Standardised sterility

The latest version of the so-called "sterile annex", published by the European Commission in May 2003 (for implementation on September 1 2003), invoked considerable backlash from the industry that had hoped to see more synergy with FDA expectations and compatibility with the ISO 14644 family of cleanroom standards. Such was the concern, that the GMP inspectors within the European Medicines Evaluation Agency (EMEA) decided that Annex 1 should be thoroughly re-reviewed to deal with controversial areas and some of the confusing text.

Points on guidance The EMEA is made up of representatives from the regulatory agencies in European countries and its economic area. Within it are a number of committees and working groups, including the Ad Hoc Inspectors Group that focuses on GMP-related guidelines and procedures, providing a forum for exchange of information and enabling harmonisation to avoid differences in interpretation across member states. In Europe the principles and guidelines of GMP are laid down by directives 91/356 for human medicinal products, and 91/412 for veterinary medicinal products. Most importantly in both directives – we have the same words in Article 3 – that tell us that manufacturers and the inspectors from the competent authorities shall refer to detailed guidance. The detailed guidance is published by the Commission in the Guide to GMP for Medicinal Products, and that is the legal basis for Volume 4 of the Rules Governing Medicinal Products in the Community.

Reasons for revisions In practice this means that the Commission is responsible for the publication, relying heavily on the technical input from the EU GMP inspectors. The EMEA is responsible for co-ordinating this input through the Ad Hoc Inspectors Group. Revisions to GMP guidance may be prompted by a number of different events: proposals from industry; a technical change; a change in a referenced standard; feedback from an inspector. The normal process within the EMEA is to first develop a concept paper to frame the discussion that will consequentially develop. This will be discussed in the Ad Hoc Inspector's Group and if there is agreement that this area needs revision, the group seeks internal volunteers to draft the guidance. Eventually the proposals are issued for public comment, the comments are considered, and final guidance is published. So what happened in the case of the Annex 1 revisions? The first edition of the EU GMP Guide, published in 1989, included an annex on the manufacture of sterile medicinal products that was similar to the guidance available in the UK 1983 Orange Guide. This UK guide was based on the superseded BS 5295: 1976 cleanroom standard. The second edition of the EU GMP Guide was published in 1992 with an unmodified Annex 1. In 2000, paragraph 42 was revised to clarify certain aspects concerning media simulations. In December 2000 the author (Gordon Farquharson) addressed the EMEA's Ad Hoc Group to explain the rationale and content of the then new EN/ISO 14644-1 & 2 standards. These standards, which replaced US Federal Standard 209E and all existing cleanroom standards in the CEN member nations, are essentials tools needed to support regulatory guidance such as Annex 1. Annex 1 contains nothing specific about cleanroom classification, metrology or evaluation of particle counting data. It relies on the generic technical standards.

Developing a compromise This explanation prompted the revision of the Annex 1 as issued in May 2003. The AD Hoc Group reviewed the ramifications of the presentation, and set out to amend Annex 1 to try to develop a compromise between the ISO/CEN standards and the EU GMP requirements. A drafting Group, comprising GMP inspectors from the UK, France and Germany, set to work in December 2001. The proposals were published for consultation and comments were received from some 60 different bodies covering the modified text (confined to section 3 and 20) as well as other parts of Annex 1. These comments were summarised by the Drafting Group and then discussed in the full Ad Hoc Group, resulting in publication of the revised version in May 2003.

The problems remaining In discussions about GMPs we are often reminded of important works in the forward to the EU GMP Guide. This states that if there are other acceptable methods, other than those described in the Guide, which are equally capable of achieving the same principles of quality assurance, then it is not intended to place any restraint on the development. However, it would be expected that alternative approaches be justified and discussed with the particular supervisory authorities involved. It is the author's opinion that this flexibility is applicable to developing options as how to achieve certain performance, but is practically impossible when it comes to setting the most basic of performance targets. Cleanroom class and air cleanliness specifications are these basic requirements that must be specified in an unambiguous way. This is where the biggest problems lie with Annex 1 right now! There is little clarity about which of the EN/ISO standards apply and how to use them. This is exaggerated because there are several specific areas where the requirements of the GMP are at variance with the generic EN/ISO 14644 cleanroom standards or the specific aseptic processing standards such as ISO 13408. The GMP should be quite clear where it is necessary to adapt the available standards to the demands of the GMP. Our industry is seeking a clear unambiguous tool that will act as an effective foundation for the delivery of safe and efficacious sterile medicinal products. The following outlines some specific proposals.

Basic cleanroom classification We need to understand some basic principles about particle counting, in particular the risks associated with the measurement of very small populations of particles using light scattering techniques. Both EN/ISO 14644-1 and superseded FS 209E have virtually identical methods for calculating the air sample volume required at each location for formal cleanroom classification:

• The formula is Vs = 1,000 x 20/Cl
• Vs is the sample size in litres
• Cl is the class limit at the largest considered particle size (particles/m3)
• 1,000 is the conversion from m3 to litres
• 20 is the minimum count deemed to be significant and acceptable in a sample if at the class limit.
• There is also a minimum requirement for a 2 litre sample size and a minimum sample time of 1 minute

More detailed guidance The following text is suggested to resolve these problems: "Grade A zones shall be monitored in operation using automatic fixed or mobile particle counting systems, and the use of such systems should be considered for Grade B zones that are likely to have a direct effect on the performance of the Grade A zone. Individual particle counters, manifold systems or point of use counters may be used. "Since it is desirable to review samples frequently to look for deviation events, sample volumes should be related to a sensible frequency (typically 1-5 minutes) and not the sample size required for classification. Additionally, when using manifold systems, the recovery of particles <5.0 micron is likely to be poor, and these larger particle sizes should not be considered. "When particles <5.0 micron are considered, individual occasional small concentration excursions should not trigger an investigation because they are unlikely to indicate an actual problem. The monitoring protocol should look for deviations in sequential samples as an indicator of a real problem." The expression of a range of flow velocities is unclear in Annex 1. Is the intent that set point or average velocities chosen from the range are acceptable, and that some variation around this set point would also be acceptable? Industry seeks the ability to select an appropriate velocity and then prove it is effective by airflow visualisation. The following is suggested: "Normally Grade A zones shall be protected by a uni-directional airflow cleanzone. (Note: these were formerly frequently referred to as Laminar Flow work stations.) Uni-directional airflow systems shall provide a homogeneous airflow at a defined velocity at a defined test plane near to the critical part of the process. The velocity should have a set point in the range 0.3-0.7m/s depending upon the application. A guidance value in open cleanroom applications is 0.45m/s +20%. "Satisfactory performance of uni-directional airflow shall be demonstrated by video recorded airflow visualisation using a neutral buoyancy aerosol or smoke. Additionally it is important to prove that the final HEPA filter is operating within its specified range. Therefore the filter face air velocity shall be measured close to the filter face (guidance value 50-100mm)." In addition to the technical issues in Annex 1 that need to be addressed, there are several areas where terminology and the use of language is inconsistent, incorrect, confusing or unnecessarily different from EN/ISO terminology. Some examples are:

• The use of the term "laminar" is now considered incorrect, and accepted standards use the term "unidirectional flow"
• Is there an intended difference between the use of "should" and "recommended"?
• The use of the word "continuous" in the context of particle monitoring is unclear
• Does the term "routine testing" mean the process of strategically or periodically undertaken formal cleanroom qualification as differentiated from the term "monitoring" which is an ongoing operational activity?
• There should be clarity about what grade of filter is required. There is some confusion in some markets.

Conclusion We have a great opportunity to refine and improve Annex 1 of the EU GMP to ensure it provides a firm foundation to our cleanroom facility design and operations. Join the debate and contribute through your professional institution or body such as ISPE, Parenteral Society or the PDA.