Note: This
document is reference material for investigators and other FDA personnel. The document
does not bind FDA, and does not confer any rights, privileges, benefits, or immunities for
or on any person(s).
I. INTRODUCTION
The Guide to the Inspection of Pharmaceutical Quality Control Laboratories provided
very limited guidance on the matter of inspection of microbiological laboratories. While
that guide addresses many of the issues associated with the chemical aspect of laboratory
analysis of pharmaceuticals, this document will serve as a guide to the inspection of the
microbiology analytical process. As with any laboratory inspection, it is recommended that
an analyst (microbiologist) who is familiar with the tests being inspected participate in
these inspections.
II. MICROBIOLOGICAL TESTING OF NON-STERILE PRODUCTS
For a variety of reasons, we have seen a number of problems associated with the
microbiological contamination of topical drug products, nasal solutions and inhalation
products. The USP Microbiological Attributes Chapter <1111> provides little specific
guidance other than "The significance of microorganisms in non-sterile pharmaceutical
products should be evaluated in terms of the use of the product, the nature of the
product, and the potential hazard to the user." The USP recommends that certain
categories be routinely tested for total counts and specified indicator microbial
contaminants. For example natural plant, animal and some mineral products for Salmonella,
oral liquids for E. Coli, topicals for P. aeruginosa and S. Aureus,
and articles intended for rectal, urethral, or vaginal administration for yeasts and
molds. A number of specific monographs also include definitive microbial limits.
As a general guide for acceptable levels and types of microbiological contamination in
products, Dr. Dunnigan of the Bureau of Medicine of the FDA commented on the health
hazard. In 1970, he said that topical preparations contaminated with gram negative
organisms are a probable moderate to serious health hazard. Through the literature and
through our investigations, it has been shown that a variety of infections have been
traced to the gram negative contamination of topical products. The classical example being
the Pseudomonas cepacia contamination of Povidone Iodine products reported by a
hospital in Massachusetts several years ago.
Therefore, each company is expected to develop microbial specifications for their
non-sterile products. Likewise, the USP Microbial Limits Chapter <61> provides
methodology for selected indicator organisms, but not all objectionable organisms. For
example, it is widely recognized that Pseudomonas cepacia is objectionable
if found in a topical product or nasal solution in high numbers; yet, there are no test
methods provided in the USP that will enable the identification of the presence of this
microorganism.
A relevant example of this problem is the recall of Metaproterenol Sulfate Inhalation
Solution. The USP XXII monograph requires no microbial testing for this product. The
agency classified this as a Class I recall because the product was contaminated with Pseudomonas
gladioli/cepacia. The health hazard evaluation commented that the risk of pulmonary
infection is especially serious and potentially life-threatening to patients with chronic
obstructive airway disease, cystic fibrosis, and immuno-compromised patients.
Additionally, these organisms would not have been identified by testing procedures
delineated in the general Microbial Limits section of the Compendia.
The USP currently provides for retests in the Microbial Limits section <61>
however there is a current proposal to remove the retest provision. As with any other
test, the results of initial test should be reviewed and investigated. Microbiological
contamination is not evenly dispersed throughout a lot or sample of product and finding a
contaminant in one sample and not in another does not discount the findings of the initial
sample results. Retest results should be reviewed and evaluated, and particular emphasis
should be placed on the logic and rationale for conducting the retest.
In order to isolate specific microbial contaminants, FDA laboratories, as well as many
in the industry, employ some type of enrichment media containing inactivators, such as
Tween or lecithin. This is essential to inactivate preservatives usually present in these
types of product and provides a better medium for damaged or slow growing cells. Other
growth parameters include a lower temperature and longer incubation time (at least 5 days)
that provide a better survival condition for damaged or slow-growing cells.
For example, FDA laboratories use the test procedures for cosmetics in the
Bacteriological Analytical Manual (BAM), 6th Edition, to identify contamination in
non-sterile drug products. This testing includes an enrichment of a sample in modified
letheen broth. After incubation, further identification is carried out on Blood Agar
Plates and MacConkey Agar Plates. Isolated colonies are then identified. This procedure
allows FDA microbiologists to optimize the recovery of all potential pathogens and to
quantitate and speciate all recovered organisms. Another important aspect of procedures
used by FDA analysts is to determine growth promotion characteristics for all of the media
used.
The selection of the appropriate neutralizing agents are largely dependent upon the
preservative and formulation of the product under evaluation. If there is growth in the
enrichment broth, transfer to more selective agar media or suitable enrichment agar may be
necessary for subsequent identification.
Microbiological testing may include an identification of colonies found during the
Total Aerobic Plate Count test. Again, the identification should not merely be limited to
the USP indicator organisms.
The importance of identifying all isolates from either or both Total Plate Count
testing and enrichment testing will depend upon the product and its intended use.
Obviously, if an oral solid dosage form such as a tablet is tested, it may be acceptable
to identify isolates when testing shows high levels. However, for other products such as
topicals, inhalants or nasal solutions where there is a major concern for microbiological
contamination, isolates from plate counts, as well as enrichment testing, should be
identified.
III. FACILITIES, EQUIPMENT, AND
MEDIA
Begin the inspection with a review of analyses being conducted and inspect the plates
and tubes of media being incubated (caution should be exercised not to inadvertently
contaminate plates or tubes of media on test). Be particularly alert for retests that have
not been documented and "special projects" in which investigations of
contamination problems have been identified. This can be evaluated by reviewing the
ongoing analyses (product or environmental) for positive test results. Request to review
the previous day's plates and media, if available and compare your observations to the
recorded entries in the logs. Inspect the autoclaves used for the sterilization of media.
Autoclaves may lack the ability to displace steam with sterile filtered air. For sealed
bottles of media, this would not present a problem. However, for non-sealed bottles or
flasks of media, non-sterile air has led to the contamination of media. In addition,
autoclaving less than the required time will also allow media associated contaminants to
grow and cause a false positive result. These problems may be more prevalent in
laboratories with a heavy workload.
Check the temperature of the autoclave since overheating can denature and even char
necessary nutrients. This allows for a less than optimal recovery of already stressed
microorganisms. The obvious problem with potential false positives is the inability to
differentiate between inadvertent medium contamination and true contamination directly
associated with the sample tested.
IV. STERILITY TESTING
On 10/11/91, the Agency published a proposed rule regarding the manufacture of drug
products by aseptic processing and terminal sterilization. A list of contaminated or
potentially contaminated drug products made by aseptic processing and later recalled was
also made available. Many of the investigations/inspections of the recalled products
started with a list of initial sterility test failures. FDA review of the manufacturer's
production, controls, investigations and their inadequacies, coupled with the evidence of
product failure (initial sterility test failure) ultimately led to the action.
The USP points out that the facilities used to conduct sterility tests should be
similar to those used for manufacturing product. The USP states, "The facility for
sterility testing should be such as to offer no greater a microbial challenge to the
articles being tested than that of an aseptic processing production facility". Proper
design would, therefore, include a gowning area and pass-through airlock. Environmental
monitoring and gowning should be equivalent to that used for manufacturing product.
Since a number of product and media manipulations are involved in conducting a
sterility test, it is recommended that the inspection include actual observation of the
sterility test even though some companies have tried to discourage inspection on the
grounds that it may make the firm's analyst nervous. The inspection team is expected to be
sensitive to this concern and make the observations in a manner that will create the least
amount of disruption in the normal operating environment. Nevertheless, such concerns are
not sufficient cause for you to suspend this portion of the inspection.
One of the most important aspects of the inspection of a sterility analytical program
is to review records of initial positive sterility test results. Request lists of test
failures to facilitate review of production and control records and investigation reports.
Particularly, for the high risk aseptically filled product, initial positive sterility
test results and investigations should be reviewed. It is difficult for the manufacturer
to justify the release of a product filled aseptically that fails an initial sterility
test without identifying specific problems associated with the controls used for the
sterility test.
Examine the use of negative controls. They are particularly important to a high quality
sterility test. Good practice for such testing includes the use of known terminally
sterilized or irradiated samples as a system control. Alternatively, vials or ampules
filled during media fills have also been used.
Be especially concerned about the case where a manufacturer of aseptically filled
products has never found an initial positive sterility test. While such situations may
occur, they are rare. In one case, a manufacturer's records showed that they had never
found a positive result; their records had been falsified. Also, the absence of initial
positives may indicate that the test has not been validated to demonstrate that there is
no carryover of inhibition from the product or preservative.
Inspect robotic systems or isolation technology, such as La Calhene units used for
sterility testing. These units allow product withdrawal in the absence of people. If an
initial test failure is noted in a sample tested in such a system, it could be very
difficult to justify release based on a retest, particularly if test controls are
negative.
Evaluate the time period used for sterility test sample incubation. This issue has been
recently clarified. The USP states that samples are to be incubated for at least 7 days,
and a proposal has been made to change the USP to require a period of 14 days incubation.
You are expected to evaluate the specific analytical procedure and the product for the
proper incubation period. Seven days may be insufficient, particularly when slow growing
organisms have been identified. Media fill, environmental, sterility test results and
other data should be reviewed to assure the absence of slow growing organisms. Also, you
should compare the methods being used for incubation to determine if they conform to those
listed in approved or pending applications.
V. METHODOLOGY AND
VALIDATION OF TEST
PROCEDURES
Determine the source of test procedures. Manufacturers derive test procedures from
several sources, including the USP, BAM and other microbiological references. It would be
virtually impossible to completely validate test procedures for every organism that may be
objectionable. However, it is a good practice to assure that inhibitory substances in
samples are neutralized.
During inspections, including pre-approval inspections, evaluate the methodology for
microbiological testing. For example, we expect test methods to identify the presence of
organisms such as Pseudomonas cepacia or other Pseudomonas species
that may be objectional or present a hazard to the user. Where pre-approval inspections
are being conducted, compare the method being used against the one submitted in the
application. Also verify that the laboratory has the equipment necessary to perform the
tests and that the equipment was available and in good operating condition on the dates of
critical testing.
The USP states that an alternate method may be substituted for compendial tests,
provided it has been properly validated as giving equivalent or better results.
You may find that dehydrated media are being used for the preparation of media. Good
practice includes the periodic challenge of prepared media with low levels of organisms.
This includes USP indicator organisms as well as normal flora. The capability of the media
to promote the growth of organisms may be affected by the media preparation process,
sterilization (overheating) and storage. These represent important considerations in any
inspection and in the good management of a microbiology laboratory.
VI. DATA STORAGE
Evaluate the test results that have been entered in either logbooks or on loose
analytical sheets. While some manufacturers may be reluctant to provide tabulations,
summaries, or printouts of microbiological test results, this data should be reviewed for
the identification of potential microbial problems in processing. When summaries of this
data are not available the inspection team is expected to review enough data to construct
their own summary of the laboratory test results and quality control program.
Some laboratories utilize preprinted forms only for recording test data. Some
laboratories have also pointed out that the only way microbiological test data could be
reviewed during inspections would be to review individual batch records. However, in most
cases, preprinted forms are in multiple copies with a second or third copy in a central
file. Some companies use log-books for recording data. These logbooks should also be
reviewed.
Additionally, many manufacturers are equipped with an automated microbial system for
the identification of microorganisms. Logs of such testing, along with the identification
of the source of the sample, are also of value in the identification of potential
microbial problems in processing.
The utilization of automated systems for the identification of microorganisms is
relatively common in the parenteral manufacturer where isolates from the environment,
water systems, validation and people are routinely identified.
Microbiologists in our Baltimore District are expert on the use of automated microbic
analytical systems. They were the first FDA laboratory to use such equipment and have
considerable experience in validating methods for these pieces of equipment. Contact the
Baltimore District laboratory for information or questions about these systems. Plants
with heavy utilization of these pieces of equipment should be inspected by individuals
from the Baltimore District laboratory.
VII. MANAGEMENT REVIEW
Microbiological test results represent one of the more difficult areas for the
evaluation and interpretation of data. These evaluations require extensive training and
experience in microbiology. Understanding the methodology, and more importantly,
understanding the limitations of the test present the more difficult issues. For example,
a manufacturer found high counts of Enterobacter cloacae in their oral
dosage form product derived from a natural substance. Since they did not isolate E.
coli, they released the product. FDA analysis found E. cloacae in
most samples from the batch and even E. coli in one sample. In this case
management failed to recognize that microbiological contamination might not be uniform,
that other organisms may mask the presence of certain organisms when identification
procedures are performed, and that microbiological testing is far from absolute. The
inspection must consider the relationship between the organisms found in the samples and
the potential for the existence of other objectionable conditions. For example, it is
logical to assume that if the process would allow E. cloacae to be present,
it could also allow the presence of the objectionable indicator organism. The
microbiologist should evaluate this potential by considering such factors as methodology,
and the growth conditions of the sample as well as other fundamental factors associated
with microbiological analysis.
Evaluate management's program to audit the quality of the laboratory work performed by
outside contractors.
VIII. CONTRACT TESTING
LABORATORIES
Many manufacturers contract with private or independent testing laboratories to analyze
their products. Since, these laboratories will conduct only the tests that the
manufacturer requests, determine the specific instructions given to the contractor.
Evaluate these instructions to assure that necessary testing will be completed. For
example, in a recent inspection of a topical manufacturer, total plate count and testing
for the USP indicator organisms were requested. The control laboratory performed this
testing only and did not look for other organisms that would be objectionable based on the
product's intended use.
Analytical results, particularly for those articles in which additional or retesting is
conducted, should be reviewed. Test reports should be provided to the manufacturer for
tests conducted. It is not unusual to see contract laboratories fail to provide complete
results, with both failing as well as passing results.
Bacteriostasis/fungiostasis testing must be performed either by the contract lab or the
manufacturer. These test results must be negative otherwise any sterility test results
obtained by the contractor on the product may not be valid. |