© World Health Organization

WHO Technical Report Series, No. 961, 2011

Annex 6
WHO good manufacturing practices for sterile
pharmaceutical products

Introduction
Following implementation of these WHO good manufacturing practices (GMP)
guidelines (1) within the context of the WHO Prequalification of Medicines
Programme, clarifying, editorial modifications have been proposed. These
changes were adopted for maintenance purposes. In order to ease reading the
full guideline has been reproduced again as an Annex to the current report of
the WHO Expert Committee on Specifications for Pharmaceutical Preparations.

WHO good manufacturing practices

for sterile pharmaceutical products
1. General considerations

2. Quality control

3. Sanitation

4. Manufacture of sterile preparations

5. Sterilization

6. Terminal sterilization

7. Aseptic processing and sterilization by filtration

8. Isolator technology

9. Blow/fill/seal technology

10. Personnel

11. Premises

12. Equipment

13. Finishing of sterile products

References

Further reading

261
1. General considerations
1.1 The production of sterile preparations should be carried out in clean
areas, entry to which should be through airlocks for personnel and/or for
equipment and materials. Clean areas should be maintained to an appropriate
standard of cleanliness and supplied with air that has passed through filters
of the required efficiency.
1.2 The various operations of component preparation (such as those
involving containers and closures), product preparation, filling and
sterilization should be carried out in separate areas within the clean area.
These areas are classified into four grades (see section 4).
1.3 Manufacturing operations are divided here into two categories:
— first, those where the product is terminally sterilized; and
— second, those which are conducted aseptically at some or all stages.

2. Quality control
2.1 The sterility test applied to the finished product should only be
regarded as the last in a series of control measures by which sterility is
assured. The test should be validated for the product(s) concerned.
2.2 Samples taken for sterility testing should be representative of the whole
of the batch but should, in particular, include samples taken from parts of
the batch considered to be most at risk of contamination, for example:
• for products that have been filled aseptically, samples should include
containers filled at the beginning and end of the batch and after any
significant interruption of work;
• for products that have been heat sterilized in their final containers,
consideration should be given to taking samples from that part of the
load that is potentially the coolest.
2.3 The sterility of the finished product is assured by validation of
the sterilization cycle in the case of terminally sterilized products, and
by “media simulation” or “media fill” runs for aseptically processed
products. Batch-processing records and, in the case of aseptic processing,
environmental quality records, should be examined in conjunction with
the results of the sterility tests. The sterility test procedure should be
validated for a given product. Pharmacopoeial methods should be used for
the validation and performance of the sterility test. In those cases where
parametric release has been authorized in place of sterility testing special
attention should be paid to the validation and the monitoring of the entire
manufacturing process.

262
 2.4 For injectable products the water for injection and the intermediate,
if appropriate, and finished products should be monitored for endotoxins,
using an established pharmacopoeial method that has been validated for
each type of product. For large-volume infusion solutions, such monitoring
of water or intermediates should always be done, in addition to any tests
required by an approved monograph for the finished product. When a sample
fails a test, the cause of the failure should be investigated and necessary
action should be taken. Alternative methods to those in the pharmacopoeias
may be used if they are validated, justified and authorized.
2.5 The use of rapid microbiological methods to replace the traditional
microbiological methods, and to obtain earlier results on the microbiological
quality of, for example, water, the environment or bioburden, could be
considered if appropriately validated and if a comparative assessment of the
proposed rapid method is performed against the pharmacopoeial method.

3. Sanitation
3.1 The sanitation of clean areas is particularly important. They should
be cleaned frequently and thoroughly in accordance with an approved
written programme. Where disinfectants are used, more than one type
should be employed. Monitoring should be regularly undertaken to detect
contamination or the presence of an organism against which the cleaning
procedure is ineffective. Interactions between different cleaning materials
should be validated. Appropriate cleaning validation should be carried out
to ensure disinfectant residuals can be detected and are removed by the
cleaning process.
3.2 Disinfectants and detergents should be monitored for microbial
contamination; dilutions should be kept in previously cleaned containers
and should only be stored for defined periods unless sterilized. Disinfectants
and detergents used in Grade A and B areas should be sterile before use.
3.3 A disinfectant programme should also include a sporicidal agent since
many common disinfectants are ineffective against spores. The effectiveness
of cleaning and disinfectant procedures should be demonstrated.
3.4 Fumigation of clean areas may be useful for reducing microbial
contamination in inaccessible places.

4. Manufacture of sterile preparations

4.1 Clean areas for the manufacture of sterile products are classified
according to the required characteristics of the environment. Each
manufacturing operation requires an appropriate level of environmental

263
 cleanliness in the operational state to minimize the risks of particulate or
microbial contamination of the product or materials being handled.
4.2 Detailed information on methods for determining the microbiological
and particulate cleanliness of air, surfaces, etc., is not given in these
guidelines.
ISO 14644-1 (2) should be used for classification of cleanliness according
to concentration of airborne particles (determination of number of sample
locations, calculation of sample size and evaluation of classification from
the data obtained). Table 1 should also be used to define the levels to be
used as the basis for monitoring clean areas for airborne particles.
4.3 For the manufacture of sterile pharmaceutical preparations, four
grades of clean areas are distinguished as follows:
• Grade A: The local zone for high-risk operations, e.g. filling and making
aseptic connections. Normally such conditions are achieved by using a
unidirectional airflow workstation. Unidirectional airflow systems should
provide a homogeneous air speed of 0.36–0.54 m/s (guidance value) at a
defined test position 15–30 cm below the terminal filter or air distributor
system. The velocity at working level should not be less than 0.36 m/s.
The uniformity and effectiveness of the unidirectional airflow should be
demonstrated by undertaking airflow visualization tests.
• Grade B: In aseptic preparation and filling, this is the background
environment for the Grade A zone.
• Grades C and D: Clean areas for carrying out less critical stages in the
manufacture of sterile products or carrying out activities during which
the product is not directly exposed (i.e. aseptic connection with aseptic
connectors and operations in a closed system).
A unidirectional airflow and lower velocities may be used in closed isolators
and glove boxes.
4.4 In order to reach the B, C and D air grades the number of air changes
should be appropriate for the size of the room and the equipment and
personnel present in it.
4.5 High-efficiency particulate air (HEPA) filters should be subjected
to an installed filter leakage test in accordance with ISO 14644-3 (3) at a
recommended interval of every 6 months, but not exceeding 12 months. The
purpose of performing regular leak tests is to ensure the filter media, filter
frame and filter seal are free from leaks. The aerosol selected for HEPA leak
testing should not support microbial growth and should be composed of a
sufficient number or mass of particles. HEPA filter patching is allowed at
the filter manufacturer and in situ operation provided that the patch sizes
and procedures follow the recommendations of ISO 1822-4 (4).

264
 Clean room and clean-air device classification
4.6 Clean rooms and clean-air devices should be classified in accordance
with ISO 14644 (2–3, 5–7).
4.6.1 Classification should be clearly differentiated from operational
process environmental monitoring. The maximum permitted airborne
particle concentration for each grade is given in Table 1.

Table 1
Maximum permitted airborne particle concentrat
Maximum permitted number of particles per m3 greater
than or equal to the tabulated size

At resta In operationb
Grade 0.5 μm 5.0 μm 0.5 μm 5.0 μm
A 3 520 20 3 520 20
B 3 520 29 352 000 2 900
C 352 000 2 900 3 520 000 29 000
D 3 520 000 29 000 Not defined Not defined
a
The “at rest” state is the condition where the installation is complete with equipment installed and
operating in a manner agreed upon by the customer and supplier, but with no personnel present.
b
The “in operation” state is the condition where the installation is functioning in the defined operating
mode and the specified number of personnel is present. The areas and their associated environmental
control systems should be designed to achieve both the “at rest” and “in operation” states.

4.6.2 For classification purposes in Grade A zones, a minimum sample

volume of 1 m³ should be taken per sample location. Referring to Table 1, for
Grade A the airborne particle classification is ISO 4.8 dictated by the limit for
particles ≥ 5.0 μm. For Grade B (at rest) the airborne particle classification is
ISO 5 for both particle sizes considered. For Grade C (at rest and in operation)
the airborne particle classification is ISO 7 and ISO 8, respectively. For Grade
D (at rest) the airborne particle classification is ISO 8. For classification
purposes ISO 14644-1 (2) methodology defines both the minimum number
of sample locations and the sample size based on the class limit of the largest
particle size considered and the method of evaluation of the data collected.
The sample volume should be determined according to ISO 14644-1 (2)
clause B.4.2. However, for lower grades (Grade C in operation and Grade
D at rest) the sample volume per location should be at least 2 litres and the
sample time per location should be not less than 1 minute.
4.6.3 Portable particle counters with a short length of sample tubing should
be used for classification purposes to avoid the loss of particles ≥ 5.0 μm.
Isokinetic sample heads should be used in unidirectional airflow systems.
4.6.4 “In operation” classification may be demonstrated during normal
operations, simulated operations or during media fills as worst-case simulation

265
 is required for this. ISO 14644-2 (6) provides information on testing to
demonstrate continued compliance with the assigned cleanliness classification.

Clean room and clean-air device monitoring

4.7 Clean rooms and clean-air devices should be routinely monitored
while in operation and the monitoring locations based on a formal risk
analysis study and the results obtained during the classification of rooms
and/or clean-air devices.
4.7.1 For Grade A zones, particle monitoring should be undertaken for
the full duration of critical processing, including equipment assembly,
except where justified by contaminants in the process that would damage
the particle counter or present a hazard, for example, live organisms and
radiological hazards. In such cases monitoring during routine equipment
set-up operations should be undertaken before exposure to the risk.
Monitoring during simulated operations should also be performed. The
Grade A zone should be monitored at a frequency and sample size such
that all interventions, transient events and any system deterioration would
be captured and alarms triggered if alert limits are exceeded. It is accepted
that it may not always be possible to demonstrate low levels of ≥ 5.0 μm
particles at the point of fill when filling is in progress, due to the generation
of particles or droplets from the product itself.
4.7.2 It is recommended that a similar system be used for Grade B zones,
although the sample frequency may be decreased. The importance of the
particle monitoring system should be determined by the effectiveness of
the segregation between the adjacent Grade A and B zones. The Grade B
zone should be monitored at a frequency and with a sample size such that
changes in levels of contamination and any deterioration of the system
would be captured and alarms triggered if alert limits are exceeded.
4.7.3 Airborne particle monitoring systems may consist of independent
particle counters; a network of sequentially accessed sampling points
connected by manifold to a single particle counter; or multiple small
particle counters located near monitoring points and networked to a data
acquisition system. Combinations of systems can also be used. The system
selected should be appropriate for the particle size considered.
Where remote sampling systems are used, the length of tubing and the radii of
any bends in the tubing should be considered in the context of particle losses
in the tubing. The selection of the monitoring system should take account of
any risk presented by the materials used in the manufacturing operation, for
example, those involving live organisms or radiopharmaceuticals.
4.7.4 The sizes of samples taken for monitoring purposes using automated
systems will usually be a function of the sampling rate of the system used.

266
 It is not necessary for the sample volume to be the same as that used for
formal classification of clean rooms and clean-air devices.
4.7.5 The airborne particle conditions given in Table 1 for the “at rest” state
should be achieved in the absence of the operating personnel after a short
“clean-up” or “recovery” period of about 15–20 minutes (guidance value),
after completion of the operations. The particulate conditions given in Table
1 for Grade A “in operation” should be maintained in the zone immediately
surrounding the product whenever the product or open container is exposed
to the environment. The “clean-up” or “recovery” test should demonstrate
a change in particle concentration by a factor of 100 within the prescribed
time (ISO 14644-3 clause B.12) (3).
4.7.6 In order to demonstrate control of the cleanliness of the various
clean areas during operation, they should be monitored for airborne particles
and microbial contamination. In addition to “at rest” and “in operation”
classification, airborne particles should be monitored periodically “in
operation” at critical locations. The sampling plan need not be the same as
that used for classification. Locations and sample sizes should be determined
based on an assessment of the process and contamination risk.
4.7.7 The monitoring of Grade C and D areas in operation should be
performed in accordance with the principles of quality risk management. The
requirements and alert/action limits will depend on the nature of the operations
carried out, but the recommended “clean-up period” should be attained.
4.7.8 Other characteristics such as temperature and relative humidity
depend on the product and nature of the operations carried out. These
parameters should not interfere with the defined cleanliness standard.
4.7.9 Examples of operations to be carried out in the various grades are
given in Table 2 (see also sections 4.12–4.20).
Table 2
Examples of operations to be carried out in the various grades
Grade Examples of operations for terminally sterilized products
(see sections 4.12–4.15 )
A Filling of products when unusually at risk
C Preparation of solutions when unusually at risk. Filling of products
D Preparation of solutions and components for subsequent filling

Grade Examples of operations for aseptic preparations

(see sections 4.16–4.20)
A Aseptic preparation and filling
C Preparation of solutions to be filtered
D Handling of components after washing

267
 4.8 To control the microbiological cleanliness of Grades A–D in
operation the clean areas should be monitored. Where aseptic operations
are performed, monitoring should be frequent using methods such as
settle plates, volumetric air and surface sampling (e.g. swabs and contact
plates). Sampling methods used in operation should not interfere with zone
protection. Results from monitoring should be considered when reviewing
batch documentation for finished product release. Surfaces and personnel
should be monitored after critical operations. Additional microbiological
monitoring is also required outside production operations, e.g. after
validation of systems, cleaning and sanitization.
4.9 Levels of detection of microbial contamination should be established
for the purpose of setting alert and action limits and for monitoring the
trends in environmental cleanliness in the facility. Limits expressed in
colony-forming units (CFU) for the microbiological monitoring of clean
areas in operation are given in Table 3. The sampling methods and numerical
values included in the table are not intended to represent specifications, but
are for information only.
Table 3
Recommended limits for microbial contaminationa
Grade Air sample Settle plates Contact plates Glove print
(CFU/m3) (diameter 90 mm) (diameter 55 mm) (5 fingers)
(CFU/4 hours)b (CFU/plate) (CFU/glove)
A <1 <1 <1 <1
B 10 5 5 5
C 100 50 25 –
D 200 100 50 –
CFU, colony-forming units.
a
These are average values.
b
Individual settle plates may be exposed for less than 4 hours.

4.10 Appropriate alert and action limits should be set for the results of
particulate and microbiological monitoring. If the action limits are exceeded
or a trend is identified in the alert limits, investigation should be initiated
and the appropriate corrective actions should be taken, as prescribed in the
operating procedures.
4.11 The area grades as specified in sections 4.12 to 4.20 should be
selected by the manufacturer on the basis of the nature of the process
operations being performed and validation runs (e.g. aseptic media fills or
others types of process simulations) are used to establish processing hold
times and a maximum fill duration. The determination of an appropriate
process area environment and a time limit should be based on the microbial
contamination (bioburden) found.

268
Terminally sterilized products
4.12 Components and most products should be prepared in at least a
Grade D environment to ensure low microbial bioburden and particulate
counts prior to filtration and sterilization. Where the product is at unusual
risk of microbial contamination (e.g. because it actively supports microbial
growth, must be held for a long period before sterilization, or is necessarily
processed mainly in open vessels), the preparation should generally be done
in a Grade C environment.
4.13 The filling of products for terminal sterilization should generally be
done in at least a Grade C environment.
4.14 Where the product is at unusual risk of contamination from the
environment (e.g. because the filling operation is slow, the containers are wide-
necked or are necessarily exposed for more than a few seconds before sealing),
the filling should be done in a Grade A zone with at least a Grade C background.
4.15 The preparation and filling of ointments, creams, suspensions and
emulsions should generally be done in a Grade C environment before
terminal sterilization.

Aseptic preparation
4.16 Components after washing should be handled in at least a Grade D
environment. The handling of sterile starting materials and components,
unless subjected to sterilization or filtration through a microorganism-
retaining filter later in the process, should be undertaken in a Grade A
environment with a Grade B background.
4.17 The preparation of solutions which are to be sterile-filtered during
the process should be undertaken in a Grade C environment (unless a closed
system is used, in which case a Class D environment may be justifiable). If
not sterile-filtered (therefore an aseptic manipulation) the preparation of
materials and products should be undertaken in a Grade A environment
with a Grade B background.
4.18 The handling and filling of aseptically prepared products, as well as
the handling of exposed sterile equipment, should be undertaken in a Grade
A environment with a Grade B background.
4.19 The transfer of partially closed containers, as used in freeze-drying,
before stoppering is completed, should be undertaken either in a Grade A
environment with a Grade B background or in sealed transfer trays in a
Grade B environment.
4.20 The preparation and filling of sterile ointments, creams, suspensions
and emulsions should be undertaken in a Grade A environment with a Grade
B background when the product is exposed and is not subsequently filtered.

269
 Processing
4.21 Precautions to minimize contamination should be taken during all
processing stages, including the stages before sterilization.
4.22 In general, preparations containing live microorganisms should not be
made, nor should containers be filled in areas used for the processing of other
pharmaceutical products. However, if the manufacturer can demonstrate
and validate effective containment and decontamination of the live
microorganisms, the use of multiproduct facilities may be justifiable. Vaccines
consisting of dead organisms or of bacterial extracts may be dispensed into
containers in the same premises as other sterile pharmaceutical products,
provided that the inactivation procedure has been properly validated.
When multiproduct facilities are used to manufacture sterile preparations
containing live microorganisms and other sterile pharmaceutical products, the
manufacturer should demonstrate and validate the effective decontamination
of the live microorganisms, in addition to precautions taken to minimize
contamination.
4.23 Validation of aseptic processing should include a process simulation
test using a nutrient medium (media fill). Selection of the nutrient medium
should be made based on dosage form of the product and selectivity, clarity,
concentration and suitability for sterilization of the nutrient medium.
4.24 The process simulation test should imitate as closely as possible the
routine aseptic manufacturing steps except where the activity may lead to
any potential microbial contamination.
4.25 Process simulation tests should be performed as part of validation by
running three consecutive satisfactory simulation tests. These tests should
be repeated at defined intervals and after any significant modification to
the heating, ventilation and air-conditioning (HVAC) system, equipment or
process. Process simulation tests should incorporate activities and interventions
known to occur during normal production as well as the worst-case situation.
The process simulation tests should be representative of each shift and shift
changeover to address any time-related and operational features.
4.26 The number of containers used for media fills should be sufficient to
enable a valid evaluation. For small batches the number of containers for
media fills should at least equal the size of the product batch. The target
should be zero growth and the following should apply:
• when filling fewer than 5000 units, no contaminated units should be
detected.
• when filling 5000–10 000 units:
— one contaminated unit should result in an investigation, including
consideration of a repeat media fill;

270
 — two contaminated units are considered cause for revalidation
following investigation;
• when filling more than 10 000 units:
— one contaminated unit should result in an investigation;
— two contaminated units are considered cause for revalidation
following investigation.
4.27 For any run size, intermittent incidents of microbial contamination
may be indicative of low-level contamination that should be investigated.
Investigation of gross failures should include the potential impact on the
sterility assurance of batches manufactured since the last successful media fill.
4.28 Care should be taken to ensure that any validation does not
compromise the processes.
4.29 Water sources, water-treatment equipment and treated water should
be monitored regularly for chemicals, biological contamination and
contamination with endotoxins to ensure that the water complies with the
specifications appropriate to its use. Records should be maintained of the
results of the monitoring and of any action taken (8).
4.30 Activities in clean areas, especially when aseptic operations are in
progress, should be kept to a minimum and the movement of personnel should
be controlled and methodical, so as to avoid excessive shedding of particles and
organisms due to over-vigorous activity. As far as possible, personnel should
be excluded from Grade A zones. The ambient temperature and humidity
should not be uncomfortably high because of the nature of the garments worn
and to reduce the risk of contamination liberated from the personnel.
4.31 The presence of containers and materials liable to generate fibres
should be minimized in clean areas and avoided completely when aseptic
work is in progress.
4.32 Components, bulk-product containers and equipment should be
handled after the final cleaning process in such a way as to ensure that they
are not recontaminated. The stage of processing of components as well as
the bulk-product containers and equipment should be properly identified.
4.33 The interval between the washing and drying and the sterilization of
components, bulk-product containers and equipment, as well as between
sterilization and use, should be as short as possible and subject to a time-
limit appropriate to the validated storage conditions.
4.34 The time between the start of the preparation of a solution and its
sterilization or filtration through a bacteria-retaining filter should be as short
as possible. A maximum permissible time should be set for each product that
takes into account its composition and the prescribed method of storage.

271
 4.35 Any gas that is used to purge a solution or blanket a product should
be passed through a sterilizing filter.
4.36 The bioburden should be monitored before sterilization. There should
be working limits on contamination immediately before sterilization, which
are related to the efficiency of the method to be used. Bioburden assay
should be performed on each batch for both aseptically filled products and
terminally sterilized products. Where overkill sterilization parameters are
set for terminally sterilized products, bioburden might be monitored only
at suitable scheduled intervals. For parametric release systems, bioburden
assay should be performed on each batch and considered as an in-process
test. Where appropriate, the level of endotoxins should be monitored. All
solutions, in particular large-volume infusion fluids, should be passed through
a microorganism-retaining filter, if possible sited immediately before filling.
4.37 Components, bulk-product containers, equipment, and any other
articles required in a clean area where aseptic work is in progress, should
be sterilized and wherever possible passed into the area through double-
ended sterilizers sealed into the wall. Other procedures that prevent the
introduction of contamination may be acceptable in some circumstances.
4.38 The efficacy of any new processing procedure should be validated
and the validation should be repeated at regular intervals thereafter or when
any significant change is made in the process or equipment.

5. Sterilization
5.1 Whenever possible products intended to be sterile should be terminally
sterilized by heat in their final container. Where it is not possible to carry
out terminal sterilization by heating due to the instability of a formulation
or incompatibility of a pack type (necessary to the administration of the
product, e.g. plastic eye-dropper bottles), a decision should be taken to use
an alternative method of terminal sterilization following filtration and/or
aseptic processing.
5.2 Sterilization can be achieved by the use of moist or dry heat, by
irradiation with ionizing radiation (noting that ultraviolet irradiation is
not normally an acceptable method of sterilization), by ethylene oxide (or
other suitable gaseous sterilizing agents), or by filtration with subsequent
aseptic filling of sterile final containers. Each method has its advantages
and disadvantages. Where possible and practicable, heat sterilization is the
method of choice. In any case the sterilization process must be in accordance
with the marketing and manufacturing authorizations.
5.3 The microbial contamination of starting materials should be minimal
and their bioburden should be monitored before sterilization. Specifications

272
 should include requirements for microbiological quality when the need for
this has been indicated by monitoring.
5.4 All sterilization processes should be validated. Particular attention
should be paid when the adopted sterilization method is not in accordance
with pharmacopoeial standards or other national standards, or when it is
used for a preparation that is not a simple aqueous or oily solution, for
example, colloidal suspensions.
5.5 Before any sterilization process is adopted, its suitability for the
product and its efficacy in achieving the desired sterilizing conditions in
all parts of each type of load to be processed should be demonstrated by
physical measurements and by biological indicators, where appropriate.
The validity of the process should be verified at scheduled intervals, at least
annually, and whenever significant modifications have been made to the
equipment. Records should be kept of the results.
5.6 For effective sterilization the whole of the material should be subjected
to the required treatment and the process should be designed to ensure that
this is achieved.
5.7 Biological indicators should be considered only as an additional
method of monitoring the sterilization process. They should be stored and
used according to the manufacturer’s instructions, and their quality checked
by positive controls. If they are used, strict precautions should be taken to
avoid any transfer of microbial contamination from them.
5.8 There should be a clear means of differentiating products that have
not been sterilized from those which have. Each basket, tray, or other carrier
of products or components should be clearly labelled with the name of the
material, its batch number and an indication of whether or not it has been
sterilized. Indicators such as autoclave tape may be used where appropriate to
indicate whether or not a batch (or sub-batch) has passed through a sterilization
process, but they do not give a reliable indication that the batch is in fact sterile.
5.9 Validated loading patterns should be established for all sterilization
processes.
5.10 Sterilization records should be available for each sterilization run.
They should be approved as part of the batch-release procedure.

6. Terminal sterilization
Sterilization by heat
6.1 Each heat-sterilization cycle should be recorded by means of
appropriate equipment of suitable accuracy and precision, e.g. on a time/
temperature chart with a suitably large scale. The temperature should

273
 be recorded by a probe situated at the coolest part of the load or loaded
chamber, this point having been determined during the validation; the
temperature should preferably be checked against a second independent
temperature probe located at the same position. Sterilization records should
be available for each sterilization run and should be approved as part of the
batch release procedure. Chemical or biological indicators may also be used
but should not take the place of physical controls.
6.2 Sufficient time should be allowed for the whole of the load to reach
the required temperature before measurement of the sterilizing time
is started. This time should be determined for each type of load to be
processed.
6.3 After the high-temperature phase of a heat sterilization cycle,
precautions should be taken against contamination of a sterilized load
during cooling. Any cooling fluid or gas in contact with the product should
be sterilized.
Sterilization by moist heat
6.4 Both temperature and pressure should be used to monitor the process.
Control instrumentation should normally be independent of monitoring
instrumentation and recording charts. Where automated control and
monitoring systems are used for these applications they should be validated
to ensure that critical process requirements are met. System and cycle
faults should be registered by the system and observed by the operator.
The reading of the independent temperature indicator should be routinely
checked against the reading on the chart recorder during the sterilization
period. For sterilizers fitted with a drain at the bottom of the chamber, it
may also be necessary to record the temperature at this position throughout
the sterilization period. There should be regular leak tests on the chamber
when a vacuum phase is part of the cycle.
6.5 The items to be sterilized, other than products in sealed containers,
should be wrapped in a material that allows the removal of air and the
penetration of steam but prevents recontamination after sterilization.
Specially designed autoclavable stainless steel containers, that allow steam
to enter and air to leave, can also be used. All parts of the load should be
in contact with water or saturated steam at the required temperature for the
required time.
6.6 Care should be taken to ensure that the steam used for sterilization
is of suitable quality (chemical, microbiological and endotoxin analysis of
condensate and physical examination of steam (such as dryness, superheat,
and non-condensable gases) and does not contain additives at a level that
could cause contamination of the product or equipment. Steam used for
sterilization should be tested regularly.

274
Sterilization by dry heat
6.7 Sterilization by dry heat may be suitable for non-aqueous liquids or
dry-powder products.
The process used should include air circulation within the chamber and the
maintenance of a positive pressure to prevent the entry of non-sterile air. If
air is supplied it should be passed through a microorganism-retaining filter
(e.g. a HEPA filter). Where sterilization by dry heat is also intended to remove
pyrogens, challenge tests using endotoxins are required as part of the validation.

Sterilization by radiation
6.8 Sterilization by radiation is used mainly for heat-sensitive materials
and products. Many pharmaceutical products and some packaging materials
are radiation-sensitive, so this method is permissible only when the absence
of deleterious effects on the product has been confirmed experimentally.
Ultraviolet irradiation is not an acceptable method for terminal sterilization.
6.9 If sterilization by radiation is done by an outside contractor, the
manufacturer is responsible for ensuring that the requirements of section
6.8 are met and that the sterilization process is validated.
6.10 During the sterilization procedure the radiation dose should be
measured. The dosimeters used for this purpose should be independent of the
dose rate and should provide a quantitative measurement of the dose received
by the product itself. Dosimeters should be inserted in the load in sufficient
number and close enough together to ensure that there is always a dosimeter
in the chamber. Where plastic dosimeters are used they should be used within
the time-limit of their calibration. Dosimeter absorbance should be read
shortly after exposure to radiation. Radiation-sensitive colour discs may be
used to differentiate between packages that have been subjected to irradiation
and those that have not; they are not indicators of successful sterilization. The
information obtained should constitute part of the batch record.
6.11 Validation procedures should ensure that consideration is given to
the effects of variations in the density of the packages.
6.12 Material-handling procedures should prevent any mix-up of
irradiated and non-irradiated materials. Each package should carry a
radiation-sensitive indicator to show whether or not it has been subjected to
radiation treatment.
6.13 The total radiation dose should be administered within a predetermined
period.

Sterilization by gases and fumigants

6.14 Sterilization by gases and fumigants should only be used for finished
products where there is no suitable alternative.

275
 6.15 Various gases and fumigants may be used for sterilization (e.g. ethylene
oxide and hydrogen peroxide vapour). Ethylene oxide should be used only
when no other method is practicable. During process validation it should
be shown that the gas has no damaging effect on the product and that the
conditions and time allowed for degassing are such as to reduce any residual
gas and reaction products to defined acceptable limits for the type of product or
material concerned. These limits should be incorporated in the specifications.
6.16 Direct contact between gas and microorganisms is essential;
precautions should, therefore, be taken to avoid the presence of organisms
likely to be enclosed in materials such as crystals or dried protein. The nature
and quantity of packaging materials can significantly affect the proces.
6.17 Before exposure to the gas, materials should be brought into
equilibrium with the humidity and temperature required by the process.
This requirement should be balanced against the need to minimize the
waiting time before sterilization.
6.18 Each sterilization cycle should be monitored with suitable biological
indicators, using the appropriate number of test pieces distributed throughout
the load. The information thus obtained should form part of the batch record.
6.19 Biological indicators should be stored and used according to the
manufacturer’s instructions and their performance checked by positive
controls.
6.20 For each sterilization cycle, records should be made of the time taken
to complete the cycle, of the pressure, temperature and humidity within
the chamber during the process and of the gas concentration. The pressure
and temperature should be recorded on a chart throughout the cycle. The
records should form part of the batch record.
6.21 After sterilization, the load should be stored in a controlled manner
in ventilated conditions to allow concentrations of residual gas and reaction
products to fall to their prescribed levels. This process should be validated.

7. Aseptic processing and sterilization by filtration

7.1 The objective of aseptic processing is to maintain the sterility of a
product that is assembled from components, each of which has been
sterilized by one of the above methods (see sections 5 and 6).
7.2 The operating conditions should be such as to prevent microbial
contamination.
7.3 In order to maintain the sterility of the components and the product
during aseptic processing, careful attention needs to be given to:

276
— the environment;
— personnel;
— critical surfaces;
— container/closure sterilization and transfer procedures;
— the maximum holding period of the product before filling into the final
container; and
— the sterilizing filter.
7.4 Certain solutions and liquids that cannot be sterilized in the final
container can be filtered through a sterile filter of nominal pore size
0.22 micron (or less), or with at least equivalent microorganism-retaining
properties, into a previously sterilized container. Such filters can remove
bacteria and moulds, but not all viruses or mycoplasmas. Consideration
should be given to complementing the filtration process with some degree
of heat treatment. Filtration alone is not considered sufficient when
sterilization in the final container is possible. Of the methods currently
available, steam sterilization is preferred.
7.5 Owing to the potential additional risks of the filtration method
as compared with other sterilization processes, a double-filter layer or
second filtration through a further sterilized microorganism-retaining filter
immediately prior to filling may be advisable. The final sterile filtration should
be carried out as close as possible to the filling point.
7.6 The fibre-shedding characteristics of filters should be minimal
(virtually zero). Asbestos-containing filters should not be used under any
circumstances.
7.7 The integrity of the sterilized filter should be verified before use and
should be confirmed immediately after use by an appropriate method such
as a bubble point, diffusive flow or pressure hold test. The time taken to
filter a known volume of bulk solution and the pressure difference to be used
across the filter should be determined during validation and any significant
differences from these during routine manufacturing should be noted
and investigated. Results of these checks should be included in the batch
record. The integrity of critical gas and air vent filters should be confirmed
after use. The integrity of other filters should be confirmed at appropriate
intervals. Consideration should be given to increased monitoring of filter
integrity in processes that involve harsh conditions, e.g. the circulation of
high-temperature air.
7.8 The same filter should not be used for more than one working day
unless such use has been validated.
7.9 The filter should not affect the product either by removing ingredients
from it or by releasing substances into it.

277
8. Isolator technology
8.1 The use of isolator technology to minimize human interventions in
processing areas may result in a significant decrease in the risk of microbial
contamination of aseptically manufactured products from the environment.
There are many possible designs of isolators and transfer devices. The
isolator and the background environment should be designed so that the
required air quality for each zone can be realized. Isolators are constructed
of various materials more or less prone to puncture and leakage. Transfer
devices may vary from single-door to double-door designs to fully-sealed
systems incorporating sterilization mechanisms.
8.2 The transfer of materials into and out of the unit is one of the greatest
potential sources of contamination. In general the area inside the isolator
is the local zone for high-risk manipulations, although it is recognized that
unidirectional airflow may not exist in the working zone of all isolators and
transfer devices.
8.3 The air classification required for the background environment
depends on the design of the isolator and its application. It should be
controlled, and for aseptic processing it should be at least Grade D.
8.4 Isolators should be introduced only after appropriate validation.
Validation should take into account all critical factors of isolator technology,
for example, the quality of the air inside and outside (background) the
isolator, sanitization of the isolator, the transfer process and isolator integrity.
8.5 Monitoring should be done routinely and should include frequent leak
testing of the isolator and the glove/sleeve system.

9. Blow/fill/seal technology
9.1 Blow/fill/seal units are purpose-built machines in which, in one
continuous operation, containers are formed from a thermoplastic granulate,
filled and then sealed, all by the one automatic machine. Blow/fill/seal
equipment used for aseptic production which is fitted with an effective Grade
A air shower may be installed in at least a Grade C environment, provided
that Grade A or B clothing is used. The environment should comply with
the viable and non-viable limits at rest and the viable limit only when in
operation. Blow/fill/seal equipment used for the production of products which
are terminally sterilized should be installed in at least a Grade D environment.
9.2 Because of this special technology, particular attention should be paid
to at least the following:
— equipment design and qualification;

278
 — validation and reproducibility of cleaning-in-place and sterilization-in-
place;
— background clean room environment in which the equipment is located;
— operator training and clothing; and
— interventions in the critical zone of the equipment including any aseptic
assembly prior to the commencement of filling.

10. Personnel
10.1 Only the minimum number of personnel required should be present
in clean areas; this is particularly important during aseptic processes. As
far as possible, inspections and controls should be conducted from outside
such areas.
10.2 All personnel (including those concerned with cleaning and
maintenance) employed in such areas should receive initial and regular
training in disciplines relevant to the correct manufacture of sterile products,
including hygiene and the basic elements of microbiology. When outside
staff who have not received such training (e.g. building or maintenance
contractors) need to be brought in, particular care should be taken over their
instruction and supervision.
10.3 Staff who have been engaged in the processing of animal-tissue
materials or of cultures of microorganisms other than those used in the
current manufacturing process should not enter sterile-product areas
unless rigorous and clearly defined decontamination procedures have been
followed.
10.4 High standards of personal hygiene and cleanliness are essential
and personnel involved in the manufacture of sterile preparations should
be instructed to report any conditions that may cause the shedding of
abnormal numbers or types of contaminants; periodic health checks for
such conditions are desirable. The action to be taken in respect of personnel
who might be introducing undue microbial hazards should be decided by a
designated competent person.
10.5 Changing and washing should follow a written procedure designed
to minimize the contamination of clean-area clothing or the carry-through
of contaminants to clean areas. The clothing and its quality should be
appropriate for the process and the grade of the working area. It should be
worn in such a way as to protect the product from contamination.
10.6 Outdoor clothing should not be brought into changing rooms leading
to Grade B and C rooms. For every worker in a Grade A/B area, clean sterile
(sterilized or adequately sanitized) protective garments should be provided
at each work session. Gloves should be regularly disinfected during

279
 operations. Masks and gloves should be changed at least every working
session. Operators working in Grade A and B areas should wear sanitized
goggles.
10.7 Wrist-watches, cosmetics and jewellery should not be worn in clean
areas.
10.8 The clothing required for each grade is as follows:
• Grade D. The hair and, where relevant, beard and moustache should be
covered. Protective clothing and appropriate shoes or overshoes should be
worn. Appropriate measures should be taken to avoid any contamination
from outside the clean area.
• Grade C. The hair and, where relevant, beard and moustache should be
covered. A one-piece jumpsuit, gathered at the wrists and with a high
neck, and appropriate shoes or overshoes should be worn. The clothing
should shed virtually no fibres or particulate matter.
• Grades A and B. Entry of personnel into Grade A areas should be
minimized. Headgear should totally enclose the hair and, where relevant,
beard and moustache. A one-piece jumpsuit, gathered at the wrists and
with a high neck, should be worn. The headgear should be tucked into
the neck of the suit. A facemask should be worn to prevent the shedding
of droplets. Sterilized, non-powdered gloves of appropriate material and
sterilized or disinfected footwear should be worn. Trouser-bottoms should
be tucked inside the footwear and garment sleeves into the gloves. The
protective clothing should shed virtually no fibres or particulate matter
and should retain particles shed by the body.
10.9 Clothing used in clean areas should be laundered or cleaned in such
a way that it does not gather additional particulate contaminants that can
later be shed. Separate laundry facilities for such clothing are desirable. If
fibres are damaged by inappropriate cleaning or sterilization, there may be
an increased risk of shedding particles. Washing and sterilization operations
should follow standard operating procedures.

11. Premises
11.1 All premises should as far as possible be designed to avoid the
unnecessary entry of supervisory or control personnel. Grade A and B
areas should be designed so that all operations can be observed from
outside.
11.2 In clean areas all exposed surfaces should be smooth, impervious
and unbroken to minimize the shedding or accumulation of particles or
microorganisms and to permit the repeated application of cleaning agents
and disinfectants, where used.

280
11.3 To reduce the accumulation of dust and to facilitate cleaning, there
should be no uncleanable recesses and a minimum of projecting ledges,
shelves, cupboards and equipment. Doors should be carefully designed
to avoid uncleanable recesses; sliding doors may be undesirable for this
reason. Swing doors should open to the high-pressure side and be provided
with self-closers. Exceptions are permitted based on egress and site
environmental, health and safety containment requirements.
11.4 False ceilings should be sealed to prevent contamination from the
void space above them.
11.5 Pipes and ducts and other utilities should be installed so that they
do not create recesses, unsealed openings and surfaces that are difficult
to clean. Sanitary pipes and fittings should be used and threaded pipe
connections should be avoided.
11.6 Sinks and drains should be avoided wherever possible and should
be excluded from Grade A and B areas where aseptic operations are carried
out. Where installed they should be designed, located and maintained so as
to minimize the risks of microbial contamination; they should be fitted with
effective, easily cleanable traps and with air breaks to prevent backflow. Any
floor channels should be open and easily cleanable and be connected to drains
outside the area in a manner that prevents the ingress of microbial contaminants.
11.7 Changing rooms should be designed as airlocks and used to provide
physical separation of the different stages of changing and so minimize
microbial and particulate contamination of protective clothing. They should
be flushed effectively with filtered air. The final stage of the changing room
should, in the at-rest state, be the same grade as the area into which it leads.
The use of separate changing rooms for entering and leaving clean areas is
sometimes desirable. In general hand-washing facilities should be provided
only in the first stage of the changing rooms.
There should not be a change of more than one grade between airlocks or
passages and changing rooms, i.e. a Grade D passage can lead to a Grade C
airlock, which leads to a Grade B changing room, which leads to a Grade B
clean room. Changing rooms should be of a sufficient size to allow for ease of
changing. Changing rooms should be equipped with mirrors so that personnel
can confirm the correct fit of garments before leaving the changing room.
11.8 Airlock doors should not be opened simultaneously. An interlocking
system and a visual and/or audible warning system should be operated to
prevent the opening of more than one door at a time.
11.9 A filtered air supply should be used to maintain a positive pressure
and an airflow relative to surrounding areas of a lower grade under all
operational conditions; it should flush the area effectively. Adjacent rooms

281
 of different grades should have a pressure differential of approximately
10–15 Pascals (guidance value). Particular attention should be paid to the
protection of the zone of greatest risk, i.e. the immediate environment to
which the product and the cleaned components in contact with it are exposed.
The recommendations regarding air supplies and pressure differentials may
need to be modified where it becomes necessary to contain certain materials,
e.g. pathogenic, highly toxic, radioactive or live viral or bacterial materials
or products. The decontamination of the facilities and the treatment of air
leaving a clean area may be necessary for some operations.
11.10 It should be demonstrated that airflow patterns do not present a
contamination risk; for example, care should be taken to ensure that
particles from a particle-generating person, operation or machine are not
conveyed to a zone of higher product risk.
11.11 A warning system should be operated to indicate failure in the air
supply. Indicators of pressure differentials should be fitted between areas
where this difference is important, and the pressure differentials should be
regularly recorded and failure alarmed.
11.12 Consideration should be given to restricting unnecessary access to
critical filling areas, e.g. Grade A filling zones, by means of a physical barrier.

12. Equipment
12.1 A conveyor belt should not pass through a partition between a Grade
A or B clean area and a processing area of lower air cleanliness, unless the
belt itself is continuously sterilized (e.g. in a sterilizing tunnel).
12.2 Whenever possible, equipment used for processing sterile products
should be chosen so that it can be effectively sterilized by steam or dry heat
or other methods.
12.3 As far as possible, equipment fittings and services should be designed
and installed so that operations, maintenance and repairs can be carried out
outside the clean area. Equipment that has to be taken apart for maintenance
should be re-sterilized after complete reassembly, wherever possible.
12.4 When equipment maintenance is carried out within a clean area,
clean instruments and tools should be used and the area should be cleaned
and disinfected again, where appropriate, before processing recommences,
if the required standards of cleanliness and/or asepsis have not been
maintained during the maintenance work.
12.5 All equipment such as sterilizers, air-handling and filtration systems,
air vent and gas filters, water treatment, generation, storage and distribution

282
 systems should be subject to validation and planned maintenance; their
return to use should be approved.
12.6 Water-treatment plants and distribution systems should be designed,
constructed and maintained so as to ensure a reliable source of water of an
appropriate quality. They should not be operated beyond their designed
capacity. Consideration should be given to including a testing programme in the
maintenance of a water system. Water for injection should be produced, stored
and distributed in a manner which prevents the growth of microorganisms, e.g.
by constant circulation at a temperature above 70 °C or not more than 4 °C (8).

13. Finishing of sterile products

13.1 Containers should be closed by appropriately validated methods.
Containers closed by fusion, e.g. glass or plastic ampoules, should be
subject to 100% integrity testing. Samples of other containers should be
checked for integrity according to appropriate procedures.
13.2 The container closure system for aseptically filled vials is not
fully integral until the aluminum cap has been crimped into place on the
stoppered vial. Crimping of the cap should, therefore, be performed as soon
as possible after stopper insertion.
13.3 As the equipment used to crimp vial caps can generate large quantities
of non-viable particulates, the equipment should be located at a separate
station equipped with adequate air extraction.
13.4 Vial capping can be undertaken as an aseptic process using sterilized
caps or as a clean process outside the aseptic core. Where this latter approach
is adopted, vials should be protected by Grade A conditions up to the point
of leaving the aseptic processing area, and thereafter stoppered vials should
be protected with a Grade A air supply until the cap has been crimped.
13.5 Vials with missing or displaced stoppers should be rejected prior
to capping. Where human intervention is required at the capping station,
appropriate technology should be used to prevent direct contact with the
vials and to minimize microbial contamination.
13.6 Restricted access barriers and isolators may be beneficial in assuring
the required conditions and minimizing direct human interventions into the
capping operation.
13.7 Containers sealed under vacuum should be tested for maintenance of
that vacuum after an appropriate, predetermined period.
13.8 Filled containers of parenteral products should be inspected individually
for extraneous contamination or other defects. When inspection is carried

283
 out visually this should be done under suitable and controlled conditions of
illumination and background. Operators doing the inspection should pass regular
eyesight checks, using personal corrective lenses (e.g. spectacles or contact
lenses) as required, and be allowed frequent breaks from inspection. Where
other methods of inspection are used, the process should be validated and the
performance of the equipment checked at intervals. Results should be recorded.

References
1. Good manufacturing practices for sterile pharmaceutical products. In: WHO
Expert Committee on Specifications for Pharmaceutical Preparations. Thirty-sixth
report. Geneva, World Health Organization, 2002 (WHO Technical Report Series,
No. 902), Annex 6; and in Quality assurance of pharmaceuticals. A compendium
of guidelines and related materials. Vol. 2. 2nd updated ed. Good manufacturing
practices and inspection. Geneva, World Health Organization, 2007; and in
Quality Assurance of Pharmaceuticals. A compendium of guidelines and related
materials. Geneva, World Health Organization, 2010 (CD-ROM).
2. ISO 14644-1. Clean rooms and associated controlled environments. Part 1:
Classification of airborne particles. Geneva, International Organization for
Standardization.
3. ISO 14644-3. Clean rooms and associated controlled environments. Part 3:
Test methods. Geneva, International Organization for Standardization.
4. ISO 1822-4. High efficiency air filters (HEPA and ULPA). Determining leakage
of filter elements (scan method).
5. ISO 14644-4. Clean rooms and associated controlled environments. Part 4:
Design, construction and start-up. Geneva, International Organization for
Standardization.
6. ISO 14644-2. Clean rooms and associated controlled environments. Part 2:
Monitoring for continued compliance with ISO 14644-1. Geneva, International
Organization for Standardization.
7. ISO 14644-5 Clean rooms and associated controlled environments. Part 5:
Cleanroom operations. Geneva, International Organization for Standardization.
8. Good manufacturing practices for pharmaceutical products: water for
pharmaceutical use. In: WHO Expert Committee on Specifications for
Pharmaceutical Preparations. Thirty-ninth report. Geneva, World Health
Organization, 2005 (WHO Technical Report Series, No. 929), Annex 3; and
in Quality assurance of pharmaceuticals. A compendium of guidelines and
related materials. Vol. 2. 2nd updated ed. Good manufacturing practices and
inspection. Geneva, World Health Organization, 2007.

Further reading
FDA Guidance for Industry. Sterile drug products produced by aseptic
processin g — cGMP. US Food and Drug Administration, 2004.
Guidance for industry. Sterile drug products produced by aseptic
processing. Japan, 2005.
Manufacture of sterile medicinal products. In: The rules governing medicinal
products in the European Union Vol. 4. EU guidelines to good manufacturing
practice medicinal products for human and veterinary use. Annex 1, Brussels, 2008.

284