VALIDATION OF STERILIZING FILTERS

A. Introduction to Filtration
The following definitions will be helpful in using this section. When filter is
used as a verb (“to filter”) it means to pass a solid–liquid mixture through a
permeable medium to cause a separation of the two. Filter when used as a noun
refers to a device for carrying out filtration, and it consists of the filter medium
and a suitable holder for constraining and supporting it in the fluid path. The
permeable material that separates solid particles from the liquid being filtered
is called the filter medium. The unit operation of filtration, then, is the separation
of solids from a liquid by passage through a filter medium. In many instances,
the filter, including the permeable medium, the means for passing liquid
through the medium, and the process piping, are all referred to by the term filter
system.
In general, filtration objectives can be separated into four basic categories:
to save solids and reject liquids, to save liquids and reject solids, to save both
liquids and solids, and to reject both liquids and solids [40].
As a filtration process proceeds, generally under an applied driving force
of pressure, solids are removed by and begin to accumulate on the filter medium.
The liquid portion continues to move through the filter medium and out
of the filter system. The separated liquid is referred to as the filtrate. The amount
of pressure applied to accomplish the filtration depends on the filtration resistance.
Filtration resistance is a result of the frictional drag on the filtrate as it
passes through the filter medium and the accumulated solids. In equation form,
Filtration rate =
pressure
resistance
(13)
Permeability is often referred to as a measure of liquid flow through a filter
system and is the reciprocal of the filtration resistance.
During filtration, as the particulate buildup continues on the filtration medium,
the filtration resistance increases, or in other words, the filtration permeability
decreases. The capacity of a system, expressed in time, volume of liquid
fed, or amount of solids fed, depends on the ability of the system to maintain
acceptable permeability.
When operating a filtration system, it is important to note the following
general relationship:
Retention × permeability = constant
Therefore, in attempting to have a certain degree of filtration efficiency or retention,
a high rate of filtration, and the lowest possible cost, it is necessary to
make a compromise with one or more of the above factors. A high permeability
or low resistance for large filtration flow rates requires a filter medium of low
retention efficiency. A highly efficient retention will have low permeability, low
flow rates, and higher filtration costs.
B. Sterile Filtration
Production of parenteral drugs requires that the product be sterile. In many
cases, terminal sterilization by heat, ethylene oxide gas, or ionizing radiation is
used to render a product sterile; however, certain products are not stable when
exposed to heat, gas, or radiation, and they must be sterilized by other means.
Filtrative sterilization is suitable in such cases. Indeed, the practice of sterile
filtration is not limited to labile preparations. Unlike the other forms of sterilization,
filtration sterilizes by the removal of the bacteria from the product rather
than by inducing a lethality to the micro-organism. Filtration is straightforward
and reliable; it removes particulate matter other than microbiological; it avoids
possible pyrogenicity owing to the presence of dead bacteria in the dosage form;
it is cost effective and energy efficient; and it allows convenient and flexible
manufacturing systems and schedules with low capital investment [41].
Sterile filtration processes are employed to sterile-filter a product prior to
filling it aseptically into its final containers. Bulk drug solutions are sterilefiltered
prior to aseptic crystallization, thus eliminating the possibility of having
organisms within the bulk drug crystals. The bulk drug can then be processed
into a dosage form aseptically or further processed to be terminally sterilized.
Other filtrative operations reduce the organism content of a final product prior
to terminal sterilizations.
As noted earlier, a highly efficient retentive media will have low permeability,
low flow rates, and higher filtration costs than other less retentive filter
media. The highly retentive filter media used for sterilization have a short useful
life because they clog very easily. Consequently, most filtration processes cannot
be efficiently or economically carried out without the use of prefiltration.
Prefiltration filter media are used to protect and thus lengthen the useful life of
the final membrane filter media by collecting the bulk of the particulate material
so that the membrane filter media must filter out only a small portion of the
particulate. Prefiltration media are normally depth-filter media having a relatively
wide pore and size distribution. A properly selected prefilter must meet
the following conditions: (1) it must be retentive enough to protect the final
membrane filter medium; (2) the prefilter assembly must not allow fluid bypass
under any condition; (3) the prefilter system must be designed to make use of
the prefilter medium; (4) it must have the best retention efficiency (with depth-filter media low pressure differentials and low fluid flux, accomplished by a
multielement parallel design, are best); and (5) the prefilter medium must be
compatible with the solution and not leach components into the solution or
absorb components from the solution. One note of caution needs to be mentioned
in reference to lengthening membrane filter media life. Organism growthrough
can become a problem if filtration takes place over an extended period
of time. During filtration, bacteria continuously reproduce by cell division and
eventually find their way through the filter medium to contaminate the filtrate.
For this reason, prolonged filtration must be avoided. The proposed CGMPs for
large-volume parenterals state that final filtration of solutions shall not exceed
8 hr [42].
Sterilization by filtration is a major unit operation used in aseptic processes.
Aseptic processes require the presterilization of all components of the
drug product and its container. Then all of the components are brought together
in a controlled aseptic environment to create the finished sterile product sealed
within its container/closure system. The level of sterility attained by an aseptic
procedure is a cumulative function of all the process steps involved in making
the product. Therefore, the final level of sterility assurance for such a product
cannot be greater than the step providing the lowest probability of sterility. Each
step in the aseptic process must be validated to known levels of sterility assurance
[43].
This section will concentrate on that portion of the aseptic process wherein
the drug product is sterilized by filtration. From the earlier discussion, sterile
filtration is perhaps a misnomer, since the “sterile” filtrate is almost always
processed further under aseptic conditions, which involves a risk of contamination
[44]. Therefore, to speak of drug product sterilization by filtration as being
as final a processing step as the steam sterilization of a product could possibly
lead to erroneous assurances or assumptions. Since a sterile filtrate can be produced
by filtration, however, we will continue to refer to the process as product
sterilization by filtration.
The primary objective of a sterilizing filter is to remove microorganisms.
The filter medium used to accomplish such an efficient retention may be classified
as one of two types—the reusable type or the disposable type.
The reusable filter media are made of sintered glass, unglazed procelain,
or diatomaceous earth (Table 8). Because these filter media may be used repeatedly
without being destroyed, they are less costly; however, the use of reusable
filter media demands that the media be cleaned perfectly and sterilized prior to
use to prevent microbial contamination and chemical cross-contamination. Even
after exacting and painstaking cleaning processes have been used on reusable
filter media, most companies using sterile filtration have decided that the risk
of contamination is still great and prefer the use of the disposable media that
are used once and then discarded. The remainder of our discussion will concern
the disposable media, often referred to as membrane filter media.
Filter media consist
of a matrix of pores held in a fixed spatial relationship by a solid continuum.
The pores allow the product solution to pass through the medium while retaining
the unwanted solid particles and micro-organisms. The size of filter medium
pores to retain micro-organisms must be quite small. The 0.20- or 0.22-μm pore
size filter media are considered to be capable of producing sterile filtrates.
The characteristics of a given membrane filter medium depend on its
method of manufacture: whether by phase separation of casting solutions, by
adhesion into an organic union of matted fibers, or by track etching of solid
films [45]. The retention of micro-organisms by the various membrane media,
while not fully understood, has been investigated by numerous researchers who
have indicated that several mechanisms are responsible. The dominant mechanism
of retention is sieve retention. Particles larger than the pore size of the
filter medium are retained on the medium, and as large particles are retained,
pore openings can become bridged and thereby effectively reduce the filter medium’s
pore size. Other possible mechanisms of retention are adsorption of the
particles into the medium itself, entrapment in a tortuous path, impaction, and
electrostatic capture. [46]. The importance of these latter retention mechanisms
has not been fully determined, and on the whole filtration sterilization is treated
as depending on the steric influences of the sieve retention mechanism. The
problem with assuming a sieve retention mechanism is that a sieve or screen
has uniform openings, whereas a membrane filter medium does not. The filter
medium has a distribution of pores, albeit narrow, rather than pores of a singlesize.
In addition, thinking in terms of a sieve or screen conjures up a vision of
precisely measured and numbered openings. Precise methods for computing
both numbers and the actual sizes of pores in a filter membrane medium are not
available.
Many approaches have been taken in an attempt to measure the size of
membrane filter media pores [47,48]. Flow measurements, both of air and of
water, have been made. Mercury intrusion under high pressure has been employed,
and pore sizing using either molecular templates or particles, including
bacteria of known size, has been tried. The numerical values for pore sizes from
these methods are based on a derivation from a particular model selected. Each
of the various models has difficulties and shortcomings, and a pore size designation
based on one method does not necessarily mean that a filter medium with
the same designated size but from a different method really is the identical size
[49]. More important, relating such a designated pore size to the membrane’s
ability to retain certain size particles may be anywhere from merely uncertain
to misleading. Therefore, a given membrane filter medium with a designated
pore size of 0.2 micron should not be thought of as “absolutely retaining all
particles greater than 0.2 micron” without challenging the medium with a known
size particulate. In fact, filter media should not be thought of as “absolute retentive”
devices at all. It has been demonstrated that under certain operational
conditions or with certain bacterial challenges, 0.2 micron–rated membrane filter
media can be penetrated by bacteria. Filter media companies do challenge
their products to ensure retention efficiency to sterility [46,50–52].
In addition to the pore size–particle size retention relationship problems
mentioned above, other factors can influence a filter medium’s retention characteristics.
Absorptive retention can be influenced by the organism size, organism
population, pore size of the medium, pH of the filtrate, ionic strength, surface
tension, and organic content. Operational parameters can also influence retention,
such as flow rate, salt concentration, viscosity, temperature, filtration duration,
filtration pressure, membrane thickness, organism type, and filter medium
area [52,53].
The complexity of the sterile filtration operation and the CGMP regulations
require the validation of sterilizing filter systems. The validation of a sterile
filtration operation can be complex, with many operational parameters and
their interactions needing to be identified, controlled, and predicted for each end
product to demonstrate that sterility is adequately achieved by the filtration process.
In the commonly used steam sterilization process, the heat parameters are
identified and in-process controls specified such that a level of sterility assurance
can be reproducibly obtained. In steam sterilization, the important parameter
of heat, measured by temperature, can be accurately measured and continuously
monitored to ensure the operational integrity of the autoclave; however,
unlike steam sterilization, filtration sterilization cannot be monitored on a continuous
basis throughout the process.
The important aspect of filtration sterilization, the membrane filter me-
dium—its pore size, pore size distribution, integrity, and capacity—cannot be
monitored during use. Therefore, the prediction that a filter membrane, given a
certain set of operational parameters, will produce a sterile filter is critical. The
only way to test a membrane filter medium’s ability to retain bacteria is to
challenge the medium with bacteria. Unfortunately, after a challenge with bacteria
the filter membrane cannot be used again. Therefore, nondestructive tests
need to be developed by which a filter can be tested as to its suitability for
bacterial retention. Consequently, the approach in filter system validation has
been to establish a reproducible relationship between a membrane’s pore size
and its bacterial retention efficiency. The thinking is that once such a relation
is established, a nondestructive physical test can be developed by which each
filter membrane medium can be tested and its bacterial retention efficiency assured.
Testing of the membrane can then be performed both before and after
use, and if the test results are satisfactory, the filtration process can be deemed
to have been carried out successfully.