Periodic Media Fills in Routine Operation

It is unlikely that any responsible regulatory body would tolerate a frequency of less than twice a year for periodic media fills. Media fills are probably the most sensitive method of detecting unexpected sources of process contamination. The regulatory standpoint coming from the principle of patient protection is that if unexpected process contamination occurs in a media fill, and is considered sufficient to compromise the sterility of past product, they would expect market withdrawal. Following this logic, the greater the frequency of periodic media fill, the lower the risk to the patient, and the lower the commercial risk to the manufacturer.

Media fills are generally done on every filling line at least twice a year. (88.5% of the respondents to the PDA's 1996 survey performed media fills at least twice a year.) Within this program it is sensible to ensure that on multicontainer filling lines every container size has been filled at least once in a reasonable time frame, say, over two years. Otherwise the possibility of unexpected contamination as it relates to a particular size may never be addressed.

It is also arguable that at least one of those sizes identified in the validation protocols as presenting the worst risks of contamination should be tested on every occasion of periodic media fills. This is usually achieved by setting up some sort of matrix approach to periodic media fills on multicontainer filling lines. Of course it is much easier on a single-container size, single-volume filling line.

Periodic media fills should be done at the end of a routine production operation. Care should be taken to run a few litres of sterile water through the filling setup, to flush out any product-related inhibitory substances, before filling the placebo. This is intended to address the two possibilities of contamination buildup in a filling room over a period of manned operation between clean-ups, and of lapse of operator discipline as a result of tiredness.

The exception to this is for aseptic filling of sterile antibiotics. Here the filling room must be cleaned up and all antibiotic traces removed before the placebo is brought in and filled. This ensures that recovery of contaminants is not inhibited. This reinforces the argument concerning the purpose of media fills as related to process contamination, rather than to product sterility. The media fill is intended to disclose process contamination, regardless of whether the contaminants would survive or die in the product.

ISO/IS 13408 (1997) recommends a two-tier approach of alert and action to limits for periodic media fills. It does not, however, elaborate on how the two levels should be applied. The action limits are listed under ISO/IS 13408 in Table 3.4. The alert limits are lower.

The divergence between the ISO/IS 13408 action limits and acceptable reality has already been discussed in relation to validation media fills. Reality for a valid aseptic process that has been transferred to routine control is that there will have been no more than one contaminated unit per media fill run, irrespective of the total number of units filled.

It is axiomatic that the periodic media fill should not generate significantly worse results than the validation media fill without some appropriate action being taken. Here it is suggested that limits for periodic media fills should be related to the results obtained in validation media fills as summarised in Table 3.5.

Table 3.5 Recommended Action Limits for Periodic Media Fills Irrespective of Total Numbers of Units Filled and Related to Results from Validation Media Fills

Numbers of contaminated units actually occurring in three successive validation media fills

Action limit (numbers of contaminated units) for marginal failures in periodic media fills

Action limit (numbers of contaminated units) for consequential failures in periodic media fills

0, 0, 0

> 1 but < 3

> 3

0, 0, 1

> 1 but < 3

> 3

0, 1, 1

> 2 but < 4

> 4

1, 1, 1

> 2 but < 4

> 4

This may appear a little radical but it is no more than common sense, and reflects what is done in many other less critical industries, than those manufacturing sterile pharmaceutical products. Fundamentally, this suggestion proposes that a new aseptic process should be developed such that there is minimal evidence of contamination and this is surely what is being done already. Then the response limits for routine media fills should be based on the process capability demonstrated in validation.

The numbers of permissible contaminated units shown in Table 3.5 are presented in a two-tier approach for which both levels demand action. The difference in the approach is in the consequences of the actions to production, to scheduling and to past product.

The term "action limits for marginal failures" is used here rather than "alert" limit, because the author's opinion is that all contaminants found in media fills merit some action, and that the use of the "alert" term detracts from this.

The action limits for marginal failures allow for the extremes of statistical variation from the validation media fill results that might be expected, without any significant change in the real contamination rate. In the author's experience, media fill contaminants are fairly rarely found in well-controlled facilities, but when they occur, most fall into this marginal category.

Identification and investigation are essential. The possibility that they may not be a statistical phenomenon should not be discounted. Bacillus spp. should, for instance, be treated with extreme suspicion in relation to the possibility of some systematic problem with nonsporicidal disinfection, or of residual air in autoclave loads, etc.

Actions from marginal failures, which do not appear to have arisen from a systematic failure of one of the systems necessary for the maintenance of asepsis, are best dealt with by counselling, retraining, and improved supervision of operators. The media fill should be repeated as soon as possible. A further media fill on the container size implicated should be scheduled into the next periodic media fill, in addition to those sizes defined by the predetermined matrix.

Successive marginal failures on the same container size should be treated as a consequential failure, as also should marginal failures on three or more successive media fills on the same filling line, irrespective of container size. Other circumstances of repeated failures within the marginal range may also be indicative of process conditions that have deteriorated from the validated condition, and should be treated as infringements of the action limits.

Table 3.5 gives action limits that are described as consequential. These limits are well beyond the expected variation seen in the validation media fills and must therefore be interpreted as indicators of real loss, or genuinely deteriorating control levels. It is reasonable to expect that the potential for any patient risk should be minimized while these failures are being resolved. Product manufactured on the filling line after the date of the media fill, and product still in the company's warehouses, should be quarantined until the failure investigation is completed.

Ideally production on the line in question should be suspended pending the outcome of the investigation. In practice it may be advantageous to the investigation for production to continue, but this decision should not be taken lightly in view of the commercial risk of possibly having to reject the product made in that period.

The most important factor in the failure investigation is the identification of the contaminants. Any microbiologist should be able to categorise identified contaminants within their most likely sources to the environment (air, dust, etc.), water, or human sources. An experienced QA microbiologist may be able to pinpoint the contaminants to their origins in the facility (e.g., nonsterile disinfectants, water leakage, worn-out garments, etc.) or to general weaknesses in control.

Environmental microbiology is not an exact science. A weakness in control will always be a systematic weakness even though it may not manifest itself every time it is tested by media fill. If a specific problem is diagnosed it should be traced, if possible, to the time it began.

The identified contaminants should be considered for their ability to survive in the products filled on the line in which the consequential media fill occurred. The importance of this information is to determine if sterility has been compromised. In a multiproduct filling line the decision might be different for different products. If sterility is compromised then product must be withdrawn from the market.

Once the failure investigation is complete and corrective or preventive action implemented, it is customary to repeat the media fill. Some would argue for revalidation of the line by repeated media fill, but this decision should be contingent upon the extent of the corrective or preventive action implemented.

In some instances of consequential failure it may not be possible to pinpoint the cause and corrective or preventive action cannot therefore be targeted. In such cases it is normal to clean, disinfect, fumigate, counsel, train and improve supervision overall. Three repeat media fills should be done to counterbalance the uncertainty of the diagnosis.

Where repeat media fills have to be done, the date of recommencement of production is a business that requires account to be taken of the uncertainty of the diagnosis of the cause of the problem. Where the source of media fill failure is quite clear, and preventive action self-evident, it is probably a reasonable risk to recommence production before the media fill incubation is complete. The commercial risk is greater where diagnosis of the problem is unclear.

The major practical issue of periodic media fills is how to respond to the results. This is less problematic (in principle, if not always in practice where deadlines have to be met) in validation than in periodic media fills. The major issues are:

• Should production on a particular filling line be allowed to continue if media fill results are unfavourable?

• Media fill results are not available until 14 days after the media fill has been conducted. What should be done with the product manufactured between these dates when results are unfavourable?

• Media fills are only done every six months. What should be done with the product manufactured since the last successful media fill when results are unfavorable?

If a company has the luxury of running terminally sterilized products on the same filling line as aseptically filled products, there is the opportunity to run the line and investigate a media fill failure while fully operational. If only aseptically filled products are filled on the line, filling should be suspended until investigations are complete and repeat media fills are satisfactory. It is exceedingly difficult to reach a firm conclusion unless the line is running. Although running a series of repeat media fills is essential, the hope is usually that they will pass rather than providing information.

It is normally recommended to "freeze" all of the product still in company control aseptically filled on a failed media fill line until investigations are complete and repeat media fills have given the line the "go ahead." This strategy, although fine in principle, usually raises significant pressure from marketing and distribution over stock-outs or impending stock-outs.

If a media fill fails and this is traceable to a failure in one of the component systems making up the sterility assurance system, there is little choice but to reject and recall back to that date, unless the regulatory bodies can be convinced otherwise. A responsible recall initiative from a company is generally less harmful than a recall requested by an inspector who discovers the matter later on. An example of this could be a tear in a HEPA filter — recall back to the last satisfactory in situ integrity test.

The outcome of the investigation of most marginal media fill failures is inconclusive, often from some human commensal microorganism shed by an operator, not necessarily on point-of-fill, possibly even when unloading stoppers from an autoclave. It would not be sensible to recall for this type of phenomenon, and in mitigation there could be some work done on the potential for the particular microorganism to survive and grow in specific products filled on that line.


Bernuzzi, M., Halls, N.A., Raggi, P. Application of statistical models to action limits for media fill trials. European Journal of Parenteral Sciences, 2: 3-11, 1997.

Commission of the European Communities (CEC). The Rules Governing Medicinal Products in the European Community. Volume IV Good Manufacturing Practice for Medicinal Products. Luxembourg: Office for Official Publications of the European Community, 1992, 1997, 2002.

Department of Health and Social Security. Guide to Good Pharmaceutical Manufacturing Practice (the "Orange Guide"). London: Her Majesty's Stationery Office, 1983.

Food and Drug Administration of the United States Department of Health and Human Services (FDA). Guideline on Sterile Drug Products Produced by Aseptic Processing. Rockville, MD: Center for Drugs and Biologics, 1987.

Food and Drug Administration of the United States Department of Health and Human Services (FDA). Guideline to Industry for the Submission Documentation for Sterilization Process Validation in Applications for Human and Veterinary Drug Products. Rockville, MD: Center for Drug Evaluation Research and Center for Veterinary Medicine, 1994..

Halls, A. Achieving Sterility in Medical and Pharmaceutical Products. New York: Marcel Dekker, 1994.

International Organization for Standardization (ISO). ISO 13408-1 Aseptic Processing of Health Care Products — Part 1: General Requirements, 1998.

Medicines Control Agency (MCA). Rules and Guidance for Pharmaceutical Manufacturers and Distributors, 1997. London: The Stationery Office, 1997.

Parenteral Drug Association (PDA). Validation of Aseptic Filling for Solution Drug Products, Technical Monograph No 2. Bethesda, MD: Parenteral Drug Association Inc., 1981.

Parenteral Drug Association (PDA). Technical Report No. 24. Current practices in the validation of aseptic processing. PDA Journal of Pharmaceutical Science and Technology, 51: Supplement S2, 1996.

Tallentire, A., Dwyer, J., Ley, F.J. Microbiological quality control of sterilized products: evaluation of a model relating frequency of contaminated items with increasing radiation treatment. Journal of Applied Bacteriology, 34: 521-534, 1971.

Was this article helpful?

0 0
Herbal Remedies For Acid Reflux

Herbal Remedies For Acid Reflux

Gastroesophageal reflux disease is the medical term for what we know as acid reflux. Acid reflux occurs when the stomach releases its liquid back into the esophagus, causing inflammation and damage to the esophageal lining. The regurgitated acid most often consists of a few compoundsbr acid, bile, and pepsin.

Get My Free Ebook

Post a comment