The revised version of Chapter 1 of the EU GMP Guidelines  came into force in 2006. Item 1.5 makes the annual Product Quality Review (PQR) compulsory for all licensed products. This continuous revision of the consistency and validity of the entire manufacturing process also includes a stability monitoring programme as listed under subitem vii.
These On-going stability tests are specified in Chapter 6 of the EU GMP Guidelines, which has also been updated in the same year.
Areas of Quality Control which previously formed only part of R&D are now included in official GMP inspections. This is the case, for example, with the qualification of the special equipment for stability testing, such as controlled storage cabinets. Standard operating procedures (SOPs) dealing with out-of-specification results, not only for batch release but during stability testing and for evaluating out-of-trend results have to be implemented.
External laboratories contracted to carry out on-going stability testing must be included in the manufacturing licence. Technical Agreements for third party analysis (which define responsibilities on both sides) must be thoroughly revised under theses premises or must take account of it when they are first drawn up. The Qualified Person employed by the manufacturing or contracting company must ensure that the tests are carried out in accordance with GMP requirements by means of an audit.
It should be kept in mind that on-going stability testing differs from other stability tests in its regulatory and legal background (see also table 1) [2,3].
However, these requirements only apply to licensed medicinal products which are currently on the market.
REQUIREMENTS FOR ONGOING STABILITY TESTING
The categorical requirements of the GMP Guidelines for on-going stability testing are that:
- All medicinal products/formulations have to be tested, with no exception in principle, for homeopathic products, herbals etc.
- All finished products and where appropriate bulk products, too, have to be tested (e.g. when stored or transported for prolonged periods). Excipients and active substances are not taken into account here.
- Requirements apply in principle to every product in every dosage and pack size or (primary) packaging type/package.
- Tests have to be carried out continuously, usually one batch a year.
- Studies are to be carried out under long-term conditions (e.g. 25°C / 60% rh) continuously over the period of the labelled shelf-life.
Tests following storage under intermediate and accelerated conditions should only be carried out if supplementary information is required at an early stage. In principle, all decisions which affect the test protocol, the frequency of testing and the choice of test samples should be product-based, targeted and based on a risk analysis (see table 2).
On-going studies are intended to prove that over the period of its labelled shelf-life and under “real life conditions”, the product remains of the quality defined in the authorisation/registration documents. Stability studies done for registration solely provide a “snap-shot”. Adverse effects such as changes in manufacture and the supply chain (even those not on a variation level) should be identified by on-going studies. However, this procedure also implies that the test protocol can be adjusted at any time to the current situation. The stability protocol does not necessarily have to comply with the ICH stability testing guidelines.
POTENTIAL SAVINGS – REALISATION AND PITFALLS
Item 6.28 of the EU GMP Guidelines specifically states that the protocol for the on-going stability programme may differ from that of the initial long-term stability protocol , giving a reduction in the frequency of testing as an example. If it has been shown that a test parameter is not critical, for example during the course of commitment (follow-up) studies or later in on-going studies which have already been performed, the frequency of testing may be reduced.
A given example (see table 3) shows that overall, there is a possible reduction in the required frequency of testing of about 50% or even more. It should be noted that at the end of the shelf life, all the test parameters should be checked again (in this case t60), in order to prove stability over the entire storage life.
If the product is manufactured in different strengths (same API with similar matrix) and pack sizes, the “bracketing”  recommendations can be applied. If there are more than two sizes for any parameter, this provides for only the extremes to be tested (see table 4).
This example shows the requirements in the guidance. If the exemplary product above, assuming an identical bulk product, is packed in blister packs of 10 tablets each with identical primary packaging material, the effect of secondary packaging – for example different secondary packaging for different foreign markets – does not have to be taken into account. Testing of only one pack size (would therefore be sufficient. The same applies too to multi-dose containers, where only the most sensitive size of container would have to be tested on a “worst case scenario” basis.
In the latter (ideal) example, the number of tests could be reduced from 9 to 2, with corresponding cost savings.
However, consideration of the individual case is always necessary. It may be necessary anyway, for example, for the first batches to be tested in accordance with the full protocol in order to provide sufficient data for trend analyses.
The combination of on-going studies with follow-up (commitment) studies is often discussed. It should be emphasised here that in principle the data which are generated in follow-up studies can also be used for review as mentioned in subitem vii in the PQR. Conversely, however, this means that tests must cover the full protocol in the dossier. Savings such as those discussed above are hardly possible in that case.
If it was agreed in the post-marketing commitment that after launch, one batch per year would be tested for three years for the follow-up study, the additional (on-going) tests could be saved for three years.
As usual, analyses for the release of a product batch can also be used as starting values for on-going studies but, possible differences in the specifications have to be observed. Thus, further on-going testing does not start until a year after the start of storage.
It is also recommended that production and hence starting dates should if possible be in the same period of the calendar year. As a result, in subsequent years there will be further synergies as a result of parallel testing.
STEPS AFTER TESTING
Suitable data analysis procedures have to be established which allow not only retrospective but also prospective evaluation. The EU GMP Guidelines require that at the individual test times it must be stated whether it can be assumed that for example the active substance content is likely to remain within the specified limits as testing continues. Here, statistical methods of regression and confidence interval analysis can be useful, although they should be used with caution.
The results of on-going stability testing have to be summarised in a report. All current results have to be compared with data from previous tests, for example from follow-up studies, and trend analyses have to be carried out (if applicable).
The results have to be incorporated and discussed in the relevant PQR.
It must also be noted that the manufacturing department (internal or contract manufacturer) and the Qualified Person responsible for market release have to be notified of the results of the on-going stability studies.
An appropriate Standard Operating Procedure (SOP) should specify what should be done in the case of out-of-specification (OOS) results and significant negative trends. If verified they must be reported to the supervisory authorities. Further considerations should also be taken into account here:
- for example a review of sensitive or relevant areas in production, to try to identify the cause and possible transmission to other batches/products;
- a critical review of the labelled stability, if necessary a recall of the affected product batch must be considered;
- measures to avoid errors should be discussed and implemented (corrective and protective actions, CAPA); in justified cases other specialist personnel, such as the Qualified Person for Pharmacovigilance, need to be involved.
After the implementation of corrective actions, the subsequent production batch should be reviewed by an on-going study in addition to the normal yearly rate of testing. The test parameters must be agreed on the basis of the individual situation.
As with PQR in general, on-going stability testing also requires intensive exchange of information between the responsible persons/relevant specialist departments. The responsibilities must be clearly laid down. If external service-providers are involved, appropriate Technical Agreements must be drawn up which define the responsibilities.
The involvement of an external service-provider may be beneficial to the marketing authorisation holder: Expensive implementation of an in-house system can be avoided by employing the existing system of the service company. Investments and fixed costs are not increased. The marketing authorisation holder can concentrate on its core competences such as marketing. Outsourcing all other functions is, in principle, possible .
The message to take away is that expenditure and hence production costs can be minimised by sensible and skilled planning of the test protocol for on-going stability testing.
Thanks to Dr. Annemarie Jasper and Christian Rieke for their helpful support.
 EU Guidelines to Good Manufacturing Practice Medicinal Products for Human and Veterinary Use (http://ec.europa.eu/enterprise/pharmaceuticals/eudralex/home4.htm)
 Guideline on Stability Testing: Stability of Existing Active Substances and Related Finished Products (CPMP/QWP/122/02, rev 1)
 Note for Guidance on Stability Testing of New Drug Substances and Products (CPMP/ICH/2736/99 corr)
 Note for Guidance on Bracketing and Matrixing Designs for Stability Testing of Drug Substances and Drug Products (CPMP/ICH/4104/00 adopted 2002 / ICH Q1D)
 Sandner, Klöpf (2006), Outsourcing in der Pharmaindustrie, Teil 1 Pharm. Ind. 68, Nr. 7, 825-828, Teil 2, Pharm. Ind. 68, Nr. 8, 932-936