This is the final part of a three part article on the new PIC/S Recommendation (PI 032). The Recommendation interprets the changes to Annex 1 of the PIC/S GMP Guide. In this part, we look a few minor issues arising from the publication.
There are a few other points in the interpretation which I feel have the potential to cause confusion. These aren’t so significant, but warrant a mention to raise awareness.
- In section 12 the Annex confirms that sample sizes for monitoring do not need to be the same as for classification. The PI 032 interpretation quite rightly points out that needing to sample a full cubic metre before activating an alarm is problematic, especially in Grade A zones. The implication, of course, is that significantly smaller sample sizes are appropriate where fast response is required.
With that in mind, the use of the phrase “sampling of 1 m3 … could be inadequate” is potentially misleading, insofar as the word “inadequate” implies not enough, when really it is too much. A more suitable term to avoid confusion would have been “inappropriate”.
- The inconsistency between sections 34 and 116 was raised in part 2 of this article. If you are unsure about which section to follow, obviously the safer and more advisable route is to stick to section 116 and avoid the use of sealed transfer trays.
In itself this issue is barely worth mentioning, but with PI 032 implying that section 12 is now superseded, it seems that the carryover of section 34 has not been noticed by the reviewers or interpreters.
Additionally, the interpretation states that section 116 is “basically equivalent” to the old section 12, yet there is a very important distinction. There is no provision in the new section for transfer in sealed containers, and it would be impossible to meet this stated requirement using sealed containers, as Grade A airflows cannot be met.
- On a positive note, I should mention that I had some issues with the PI 032 interpretation of clause 121, however, the revision PI 032-2 has largely addressed those concerns. Originally the document could be interpreted to suggest that you could mitigate an ineffective rejection system by improving the quality of the background environment. The revision clarified these comments to put the emphasis on monitoring and the degree of human intervention required in the rejection system.
Finally, it should be noted that there are 20 or so interpretations in the PIC/S recommendation and I have highlighted 4 or 5 which I think could be reworked to varying degrees. The remainder generally provide sound advice, clarity and definitions that the user can employ with confidence.
However, the issues raised should be addressed by PIC/S – especially those covered in parts 1 and 2 of this article. Given that PIC/S was prepared to issue version 2 only a month after version 1, let’s hope that version 3 may not be too far away.PIC/S Interpretation of Annex 1 – Clarifying or Confusing? Pt 2
January 18th, 2010
Following on from my previous blog entry, this article looks at a second significant issue with the PIC/S Recommendation (PI-032) which interprets the changes to Annex 1 of the PIC/S GMP Guide.
The second, less critical issue I have relates to section 120 on capping environments. The Annex states that capping can be performed as either an aseptic process, or as a clean process outside the aseptic core. However, if uncapped vials are to be removed from the aseptic core, the Annex specifies that a “Grade A air supply” must be maintained until the vials are capped. The definition of Grade A air supply has been helpfully clarified in the revision of PI 032.
While this requirement no doubt creates a significant cost impost for new facilities and where the requirement is currently not met, the reasons for it are understandable. There are, however, two areas for concern:
- The requirement for Grade A air supply has presumably come about because uncapped vials are considered non-integral, and therefore potentially at risk of being contaminated. Grade A air supply rather than “Grade A conditions” have been specified, perhaps because of the difficulty and expense in installing, maintaining and qualifying such systems couldn’t be justified in light of the risk. Even so, a risk has been identified and measures provided to mitigate that risk.
With this is mind, why would the interpretation advocate the use of unsterilized caps on these non-integral vials? Yes, it’s understood that the process at this stage is not aseptic, but after having deemed the risk high enough to mandate that manufacturers spend many thousands of dollars on maximising air quality, why is PI 032 prepared to expose these same ‘non-integral’ vials to potential contamination from unsterilized caps?
- Another area of confusion is related more to the Annex than the interpretation. section 34 states that partially stoppered vials may be transferred to a freeze dryer either under GradeA conditions, or in sealed transfer trays in a Grade B environment. section116 covers the same topic, but omits the sealed transfer trays. section 120, on the transfer of fully stoppered vials to a capping station also does not provide scope for transfer in sealed trays.
If these sections are taken at their word, the Annex allows the transfer of partially stoppered vials in sealed transfer trays, but does not allow the same for fully stoppered vials. Logically, this doesn’t make sense. (The section 34/116 inconsistency is discussed further in part 3)
To sum up, I would expect that caps on aseptically filled vials would be sterilised before introduction into the capping process, regardless of whether the process is truly aseptic, or simply ‘clean’. The impost of sending a bag through the autoclave is surely next to nothing compared with that of complying with the environmental requirements of the Annex, but maintains a consistent approach to protection of ‘non-integral’ vials.
Similarly, despite the retention of section 34, it seems clear that the intention of the guide is to move manufacturers away from sealed containers and I would recommend against their use. It would surprise to see section 34 modified or removed from future revisions.
The final part of this article will discuss some minor issues in the interpretation.
January 18th, 2010
This is the first part of a three part article critiquing the new PIC/S Recommendation on Annex 1 of the PIC/S GMP Guide.
Last month, the PIC/S committee released their interpretation of the 2008 changes to the GMP Guide Annex 1 (manufacture of sterile medicines). This document, PI 032-1, was agreed by PIC/S in November last year after a committee meeting in Uppsala, Sweden. Following some apparent concerns about the publication, a revision (PI 032-2) was published on January 8, 2010. But if the original document had issues, has the revision addressed them all?
In my opinion, while the revision has provided some modest improvements in clarity and added one or two useful pieces of information, there are still a number of areas for concern. There are statements within the interpretation that, rather than providing clarity, create more confusion for manufacturers. In both the original version and the revision, there are two interpretations which I find particularly confusing.
In this first part of the article I will address the most critical issue, which relates to section 10 of Annex 1. The text states (emphasis and comment added):
“It is recommended that a similar system (to that employed for Grade A zones) be used for Grade B zones although the sample frequency may be decreased. The Grade B zone should be monitored at such a frequency and with suitable sample size that changes in levels of contamination and any system deterioration would be captured and alarms triggered if alert limits are exceeded.”
The PI 032 interpretation states in full:
“Continuous monitoring (see definition under interpretation to section 9) is expected while not fully integral containers are handled in the B zone, e.g. partially stoppered vials within a laminar air flow mobile unit prior to lyophilisation. Manifold systems might not be suitable for Grade B Zone monitoring due to a lack in responsiveness.”
I see at least 2 problems with this interpretation:
- PI 032 applies the same definition for continuous monitoring for both Grade A and B zones (a system which is able to pick up ANY potentially occurring event of an unusual number of particles). Interestingly, the term “continuous monitoring” is not used anywhere within Annex 1, least of all in the relevant sections 9 and 10. The definition described in PI 032 most closely matches the description of the monitoring requirements of Grade A zones, BUT Annex 1 clearly differentiates between monitoring requirements for Grade A and B.
- Grade A: The grade A zone should be monitored at such a frequency and with suitable sample size that all interventions, transient events and any system deterioration would be captured and alarms triggered if alert limits are exceeded
- Grade B: The grade B zone should be monitored at such a frequency and with suitable sample size that changes in levels of contamination and any system deterioration would be captured and alarms triggered if alert limits are exceeded.
My interpretation of the differences here is that Grade B monitoring need not capture every single unusual event, but rather be able to detect system deterioration and or gross change in contamination levels. Surely if the intention of the code was that every single spike must be captured, terminology much more akin to the Grade A definition would have been applied.
I believe those interpreting Annex 1 have erred in suggesting that monitoring for Grade A and for Grade B environments can be grouped together like this. In my opinion, continuous monitoring in Grade A and Grade B zones is NOT intended to be the same.
- PI 032 states “Continuous monitoring … is expected while not fully integral containers are handled in the B zone”. Given the absence of any other information, this implies that this is the only condition under which “continuous” monitoring is required in B zones. Having understood and accepted the term “continuous monitoring”. I find this interpretation quite strange for the following reasons:
- Firstly, Annex 1 provides no direction on what activities should be monitored in Grade B areas. The only concession given is that the effectiveness of segregation between Grade A and B zones should determine the importance of the particle monitoring. Surely PI 032 goes beyond interpretation to place a specific condition which cannot possibly be inferred from Annex 1 itself.
- Secondly, what exactly does “while not fully integral containers are handled within the B zone” mean? PI 032 gives an example of partially stoppered vials within a mobile laminar flow (must be Grade A according to section 116), but how is that being “handled within the B zone” anymore than vials being processed within a fixed laminar flow? There are more questions here, but the confusion is evident.
In trying to reconcile why PIC/S have published this interpretation of section 10, I wonder if they have just been too brief, omitting important additional information, or perhaps they have misinterpreted the intent of the original text. To me, the text appears to require monitoring of Grade B environments for the duration of aseptic and other critical processes (continuous, if you like), albeit at a reduced sampling frequency from Grade A monitoring. But this is not what the interpretation says.
We need further elaboration from PIC/S on this subject to remove the confusion.
Part 2 of this article will deal with concerns about section 120.
January 16th, 2010
An interesting press release was issued by PIC/S last month, commenting on the results of a PIC/S assessment team visit to the FDA: ” more progress was achieved during these
two weeks than during the previous three years of assessment. The team assessed the US GMP inspection system and also observed inspections carried out by FDA inspectors. It also met with FDA Commissioner, Dr. Margaret Hamburg, who stated that PIC/S membership was a top priority for the FDA.”!
This is exciting news for pharmaceutical and medical device manufacturers who must comply with multiple regional regulations and country-specific regulations. At long last, there may be an end in sight to the wasting of valuable time and resources in complying with multiple regulations rather than building quality and patient safety into the product. A global GMP standard may actually be achieved in my lifetime!
Whilst it’s great news that the FDA is moving closer to PIC/S membership, it would be a real shame they throw out all their regulations / guidance documents in the process. For example the draft “Process Validation Guidance” issued by the FDA last year provides a refreshing look at product development, commissioning and control. It outlines a whole life cycle development approach to product quality and risk management. So much better than the old three golden batches, how can three consecutive batch “prove” that your process is control! If the FDA does become a PIC/S member, I hope that PIC/S can pick up the best of the FDA’s work and integrate it into their own regulations and guidance documents.
It is interesting to note that there are now thirty seven PIC/S member countries. The membership applications of Indonesia / NADFC, Slovenia / JAZMP and Thailand’s FDA were recently reviewed, potentially expanding the number further. The next assessment for membership will be Latvia’s State Agency of Medicines (ZVA).
December 1st, 2009