Elements of

Biopharmaceutical
Production

Qualification of a
Chromatographic Column
Why and How to Do It
Anurag S. Rathore, Robert M. Kennedy, J. Kevin ODonnell,
Ivars Bemberis, and Oliver Kaltenbrunner

reparative chromatography is the dominant

purification technique in the production of
biological compounds, especially for
bioseparations. The column used is central
to the performance of the chromatographic
step. To ensure robust and reproducible column
performance, columns need to be qualified in an
acceptable fashion before each use.
Some of the critical elements of column qualification include identifying leachables and
extractables from resins, product use, and column
parts and accessories (such as residuals from
manufacturing and shipping). Identity tests must
be developed for chromatography resins, such as
base matrices (FTIR spectroscopy, for example);
functional groups and substitution levels (such as
colorimetric titration); and porosity tests, or if
necessary, function tests (such as dynamic capacity, selectivity, or real runs). Column qualification also involves writing protocols for training
plant personnel and protocols for column hardware, packing, and operation. And qualification
requires protocols for measuring bed integrity,
such as background buffers, tracers, and injection

FDA has become more knowledgeable about process

chromatography and more demanding about column
qualification. In the latest installment of BioPharm
Internationals Elements of BioPharmaceutical
Production, five industry experts share their insights
on how to qualify a chromatographic column. Their
advice: Write unambiguous SOPs. Focus on
reproducibility of column packing. Choose appropriate
metrics. And analyze your testing procedures to reduce
the chance of erroneous results.

30

BioPharm International

MARCH 2003

volumes (see the Measuring Column Bed

Integrity sidebar for further information and
equations that help confirm the quality of chromatographic operations).
This BioPharm International series on the
Elements of Biopharmaceutical Production
presents the viewpoints of industry experts on
issues commonly encountered in developing and
manufacturing biopharmaceuticals. The series
has previously covered process validation (1) and
the optimization and scale-up of preparative chromatography (2). In this article, our series coordinator has brought together four other experts
representing major chromatography media and
equipment manufacturers and from biopharmaceutical companies to offer their viewpoints,
insights, and experience on various facets of
chromatography column qualification.

ROBERT M. KENNEDY
Qualifying Protocols and Hardware
Qualification is the
part of process validation that ensures
that the equipment
and protocols are capable of fulfilling the
specified requirements. Column qualification is documentation of work done
to ensure that the column and the protocols written to direct operation of the column are appropriate for the purification step intended. Column
qualification can be divided into two broad sub-

Measuring Column Bed Integrity

After a column is packed, integrity of the column bed needs to
be measured to confirm the quality and consistency of the
chromatographic operations. Several measures are commonly
used for this purpose.

(a)

tr

Number of Plates (N)

h
W1/2

To calculate the number of theoretical plates for a column, use:

N = 16

tr
Wb

1/2 h

Injection

where tr is the retention time of the probe molecule, and Wb is

the width of the peak at the baseline.

(b)

Alternatively, use
N = 5.54

tr
W 1/2

1/10 h

Injection

in which W1/2 is the width of the peak at one half the maximum
height (figure a). Both of these equations are used widely for
estimating N and yield very similar results.
For comparing two different columns, number of plates per
meter can be calculated using:

Tailing Factor (T)

The tailing factor of the peak can be defined as:

T=

plates
= N
meters
L

in which W0.05 is the width of the probe peak at 5% of full height,

and f is the distance from the leading edge to the midpoint of the
peak.

where L is the total length of the column.

Height Equivalent to the Theoretical Plate (HETP)

This equation also normalizes column performance to the length

of the column, so it can be used to compare performance:

Another equation that uses a variant of HETP is reduced plate

height h, which normalizes HETP for particle diameter dp, and is
defined as:

ject areas: column hardware qualification and

column protocol qualification.
In column hardware qualification, the column size
and manufacturing materials need to be documented. The intended packing material needs to
be associated with a particular column. Resin
qualification and the workflow associated with
resin is another topic to be documented. The location in the plant needs to be established. A
schematic needs to diagram the column and its
associated peripheral equipment. Peripheral
equipment, such as tubing and hoses, pressure
gauges, clamps, and gaskets needs to be documented. Spare parts must be associated with a
particular column as well.

Asymmetry (As)

Asymmetry is defined as:

As = b
a

HETP = L
N

h = HETP
dp

W 0.05
2 f

where a is the distance from the leading edge of the peak to the
midpoint of the peak, and b is the distance from the midpoint of
the peak to the trailing edge (figure b). As is routinely
determined at 5% or 10% of maximum peak height. As values
between 0.8 and 1.4 are typically suitable. As shown in the T
and As equations, As is not the same as peak tailing.

A reference to the process workflow diagram

should call out all the details of an individual column. The column can be defined by its serial
number, its asset property number, or its unit
operation number. The construction materials
should be verified as part of the overall validation
program, and records of those materials need to
be kept in the validation record. Leachables and
extractables from those materials need to be verified as well.
If the column is mobile; the process work diagram should show how the column travels, that
is, where it is stored, where it is packed, how it is
transported, where it is tested, and how it is
removed from service and transported out of the
operations area.
BioPharm International

MARCH 2003

31

The schematic diagram shows all the associated lines into and out of the column. The process
workflow diagram indicates the tanks that are
connected to the inlet side and which tanks are
connected to the outlet. Product is collected into
one of the tanks on the outlet. People responsible
for the process need to know the location and the
condition of the product at all times. Qualification of these tanks and fluid path transfers is
another part of hardware qualification. Included
in the validation package for the column is the
engineering drawing of the column. Spare parts
are associated with an individual column, and
engineers performing the scheduled maintenance
on the column can be expected to require access
to this part of the validation record to ensure that
the appropriate spare parts are used in maintenance and reordered once they are consumed.
Documentation of the gauges used to operate a
column should include the calibration record of
the gauge as well as the serial numbers or asset
number to ensure that the correct gauge is in
place during operation.
In column protocol qualification, process development work can contribute to the validation of the
chromatography step by providing data on the
stability of the chromatography resin under equilibration, operating, regenerating, cleaning and
sanitizing, and storage conditions. To operate the
column, an associated set of standard operation
protocols (SOPs) needs to be written and tested:
SOPs must be unambiguous. Regulatory requirements exist to ensure that operators are trained
and regularly retrained to maintain their competence in performing the operation.
Each column in the purification process needs
to have SOPs for column packing and testing.
Once accepted for use, each column needs SOPs
for equilibration, running (that is, applying feed
to a collection of product), regenerating, and
cleaning. SOPs are needed to direct the collection
of data for use in ascertaining performance quality, to document how preventive maintenance
should be done, and to instruct on how material
should be stored.
Assistance with writing many of these SOPs is
often available from the appropriate equipment
and chemical suppliers. The suppliers technical
department can be a valuable contributor to inhouse writers of these protocols. Chromatography
suppliers can contribute to SOPs on column
packing, performance quality, maintenance, and
storage, for example. Engineering companies can

Elements of
Biopharmaceutical
Production

32

BioPharm International

MARCH 2003

provide assistance with the process workflow diagrams and advice on data storage related to the
physical column and the drawings and parts.
Purification by chromatography is at the heart
of biopharmaceutical production. Column qualification is a critical activity in the assurance of
quality in the production area.
Robert M. Kennedy is department director, separation
sciences, Amersham Biosciences Corporation, 800 Centennial Avenue, Piscataway, NJ 08855, 732.457.8438,
[email protected].

J. KEVIN ODONNELL
The Science of Column Packing
Recent FDA inspections have shown an increased
interest in the packing
characteristics of
process-scale, liquid
chromatography
columns. That is manifested by the
agencys emphasis on
reproducible packing
of chromatography
columns by pharmaceutical and biotechnology companies. Although
process-scale column packing is still, in many
cases, more of an art than a science, FDA by its
inquiries at inspections is encouraging companies to move toward a column-packing science. A
large part of that science is how a column is characterized or qualified after it is packed.
Resins. Many variables influence the science of
column packing. The first variable is the chromatographic resin. A chromatographic resin is
selected primarily for its ability to separate the
molecule of interest from its impurities, not on
how well it packs into a manufacturers column.
Many different resins are used in process chromatography, and all have individual columnpacking characteristics.
Some resins are soft and cannot withstand
even moderate pressures. As a result, when
columns are packed with this group of resins,
2030% more resin often has to be added to get
the desired column volume at standard operating
pressures. Semirigid resins, including those based
on silica, are better able to withstand normal
operating pressures and require only 515% extra
resin to get to the specified volume. Silica resins
are pressure resistant, so they can be used in
columns with extremely high backpressures.
Silica resins require little additional resin for efficient column packing.

Resin slurry concentration is often overlooked when packing process-scale columns. For
unknown reasons, resin can pack differently
when charged to the column at different concentrations. A slurry concentration that works in one
column may not work as well in another manufacturers column. Resin manufacturers generally
have a good idea of what slurry concentration
works best with different column technologies.
Column packing. Traditional columns with
adjustable upper adaptors can be packed reproducibly with repeated practice. The caveat here is
that it may take several attempts to get the first
optimal pack. Once the technique is determined,
subsequent packing can proceed more efficiently.
Newer process-scale column designs, which
include dynamic axial compression and selfpacking technologies, are generally more user
friendly and can be packed successfully on the
first try. However these columns require more
up-front training, and the hardware is usually
more expensive because it includes additional
equipment specially designed to assist in column
packing. Expanded-bed columns may be the easiest to pack; however they also require more supervision while running.
Ideally, all resins should be defined before
packing, eliminating any small fragments that can
be generated by shipping and handling resin.
Trace amounts of fines may not have any effect
on the actual chromatography, but can eventually
lead to occlusion of the bottom screens or frits.
Such occlusion would ultimately create increased
backpressure, which would require a reduced
flow rate. As a result, the column would have a
reduced throughput.
The mobile phase plays an important role in
packing process columns. Generally the solvent
that packs the resin most tightly is preferred.
Placing and securing the upper adaptor directly
on the top of the settled bed minimizes swelling
when the mobile phase is changed. Exceptions to
this method include some polymeric reversedphase resins that can swell 1525% in organic
solvents. Under those conditions, a fixed adaptor
or column housing could fail, with resulting obvious safety concerns. Allowing a void above the
resin bed or using a dynamic axial compression
column that moves with the shrinking and
swelling of the resin prevents that.
Test result variations. Once a column is packed,
a series of standard tests should be conducted to
evaluate performance and qualify the column.
The results of these tests do not predict success in
the actual chromatography step; however, the
results are useful for column packing repro-

Troubleshooting
Performance Evaluation
As
0.8

Overpacking the column

Packing at too high of a pressure
Column bed cracking
As1.4

Column not packed tight enough

Clogged screens or frits at top or bottom of
the column
Air pockets in column hardware void spaces
Poor injection technique
High HETP

Injection sample or detector too far from

column
Column not packed efficiently
Low HETP

Probe molecule retained on column because

of interaction with functional group or
backbone

ducibility. It is this reproducibility that most

interests FDA.
Usually a small, unretained probe molecule is
used for standard column tests. Most chromatographers use a UV-absorbing molecule such as
p-aminobenzoic acid (PABA) or acetone.
Alternatively, concentrated sodium chloride
spikes are monitored using a conductivity detector. Many more companies use sodium chloride
because it is almost always used in the mobile
phase of applications that dont use reversedphase chromatography (RPC).
Injection volume becomes more important as
particle size decreases. A 0.1% column volume
injection is a good starting point. The flow rate
for such a test should be 6080 cm/hr for best results. Data can be analyzed by computer if
recorded digitally, or if not, data can be calculated by hand.
The location of the injection loop is important and should be located as close to the column
inlet as possible. The detector at the bottom of the
column should also be located as close to the column as possible. Sometimes neither suggestion is
possible. In such cases, column performance will
be noticeably less than what is observed at bench
scale. That might not be interpreted as a negative
result as long as the root cause is known, and the
results are consistent. In future column-packing
tests, similar results must be observed.
Fronting can occur if sodium chloride is
injected onto an ion-exchange resin with a water
mobile phase. Adding salt to the mobile phase
BioPharm International

MARCH 2003

33

(0.1 M NaCl, for example) usually eliminates that

problem. With hydrophobic-interaction chromatography (HIC) or RPC resins, using acetone
or PABA can generate tailing or higher-than-normal height equivalent to a theoretical plate
(HETPs) because of interaction with the functional group or resin backbone. Adding small
amounts of organic solvent (1020%) to the mobile phase eliminates that interaction (see Table 1
for troubleshooting performance evaluations).
Operators on the manufacturing floor should
know how column performance affects the separation of the target molecule. In some cases, lessthan-optimal column performance values may be
tolerated because of the columns function. For
instance, higher HETPs may be acceptable when
the column is used for a simple capture step with
a step elution. Tailings may also be acceptable if
no closely eluting impurities are found and the
extra volume is handled easily in subsequent unit
operations.
FDA is now more knowledgeable about operations using process-scale column chromatography. It is now incumbent on pharmaceutical and
biotechnology companies to address FDA
inquiries with more appropriate scientific
responses. One of the ways to do this is through
reproducible column packing as indicated by
comparing the performance of one column pack
to another. That, in turn, helps to predict repeated
success in purifying the target molecule.

Elements of
Biopharmaceutical
Production

J. Kevin ODonnell is the technical service manager at

Tosoh Biosciences LLC, 156 Keystone Drive, Montgomeryville, PA 18936, [email protected].

IVARS BEMBERIS
Choosing the Right Metric
To ensure robust
chromatographic performance, it is essential to have a proper
packing and qualification protocol. Because chromatographic purification
typically involves
several columns,
packing and qualifying columns is an important
process validation concern. An effective packing
and qualification protocol typically consists of a
prepacking checklist, a description of the medias
properties, choosing the right metric for column
qualification, and qualifying the packing protocol.
Prepacking checklist. A columns mechanical
robustness must be ensured before packing. That
34

BioPharm International

MARCH 2003

should include an inspection of all column seals,

bed supports, and bolts to ensure they are fit for
use. Hydrophobic bed supports, such as polyethylene, must be prewetted and checked for uniform wetting. Before packing, the column should
be assembled and pressure tested with water-forinjection (WFI) to a pressure in excess of the
highest operating pressure anticipated during the
process.
Media properties. Understanding the flow and
pressure characteristics is essential to reliable
packing. Determining the compressibility of the
media to be packed is most critical to ensuring
that the proper amount of media is packed into
the column. As always, the conditions of packing
must exceed those to be experienced during the
process to ensure bed stability and an extended
cycle life for the packed column.
Choosing the metric. An appropriate metric must
be chosen for qualifying the column. Conventional practice is to measure the HETP and the
asymmetry (A s). However, the usefulness of a
method can be realized only if the method can be
repeated with sufficient precision to allow it to be
used to compare the packing performance over a
period of time. HETP and As test results can be
applied in many ways to aid in the overall control
of the chromatographic process, from initial
packing qualification to use as a periodic inprocess check of bed stability. HETP and As tests
are, with increasing frequency, used to verify bed
integrity throughout the service life of a packed
column. In the biotechnology industry, where
processes are frequently campaigned, the use of
HETP and As testing to requalify a column coming out of storage has particular merit.
Choosing the marker. Usually, pulse injection
of a nonreactive marker is used to qualify a column. The marker must be chosen carefully to
ensure that it is representative of the process and
because it affects qualification. The marker can
be an internal standard, selected to mimic the
protein to be processed. Or the marker can be a
generalized, external standard, such as acetone,
NaCl, or NaOH.
Pulse injection is the favored packing method
because it uses a small injection volume. However, because precise, nondiluted injections are
needed, that small volume places a great burden
on the test. The injection method must be refined
to ensure it is precise, repeatable by many operators, and includes this desideratum: It is strongly
suggested that the qualification test be carried
out using an automated skid.

Product Concentration

100

50

Percent

dC/dV

0
Feed Volume

Figure 1. Frontal analysis curve and calculated peak

36

BioPharm International

MARCH 2003

Pulse injection tests are intended for assessing

the uniformity of the packed bed, not for assessing aspects of the hold-up volume of the HETP
test rig, the inlet, or the flow cell. Therefore, the
injection volume should be sufficiently large so
that it remains undiluted by such hold-up volumes. Generally, media vendors suggest injection
volumes of 12.5% of the packed-bed volume.
For consistency, Millipore suggests that the volume be fixed at 1% of the total column volume.
For a column with a tube height of 50 cm, the
injection volume should be equivalent to the volume contained in 0.5 cm or 5 mm of the column
to be tested. (For a column with a 45-cm diameter
and a 50-cm height, the injection volume should
be 750 mL.)
In its purest sense, the qualification procedure
is really an integrity test of the packed bed.
Postpacking testing is advised as a means to periodically validate column integrity. The repeatability of HETP and As data over the life of the
packed column gives assurance that the packed
bed remains stable and the column continues to
be suitable for service.
An alternate metric for column qualification. In normal chromatographic operations, the
column is exposed to a variety of buffers during
equilibration, loading, washing, eluting, regenerating, and cleaning. During these steps, the
column undergoes a number of buffer exchanges.
Data from these steps can also be used for column qualification. The process is similar to the
frontal analysis used for determining media
capacity. However as a qualification technique,
data from buffer exchanges have the advantage of
requiring no process piping changes to run the
test. So the potential for operator-induced variability is greatly reduced.

Frontal curves and pulse injection peaks dont

look the same, but they can provide precisely the
same data. A uniform frontal curve indicates that
the packing is of high quality, the same as a narrow, symmetrical injection peak. The frontal peak
data can be manipulated to compare it with a
pulse injection peak. Taking the first derivative of
the frontal curve produces the same curve as its
analogous injection peak (see Figure 1).
Qualification of the packing protocol. A packing
protocol must be rigorous to ensure good packing. Figure 2 shows the development of a packing
protocol for Sephacryl S-200 HR resin a gel
filtration resin, which is a challenge to consistently pack well. The acceptance criteria required
that Millipore demonstrate five repeat packings
(in an IsoPak 630-mm column) meeting the performance specification: HETP  0.02 cm and
As  0.81.2. Client operators would then use
the same packing protocol to pack the column to
within the performance specification. The data in
Figure 2 confirm that the seven packings all met
the qualification (and process) specifications.
A rigorous packing protocol ensures good
packing, which in turn ensures that columns can
be well packed by any capable operator. Variability in packing methods must be understood, and
the test method must be properly set to the qualification specifications. Some conservatism is suggested in finalizing the specifications because
testing may be conducted on initial packing as
well as throughout the service lifetime of the column and gel. Robust packing methods make
qualification test results more consistent and
reproducible.
Ivars

Bemberis

is

technology

manager,

chromatography, Millipore Corporation, 80 Ashby

Road,

Bedford,

MA

01730,

781.533.8452,

fax 781.533.3110, [email protected].

OLIVER KALTENBRUNNER
Preventing Erroneous Results
Chromatography
column packing integrity is usually
tested before process
use of a column.
Typically, a pulse of
a noninteracting sample tracer is applied
while flowing a column signal (which
alters the initially rectangular tracer) under defined conditions. An analysis of the response signal of the pulse sample can be used to assess

1.4

0.020

HETP (cm)

1.2
0.015

1
0.8

0.010

HETP (cm)
As

0.6
0.4

0.005

0.2
0.000
3

4
Pa
ck
5
Av
er
ag
e
R
un
1
R
un
2

Pa
ck

Pa
ck

Pa
ck

Pa
ck

Packing Number

Average Results:
HETP  0.014 cr
As
 1.082

Figure 2. Summary of Sephacryl S-200 HR packing qualification

packing quality by determining the number of
theoretical plates (N) of the column. N can be
interpreted as the signal-to-noise ratio of the
retention time, that is, the ratio between retention
time and the peak width of the tracer pulse (see
Figure 3). By relating N to column height, the
HETP can be calculated and compared to the
resin bead diameter to calculate the reduced plate
height h. Chromatographic theory based on the
van Deemter equation (2) provides a clear interpretation for the meaning of h as
h = A + B + C Pe
Pe
where h is influenced by the quality of the packing (A), diffusivity of the tracer molecule (B), and
the mass transfer characteristics of the resin
beads (C). For column qualification, test conditions have to be chosen to maximize the contribution of the packing quality and minimize the contribution of diffusivity and mass transfer
resistance.
Making diffusivity invariable. To optimize column packing for qualification, the A term needs
to be maximized by minimizing the B and C
terms. Whereas in typical protein chromatography the B term can be ignored because the diffusion of macromolecules is insignificantly small
compared to the other contributions, the same is
not true for small molecule tracers. As demonstrated by Knox, h has its minimum value at
Pe  5. Because h grows quickly at Pe
5 and
increases gradually at Pe  5, the flow rate region between 5
Pe
10 gives the most stable
results for HETP testing (4). The definition of the
flow rate range based on the Pclet number indicates that, for reproducible and transferable test
results, the diffusivity of the tracer molecule has
to be invariable.

38

BioPharm International

MARCH 2003

Additional information can be gathered when

analyzing the A s of the peak. Typically, a
fronting peak indicates a compression that is too
low (with an increased risk of column channeling), whereas peak tailing indicates excessive
packing compression. However, interpretation of
As is normally not that simple because most extra
column equipment introduces exponential
washout behavior that also manifests as peak tailing. Nevertheless, for a given column in a defined
test setup, a change in peak As can typically be
traced back to over or under compression of the
packing.
Accounting for broadening. In large-scale
chromatography, less-than-ideal theoretical
HETP results are typically accepted because the
chromatographic setup in a manufacturing plant
is designed to handle larger volumes and flow
rates than were used for the column qualification
test. The broadening introduced by switching
valves and the monitoring unit can be significant
for an integrity test. To assess a specific equipment setup, a UV detector can be put before the
monitoring assembly. The alteration of a UV
transition signal on its pass through the assembly
can be used for qualifying the monitoring setup.
In our measurements, typically the apparent mixing volume of an assembly is between 20% and
40% of its actual volume. The volume of a
3
/4-inch OD monitoring assembly, including a
column-switching valve, can easily exceed
200 mL with about 50 mL of apparent mixing
volume. With a simple estimate of retention for a
nonretained small tracer molecule of about
0.95 column volumes, and an ideal HETP of
about 5 particle diameters, we can easily estimate
the expected band broadening () in the packed
bed:
2
2

Column 2 =

VR
=
N

2
d
4

L 0.95
L
5 dp

Comparing this column broadening to external

broadening of the system reveals that, in many
cases, the monitoring equipment used for
preparative runs significantly affects the column
qualification result. Ignoring this aspect can
cause problems when setting column acceptance
criteria during process transfers. Difficulties are
typically exacerbated when packing with a resin
with a small particle size, for which the packing
quality process requirements are higher and, at
the same time, the relative negative effect of the

Elements of
Biopharmaceutical
Production

Corresponding author and series

coordinator Anurag S. Rathore is at
Pharmacia Corporation, GG3K, 700
Chesterfield Parkway North,
Chesterfield, MO 63017,
[email protected];
Robert M. Kennedy is department
director in separation sciences at
Amersham Biosciences Corporation,
J. Kevin ODonnell is technical
service manager at Tosoh Biosciences
LLC, Ivars Bemberis is technology
manager at Millipore Corporation,
and Oliver Kaltenbrunner is in
process engineering at Amgen Inc.

40

BioPharm International

MARCH 2003

setup is also higher. Obviously the tracer application technique can have an even bigger effect on
the test result, and a defined and reproducible
sample application method or device needs to be
established.
Preventing erroneous conclusions. Once
reproducible test methods are in place, HETP
results can be used to accept or reject columns for
processing. Acceptance criteria should be set primarily by considering the best practice for column packing and the available equipment, then
establishing criteria that are specific to that system. Consequently, there is not necessarily a
close correlation between a columns qualification test result and its process performance.
However, as in any acceptance test, there are two
possible erroneous conclusions.
The most likely error is that a column fails the
test criteria but would not fail to meet process
requirements. This is the typical error to expect if
the testing operation failed to follow the required
best practices. It causes an operational problem of
producing extra work and unnecessary repacking,
but it does not set the process at risk. Setting
appropriate acceptance criteria, reproducible test
methods, and operator training can minimize the
risk.
The second type of erroneous result is to
accept a column that is not fit for the process.
This obviously sets the process at risk and needs
to be avoided. Reasons for this error can be column failure occurring at a particular flow rate, or
viscosity of a process feed while the test conditions allow good packing integrity.
Bed stability is not only a function of the bed
height and compression, but also of the height-todiameter ratio of the packing. Identical linear velocities cause increasing bed compression with
increasing column diameter. This leads to decreasing maximum operating linear velocities,
with increasing column diameter (5). If the
process flow rates are developed using small column diameters without consideration of potential
flow rate limitations at large column diameters,
bed instabilities can occur that are not evident
during typical column qualification conditions.
Consequently, test conditions have to be evaluated column by column, considering the process
conditions at the respective scale.
Oliver Kaltenbrunner is in process engineering,
MS: 30W-2-A, Amgen Inc., One Amgen Center Drive,
Thousands

Oaks,

CA

91320,

fax 805.499.5008, [email protected]

805.447.2641,

ANURAG RATHORE
From the Series Coordinator
BioPharm International and I hope that
the insightful comments from these authors are useful for biochemists and engineers
practicing preparative
chromatography in the
biopharmaceutical industry. We conclude
this article with a caution about some common
pitfalls found in preparative chromatography:
Packing procedures need to be optimized for
each different type of resin to get reproducible
packing;
Measures of bed integrity need to be carefully
chosen along with their specifications to
guarantee robust column performance;
The tracer used for estimating HETP or As can
interact with column packing, yielding
erroneous results;
Flow rate or particle size effects must be
eliminated when qualifying column packing to
achieve consistent results;
Function tests might be necessary, and lot-to-lot
variations in chromatography media may be
unacceptable for some applications if the
process is very sensitive; and
Resin qualification procedures need to be
carefully chosen because excessive testing can
be expensive and should be reduced gradually
through reexamination.
Anurag Rathore , series coordinator, is at
Pharmacia Corporation, 700 Chesterfield Parkway North,
Chesterfield, MO 63017, 636.737.6790, fax 636.737.7281,
[email protected]. BPI

REFERENCES
(1) Rathore, A.S. et al., Process Validation: How Much
to Do and When to Do It, BioPharm 15(10), 1828
(October 2002).
(2) Rathore, A.S. and Velayudhan, A., Guidelines for
Optimization and Scale-Up in Preparative
Chromatography, BioPharm International 16(1),
3442 (January 2003).
(3) Van Deemter, J.J., Zuiderweg, F.J., and Klinkenberg,
A., Longitudinal Diffusion and Resistances to Mass
Transfer as Causes of Nonideality in
Chromatography, Chem. Eng. Sci. 5, 271289
(1956).
(4) Kaltenbrunner, O., Watler, P., and Yamamoto, S.,
Column Qualification in Process Ion-Exchange
Chromatography, Progress in Biotechnology:
Bioseparation Engineering, Vol. 16, I. Endo et al.
Eds. (Elsevier Science, Amsterdam, The Netherlands,
2000), pp. 201206.
(5) Stickel, J.J. and Fotopoulos, A., PressureFlow
Relationships for Packed Beds of Compressible
Chromatography Media at Laboratory and
Production Scale, Biotechnol. Prog. 17(4), 744751
(2001).