Bulk Sterile Vessel
Bulk Non-Sterile Vessel
, r Filled Units
Bulk Sterile Vessel
Bulk Non-Sterite Vessel
Figure 3.2. Simplified representation of aseptic filling of a liquid dosage form in ampoules.
At first glance it seems that aseptic filling of liquids is a less complicated example than solid dosage forms, as there is no need for an extra filling stage — TSB is both placebo and recovery medium. Indeed, all of the provisions to the conduct of media fills also apply.
• Inert gas sparging should be replaced with compressed air.
• The volume of medium filled may be reduced. In a 1996 Parenteral Drug Association (PDA) survey of aseptic manufacturers, some 34% respondents did not fill the same volume of media as they filled of product in routine production (PDA, 1996).
• All contaminating events must be simulated. In modern, high-speed, tunnel-linked ampoule filling lines, this often results in as many as 10,000 or 20,000 units filled just to give enough time to simulate everything.
• The duration of the media fill has to be long enough to simulate everything needed but not so long as to create problems with incubator space.
A divide has arisen among sterile liquid manufacturers. Traditionally, aqueous-liquid media fills were done by taking vessels of autoclave presterilized TSB into the filling room, connecting them one by one to the filling line, and then filling the ampoules or vials, etc. Some manufacturers, however, interpret regulatory pressure to "simulate the whole process" to mean that they must take dehydrated culture medium as their starting point, make it up in their manufacturing areas, pass it through the process sterilising filters and then connect to the filling room and fill ampoules or vials. Both approaches have some advantages and disadvantages.
The origins of the "traditional approach" lie in older, slow-speed technology, when a regulatory-satisfactory media fill could often be achieved by filling as few as 1000 units. There would be sufficient laboratory autoclave capacity to sterilize sufficient media in aspirators, or large vessels that could then be brought to the filling machine.
One aseptic connection would have to be made between the media vessel and the filler. However, in routine operation, there would most likely be other additional aseptic connections, e.g., between the downstream side of the sterilizing filter and the sterile holding vessel. Very few older processes have the SIP systems addressing the whole line — from filters to filling needles — now developed for newer processes. Inevitably, aseptic connections would also be required and probably not be simulated by the traditional approach to the media fill. Conscientious manufacturers might simulate these aseptic connections separately to the media fill; others would ignore them.
With the advent of high-speed filling lines and the need for larger numbers of filled units, laboratory autoclave capacity often became a limiting factor in complying with regulatory requirements on numbers of units filled. The question was inevitably asked as to why autoclave sterilization was necessary for media, when there was a perfectly good sterilizing process (filtration) used for the product. So filtration through the production filtration setup came a fairly commonplace practice.
1. By taking dehydrated media through all the stages of dispensing and compounding in production vessels, every potential for contamination of the media is taken into account. There is some confusing logic in this contention.
• Dehydrated microbiological media is most usually heavily contaminated with microorganisms reaching levels of around 104 colony-forming units (cfu)/g. Raw materials for aseptic manufacture are invariably specified to be within standards of contamination of no more than 103 cfu/g and rarely ever approach those limits. Compounding areas must be restricted and microbiologically controlled — they are a medium-level clean room. Operators in compounding areas must wear dedicated footwear, clean overalls, head covers and gloves. At least twice a year, in the name of QA and regulatory requirements, the notion of bringing nonsterile, dehydrated microbiological media through these areas makes a mockery of the other enforced controls. If simulation of the filtration process is thought to be valuable to the media fill, it is sensible to have the dehydrated media sterilized by gamma radiation, or for prepared media to have been autoclaved before it is brought into the compounding areas.
• The media fill is intended to detect weaknesses in aseptic processing. Compounding is intended to be sufficiently clean to prevent increases in contaminants or of their byproducts (e.g., endotoxins) resulting from conditions in the manufacturer's premises, but it is not an aseptic process. The media fill should not be seen as an instrument for detection of problems in nonaseptic manufacture; there are simpler and more straightforward methods to achieve that end.
2. The media follows exactly the same route as the product and is therefore an exact simulation of the process, including the risks associated with sterile filtration. Indeed there is some contention that the media fill validates sterile filtration — it does not.
There is a totally independent regulatory requirement for sterile filtration to be validated by a bacterial challenge test that is specified in detail and relates to the way filters, particular microorganisms in particular concentrations, and specific products interact. Sterilizing filters are not intended to retain microorganisms at particularly high challenge levels and at the viscosity of microbiological media. The newer approach to simulating the challenges to the product is probably fairer than the traditional approach, but its limitations must be recognized.
Was this article helpful?