Paradise Lost

When isolators were introduced into the pharmaceutical industry they were properly viewed with some degree of skepticism. The early designs were relatively crude in appearance and certainly lacked sophistication. With a few years of technology development and successful operational experience it seemed that the isolator would change the way in which sterile products were made across the world. 

I was bold enough to predict the rapid demise of manned cleanrooms as the highly capable isolator proved its superiority both operationally and financially.[1] The isolator was expected to be the paradigm changer that the pharmaceutical industry needed to attain the next level of performance and product safety. The virtual elimination of contamination compounded with expected lower costs would create an operational paradise. A number of unanticipated changes to isolator designs occurred on the way to that rosy future that has dramatically lessened the expected impact. This article will review the ways in which the vision of the future envisioned in 1995 has been diminished and outline changes to current practices in isolator and barrier that would enable the industry to fully realize the potential in isolation technology.

The essential difference between isolators and manned aseptic processing area is the absence of personnel from the operating environment. The operator is universally recognized to be the largest contributor to microbial contamination in conventional aseptic processing. First, the operator carries on/in them a population of microorganisms of greater than 10¹⁴ CFU. Second, these microorganisms must be somehow contained within their gowning materials.  Third, microorganisms from the operator are continuously dispersed into the environment because their gowning materials and methods are not absolute. 

Manned aseptic environments especially those locales where exposed sterile items are handled have been specifically designed to address the microbial contamination threat associated with the operators required presence. The predominant design elements used to control manned environments include:

• Unidirectional (laminar) airflow – to provide a sweeping action and avoid re-circulation of air over the sterile materials.
• A defined air velocity (90 FPM ± 20%) – to avoid potential air turbulence that might disrupt the desired unidirectional flow.
• A large number of air changes – a consequence of the expected air velocity.
• Monitoring of pressure differentials - to assure that the air flows in the direction away from the critical environments where sterile materials are handled.
• Decontamination of the environment – post-batch and periodic sanitization of the non-product contact surfaces of the equipment and cleanroom.

The design features and monitoring practices outlined above are a substantial part of the expected norms when using manned aseptic processing. These design components are all intended to reduce the adverse impact of microbes and particles derived from the operating personnel who are the acknowledged primary contamination source.

However, aseptic isolators were specifically designed to exclude personnel from the environment in which sterile materials are exposed, and it is appropriate to question whether measures intended for use with aseptically gowned personnel are necessary in an environment in which they are not present. The first isolators used in this industry demonstrated superior performance when compared to manned aseptic environments yet they lacked two primary design components commonly associated with those manned environments:

• They employed turbulent airflow delivered through HEPA filter cartridges remote from the isolator chamber (unidirectional flow is used in cleanrooms to mitigate the impact of the personnel).
• Air returns were located in the ceiling of the isolator chambers (floor level returns are used in cleanrooms to prevent re-entrainment of potential contaminants at work height).

The absence of these and other cleanroom design features in these early isolators had no adverse effect on their operational performance.[2] The expected operational advantages of isolators in aseptic processing projected at that time were not contingent on any refinement of the basic designs. The first isolator-based aseptic fill lines installed evidenced performance far exceeding that of any manned cleanroom, yet they did not include any of the accoutrements of manned aseptic filling operations!

The promise of isolation technology was superior aseptic processing performance at a fraction of the operating cost of traditional manned operations.  The simplicity of these early isolator systems also suggested easy fabrication, short lead times, lower facility costs and a comparatively easy qualification / validation.  The future for isolation technology appeared to be near limitless. 

Paradise Lost – Complications ensued and opportunity missed

Regrettably, the expected ‘paradise’ of isolators for aseptic processing was never fully realized.  Despite evidence that comparatively simple isolator designs were capable of outstanding performance aspects of cleanroom design began to appear in 2^nd generation isolators. The wrong-headed notion that an isolator was little more than a small cleanroom requiring all of the accoutrements of cleanroom design.[i] Unidirectional (also called laminar flow) air is a requirement in manned cleanrooms of ISO 5 and better classification that serves to reduce the dispersion of personnel derived contamination into critical locales by minimizing the formation of eddy’s and moving contaminated air to low wall returns. Unidirectional flow patterns are rarely absolute even in the best cleanrooms. Horizontal surfaces of process equipment and the presence of gowned personnel preclude anything truly resembling unidirectional air. The absence of the primary contamination source, the human operator, when using isolation technology largely mitigates the contamination risk without the need for a specific air direction.  Sterility test isolators (which only rarely employ unidirectional air flow) and the 1^st generation isolators demonstrated environmental performance equivalent to that of the more complex isolator designs that include unidirectional flow.  Particle generation from equipment operation and component handling with modern filling and stoppering equipment is a lesser concern and can be readily controlled by means other than airflow direction.

The consequences of this perceptual error are myriad as it had a negative ripple effect on the design of isolator systems: the seemingly simple introduction of unidirectional airflow into isolation technology required substantial physical changes with unfortunate adverse consequences.

• The isolator HVAC system became both larger and more complex to move additional air - increasing both initial and routine operational costs; with fabrication, qualification and validation efforts becoming more extensive as well.
• Limited access for cleaning because of the larger size of the overall system – made extended operation more difficult and increased changeover times between products by extending both decontamination and aeration cycle times.
• Adding unidirectional flow required the use of return air ducts at or near the floor of the room to avoid turbulence at the level of exposed sterile materials. These are difficult to clean locations without opening of the isolator.
• The isolator and its air handling system grew to a size that allowed for final installation only at the operating system eliminating pre-shipment FAT testing, and increased overall facility dimensions and costs.
• Required opening of the isolator at the completion of the batch for cleaning / changeover as portions of the isolator internals were no longer easily accessible. This resulted in increased changeover and cleaning periods and restricted campaign operations.

In parallel with the unidirectional flow designs cited above, maintenance of 90 FPM (0.45 m/s) ±20% air flow velocity at the HEPA filter face was often instituted. The exactness of the expectation belies its arbitrary nature, obscure origin and unknown utility as an environmental control measure. Unidirectional flow is possible at velocities above and below this range.  The consequences of it and its lack of utility of in isolators are the identical to those for unidirectional air. 

One of the main advantages of isolation technology is the ability to decontaminate the interior surfaces by automated means. This practice replaces the manual disinfection procedures that are prevalent in manned cleanrooms and is more effective as it virtually eliminates human error or oversight in execution. Automated systems for decontamination provides for the treatment of surfaces and objects that are not readily accessible. Given the closed design of most early isolators, and the availability of an automated capability, some early practitioners endeavored to ‘sterilize’ rather than decontaminate them. That such a measure was never possible, nor, necessary in manned cleanrooms for successful usage was not considered. The closed design of isolators and the availability of a reliable means for antimicrobial treatment perhaps encouraged this excessive practice. Whether this procedural addition would provide a measurable (or necessary) improvement in environmental control or patient safety was not considered. More treatment was believed to be better than less. This unnecessary raising of the performance bar appears benign, but triggered added complications in both validation process execution and routine operation. 

• In order to accomplish ‘sterilization’, the number of biological indicators placed and the population of each biological indicator were increased.[ii] The increase in biological indicator population had the greatest adverse effect due to positive results largely associated with difficulties in preparation of biological indicators.[3] 
• Due to the increased biological indicator population, there was a commensurate increase in the duration of the decontamination dwell time to destroy them.
• Increases in the exposure period meant that items exposed to the process would have greater exposure to the principal decontaminating agent – H₂O₂.
• Increased exposure of items in the enclosure led to extended aeration times post-exposure due to increased H₂O₂ adsorption by some materials.
• In some instances the operational life of polymeric materials used in isolator construction was shortened due to repeated extended exposure to H₂O_2.
• As the primary purpose of most isolators is separative aseptic operation, their designs were often sub-optimal for decontamination, resulting in lengthy decontamination cycles.

The greatest failing in decontamination was the complete rejection of regulatory and industry recommendations with respect to the expected process objective.  FDA, PIC/S, USP, PDA and others had all issued guidance documents that recommended a lesser treatment using a lower population on the biological indicators.[4]^,[5],[6],[7]

The use of a potent sporicidal compound in the decontamination of isolators and their closed configuration during the process led to concerns relative to the integrity of the system.  The intent of leak testing is to confirm minimal operator exposure to H₂O₂ during the decontamination process. Here too, a seemingly useful consideration has been elevated to extremes. Initially a qualitative test, leak testing quickly became a quantitative metric that was both increasingly complex and overly rigorous. While it is readily acknowledged that aseptic cleanrooms continuously leak air to their surroundings, and aseptically gowned personnel are ‘the’ source of contamination, the idea that an isolator system should leak at all became problematic. The futility of leak testing was perhaps best addressed by Staerk and Sigwarth who evaluated a variety of leak test methods on isolator gloves and showed that the level of detection for all was orders of magnitude larger than the typical microorganism.[8] Nevertheless frequent glove leak testing is a de facto requirement for present day fill isolators. The following adverse consequences have resulted from this largely unnecessary precaution:

• Increased cost of fabrication for the isolator system to eliminate even the smallest of leaks.
• Extended times during initial qualification and routine operation to check for leaks, and remediate them where possible.
• Increased cost for the purchase, calibration and maintenance of glove leak testing equipment.
• Increased downtime between operating runs spent in leak testing gloves on the isolator.

That cleanrooms operate successfully with continual leakage has apparently never been given adequate consideration. More importantly, the need for adherence to proper aseptic technique inside an isolator should always be respected. This measure is sufficient to maintain asepsis in cleanrooms where operator routinely shed significantly more microorganisms than could ever be present in an isolator, and their glove/gown integrity has never been considered absolute.

Paradise Delayed - Having your cake and eating it too!

In the early 1980’s, the author encountered isolators and became a strong proponent of the technology. I believed that the physical separation of personnel from the critical aseptic environment would revolutionize aseptic processing. By removing the major source of viable and non-viable contamination from proximity to sterile materials an unmatched level of performance would be realized. When isolators were still a novelty there were a myriad of design options, and isolator systems were implemented without major difficulty. As the cleanroom relevant concerns were added to isolator designs implementation began to slow. I heard statements such as, “It’s taken XYZ more than 3 years to validate their filling isolator.  What makes you think we can do it at all?”

The over-specification of isolator system designs caused by the unnecessary imposition of clean room concepts resulted in a surprising outcome. A less capable technology was touted as an acceptable substitute. Restricted Access Barrier Systems (RABS) were introduced as the best of both worlds. They would deliver isolator like performance with the simplicity of a cleanroom.  RABS are actually highly evolved cleanroom designs that rely on some isolator like design elements, but eliminates those believed to be particularly challenging such as unidirectional air, automated decontamination, and leak testing. RABS advocates were often employed at firms that had experienced isolator technology implementation difficulties, while others were those without actual isolator experience that were swayed by the isolator ‘war stories’ that were frequently heard. RABS lack a singular description and installations vary in sophistication from those that certainly match isolator performance to less well evolved designs that are little more than gloves installed on a partial barrier.

As a full-time consultant I have visited many different aseptic filling installations. To those that have implemented RABS in the best possible manner I must acknowledge their proficiency. To those that operate less capable RABS systems I must question the technology decision. Without extreme diligence in system design, RABS can be disappointing in reality. I have observed many RABS designs that are only marginally better than the cleanrooms they were intended to displace. That these firms have invested in a technology that is decidedly second place in aseptic capability when done less than perfectly is most disappointing.

Paradise Found - Keep it Simple

An oft quoted adage is the KISS principle or “Keep it simple, stupid”. This is perhaps the best approach to undertake with any aseptic processing. The acknowledged weakness in aseptic processing is the contamination derived from the human operator. The simplest means to prevent adventitious contamination from personnel is separation of the operator from the critical zone.  This was understood more than 50 years ago before the advent of HEPA filters when gloveboxes were used for the manual filling / assembly of sterile products. These systems operated without air filtration, automated decontamination and means for easy transfer of materials across the separative divide. They were successful in spite of operational limitations of today because they removed the major source of contamination from proximity to sterile materials and surfaces. They could not be cleanrooms (something that was yet to be invented) and yet these gloveboxes were ‘best available technology’ for their time.

In considering isolator designs, our industry must carefully weigh forcing cleanroom design elements upon them. The very first aseptic isolators were operationally successful and had more in common with the gloveboxes of 1940 than a contemporary cleanroom. These early isolators may have looked primitive and unsophisticated to today’s industry, but their performance was nothing less than stellar and led to the isolators of today. I continually encounter individuals and firms that cite the ‘isolator problems’ as justification for use of less capable systems. The message to these is to design an isolator system that separates the operator first, and then weigh the addition of cleanroom design features with the understanding that adding features adds complexity, size, cost and time to the project and likely has no impact  on isolator performance.  No regulator has mandated that isolators be designed to cleanroom standards, and the more we devoid ourselves of that miss-direction the easier will be the implementation of what should be the globally acknowledged superior technology of isolation.

REFERENCES