models from the same manufacturer may have differences in construction that will challenge a method in ways that will prevent a successful transfer. Some exploratory testing may be valuable, prior to transfer, to determine the compatibility of the instruments in the receiving lab and the method.

Smith (Agilent): Method transfer can go smoothly or can result in significant problems. The robustness of the analytical methodology being transferred and the experience of the ‘receiving’ laboratory have a significant impact on the success of the transfer. Differences in instrumentation, culture, and ways of working also contribute to possible challenges, and therefore, need to be considered. The analytical technology transfer (ATT) usually follows an analytical protocol, where results obtained between the ‘receiving’ laboratory are compared to those obtained by the ‘donor’ laboratory transferring the technology.  Subjective tests such as color and appearance can cause disproportionate problems unless the specification is carefully worded. Tests with low level impurities require careful evaluation using % absolute differences, rather than student’s t-tests in the ATT protocol. In part, cultural and ‘ways of working’ differences can be overcome through training and short-term secondment, so that analysts in the ‘receiving’ laboratory acquire as much knowledge of the methodology as possible.

For transfer of HPLC methods, generally, fewer problems are experienced where both laboratories have the same analytical equipment. However, this is not always possible and ‘hidden’ differences in instrumentation (such as gradient formation, delay volume, or signal to noise) can contribute towards problems with transfer. Testing or qualifying the equipment in the same way can help identify differences in performance between analytical equipment.

References
1. FDA, Guidance for Industry, Analytical Procedures and Methods Validation for Drugs and Biologics, Draft Guidance (CDER, Rockville, Md.,  February 2014).
2. PDA, Technical Report 57 (PDA, 2012).
3. ICH, Q2(R1): Validation of Analytical Procedures: Text and Methodology (ICH, November 2005).

About the Author
Susan Haigney is the Managing Editor for BioPharm International.