Isolators and separative devices

Jeff Granger of Pharmagraph looks at particle monitoring of isolators and separative devices to meet the EU-GMP Annex 1

Jeff Granger of Pharmagraph looks at particle monitoring of isolators and separative devices to meet the EU-GMP Annex 1.

The revision of the EU-GMP Annex 1 in September 2003 changed the requirements for non-viable particle counting in pharmaceutical Grade A and B areas. While there has been much discussion of how the new requirements should apply to cleanrooms, little has been said about how the guideline should be applied in the case of isolators and other "separative devices", such as microbiological safety cabinets and laminar flow benches. This article offers some guidance on the approach we should take to ensure compliance.

The designs of isolators and safety cabinets are many and varied. Different manufacturers each offer their own unique designs, and different designs are relevant to different applications. Broadly however we can identify three groups of enclosures, those that simply bathe the process/product in clean air (e.g. horizontal laminar flow benches), those that also provide a reasonable level of protection to the operator (e.g. a Class 2 microbiological safety cabinets), and those that completely isolate the product from the operator and surrounding environment by means of a physical barrier (isolators of various types). Each device offers a different challenge when attempting to devise a suitable particle monitoring strategy.

Why, you may ask, should we concern ourselves with particle monitoring now when previously it was rarely considered to be of importance? The answer lies in the changes to the EU GMP Annex 1, which itself reflects a changing attitude in the industry to the benefits particle monitoring.

"A continuous measurement system should be used for monitoring the concentration of particles in the grade A zone, and is recommended for the surrounding grade B areas.

For routine testing the total sample volume should not be less than 1m3 for grade A and B areas". (EU-GMP Annex 1 Sept 2003).

The basic requirement is that all Grade A areas should be continuously monitored during production for the presence of >0.5µm and >5.0µm particles. It therefore follows that if an isolator or safety cabinet is deemed to be a Grade A area then it should be continuously monitored to be in compliance with the EU-GMP.

This raises a number of issues:

    1. How can we physically sample the air inside a cabinet?
    2. What type of particle counter is suitable?
    3. Are there any safety considerations?
    4. Will the processes in the cabinet have an adverse affect on the particle counter?
    5. Will the particle counter have an adverse affect on the processes inside the cabinet?

The sampling of airborne particulates from within an enclosed controlled environment is a challenge in itself. Most 1cfm flow rate counters on the market would be physically too big to locate within the restricted dimensions of the typical isolator. Even if it would fit, it is not possible to sanitise the inside of the counter, which would potentially leak non-sterile air into the work area. Some smaller, hand-held counters such as the Met One HHPC6 have been used, but in continuous operation would require frequent refurbishment of the sample pump which is not rated for continuous operation.

If it is not practical to put a particle counter inside an isolator, the only other option is to draw our air sample out of the isolator to monitor its cleanliness. At this point we need to take a step backwards and consider all the safety issues, including a formal a risk assessment. We are now proposing that we take our carefully designed isolator or microbiological safety cabinet, which has been certified to a known level of containment, and we are going to make a hole in the fabric of the device, thereby breaching its containment, and we are deliberately going to draw air outside the enclosure.

Clearly, for some hazardous materials, we would rapidly conclude that this is simply not worth the risk. However, for more "routine" pharmaceutical products we can design a system which is both safe and meets the sampling requirement of Annex 1. If the sample air is deemed to be potentially hazardous then suitable precautions must be taken to ensure that the exhaust air is handled properly.

On recent projects we have implemented two different solutions. In one pharmaceutical production facility it was deemed adequate to duct the exhaust from the air sampling pump of the particle counter back into the building air handling unit, which itself was filtered before discharging into the open air. In a hospital production facility on the other hand, the requirement was to sample from both cytotoxic and radio pharmaceutical isolators. In this case the system design incorporated a filter capsule plumbed into the vacuum line within the production area. This ensures that no hazardous material could exit the facility, and then as a belt-and braces backup, the exhaust is ducted back to the air handling system.

At first sight, particle counting and isolators do not make good bed fellows. Firstly, the very nature of the manipulations that occur within an isolator or safety cabinet often involve the spraying of sanitising agents such as ethanol or IPA. Any aerosol that enters a particle counter will be counted, and more likely than not will trigger an alarm. This will then need to be acknowledged and explained away. Further more, getting any aerosol in the sensor of a laser particle counter is likely to "watermark" the optics over a period of time which will require a service visit to resolve.

It is now accepted that whilst continuous particle sampling during production is normally required, there are instances where this is simply not practical. The accepted compromise is baseline particle monitoring, where we continuously sample the air quality in the isolators when production is NOT taking place, and shut off the monitoring system before production begins. What this achieves is to provide a constant level of assurance that no leaks have developed in the fabric of the enclosure or in the filters for a 14-hour period before production starts. The particle monitoring system can then be partly or completely shutdown from the enVigil monitoring software, and the sampling probes are capped of with bright red caps as a visual check that it is safe to begin production. When production ends and after the isolator is cleaned down the caps are removed and sampling restarts. It is then possible to confirm that the isolator has again cleaned down to within specified limits and no leaks are detectable. This provides both a before and after production baseline and assurance that the isolator is in good working order and also falls in line with the Annex 1 requirement for the area to be monitored both before and after production.

One project on which we are currently involved utilises the popular gassing isolator from Amercare. This has a high rate of air flow, and is a turbulent flow design rather than the more usual laminar flow system. Turbulent flow sampling is ideal from a particle monitoring perspective, since contamination from any leak is easily detected by the particle counting system.

Furthermore, the manufacturer has designed in a choice of two sampling ports at the air inlet port and also at the outlet port. Sampling at the outlet is favourable in order to monitor worst case situation, providing the production processes themselves do not generate excessive levels of particulate contamination.

On projects that use laminar flow cabinets it is important that sampling should be isokinetic, and to this end Pharmagraph have designed a range of sampling probes that will integrate with isolators and microbiological safety cabinets from most of the major manufacturers.

The particle sensors themselves should ideally be full 1 cfm flow rate pharmaceutical point of use sensors. The Pharmagraph CPC1 has been designed specifically to meet this requirement, and when integrated with the enVigil software package provides accost effective, validated solution to demonstrate continued compliance with EU-GMP Annex 1.

With multiple sensors on site (one for each cabinet), cost of ownership is an important issue, and the Pharmagraph environmental monitoring system has been specially developed to be the lowest maintenance cost system currently available. This is achieved by utilising the Met One Long Life Laser ("L3") technology which offers a 10-year MTBF rating on the laser diode, together with "Smart Socket" sensor connections which allow fast swap outs of particle count sensors to minimise down time. In conclusion, the requirement to continuously sample particulates during production from within an isolator or MSC might seem daunting at first, but with the right design it can be achieved successfully. Pharmagraph have now successfully implemented a number of such applications and are happy to advise on new projects.

Featured Companies

Pharmagraph (more information, website)