There is a ‘c’ in ‘cGMP’ for a reason. Technology is always advancing; new methods are constantly being developed to improve quality, reliability, safety and reduce risk in pharmaceutical processing.
Some years ago Getinge La Calhène (through its then subsidiary Linac Technologies), introduced the SterStar, a system that uses an array of electron beams to biodecontaminate the external surfaces of tubs containing sterile ready-to-fill/ready-to-use (RTU) syringes as they pass into a barrier isolated filling line.
It was a revolution at the time, designed to feed high-speed production lines operating at up to 6 tubs/minute, or about 600 syringes/min.
The tubs pass through a ‘curtain’ of e-beam radiation as they feed continuously into the filling line, treating all the external surfaces of the tub uniformly and assuring a 6-log sporicidal reduction.
This had to be achieved without the e-beam penetrating the tub or beyond the first layer of the Tyvek lid, as to do so would negatively affect the glass containers inside.
A key identified challenge (and validation test) at the time was to assure this same 6-log reduction under the edge of the Tyvek lid, the critical area where the Tyvek is adhered to the tub flange.
When the lid is subsequently removed before filling, this flap is immediately adjacent to, and passes over the open necks of the sterile containers within the tub – hence the focus on this area.
E-beam has become an accepted solution for high speed pharmaceutical filling lines.
Getinge has installed more than 20 of these systems since their introduction in 2002, and several other industry vendors have introduced competitive e-beam systems following a similar concept.
An e-beam system is a complex piece of equipment. Typically, three electron beam accelerators are arranged in a triangular pattern to create the ‘curtain’ through which the tub passes on a conveyor system.
When e-beams hit metal, X-rays are produced, so the whole system must be shielded with lead, and the conveyor must be arranged in a labyrinth for operator safety.
Furthermore, during the process, some of the oxygen (O2) in the air is ionised and transforms to ozone (O3), which must be removed from the system by ventilation.
Plus, the area downstream of the e-beam must be maintained as a Grade A zone and must be bio-decontaminated (as with the downstream isolator).
The systems are large (greater than 1m x 2m x 3m), heavy (3 tonnes plus) and are a substantial investment. They have a large footprint and are relatively utility intensive.
Although designed for a maximum throughput of 6 tubs/min, the physical dimensions, equipment requirements and investment would not be significantly different for a lower speed system: the e-beam accelerators (and all supporting engineering) are dimensioned for the tub surface area and geometry, not the speed of the conveyor.
The footprint and investment cost of an e-beam system are proportionate when considering a high capacity/high speed production line (400–600 pcs/min). However, it is difficult to justify such a system for lower volume production, particularly when considering 100–200 pcs/min or 1–2 tubs/min.
Current solutions for lower speed systems (<4 tubs/min)
Automated chemical processes have been used for biodecontamination of tubs as they enter an isolated filling line.
Vaporised hydrogen peroxide (HPV), which is commonly used for isolator surface biodecontamination, has been used in lock-chambers for the transfer of RTU tubs; however, it has been demonstrated that HPV is not effective under the critical Tyvek flap.
Also, as the process is relatively slow, large lock-chambers are required to meet line-speed requirements. Additionally, HPV is a strong oxidising agent that diffuses through the Tyvek and may condense, leaving residues on the primary packaging containers.
Oxidising agents are known to be detrimental to pharma products, particularly large molecule biologics.
Low temperature and low pressure gas plasma has also been used, but this is a slow process and the manufacturer claims only a 4-log sporicidal reduction, and not the 6-log provided by e-beam.
An alternative to an automated system is to use combined manual handling and chemical decontamination followed by automated ‘no-touch’ transfer:
- Manual removal of outer bag (RTU tubs are available double-bagged)
- Manual decontamination (spray-wipe) of external surfaces of inner bag as it is manually passed into a Grade A unidirectional airflow system with automated transfer system
- Automatic removal of the inner bag and mechanical transfer into the filling line isolator (no manual operator intervention)
Questions and risks with such a process would include:
- The RTU container vendor guarantees the sterility of the primary packaging containers within the tub. Is there a similar guarantee (validation data) of the sterility of the outer surfaces of the tub within the bags?
- Is the integrity of the bag(s) guaranteed? For what shelf life?
- How reliable and repeatable is a manual decontamination process (e.g. alcohol spray/wipe)?
A manual process inherently carries greater risk than an automated process. While initial capital cost may be lower, the ongoing cost of the manual operation (personnel) and increased revalidation and environmental monitoring costs would need to be considered.
The future
If we embrace the ‘c’ and have demonstrated a low(er) risk solution for introduction of RTU primary containers into high speed filling lines, then logically we should strive to have the same acceptance criteria for lower speed lines.
What is the rationale for accepting higher risk in the same application? This is a particularly relevant question considering that high-speed lines constitute only a small fraction of new filling installations.
Over the past year Noxilizer, Baltimore, MD, US, has worked with Getinge La Calhène to develop a solution for low speed filling lines that matches the performance of e-beam systems: 6-log sporicidal reduction on all external surfaces, including under the critical Tyvek lid flap.
The lock-chamber is compact and self-contained with a footprint of approximately 1.2sq m
Noxilizer bio-decontamination technology is based on the use of nitrogen dioxide (NO2) as a sterilant. Its effectiveness is proven: it is used for the terminal sterilisation of medical devices (the FDA has cleared a medical device using NO2 sterilisation). Data to support the sterilant efficacy claims are available.
As a gas and not a condensing vapour, NO2 readily diffuses uniformly and is easily removed by dilution aeration.
White Papers and other publications are available comparing NO2 and HPV, but the primary conclusions with regard to biodecontamination processes may be summarised as follows:
- Biodecontamination total process time (TPT) with NO2 is less than 50% of the TPT with HPV
- 6-log sporicidal reduction (with G. stearothermophilus as indicator organism)
- Aeration to 1ppm
- The majority time saving is in aeration
- Rapid aeration to ppb levels (just a few minutes longer than to 1ppm)
- Important when oxidation-sensitive biologics are to be filled
- ‘On-demand’ generation of sterilant
- Noxilizer generates NO2 at point of use from air using electrical energy
- No manual handling, storage and transportation of hazardous chemicals
- NO2 more readily penetrates into cavities/complex geometries than HPV
- Less disassembly of equipment prior to biodecontamination
To demonstrate 6-log sporicidal reduction under the tub flap (the worst case challenge), Noxilizer directly inoculated the seam area with liquid spore suspension and allowed it to dry, effectively creating a representative biological indicator or ‘tub BI’ with 106 spores/tub.
These BIs were then exposed to the biodecontamination process and recovered. All BIs were inactivated.
The TPT for this process was 15 min (1 min humidification, 8 min dose/dwell and 6 min aeration to 1ppm).
For reference, the same test was performed using a standard HPV process. The BIs were not inactivated. Furthermore, the TPT (at 43 min) would not match tub throughput requirement of even a low speed filling line without a large and ergonomically unmanageable lock-chamber.
Both NO2 and HPV will readily diffuse through Tyvek film, and will come into contact with the primary containers.
As a gas (and unlike HPV), NO2 will not condense on the containers and will be readily removed in the aeration process.
It is also worth noting that NO2 is a constituent of normal room air at levels of around 50ppb, so the primary containers are already continuously exposed to NO2.
The process has subsequently been integrated with a Getinge La Calhène lock-chamber and was introduced to the industry at the recent Interphex trade show in New York.
The design parameters of the lock-chamber are to bio-decontaminate 20 tubs in 20 min, an average of 1 tub/min. This allows 15 min for the process and 5 min for load and unload. The operating concept is:
- Manually load 20 tubs on a rack
- Insert the rack into the chamber
- Perform automatic biodecontamination
- Extract the rack to a delidding isolator (Grade A)
- Offload the tubs
- Return the rack to the chamber for the next process to start
- Remove the Tyvek lid within the delidding isolator and pass via mousehole to the filling line
In conclusion, the innovation meets the criteria set out in the challenge. It is relatively compact, uses minimal utilities, and has an investment cost that is proportionate for a low-speed filling line.
It fulfils the performance requirements to feed low speed lines, with 6-log sporicidal reduction of all surfaces in an automated, easily validated process, with very low, non-oxidising residuals (close to environmental levels) at a proportionate investment cost.