EMA publishes final Version of Q&A Paper on Production of "cold" WFI

After the change to the European Pharmacopoeia, it is possible to produce water for injection by other methods than distillation since April 1st 2017. The additional guideline, which the European Medicines Agency (EMA) published as a Question & Answer draft document in 2016, has now been published in its final version.

The original separation of the Q&A document into two parts remained. In the first part, questions relating to the production of WFI with membrane technologies are answered. The second part is about biofilms, their background and approaches for their prevention.

Part 1, which is essential for system design and operation, is now more precise. Some inconsistencies, especially concerning sanitization, have been eliminated. The seven questions of part 1 also stayed the same, only the answers were revised in some parts:

1. According to the Pharmacopoeia, the supervisory authority has to be informed before the use of cold-generated WFI. Who is the supervisory authority?

For manufacturers in the EU, it is clearly the supervisory authority responsible for GMP oversight. If the manufacture takes place in a third country, it is the GMP authority in the country of the importer.

2. What are the greatest concerns in regards to the use of reverse osmosis (RO) to manufacture WFI?

The greatest concerns relate to the microbiological quality of the produced water, meaning the number of microorganisms and the endotoxin load as well as their measurement. Reverse osmosis systems are usually operated at room temperature, which is ideal for the formation of a biofilm. A new addition says that it has to be proven that the purification process (RO coupled with suitable techniques is given as an example here) is equivalent or better than distillation.

3. Which are the most important elements concerning the design of a (water) system?

This is the biggest and primary part of the Q&A paper. EMA's explanations include: control strategy, materials of construction, pre-treatment of the water, RO membranes, additional techniques for supplementing the RO, TOC measurement, conductivity measurement and sanitisation of the system. Within the control strategy, measures should be defined to evaluate the potential for the formation of biofilms as well as measures to minimize the formation of biofilm. As for materials of construction: it should be made sure that the generation and distribution system allow routine thermal and/or chemical sanitisation. It should be noted that during pre-treatment, microorganisms that enter the RO system through untreated water find favourable growth conditions on the large surface of the membrane. Therefore, organic particles (which may serve as nutrient for microorganisms) should be removed. It is also important that existing chlorine or other oxidising substances may damage the membrane of the RO module. Electrochemical monitoring of those substances before the RO membrane is therefore recommended. The RO membranes themselves should be suitable for thermal and chemical sanitisation. The integrity of the membranes should be tested routinely. Further methods mentioned for the coupling with reverse osmosis are double reverse osmosis; nano, micro or ultrafiltration or electro-deionisation (EDI). There is likely no way around an online total organic carbon (TOC) and conductivity measurement. The document demands a risk-based approach for the number and location of measurement points, e.g. in the feed water, after pre-treatment, after the RO membranes, before the delivery to the storage tank and of course, as is often installed by default, in the return loop back to the storage tank. The system should allow draining if the water is non-compliant or to recirculate through the purification process, both of which cases should be recorded and monitored by the quality system.

4. How is qualification supposed to look?

Amongst other things, a maximum time limit for membranes should be established during qualification. A destructive analysis on membranes can be sensible to ensure the absence of a biofilm.

5. At which point during qualification and routine should sampling be performed?

The sampling locations are listed as expected: feed water, during pre-treatment, pre and post RO membrane, after the final purification step, in the storage tank, at all user point and in the return loop. In the part about routine sampling, the document lists the daily sampling of users (who received water on a respective day), the recircling (return loop), but also the sampling before and after the RO membranes.

6. What testing (analyses) should be performed during the qualification and the routine?

Apart from the tests described in Ph.Eur. Monograph 169, rapid microbiological methods are mentioned several times here. The use of rapid microbiological methods should be part of a primary control strategy for the system.

7. What is expected in regards to preventive maintenance of a (reverse osmosis) system?

During maintenance, a change of the filters, gaskets, seals and RO membranes in previously defined intervals is expected, as well as their replacement if damaged. Detailed (visual) inspections for possible biofilm growth should also be part of the maintenance. The performance and separating behaviour of the RO membranes should be assessed as part of a routine, as well.

The explanations above show that membranes and reverse osmosis systems as well as their impact on the microbiological quality of water and WFI are in the focus of the EMA.

On the EMA website you can find the final Q&A document about the production of WFI by non-distillation methods.

Go back


Stay informed with the GMP Newsletters from ECA

GMP Newsletter

The ECA offers various free of charge GMP newsletters for which you can subscribe to according to your needs.

To subscribe, please click here.