Introduction
This document provides guidance to determine if validation of re-work and / or re-processing steps is required for Active Pharmaceutical Ingredient (API) processes.
This guidance provides recommendations for evaluating the potential impact on product quality to determine if a given re-work or re-process step requires validation.
This guidance applies to critical steps of API manufacturing via chemical synthesis and the isolation steps for APIs produced by classical fermentation. It does not apply to biopharmaceutical API manufacturing or drug product manufacturing. Recovery of materials by 2nd crop production is not included within the scope of this guidance.
Compliance with regulatory filings, consideration of impact on product stability, and
deviation investigation activities should be evaluated, but are outside of the scope of this guidance. For further information regarding regulatory filing compliance, stability and deviation investigation, please refer to relevant guidances.
Recommendations & Rationale
To clarify validation considerations for re-processing and re-work steps, definitions of these terms are listed below:
Re-process – Introducing an intermediate or API, including one that does not conform to standards or specifications, back into the process and repeating a crystallization step or other appropriate chemical or physical manipulation steps (e.g., distillation, filtration, chromatography, milling) that are part of the established manufacturing process.
Continuation of a process step after an in-process control test has shown that the step is incomplete is considered to be part of the normal process, and not re-processing.
Re-work – Subjecting an Intermediate or API that does not conform to standards or specifications to one or more processing steps that are different from the established manufacturing process to obtain acceptable quality intermediate or API (e.g., re-crystallizing with a different solvent).
The determination of whether a remediation procedure is considered to be re-processing or re-working should be done on a case-by-case basis. This determination should be performed relative to commitments and descriptions of the process in the regulatory filing. Subsequent to this determination, the potential impact of the re-processing or re-work step(s) on the quality of the final API should be identified. Additional evaluation may be needed to determine if validation of the re-work or re-processing step is required. Not all re-works will necessarily require validation (e.g. re-work of non-critical, non-registered intermediate process step), but they do require consideration of validation.
ICH Q7A require that re-work steps be validated for those portions of a process that are critical to API quality. There can be steps after introduction of API Starting Material that are not critical steps and rework may not require validation for those steps.
Validation of re-work and/or re-processing steps can be concurrent and may only include 1 batch since re-work and/or re-processing steps are typically not conducted frequently and there may never be more than 1 batch.
Where validation of the remediation is necessary, a concurrent validation approach can be suitable. In cases where there is already a concurrent validation exercise in existence for a given remediation procedure, further validation runs conducted using this remediation procedure can be added to the existing validation exercise provided suitable justification for inclusion and possible scope expansion is provided.
The following questions are provided for guidance when conducting an impact assessment, in order to determine what aspects of validation may be needed. An impact assessment conducted to determine whether validation is required or not should be documented and approved as part of the deviation investigation.
The flow-chart provided on the pages below is intended to provide guidance for the determination of validation activities for re-processing and re-work procedures.
Frequency of failure requiring re-processing/re-work. How often has use of the proposed remediation process been needed? If the number of production batches over a period of time that have needed this remediation exceed that allowed by site policy for failure rate, one should ask if control of the routine production process is still in a validated state.
If it is an intermediate that is to be re-processed or re-worked, is it produced prior to the introduction of the API starting material (as described by ICH Q7A)? If yes, then validation of the step is not necessary, and by extension validation of the re-process or re-work to remediate the quality of the intermediate is also unnecessary.
If the material is the product of a step in the process containing one or more Critical Process Parameters (CPPs), validation of the proposed remediation may be needed. Is the re-work /re-process step covered by an existing validation study? If yes, then validation of the proposed procedure may
have already been addressed. For example, re-processing steps that address chemical issues (e.g. recrystallization to reduce normal process impurities), as well as re-processing steps to remove physical contamination (assuming it contains a filtration step to remove particulates). In cases where a previously validated re-work or re-process method had been previously validated for a different purpose, validation is needed to demonstrate that the proposed remediation procedure is effective in achieving the new intended quality improvement.
Is the material requiring remediation exposed to processing conditions that it may not normally experience? For example, is there greater potential for over-reaction of the product (e.g. formation of by-products or degradation of the material)? Is the exact unit operation being used or is it a variation there of? What are the potential impacts to the physical and chemical characteristics of the product?
There are several factors that should be considered in answering these questions.
- Repeating chemical reactions. This is generally considered a higher risk scenario than a simple recrystallization or physical re-processing, for example. Points for consideration here include:
- What side-reactions may occur as a result of repeating chemical reactions? A substance can sometimes be subject to over-reaction when exposed to the reaction conditions in which it was made.
- Are there concentration differences used in the chemical step(s) and if so what is the potential impact?
- If critical impurities are increased above typical maximum levels in the repeated step, are they purged from the final product by subsequent down-stream steps?
- If critical impurities are reduced below typical minimum levels in the repeated step or subsequent down-stream steps, what is the impact on the final product quality?
- Repeating isolation procedures. These are generally lower risk situations, but the potential impact on impurity profiles, solvent content profiles, and physical quality attributes of the product should be considered, for example, is there a difference in using clean solvent versus mother liquor wash with respect to crystal form or other physical properties?
Repeating physical processing steps:
What is the potential impact on the physical and chemical characteristics of the product? An assessment (which may include small scale trial) may be needed to consider the impact on impurity profiles/levels, crystal morphology and/or particle size distribution of the product.
In general when assessing the proposed procedure, if one or more of the above factors can potentially impact the quality of the product then equivalency of the material prepared by the remediation procedure should be demonstrated. For further guidance on equivalency, please refer to relevant guidance.
It may be necessary to extend evaluation of the product to its performance in subsequent steps of the process, including the drug product manufacturing process. This evaluation can be conducted via change control and doesn’t automatically require a validation exercise.
Is the material prepared by the remediation procedure isolated as a solid or not? While this may not be important to the determination of whether or not validation of the remediation procedure is needed, it may influence where in the process to evaluate the effectiveness of the remediation procedure.
Is there an impact on an existing critical process parameter (CPP), or introduction of a new CPP? A change to an existing CPP should be evaluated for impact on validated control of the process, while introduction of a new CPP requires validation of this control. For further information on CPPs, refer to relevant guidance.
Are there changes in equipment used in the process? If yes, validation activities may be required, per normal site change management system.
What potential impact is there on product homogeneity? If the batch size is outside the validated batch-size range, then a homogeneity study may be required to show that product prepared at the modified scale is homogeneous. For further guidance on homogeneity, please refer to relevant guidance.
Is it intended that the material will be re-processed or re-worked more than once? If so, an evaluation of the cumulative potential impact of multiple re-process or re-work steps on the same batch is required. The evaluation, including requirements for validation should include the considerations indicated in this section. If multiple iterations are proposed, one should determine if alternative methods for remediation have been considered.
Batches that have different reasons for re-processing or re-work may be combined to perform the proposed remediation, if there are data and/or rationales (such as blend uniformity) available to support that the re-processing or re-work would be effective on each individual batch. Validation considerations for these combined batches can be determined using the flow-chart provided. Per ICH Q7A, “Out of specification batches should not be blended with other batches for the purposes of meeting specifications”.
Examples of Validation Considerations for Re-processing and Re-work Steps
The following examples of validation considerations of re-processing and re-work are provided for guidance:
Example 1.
Batch A of a final API is found to contain an unacceptable amount of a known process-related impurity. Investigation reveals the root cause of the problem, and the corrective action for Batch A is to repeat the final product crystallization but filter (isolate) the batch at a warmer temperature than usual to improve the efficiency of removing the elevated impurity.
The effectiveness of this procedure was demonstrated by lab studies that showed the impurity was diminished by the modified crystallization that is within the proven acceptable ranges and registered ranges for the process. The modified isolation temperature was not originally identified as a critical parameter for the process, but is now considered a CPP. This crystallization at a higher temperature has not been used before on commercial scale and has not been validated.
Here, validation of the proposed remediation procedure should be done, because the isolation temperature is critical to product quality. Also, since the isolation of the crystals at a warmer temperature is likely to provide a lower than usual product yield from the crystallization, the size of the re-processed batch may fall outside the validated batch size range. If this is the case, homogeneity of the product batch should be included as part of the validation activities to show the smaller batch size on the same size filter still provides a homogeneous product.
Other considerations that may need evaluation for this example could include the potential impact of different temperatures for generating other impurities and the potential for different physical characteristics (such as crystal morphology or particle size distribution).
Example 2
Batch B of a final API is found to contain an unacceptable amount of a known process-related impurity. Investigation of the incident reveals the root cause of the problem, and the corrective action for Batch B relies on process knowledge that indicates an elevated amount of this impurity cannot be removed by the normal product purification procedure. During process development, an alternate crystallization solvent mixture was shown to effectively control elevated amounts of this impurity. Use of this procedure did not become part of normal processing because of poor product yield from the solvent mixture. This re-work procedure has not been used before during commercial-scale manufacturing and therefore has not been previously validated.
This example would require validation in order to evaluate the potential impact of the remediation processing on impurity profile. The validation study should at a minimum, include evaluation of equivalence to historical product quality results, especially for process impurities such as solvents and process-related impurities. Because in this example the yield is expected to reduce the output batch size below the validated range, homogeneity testing of the final product should also be included in the validation.
Other considerations that may need evaluation are similar to those for Example 1:
- Is there any potential for new impurities (or differing impurity levels) above the qualification threshold because of the use of a different crystallization solvent?
- Is there a potential for different physical characteristics such as crystal morphology or particle size distribution?
- What is the impact on the drug product manufacturing process, or on the quality of the drug product?
Example 3
Batch C of a final API has been contaminated with mineral oil as a result of a leak from a filter dryer agitator seal. The contaminated material has been filtered and the cake washed with solvent per the process instructions. The next step of routine manufacturing would be to dry the product, prior to discharge. The remediation proposed involves washing the material with additional solvent, in order to remove the mineral oil, before drying and discharging the material per the routine batch production.
There are no validation requirements for this re-process; however the acceptability of the mineral oil removal would have to be assessed as part of the deviation investigation.