Cytometry Part B (Clinical Cytometry) 82B:917 (2012)

ISHAGE Protocol: Are We Doing It Correctly?

Alison Whitby,
1
Liam Whitby,
1
Matthew Fletcher,
1
John T. Reilly,
1
D. Robert Sutherland,
2
Michael Keeney,
3
and David Barnett
1
*
1
UK NEQAS for Leucocyte Immunophenotyping, Department of Haematology, Royal Hallamshire Hospital,
Shefeld S10 2QN
2
Department of Laboratory Hematology, University Health Network, Toronto, Canada
3
Hematology/Flow Cytometry, London Health Sciences Centre, London, Canada
Background: Flow cytometric CD34
1
stem cell enumeration is routinely performed to optimize timing
of peripheral blood stem cell collections and assess engraftment capability of the apheresis product.
While a number of different ow methodologies have been described, the highly standardized ISHAGE
protocol is currently the most widely employed, with 204/255 (81%) international participants in the UK
NEQAS CD34
1
stem cell enumeration program indicating their use of this method. Recently, two labora-
tories were identied as persistent poor performers, a fact attributed to incorrect ISHAGE protocol usage/
setup. This prompted UK NEQAS to question whether other laboratories were making similar errors and,
if so, how this might affect individual EQA performance.
Methods and Results: In send out 0801, where two stabilized samples were issued, the EQA center
surveyed 255 participants with ow analysis data and subsequent results collected. One hundred and
ninety-six laboratories returned results with 103 returning dot plots. Eighty-three out of one hundred and
three stated that they used the ISHAGE protocol gating strategy but 43% (36/83) were incorrectly set-
up. Analysis of the data showed those incorrectly using single platform ISHAGE gating strategy were
twice as likely to fail an EQA exercise compared to those using the protocol correctly. This failure rate
increased two fold when incorrect ISHAGE protocol was used in a dual platform setting.
Conclusion: This study suggests a widespread fundamental lack of understanding of the ISHAGE proto-
col and the need to deploy it correctly, potentially having signicant clinical implications and highlights
the need to monitor participants rigorously in their deployment of the ISHAGE protocol. It is hoped that
once these ndings have been disseminated, performance can be improved. VC 2011 International Clinical
Cytometry Society
Key terms: CD34; stem cells; ISHAGE; quality control; quality assessment; EQA
How to cite this article: Whitby A, Whitby L, Fletcher M, Reilly JT, Sutherland DR, Keeney M, Barnett D. ISHAGE
protocol: Are we doing it correctly? Cytometry Part B 2012; 82B: 917.
Over the last 15 years or so, cytokine-mobilized pe-
ripheral blood stem cells (PBSC) have largely replaced
bone marrow as a source of hematopoietic stem cells
(HSCs) in the majority of autologous and (in an increas-
ing proportion) allogeneic PBSC transplants (1,2). The
HSCs in marrow and peripheral blood, which are re-
sponsible for multilineage engraftment in the transplant
setting express the cell surface marker CD34 (3,4). Flow
cytometric enumeration of CD34

cells provides a rapid

means of measuring this clinically useful surrogate
marker of graft adequacy in all sources of HSCs and
most transplant centers now determine graft adequacy
based on the number of CD34

cells/Kg of patient body

weight. In addition to determining yield, the number of
CD34

cells mobilized to the peripheral blood is also a

predictor of the success of apheresis (5) and can be
used to monitor, on-line, the yield of CD34

cells (6).
*Correspondence to: Dr. David Barnett, Deputy Director and Consul-
tant Clinical Scientist, UK NEQAS for Leucocyte Immunophenotyping,
4th Floor, Pegasus House, 463a Glossop Road, Shefeld S10 2QD,
England.
E-mail: d.barnett@btconnect.com
Received 18 March 2011; Revision 6 July 2011; Accepted 8 July
2011
Published online 19 July 2011 in Wiley Online Library
(wileyonlinelibrary.com).
DOI: 10.1002/cyto.b.20612
Original Article
VC 2011 International Clinical Cytometry Society
Although simple ow methods for counting CD34

cells date back to the late 1980s (713), a standardized

multi-parameter protocol that was based on the struc-
tural characteristics of the CD34 molecule and the epi-
topes detected by various CD34 monoclonal antibodies
did not emerge until 1994 (14). The latter used the maxi-
mum information available of four parameters; forward
and side light scatter and the intensity of CD34 and
CD45 staining and subsequently formed the basis of a
Clinical Guideline for CD34

cell quantitation in periph-

eral blood (PB) and PBSC for the International Society for
Hematotherapy and Graft Engineering (ISHAGE) (15).
Subsequent developments included the incorporation of
a viability dye 7-Amino Actinomycin D (7-AAD) and uo-
rescent counting beads, in a Single Platform (SP) variant
(16). The addition of counting beads eliminates the
potential introduction of errors in calculating the abso-
lute CD34

cell count inherent in two platform (cytome-

ter plus hematology analyzer) methodologies (1618).
The essential components and technical details of the
SP ISHAGE with viability Protocol are embodied in sev-
eral National (19,20), Regional (21), and International
Guidelines (2224). The single platform methodology
has been validated on a wide range of instruments
employing a variety of different types of uorescent
counting beads (25).
The use of SP ow cytometric technology has
improved standardization globally of this procedure
(18). Indeed, using the SP ISHAGE approach, CVs of
<10% have been achieved and data from within the UK
NEQAS for Leucocyte Immunophenotyping CD34

stem
cell enumeration program also indicates that the SP ISH-
AGE protocol consistently produces the lowest inter-lab-
oratory CVs of all the gating strategies used (19,2628).
As a result, this protocol has, for a number of years,
been recommended as the method of choice for labora-
tories in the United Kingdom undertaking CD34

stem
cell analysis (19). Indeed, at the time of this study, 81%
(204/255) of the international participants in our EQA
program use the ISHAGE protocol on a routine basis.
As a requirement for continued accreditation, under
ILAC G13:2000 standards for external quality assessment
program operation, it is mandatory to have a strategy in
place to monitor all UK laboratories for persistent unsat-
isfactory performance (29). This is now required on an
international basis following the introduction of ISO
17043 standards (30). UK NEQAS employs a rolling
cumulative score of three samples to determine the per-
formance of participants in the CD34

stem cell enu-

meration program. During the course of this, two
laboratories were noted to have persistent unsatisfactory
performance that, when investigated further, was shown
to be as a result of the incorrect deployment of the ISH-
AGE protocols gating strategy. This prompted us to
investigate if other laboratories were also making similar
errors and, if so, to assess how these might affect indi-
vidual EQA performance.
We report here the retrospective analysis of UK
NEQAS participants gating strategy within the CD34

stem cell enumeration program for trial 0801. Labora-

tories were invited to submit ow dot plots of their
analyses for further examination and, if they actually
used the ISHAGE protocol, their gating strategies were
compared to those described by the authors of the
original ISHAGE Guidelines (15,16,2225). For those
participants identied as using the ISHAGE protocol
incorrectly, we then compared their performance with
those who were using the ISHAGE gating strategy
correctly.
MATERIALS AND METHODS
The UK NEQAS for Leucocyte Immunophenotyping
CD34

stem cell enumeration program issues two sta-

bilized samples on a bimonthly basis (26). Waste pe-
ripheral blood samples, rich in CD34

stem cells,
were obtained from consented patients who had
undergone stem cell apheresis following stem cell mo-
bilization. Samples were stabilized, using a procedure
previously described (26,31), and added to stabilized
blood to achieve CD34

stem cells counts routinely

found prior to apheresis. Following stabilization, sam-
ples remain stable for up to 6 months at room tem-
perature or up to 3 years at 4

C and are compatible

with whole blood lysis techniques, multiparameter
analysis and sequential gating strategies that are rou-
tinely used in CD34

stem cell enumeration (26,31).

All 255 participants were issued with two samples
and requested to submit their absolute (cells/ll) and
percentage CD34

stem cell counts for each sample

for subsequent data analysis and performance assess-
ment. Flow analysis dot plots and gating strategies
provided by participants were then compared to the
originally published protocols. Since the ISHAGE gat-
ing protocol was the most commonly used, this study
was focused on this group.
Returned absolute count results were assigned a score
between 0 and 50 depending on the values in relation-
ship to the median and centile values (5th, 10th, 25th,
75th, 90th, and 95th centiles). The score increases the
further a value is from the median (e.g., 0 points are
assigned to results falling between the 25th and 75th
centile; 20 points for those between the 10th and 25th
centile, or 75th and 90th centile; 35 points for those
between the 5th and 10th centiles, or the 90th and 95th
centiles; 50 points if the participants result exceeds the
Table 1
Gating Strategies Used by all Participants in the UK
NEQAS for Leucocyte Immunophenotyping CD34

Stem
Cell Enumeration program, Compared to Those Used by the
103 Participants That Returned Dot Plots for the Study
Gating strategy Study participants (%) All particpants (%)
ISHAGE 83 81
STEMKIT 4 8
PROCOUNT 5 4
MILAN 5 6
BENDER 1 1
10 WHITBY ET AL.
Cytometry Part B: Clinical Cytometry
5th or 95th centiles). A rolling window of three samples
is used to monitor participant performance and cumula-
tive scores 100 are classied as unsatisfactory perform-
ance and remedial action is undertaken. A performance
score below 100 is deemed satisfactory. CD34

absolute
counts from participants using ISHAGE methodologies
were analyzed further with those using the protocol cor-
rectly being compared to those using the protocol incor-
rectly. Furthermore, single platform users (the predicate
method) were compared to dual platform users.
FIG. 1. Application of ISHAGE protocol gating strategy for single platform use as detailed in Sutherland et al. (24,25). This shows the correct
placement of each region when using a BD FACSCanto II ow cytometer. It also shows the most frequently omitted region P5 in the rst dot plot.
Note that the viability gate P8 must be fully opened as shown in plots 7 and 8 for the analysis of stabilized samples regardless of reagent set or cy-
tometer type used. Additionally, the duplicate lymph-blast region P4 (plots 4 and 6) is slightly extended in the SSC dimension to accommodate the
stabilized CD34

cells that lose a little forward scatter and gain a little side scatter. Also note for TruCount-based ISHAGE protocols, debris gate P7
is established to remove uorescent debris that contaminates the CD45 gate P1 (plot 1). If this debris is not removed either during acquisition or
during analysis, an incorrect absolute CD45 count will be obtained (25).
ISHAGE PROTOCOL: ARE WE DOING IT CORRECTLY? 11
Cytometry Part B: Clinical Cytometry
RESULTS
Of the 255 participant laboratories in trial 0801 (sam-
ples 128 and 129), 103 (40%) returned ow dot plots as
requested. Each set of plots was examined and the gat-
ing strategy used determined. Table 1 details the gating
strategies used by the participants that returned dot
plots compared with all participants in the program. It
should be noted that Table 1 shows the separation of
the ISHAGE protocol gating strategy users and the users
of Beckman Coulters Stem-Kit and the Becton Dickin-
son Procount Progenitor Cell Enumeration Kit. This
study is based on ISHAGE users who have used manual
acquisition and analysis protocols only. Of those that
returned dot plots, 81% (83/103) stated that they used
the ISHAGE protocol gating strategy. These were then
subject to further analysis to ascertain if the gating strat-
egy was being applied correctly. We identied that 47/
83 (57%) participants had applied the ISHAGE protocol
gating strategy correctly. Figure 1 shows the correct ISH-
AGE protocol gating strategy using BD Biosciences Tru-
Count tube-based ISHAGE protocol (25) on the
FACSCanto 2 instrument. Figure 2 shows the same sam-
ple stained with Beckman Coulters Stem-Kit-based ISH-
AGE protocol on an FC500 cytometer using manual data
acquisition and analysis (16). Note the inclusion of the
extra viability plot in both Figure 1 plot 7 and Figure 2
plot 9. This plot displays all CD34

events and is of
use in samples potentially containing dead cells such as
post thawed samples etc (16,24,25). It is important to
note that viability testing is not of use in the analysis of
FIG. 2. Application of ISHAGE Protocol Gating Strategy for single platform using Stem-Kit
TM
as detailed in Keeney and coworkers (16,23,24). This
shows the correct placement of each region when using a Beckman-Coulter FC500 ow cytometer. It also shows the most frequently omitted region
Ly in plot 1. Data acquired and analysed manually to optimize gate setting for stabilized cells. Note that the viability gate J (plot 7) is opened as per
Figure 1 so that none of the stabilized cells that stain with 7-AAD are excluded. Note also that duplicate lymph-blast regions D and E (plots 4 and
6, respectively) are extended in the SSC dimension to accommodate the altered light scatter characteristics of stabilized CD34

cells compared to
fresh ones.
12 WHITBY ET AL.
Cytometry Part B: Clinical Cytometry
stabilized cells owing to the fact that the stabilization
process causes the permeabilization of the cell mem-
brane allowing for the uptake of the viability dye giving
the appearance of nonviable cells.
Thirty-six out of eighty-three (43%) participants were
identied as using the ISHAGE protocol gating strategy
incorrectly (Table 2). The most common error, seen in
31% (26/83) was the omission of the lymphocyte-gating
region (designated P5 in plot 1 of Fig. 1 and region
Ly of plot 1 Fig. 2). The lymphocytes gated in P5 are
used to place optimally the lymph-blast region P4 on
plot 6 and its linked duplicate P4 region on plot 4 of
Figure 1. Similarly, for the FC500 template (Fig. 2) the
lymphocytes gated from Ly are used to set optimally
the lymph-blast region E (plot 6) and it is linked dupli-
cate region D (plot 4). The lymph-blast regions P4 and
E/D are drawn and set in this manner because CD34

cells from all sources of HSCs including bone marrow

cluster within this gate when it is drawn correctly
(15,16,2225). Thus, bona de CD34

cells that fulll

the Boolean gating criteria of P1, P2, and P3 (Fig. 1), or
A, B, and C, (Fig. 2) are displayed on a FSC/SSC plot
to conrm that these events fall into the generic
lymph-blast region P4 (Fig. 1 plot 4) or region D
(Fig. 2 plot 4).
Six percent of laboratories (5/83) who claimed they
were using the ISHAGE protocol gating strategy cor-
rectly had omitted plot 3 in which all CD34

events
from region P2 (or B) on plot 2 are clustered by back-
scattering to CD45 versus side scatter. This gate is criti-
cal to the proper deployment of the ISHAGE protocols
as it allows the delineation of bona de CD34

cells
that characteristically cluster on this plot from other
events that do not. Failure to deploy this gate would
generally lead to a falsely high CD34

cell count, as
nonspecically stained events would not be removed.
Interestingly, the ve laboratories that omitted plot 3
entirely also omitted the P5 (or Ly) region on plot 1 as
detailed above. A further two participants stated that
they were using the ISHAGE protocol gating strategy
but had slightly altered the protocol. For example, one
laboratory used a CD45/CD34 plot to gate on CD45
dim
events instead of using the CD45/SSC plot (Fig. 3).
As a result, a signicant number of CD34

events were
missed.
Another participant incorporated anti-CD3 into the
gating strategy to identify the lymphocyte region and
used this to establish their region P7 (Fig. 4). However,
the latter was drawn around the light scatter characteris-
tics of the CD3

lymphocytes (i.e., a duplicate of region

P6 on plot 6), rather than being drawn as a generic
lymph-blast region as shown on plot 4 of Figures 1 and
2 to include all types of CD34

cells, as described
above. Because of this modication some CD34

cells
were excluded to the right of region P7 (Fig. 4 plot 4).
Two participants who submitted their dot plots stated
that they were using the ISHAGE protocol gating strat-
egy but appeared to be using a modied Bender (9)
type protocol (Fig. 5). The Bender protocol (9) involves
measuring CD34

cells against a denominator of CD45

events. Finally, one laboratory claimed they were using

the ISHAGE protocol gating strategy but examination of
their plots showed that it did not resemble any previ-
ously published gating strategy (ISHAGE, Bender, Milan
etc.) and seemed to be a protocol they had devised in-
house.
We then examined how incorrect use of the ISH-
AGE protocol gating strategy would impact EQA per-
formance. Generally, the idea of EQA is to ensure
continued improvement, and thus it is important that
any monitoring system constantly has 5% of users that
would fall outside consensus. In this situation, our
existing scoring system uses results from all ISHAGE
single platform users (whether they are using this
incorrectly or correctly) to generate the centiles as
described above. Using this approach, we identied
that 13% had results outside the 5th and 95th centiles
(Table 3). However, we now know that at least 41%
of all users were applying the ISHAGE protocol gating
strategy incorrectly (this value may be higher or lower
as not all users returned ow dot-plots for examina-
tion) thus we recalculated the centiles using only
those laboratories we had identied as using the sin-
gle platform ISHAGE protocol gating strategy correctly
(dened as the predicate) and then re-examined the
other strategies to this predicate method. Comparing
those using single platform ISHAGE correctly to the
new centiles, 11% were outside the 5th and 95th cen-
tiles. However, those using single platform ISHAGE
and dual platform ISHAGE incorrectly had 19% and
22% outside the 5th and 95th centiles equivalent to
being 1.7 and 2 times more likely to fail an EQA exer-
cise, respectively.
DISCUSSION
The ISHAGE protocol, rst described in 1996 (15),
was designed to ensure accurate and reproducible enu-
meration of CD34

stem cells using a Boolean gating

strategy and specically validated anti-CD45 and anti-
CD34 conjugates. It was later modied to incorporate
absolute counting and viability assessment (16). In the
UK NEQAS for Leucocyte Immunophenotyping CD34

Table 2
Results of ISHAGE Gating Strategy Analysis (With
Reference to Fig. 1 and Fig. 2)
Number in
group (%)
Correct application of ISHAGE protocol
gating strategy
47 (57)
Omission of P5 region in P1 region 26 (31)
Omission of Low SSC/CD45 dim plot
(plot 3)
5 (6)
Wrong parameter 1 (1)
Wrong antibody 1 (1)
Nonuse of ISHAGE gating strategy 2 (2)
Nonuse of any previously published
gating strategy
1 (1)
ISHAGE PROTOCOL: ARE WE DOING IT CORRECTLY? 13
Cytometry Part B: Clinical Cytometry
FIG. 4. Example of an incorrect ISHAGE Protocol Gating Strategy set up showing how this laboratory used CD3 plot to identify lymphocytes instead
of CD45/SSC. This laboratory also failed to design properly the lymph-blast region with the backscattered CD3

cells.
FIG. 3. Example of an incorrect ISHAGE Protocol Gating Strategy set up showing how this laboratory used a CD45/CD34 plot to gate on CD45
dim
events instead of using the CD45/SSC plot that resulted in missing CD34

events. This laboratory had poor performance issues that were resolved
once ISHAGE Protocol Gating Strategy was applied.
stem cell enumeration program, the ISHAGE protocol is
reported to be used by at least 81% of participants
routinely. We have previously reported that a single plat-
form ISHAGE approach would enable greater standardi-
zation and lower inter-laboratory variation (18,26,28).
During the course of our EQA send outs, we identied
two laboratories that were classied as persistent unsat-
isfactory performers and required remedial action. Inter-
estingly, both these sites had the same systematic errors
in the set up of the ISHAGE protocol gating strategy,
and both sites had these protocols established by the
same individual. Once these errors had been rectied,
FIG. 5. Example of an incorrect ISHAGE Protocol Gating Strategy set up showing that the participant was actually using a Bender type gating
strategy (9) and not the ISHAGE Protocol Gating Strategy as claimed.
Table 3
Effect of Gating Strategy on Participant Performance (See text for description)
Scoring method
Participant
method
Percentage
in group
scoring
50 points
Percentage
in group scoring
35 points
Percentage
in group
scoring
20 points
Percentage
in group
scoring
0 points
All Users All 13 10 29 48
Correct ISHAGESingle
platform
All 18 12 27 43
Correct ISHAGESingle
platform
Correct ISHAGESingle
platform
11 11 31 47
Correct ISHAGESingle
platform
Incorrect ISHAGESingle
platform
19 12 15 54
Correct ISHAGESingle
platform
Correct ISHAGEDual
platform
10 20 20 50
Correct ISHAGESingle
platform
Incorrect ISHAGEDual
platform
22 0 33 45
ISHAGE PROTOCOL: ARE WE DOING IT CORRECTLY? 15
Cytometry Part B: Clinical Cytometry
the laboratories performances improved. However, this
nding prompted us to survey all participants to estab-
lish if systematic errors had been incorporated by any of
those sites claiming to be following the ISHAGE
protocols.
For send out 0801, we issued two samples to 255 par-
ticipants and asked them to enumerate the CD34

cells.
196 laboratories returned data for samples 128 and 129
that had overall mean CD34

cell counts of 9 and 47

cells/ll, respectively. In addition, each participant was
asked to submit their results, state their gating strategy
used and provide examples of the ow analysis dot plots
showing the gating strategy for each sample. A total of
103 laboratories returned dot plots (53% of those who
submitted results) that allowed us to undertake detailed
analysis of how data was affected by the gating strategy
employed. Of these, 83/103 (81%) stated they used the
ISHAGE protocol gating strategy. However, examination
of the submitted dot plots showed that 36 (43%) who
believed they were using the ISHAGE protocol gating
strategy had this set up incorrectly. The major error
identied was the omission of the region used to gate
lymphocytes. While it is not necessary to include all
lymphocytes, it is important to include those with the
lowest side scatter, as shown in plot 1 of each gure. It
is the backscatter of these gated lymphocytes that sets
the lower light scatter boundary of the generic lymph-
blast region. When set properly, this region can remove
dead/dying and most apoptotic cells as well as platelet
aggregates that can contaminate some fresh samples
(15,16,2225). Thus, using the ISHAGE Boolean gating
strategy, as originally described, allows the most accu-
rate determination of CD34

cell counts.
A number of laboratories also omitted plot 3 in addi-
tion to failing to dene the lymph-blast region as
described above. The back-scattering of CD34

events
from plot 2 to a CD45 versus side scatter plot is a den-
ing characteristic of the ISHAGE protocol and is used to
delineate bona de CD34

cells that cluster on this plot

from non-specically stained events that do not (14,15).
Other laboratories incorporated additional plots that
were not included in the original ISHAGE gating strategy
at the expense of plots that are required to conform to
this standardized protocol. Yet other laboratories
included additional antibodies, such as anti-CD3. Finally,
there were sites that claimed to be using the ISHAGE
protocol gating strategy but were using methodologies
difcult to categorize.
As the ISHAGE protocol-gating strategy was speci-
cally designed to improve the ease and accuracy of
enumerating CD34

stem cells, it was surprising that

43% of UK NEQAS for Leucocyte Immunophenotyp-
ing CD34

stem cell enumeration participants failed to

implement the protocol correctly. Additionally, an unex-
pected nding of this study was that some laboratories
had introduced additional plots or antibody combina-
tions that were not specied in the ISHAGE protocol.
We thus examined the effect that deploying the ISHAGE
protocol incorrectly would have upon EQA perform-
ance. We found that the incorrect use of the ISHAGE
gating strategy did not signicantly affect the overall CV
of the groups. This may be because of small group sizes,
or the fact that as CV is calculated from the upper and
lower quartile points of a population, that is, the middle
50%, it did not reect any spread outside of this range
and the outliers that were present in the incorrect ISH-
AGE cohort were not reected in the CVs. To investigate
this further, we undertook analysis of performance
within the EQA program, which has been statistically
designed to highlight differences from consensus values.
We examined the performance of laboratories in a
rolling three-sample window. The median and inter-quar-
tile ranges were recalculated for the whole group using
only those laboratories that were correctly applying the
ISHAGE protocol in a single platform approach. All other
results were then compared to this data to obtain the
performance scores. This revealed that those laborato-
ries incorrectly using the ISHAGE protocol gating
strategy, in either a single platform or dual platform
approach, were up to two times (incorrect use of dual
platform ISHAGE) more likely to fail an EQA exercise
when compared to those laboratories that were using
the ISHAGE protocol correctly. This will in turn have
direct clinical implications, because CD34

stem cell
enumeration is used as a trigger level to commence har-
vesting a mobilized patient. If results are underestimated
then the optimal harvest date will be missed and more
mobilization therapy will be required. Overestimation of
results will cause harvests to be performed prematurely,
leading to repeat procedures being required to obtain
sufcient cell numbers for re-engraftment. Both of these
alternatives have direct cost implications for the health
care provider and can be stressful and discomforting for
the patient undergoing treatment. Analysis of the data
showed that whilst laboratories performing the ISHAGE
protocol incorrectly were up to twice as likely to fail an
EQA exercise (as stated earlier), there was no particular
bias associated with incorrect usage of the gating strat-
egy with equal numbers of incorrect laboratories overes-
timating and underestimating the CD34

absolute count.
Following this analysis, several centers that fell into the
incorrect ISHAGE category were contacted directly,
and feedback was received by UK NEQAS that the use
of the incorrect ISHAGE strategy had indeed caused
these situations to occur, which had directly affected
patient care.
We have previously reported that single platform
ISHAGE approach is the best approach for enumerating
CD34

stem cells (18,26) and, when properly deployed,

is not sensitive to instrument platform used or bead
type employed (16,2225). It was also shown that the
use of stabilized samples, targeted training of the single
platform ISHAGE protocol, and strong technical support
led to much-improved inter-laboratory CVs <10% (28).
However, this study shows that this highly standardized
protocol that has been amply demonstrated to work
on a variety of instrument types (15,16,2225) is not
16 WHITBY ET AL.
Cytometry Part B: Clinical Cytometry
being applied correctly across all sites and this should
be addressed.
The enumeration of CD34

stem cells is now routine

practice in many clinical laboratories but training and
education in the use of standardized protocols should
be a requirement for laboratory accreditation for this
critical test. As a result, UK NEQAS for Leucocyte Immu-
nophenotyping re-issued the ISHAGE protocol, which
resulted in one laboratory contacting us to assist them
in establishing correctly the ISHAGE protocol gating
strategy because their performance was unsatisfactory.
However, 1 year on, they are having problems again and
this has been attributed to their deployment of an incor-
rect ISHAGE protocol gating strategy again (data not
shown). We intend to undertake the study again later to
ascertain if continued education and technical support
has improved the correct usage of the ISHAGE protocol
gating strategy.
ACKNOWLEDGMENTS
The authors thank all of the participants in the UK
NEQAS CD34

Stem Cell Enumeration program for their

continued support, and they would specially thank all of
those participants who returned dot plots for review
without whom this data could not have been generated.
LITERATURE CITED
1. Buckner CD. Autologous bone marrow transplants to hematopoi-
etic stem cell support with peripheral blood stem cells: A historical
perspective. J Hematother 1999;8:233236.
2. To LB, Haylock DN, Simmons PJ, Juttner CA. The biology and clini-
cal uses of blood stem cells. Blood 1997;89:22332258.
3. Sutherland DR, Keating A. The CD34 antigen: Structure, biology
and potential clinical applications. J Hematother 1992;1:115129.
4. Lanza R, Healy L, Sutherland DR. Structural and functional features
of the CD34 antigen: An update. J Biol Regul Homeost Agents
2001;15:113.
5. Chapple P, Prince HM, Quinn M, Bertoncello I, Juneja S, Wolf M,
Januszewicz H, Brettell M, Gardyn J, Seymour C, et al. Peripheral
blood CD34 cell count reliably predicts autograft yield. Bone Mar-
row Transplant 1998;22:125130.
6. Luider J, Brown C, Selinger S, Quinlan D, Karlsson L, Ruether D,
Stewart D, Klassen J, Russell JA. Factors inuencing yields of
progenitor cells for allogeneic transplantation: Optimization of
G-CSF dose, day of collection, and duration of leukapheresis. J
Hematother 1997;6:575580.
7. Siena S, Bregni M, Brando B, Ravagnani F, Bonadonna G, Gianni AM.
Circulation of CD34 hematopoietic stem cells in the peripheral
blood of high-dose cyclophosphamide-treated patients: Enhance-
ment by intravenous recombinant human granulocyte-macrophage
colony-stimulating factor. Blood 1989;74:19051914.
8. Siena S, Bregni M, Belli N, Brando B, Ravagnani F, Gandola L, Stern
AC, Lansdorp PM, Bonadonna G, Gianni AM. Flow cytometry for
clinical estimation of circulating hematopoietic progenitors for auto-
logous transplantation in cancer patients. Blood 1991;77:400409.
9. Bender JG, Unverzagt K. Flow cytometric analysis of peripheral
blood stem cells. J Hematother 1993;2:421430.
10. Kreissig C, Kirsch A, Serke S. Characterization and measurement of
CD34-expressing hematopoietic cells. J Hematother 1994;3:263289.
11. Lowdell MW, Bainbridge DR. External quality assurance for CD34
cell enumeration-results of a preliminary national trial. Royal Micro-
scopical Society Clinical Flow Cytometry Group QA Schemes. Bone
Marrow Transplant 1996;17:849853.
12. Johnsen H, Knudsen LM. Nordic ow cytometry standards for
CD34 cell enumeration in blood and leukapheresis products:
Report from the second Nordic workshop. Nordic Stem Cell Labo-
ratory Group (NSCL-G). J Hematother 1996;5:237245.
13. Gratama JW, Kraan J, Levering W, Van Bockstaele DR, Rijkers GT,
Van der Schoot CE. Analysis of variation in results of CD34
hematopoietic progenitor cell enumeration in a multicenter study.
Cytometry 1997;30:109117.
14. Sutherland DR, Keating A, Nayar R, Anania S, Stewart AK.
Sensitive detection and enumeration of CD34 cells in periph-
eral and cord blood by ow cytometry. Exp Hematol
1994;22:10031010.
15. Sutherland DR, Anderson L, Keeney M, Nayar R, Chin-Yee I. The
ISHAGE Guidelines For CD34 Cell Determination By Flow Cytom-
etry. International Society of Hematotherapy and Graft Engineering.
J Hematother 1996;3:213226.
16. Keeney M, Chin-Yee I, Weir K, Popma J, Nayar R, Sutherland DR.
Single platform ow cytometric absolute CD34 cell counts based
on the ISHAGE guidelines. International Society of Hematotherapy
and Graft Engineering. Cytometry 1998;34:6170.
17. Marti GE, Johnsen HE, Sutherland DR, Serke S. A convergence of
methods for a world wide standard for CD34 cell enumeration.
Letter to the Editor. J Hematother 1998;7:105109.
18. Barnett D, Granger V, Whitby L, Storie I, Reilly JT. Absolute CD4
T-lymphocyte and CD34 stem cell counts by single platform ow
cytometry: The way forward. Br J Haematol 1999;106:10591062.
19. Barnett D, Janossy G, Lubenko A, Matutes E, Newland A, Reilly JT.
For and on behalf of the BCSH. Guideline for the ow cytometric
enumeration of CD34 haematopoietic stem cells. Clin Lab Haema-
tol 1999;21:301308.
20. Gutensohn K, Carrero I, Krueger W, Kroeger N, Schafer P, Luede-
mann K, Kuehnl P. Semi-automated ow cytometric analysis of
CD34-expressing hematopoietic cells in peripheral blood progenitor
cell apheresis products. Transfusion 1999;39:12201226.
21. Gratama JW, Orfao A, Barnett D, Brando B, Huber A, Janossy G,
Johnsen HE, Keeney M, Marti G, Preijers F, et al. Flow cytometric
enumeration of CD34 hematopoietic progenitor cells. Cytometry
1998;34:128142.
22. Sutherland DR, Keeney M, Gratama JW.Enumeration of CD34

Hematopoietic Stem and Progenitor Cells. In: Robinson JP,Darzyn-

kiewicz Z,Dean PN,Dressler LG,Rabinovitch PS,Stewart CS,Tanke
HJ,Wheeless LL, editors. Current Protocols in Cytometry. New York:
Wiley; 2003. p6.4.16.4.23.
23. Gratama JW, Kraan J, Keeney M, Mandy F, Sutherland DR, Wood BL.
Enumeration of Immunologically Dened Cell Populations by Flow
Cytometry; Approved GuidelineSecond Edition. Clinical and Labo-
ratory Standards Institute, (formerly NCCLS). Document H42-A2
Volume 27 No.16, published June 2007.
24. Sutherland DR, Keeney M.Enumeration of CD34 cells by Flow
Cytometry. In: Aremen EM,Loper K, editors. Cellular Therapy: Prin-
ciples, Methods and Regulations.An American Association of Blood
Bankers Cell Therapy Technical Manual,Bethesda MD; 2009.
pp538554.
25. Sutherland DR, Nayyar R, Acton E, Giftakis A, Dean S, Mosiman VL.
Comparison of two single-platform ISHAGE-based CD34 enumera-
tion protocols on BD FACSCalibur and FACSCanto cytometers. Cyto-
therapy 2009;11:595605.
26. Barnett D, Granger V, Storie I, Peel J, Pollitt R, Smart T, Reilly JT.
Quality assessment of CD34 stem cell enumeration: Experience of
the United Kingdom National External Quality Assessment Scheme
(UK NEQAS) using a unique stable whole blood preparation. Br J
Haematol 1998;102:553565.
27. Gratama JW, Braakman E, Kraan J, Lankheet P, Levering WH, Van
Den Beemd MW, Van der Schoot CE, Wijermans P, Preijers F. Com-
parison of single and dual-platform assay formats for CD34 haema-
topoietic progenitor cell enumeration. Clin Lab Haematol
1999;21:337346.
28. Barnett D, Granger V, Kraan J, Whitby L, Reilly JT, Papa S, Gratama
JW. Reduction of intra- and interlaboratory variation in CD34
stem cell enumeration using stable test material, standard protocols
and targeted training. CD34 Task Force of the European Working
Group of Clinical Cell Analysis (EWGCCA). Br J Haematol
2000;108:784792.
29. Sierra-Amor RI. Accreditation process of a prociency testing pro-
vider according to ILAC G13: 2000/ISO G43-1: 1999. Clin Biochem
2009;42:306.
30. The International Organization for Standardization. International
Standard ISO/IEC 17043. First edition 2010. Conformity assess-
mentGeneral requirements for prociency testing. Reference num-
ber ISO/IEC 17043: 2010(E).
31. Barnett D, Granger V, Mayr P, Storie I, Wilson GA, Reilly JT. Evalua-
tion of a novel stable whole blood quality control material for lym-
phocyte subset analysis: Results from the UK NEQAS immune
monitoring scheme. Cytometry 1996;26:216222.
ISHAGE PROTOCOL: ARE WE DOING IT CORRECTLY? 17
Cytometry Part B: Clinical Cytometry