Dated April 2014 the WHO published a proposal for a revision of appendix 7 (non-sterile process validation) as a supplementary to the Guidelines on Good Manufacturing Practices. The reason given for the revision is compliance with actual GMP requirements. Explicitly mentioned are quality risk management principles and quality by design principles with references to WHO and ICH requirements. The text is mainly applicable to non-sterile finished pharmaceutical dosage forms. According to the draft similar approaches may be applicable to active pharmaceutical ingredients (APIs) and sterile products. The WHO now also puts the focus of validation on the life-cycle of process validation which consists of:
A risk-based approach and the use of in-line, on-line and/or at-line controls is recommended expressly.
The 16-pages-draft consists of 7 chapters and the references:
1. Background and scope
4. Process design
5. Process qualification
6. Continued process verification
7. Change control
A part of the terms in the glossary has been taken over verbatim from the ICH Guidelines and the FDA Process Validation Guidance. It is striking that the term process validation is quoted in the glossary still in the old version ("documented evidence that a process..."). The reference to the life-cycle approach is missing completely here. The matrix approach can be found as a term but without any further explanations.
In the introduction reference is made once more to the new approach pointing out critical steps and parameters. A risk assessment approach should be followed to determine the extent of variations in the starting materials and in the process. The aim is that the manufacturing process is under control before a product is placed on the market. The following words which are put in brackets are highlighted using a yellow background: "Clarify under control?".
Flow diagrams are recommended as being very helpful within the scope of a process validation and may be amended as part of the validation documentation. When applying a quality risk management, the steps preceding and following that operation should also be considered. Process performance verification is also referred to as part of the validation in the introduction without the term being clarified, however.
Generally, the validation should cover all manufactured strengths of a product. But a matrix approach may be acceptable based on appropriate risk assessment. The term matrix approach is listed in the glossary. It is not defined, however (see above). The extent of validation at each manufacturing site should also be based on risk assessment.
Different approaches to process validation are addressed in the introduction. Traditional process validation consists of prospective and concurrent validation. Reference is made furthermore to the 3-step- process (process design, process qualification and continued process verification). The combination of traditional process validation and the 3-step-process is mentioned as further possibility. A table illustrates once more the different phases of the new approach. Process validation with its 3 steps accompanies almost the complete product life-cycle, and change control and GMP in general, too.
The chapter Process design indicates that he focus of validation now also includes development under the life-cycle approach. In the draft process design is understood as: design of experiments, process development, clinical manufacturing, pilot scale batches and technology transfer. At the stage of the process design framework conditions are developed (such as selection of materials and of the manufacturing process and so on) that will then lead to a control strategy.
It is expressly mentioned that some process validation studies may be conducted on pilot-scale batches (corresponding to at least 10% or 100 000 units whichever is the greater). In case of smaller batch size it may be necessary to determine production-scale validation data.
Process qualification and continued process verification should in any case always be linked to process design and the batches used for bioequivalence testing.
Generally, reference is made to an appropriate documentation.
A development report and/or a technology transfer document, formally reviewed and approved by research and development personnel is required. A further "approval" ("formally accepted") of the documents is required to be carried out by manufacturing, engineering and quality personnel. Then examples are given for the content of such documents. Mentioned are, inter alia, specification of approved suppliers but also bills of (starting) materials.
Process qualification requires that personnel, premises, utilities, support systems and equipment should be appropriately qualified before processes are validated. As concerns the qualification of equipment (DQ, IQ, OQ, PQ) reference is made to Annex 2 of the document no. 970.
Interestingly, the draft mentions the manufacture of at least three batches as traditional approach. But now it is indicated expressly that the number of batches to be manufactured should be based on risk assessment. Examples for points to be taken into consideration are variability of materials, product history, where the product is being transferred from. The stage at which the product is considered to be validated and the basis on which that decision was made should be defined. This includes a justification for the number of batches used based on the complexity and expected variability of the process.
It is remarkable how much in detail (16 points) the contents of a validation protocol are addressed. Fortunately, no timetable is required in the validation protocol. A validation report should reflect the validation protocol. It is also welcome that a dual validation protocol report can be used if they are designed to ensure clarity and if sufficient space for recording of results is available. And, naturally, deviations should be clearly addressed and examined.
A risk assessment should be performed for the change in batch size from scale up to commercial batch size.
It is explicitly mentioned that the process should be verified on commercial-scale batches prior to marketing of the product. Results of relevant quality attributes, for example of incoming materials or IPC material, should be collected. Extensive in-line and/or online and/or at-line controls should be used to monitor process performance and product quality in a timely manner. Process analytical technology applications (PAT) and multivariate statistical process control (MSPC) can also be used. It is enigmatic why parts of this chapter are highlighted in yellow.
Continued process verification
The continued process verification aims at providing evidence that a state of control is maintained. According to the draft scope and extent of process verification will be influenced by a number of factors
Manufacturers should describe the appropriateness of the verification strategy in a protocol. Process parameters and material attributes that will be monitored as well as the validated analytical methods that will be employed should be listed.
The following should be defined:
Furthermore, any statistical models or tools used should be described. There is also a reference to continuous processing stating that the stage where the commercial process is considered to be validated should be defined clearly if continuous processing is employed.
The draft states that periods of enhanced sampling and monitoring may help to increase process understanding as part of continuous improvement. Process trends (such as the quality of incoming materials) should be assessed in order to verify the validity of the original process validation or to identify required changes to the control strategy.
The extent and frequency of continued process verification should be reviewed periodically and modified if appropriate.
The life-cycle of a product should be accompanied by an accompanying change management. This is valid also for existing systems or processes. Sufficient data should be generated to demonstrate that the changed process will result in a product of the desired quality, consistent with the approved specification.
Furthermore, the necessity of a thorough documentation and a regulatory approval, where appropriate (variation), is stated.
There can be found many "old friends" among the listed references:
The deadline for comments was strikingly short. It ended already at the end of April. The document containing the incorporated comments will be presented at the forty-ninth meeting of the WHO Expert Committee in October 2014. You can find the document in the ECA Website Members' Area.
The draft of the document is oriented strongly towards the FDA Guidance on Process Validation. The validation life-cycle for example is directly comparable with the life-cycle requested by the FDA. Hence, the qualification of apparatuses is part of the process qualification as is the case in the FDA Guidance. (Unfortunately) there are almost no indications as to the handling of legacy processes. Only one statement in the chapter Change management includes existing systems and processes. There is still mentioning of the traditional approach with 3 validation batches but the number of validation runs is to be defined on the basis of a risk assessment. Parts of the sometimes very detailed requirements concerning the content of the validation protocol and concerning changes which might trigger a revalidation are interesting. Some rudiments from the developing stage (such as comments) indicate that the review prior to the publication of the draft was not really optimal. The deadline for comments was surprisingly short.