Leukemia-associated changes identified by quantitative flow

From www.bloodjournal.org by guest on November 14, 2014. For personal use only.
1996 87: 1162-1169
Leukemia-associated changes identified by quantitative flow
cytometry. IV. CD34 overexpression in acute myelogenous leukemia
M2 with t(8;21)
A Porwit-MacDonald, G Janossy, K Ivory, D Swirsky, R Peters, K Wheatley, H Walker, A Turker,
AH Goldstone and A Burnett
Updated information and services can be found at:
Articles on similar topics can be found in the following Blood collections
Information about reproducing this article in parts or in its entirety may be found online at:
Information about ordering reprints may be found online at:
Information about subscriptions and ASH membership may be found online at:
Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American
Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036.
Copyright 2011 by The American Society of Hematology; all rights reserved.
From www.bloodjournal.org by guest on November 14, 2014. For personal use only.
Leukemia-Associated Changes Identified by Quantitative Flow Cytometry.
IV. CD34 Overexpression in Acute Myelogenous Leukemia M2 With t(8;21)
By Anna Porwit-MacDonald, George Janossy, Kamal Ivory, David Swirsky, Rowayda Peters, Keith Wheatley,
Helen Walker, Alev Turker, Anthony H. Goldstone, and Alan Burnett
During theimmunodiagnosis of517 casesof acute myelogenousleukemia(AML)entered
into the Medical Research
Council (MRC) AMLIO trials, we have observed the CD34
precursor cell antigen more frequently in AML of M2 morphology, especially in the 84% of cases with thet(8;21) chromosomal translocation, than in any other French-AmericanBritish classification group. CD34 expression was then
quantified (using QlFl and Quantum Simply Cellular beads
[Flow CytometryStandards, ResearchTriangle Park, NCI and
CD34+ standard cells). When CD34 antibody-binding capacity (ABC) of normal bone marrow
(BM) precursors and leukemic blasts wascompared, it was shown thatAML M2 cases
with t(8;21) not only had thehighest percentages of CD34+
blasts, but in >80% of CD34+cases the individual blasts
expressed higher than normal levels of CD34 antigen (>60
x I O 3 ABC per cell). In addition, in 73% of this group CD34
antigen wasoverexpressed in an asynchronous combination
with cytoplasmic myeloperoxidase (MPO). Other signs of
asynchrony included high CD34 expression with CD15 andl
or CD56,as well as aberrant combinations of CD13 with
terminal deoxynucleotidyl transferase (TdT) and CD19.
These findings demonstrate that asynchrony is identifiable
in virtually every case of AML with t(8;21), although it does
not always involve the
same antigens. M2 cases with t(8;21),
mostly CD34+, had a 100% remission rate and 71% 5-year
survival rate; other patients with CD34+ or CD34- AML
showed 69% and 84% remission rates and 31% and 36%
5-year survival rates, respectively. Consequently, individual
markers such as CD34 should be interpreted in relation to
other features such as chromosomal changes. Thesesimple
methods, which are well suited t o quantify the expression
of ligands, are a useful contribution t o diagnosis: 60% t o
65% of M2 cases with t(8;21) are rapidly identified byCD34
overexpression alone. This aberration, together with the
other signs ofasynchrony seen at presentation, can be used
t o search for residual leukemia after therapy.
0 1996 b y The American Society of Hematology.
localized Sudan black B and myeloperoxidase (MPO) staining in the Golgiregion in the absence of monocytic involvement. In this early study, the nuclear immaturity amid signs
of cytoplasmic maturation was also noted7-but the techniques were not yetavailable for afulldocumentation of
maturation asynchrony.
However, recent immunologic methods arefully suited to
perform such an analysis for three main reasons: ( l ) a range
of differentiationantigens,bothmembrane-associated
intracellular, are available to investigate stages of myeloid
theprecisesequence of
development in the normalmyeloid lineage'""; (2) these
markers can be studied with monoclonal antibodies (MAbs)
labeled with various fluorochromes in three- and four-color
combinations'","; and (3) expression of these antigens can
be quantified as antibody-binding capacity (ABC) alongthe
clear intentionof
seen onmalignant
cells."'~'' Indeed, previous phenotypic studies on AML M2
with t(8;21) have already pointed to an aberrant CD19 expression''-Is together with interleukin-5 responsiveness'' and
CD56I4.I6and CD34"-" display.
To apply these recent developments, in this study we selected cases of AML M2 with t(8;21) to analyze precursor
cell-associated markers such as CD34 or terminal deoxynucleotidyl transferase (TdT) in combination with differentiation antigens including MPO, CD13, and CD15. We then
compared the results with those observed in other cases of
AML andin fetalbone marrow (BM). Furthermore,we quantified the expression of these molecules, since it has been
demonstrated that the intensity of CD34 expression in normal BM precursor cells is variable,I7 and only CD34'"" precursors coexpress cytoplasmic MPO.I8 Such an analysis of
CD34 expression, in the context of other morphologic and
cytogenetic features, is of clinical relevance because several
investigators have previously indicated that the expression
of CD34, when investigated as an independent parameter,
HE RECENT FOCUS onuniform chromosomal aberrations in acute myelogenous leukemia (AML), and the
confirmation of these findings by molecular probes, has identified large groups of patients whose malignancies share
common pathogenetic featuresand representwell-defined
prognostic groups."3 Frequently, these cohorts also display
characteristic morphologic features such as a parody of promyelocyte development in acutepromyelocyticleukemia
with t( 15; 17), myelodysplasia with micromegakaryocytes
and lack of Auerrods in AML with 5q abnormality, or
eosinophilia in AMLwith i n ~ 1 6 . ~In
- ' another form of AML,
so typical that
M2 with t(8;21), the morphologic changes are
the aberrant morphology has guided the identification of this
subgroup at the time of its cytogenetic discovery.' The main
features of this disease are prominent Auer rods, heraldinga
disturbed maturation of the granulocytic lineage, and typical
From the Department of Clinical Immunology, Royal Free Hospital School of Medicine, London; Department of Haematology, Royal
Postgraduate Medical School, London; Medical Research Council
(MRC) Clinical Trial Service Unit, Radchffe Infirmary, Oxford; Department of Haematology, University College Hospital,
London; and
Department of Haematology, Universiiy of Wales, Cardia UK.
Submitted May 22, 1995; accepted September 14, 1995.
Supported by grants fromthe Leukaemia Research Fund (28/89)
and the Medical Research Council of Great Britain (SPG8417830)
as part of the service to the MRC United Kingdom Acute Lymphoblastoid Leukaemia and Acute Myelogenous Leukemia trials.
A.P.". and G.J. have contributed equally to this paper.
Address reprint requests to Prof George Janossy, DSc, Department of Clinical Immunology, Royal Free Hospital School of Medicine, London NW3 2QG, UK.
The publication costsof this article were defrayedin part by page
charge payment. This article must therefore be hereby marked
"advertisement" in accordance with 18 U.S.C. section 1734 solely to
indicate this fact.
0 1996 by The American Society of Hematology.
Blood, Vol 87, No 3 (February l),
pp 1162-1169
From www.bloodjournal.org by guest on November 14, 2014. For personal use only.
appeared to be predictive of poor clinical o u t ~ o m e . 'How~-~~
ever, this finding seems to contradict the fact that AML M2
with t(8;21) is a favorable prognostic category'*3323s24
CD34 positivity of the blasts."-13
In this study, a total of 517 patients entered onto Medical
Research Council (MRC) AMLlO therapeutic trials were
studied. We describe the overexpression of CD34 antigen in
AML M2 with t(8;21) and reveal the characteristic asynchrony when CD34 is studied in combination with MPO.
This asynchrony also coincides with other aberrant phenotypic features, contributing to the differential diagnosis and
allowing the immunologic monitoring of patients for the
presence of residual disease.
Selection ofpatients. Immunodiagnostic, cytochemical, and cytogenetic results were available on 517 BM and blood samples from
patients with AMLaged 17 to 55 years (mean, 47.5y). Clinical
follow-up data were available from the adult AMLlO trial of the
MRC of Great Britain. The hematologic diagnosis followed the
guidelines of the French-American-British classification based on
morphology and cytocherni~try.~~
Routine immunophenotyping results were obtained by immunofluorescence microscopy or flow cytometry as described later. The treatment included four courses of
intensive chemotherapy consisting of daunorubicin, cytarabine, and
thioguanine 3 + 10 and 3 + 8 + 5 ; methotrexate, adriamycin,
cyclophosphamide, and etoposide; and MidAc (mitozantrone, daunorubicin, cytosine arabinoside).26Patients with a matched sibling
were allocated to allogeneic BM transplantation (BMT). Those who
lacked a donor were randomized to receive either no further treatment or an autologous BMT as the fifth course. The results of
stratification according to the three choices (ie, chemotherapy and
allogeneic or autologous BMT) are available from the UK Trial
Office, but are omitted here since they do not influence the study
Expression of CD34 antigen was quantified in about a third of
the trial cases who had CD34' leukemia. This group included 53
representative patients from various French-American-British categories of AML, 11 cases with t(8;21) and 42 without t(8;21). Finally,
a selected group of 18 cases of AML M2 with t(8;21) were further
investigated to analyze the various signs of phenotypic asynchrony
with a panel of markers used in two- and three-color immunofluorescence (IF) tests.
Controlsamples. Fetal BMwas obtained from fetal long-bone
samples provided by the MRC Tissue Bank, Royal Postgraduate
Medical School, London, under the ethical guidelines and permission
of the Ethical Committees of both the sending and receiving institutions. The gestational ages, 8 to 18 weeks, were determined from
the crowdrump length. BM cells were flushed from the bone cavity
with Hanks balanced salt solution, (Gibco, Uxbridge, UK). Normal
BM samples were also received from children with idiopathic thrombocytopenia for diagnosis without evidence of BM pathology. KG1
cell line was grown under standard conditions.
Immunodiagnosis. Mononuclear cells were separated on FicollIsopaque density gradient 1.077 g/mL (Lymphoprep; Nycomed,
Oslo, Norway). The stored cells were rapidly thawed by diluting in
Hanks balanced salt solution at 20°C and washed. Cell suspensions
(2.5 to 5 X lo5cells per 50 pL) were incubated for 10 to 15 minutes
at20°C with MAbs conjugated with fluorescein isothiocyanate
([FITC], green, fluorescence 1: Fll), phycoerythrin ([PE], orange,
F12). or Peridinin chlorophyll A (red, F13) used as premixed cocktails at pretitrated optimal conditions (Table 1). After washing in
phosphate-buffered saline, the stained cells were fixed in 0.5% para-
Table 1. lmmunophrnotypes of Leukemic Blasts
in 18 Cases of AML With t(8;21)
Case No.
Results are given as percentages of positive cells within the blast
cell populationanalyzed by flow cytometryor the APAAP irnmunoenzyme method.
Abbreviation: NT, not tested.
* MAbs were directly conjugated as CD13-FITC (WM47; from Dr K.
Bradstock, Westrnead Hospital, Sidney, Australia), CD15-FITC (C3D-1;
Dakopatts), HLA-DR-perCP (RFDR-2; RFHSM), CD14-PE(UCHM1;from
Professor P. Beverley, ICRF, London, UK),CD34-FITC(BI-C35; Seralab, Crowley, UK),CD34-PE(QBEND;
Professor M.F.Greaves,LRF
Labs, London, UK), CD56-PE(Leul9; Becton Dickinson, UK), anti-MP0
(Dakopatts), and H-TdT6-FITC (Supertech, Bethesda, MD) and used
on two- and three-color combinations.
t Strongly positive when tested by cytochemistry.
formaldehyde. Isotype-matched FITC- and PE-conjugated murine
Igs were used as negative controls. Five-parameter flow cytometry,
ie, three-color IF with forward (FSc) and side (SSc) scatters, was
performed on a FACScan with Paint-a-gatel' and Lysys-II softwares
(Becton Dickinson, Oxford, UK). List mode files were also transferred to IBM-compatible computer by HP-Reader and analyzed
with Winlist software (Verity Software House, Topsham, MI). The
percentage of positivity for various MAbs was determined after
gating blast populations by their FSclSSc features. Cases were regarded as CD34' if more than 20% of cells in the blast gate expressed this antigen.
Detection of cytoplasmic and nuclear antigens. During routine
immunodiagnosis, detection of cytoplasmic M P 0 and nuclear TdT
was performed on cytospins using anti-MP0 MAb (Dakopatts Ltd,
High Wycombe, Bucks, UK) with alkaline phosphatase anti-alkaline phosphatase (Dakopatts), and rabbit anti-TdT serum with goat
anti-rabbit Ig labeled with rhodamine (TRITC) as a second layer
(Supertech, Bethesda, MD). M P 0 and TdT were also detected with
direct IF using anti-MPO-FITC (Dakopatts) and anti-TdT-FITC
(HTdT-6; Supertech) after fixing and permeabilizing the cells with
Permeafix (OPF; Ortho Diagnostic Systems, Inc, Raritan, NJ). The
cells (2.5 to 5 X 105/50 pL) were fixed for 30 minutes at 20°C in
800 pL OPF diluted in a 1:l ratio with distilled water," and after
washing in PBS, they were incubated with 5 pL anti-MPO-FITC or
anti-TdT-FITC for 30 minutes at 20°C. The cells were labeled for
membrane CD34-PE with QBend-PE (kindly donated by Professor
M. Greaves, LRF Leukaemia Laboratories, London, UK) before
fixation and then stained for intracellular antigens (see above). Cyto-
From www.bloodjournal.org by guest on November 14, 2014. For personal use only.
plasmic CD22 (RFB4; Royal Free Hospital), CD79 (mb-l; kindly
provided by Dr D.Y. Mason, John Radcliffe Hospital, Oxford, UK;
and Dakopatts), and cytoplasmic CD3 (Leu 4-PerCP, BD) were
investigated by the same methods.
Quantifation of CD34 expression. FACScan was set on logarithmic scale. Instrument settings and channel compensations were controlled by calibrated fluorescence reference beads Quantum (Flow
Cytometry Standard, Research Triangle Park, NC). Intensity of
CD34 expression was measured as ABC per cell on standard cells
with indirect IF using the QIFI kit (Biocytex, Marseille, France) as
described previously.28 Boththe first layer, CD34 MAb (BI3C5;
Seralab Ltd, Crowley Down, UK), and the affinity-purified second
layer, goat anti-mouse Ig (GcuMIg-FITC; Southern Biotechnology
Associates, Birmingham, AL), were usedat pretitrated saturated
conditions. One standard was CD34+ AML blasts taken from a single
patient at the start of the study (PJ-AML), frozen in small aliquots,
and thawed with more than 90% viability whenever needed throughoutthe investigations. These PJ-AML blasts expressed 64 X IO’
CD34 mean ABC per cell when tested with the QIFI test. The other
standard was a subclone of the KG1 cell line expressing 410 X 10’
CD34 ABC per cell, a value at the upper limit of the measured
range. At the same time, QIFI tests were also performed on CD4’
normalblood lymphocytes withthe same GaMIg-HTC second
layer, producing 47 to 57 x IO3 CD4 ABC.’’ These three values
together represented the framework to render the quantitative IF
tests comparable through the span of studies. Atthenext
directly conjugated CD34 MAbs, CD34-FITC (BI-3C5; Seralab) and
CD34-PE (QBend, see above), were usedat saturating level in a
final concentration of 5 pW100 pL on the same cellular standards
(PJ-AML and KGl). The bead standards used for direct IF (Quantum
Simply Cellular; QSC) from Flow Cytometry Standards (Research
Triangle Park, NC)” were recalibrated to match the known values
of 64 X IO3 and 410 X IO3 per cell on PJ-AML and KGl, respectively. This was a necessary step to maintain reproducibility between
different directly conjugated CD34 reagents. Values for ABC were
determined on Winlist with the QuickCal method following linearization of the mean fluorescence intensity.” Thus, CD34-PE provided higher M H values than CD34-FITC, but similar ABC per cell.
These two reagents showed an excellent correlation on the various
cases of AML (R2 = .99; not shown) and could therefore be interchangeably used in two- and three-color IF combinations.
Statistics. Life-table analyses with Kaplan-Meier curves were
prepared at the MRC Clinical Trials Service Unit, Cancer Studies
Unit, Oxford, UK, using the Statistical Analysis System (Cary, NC)
and the Mantel-Haenschel chi-square test. The nonparametric MannWhitney (Itest was also applied.
Quantitation of CD34 antigen density in normal BM.
The mean percentage of CD34+ cells in the mononuclear
fraction of fetal and normal BMwas6.2% (range, 1% to
20%; n = 6). There was a difference in CD34 binding capacityof small- to intermediate-size cells (mean, 26.5 X 10’
ABC per cell) versus larger cells (mean, 42 X lo’ ABC per
cell), as determined by gating cells of differing sizes on the
scatter plot and calculating their CD34 ABC values. This
was confirmed by double staining with CD10 MAb: CD34+/
CD10+ cells (B-cell precursors) displayed lower CD34 ABC
(mean, 29.5 X lo’ ABC per cell) than CD34+/CD10- cells
with larger size that include early myeloid precursors (mean,
44 X lo3 ABC per cell). Very few, if any, cells could be
observed with more than 60 X 10’ CD34 ABC per cell; this
value can be considered the upper limit of CD34 expression
on precursor cells in fetal and normal BM. Double staining
for CD34 and MP0 in normal and fetal BM showed that
among CD34+ cells, the MPO- majority population (R2 gate
in Fig lb) expressed higher levels of CD34 (median, 25 X
lo3 CD34 ABC per cell; Fig IC) than the rare MPO+ forms
(R3 gate in Fig lb; median, 16 X lo’ CD34 ABC per cell;
Fig Id).
Incidence of CD34 positivity in AML. When grouped by
French-American-British criteria, among 5 17 cases of AML
tested, 103 (19.9%) were MOM1, 152 (29.4%) M2, and 65
(12.6%) M3, and 174 (33.8%) showed signs of monocytic
differentiation, including 121 cases of M4and 53 of M5
(Fig 2), with 21 cases (4.1%) in other minor categories (not
shown). Itwas found that CD34 positivity, identified by
more than 20% CD34’ cells within the blast population, was
less frequent in AML M3 and M5 (18% and 28% of cases,
respectively) than in the other AML subtypes (39% to 42%
of cases); nevertheless, these differences were not statistically significant (Fig 2).
The incidence of CD34 expression was checked among
M2 cases with t(8;21). In this group, higher CD34 expression
was seen than in the other cases with M2 morphology without t(8;21) ( P < .001). In 32 of 38 t(8;21) cases tested
(84.2%) more than 20% leukemic blasts were CD34+, with
a mean of 68.5% (range, 20%to 98%). Nevertheless, six
cases of M2 with t(8;21) representing 15.8% of t(8;21) cases
were CD34-.
Quantitation of CD34 antigen expression in AML. Since
both microscopic and flow-cytometricobservations indicated
that in a number of AMLM2 cases CD34 staining was
particularly strong, the CD34 ABC was quantified in about
one third of all CD34’ leukemias entered into the trial. These
were morphologically characterized as seven MO, 18M1,
eight M2 without t(8;21), l1 M2 with t(8;21), and nine with
monocytic differentiation (M4/5). CD34 ABC was significantly higher in AML with t(8; 2 l) than in any other group
of AML (Fig 3); six of 11 cases displayed values more than
twofold higher (300 to >400 X lo3 ABC per cell) than the
highest CD34 ABC (144 X lo3 ABC per cell) found in other
cases of AML without t(8;21). In nine of 11 cases (82%),
CD34 ABC was higher than the upper limit of the CD34
range for normal BM (60 X 10’ ABC per cell).
CD34 ABC was also higher than the mean CD34 expressionseen in normalBM (30 X loq ABC) insevenof 18
(38%) M1, three of eight (37.5%) M2 without t(8;21), and
three of nine (33%) AML. Nevertheless, only seven of 42
cases (16.7%) of AML without t(8;21) had higher than the
maximal levels of CD34 expression seen in normal fetal BM
(Fig 3).
Aberrant features of myeloblasts in AML M2 with t(8;21).
A total of 18 cases were investigated in further detail, all of
which showed t(8;21)-associated cytologic and cytochemical features’such as M2 morphology displaying varying
degrees of granulocytic maturation, with Auer rods seen in
all cases. There was no evidence of any monocytic component on either morphology or esterase staining or any demonstrable erythroid or megakaryocytic abnormalities. The
blasts also displayed the typical nuclear indentation with a
prominent Golgi region. Granulocyte maturation was abnor-
From www.bloodjournal.org by guest on November 14, 2014. For personal use only.
mean ABC value (x103 CD34 ABClcell):
Fig 1. Quantitative assessment ofCD34 expression in normal infant BM la
t o d)and in
AML M2 with
t(8;21) (e t o h).Values in the gateR2 in normal BM
(b) correspond to the mean values of 25 x l o 3ABC per cell and
large cells. Only some of these
cells are MPO-positive (16 x lo3
CD34 ABC per cell). As a marked
contrast, MPO-positive cells
(gated in R51 in AML with t(8;21)
were as high as 165 x lo3 CD34
ABC per cell, above the normal
range. R5 in f can be used as a
live gate t o search for residual
aberrant myeloblasts in patients
during therapy.
Forward Scatter 3
mal in all cases, with asynchronously poor nuclear maturation and cytoplasmic hypogranularity from the myelocyte
stage onward. Sudan black B staining revealed a typical
localized positivity in the Golgi region within the blasts, and
a small proportion of blasts were chloroacetate esterasepositive. The eosinophils, where present, usually showed
variable granule size and amphophilic staining.
When studied by flow cytometry (Table 1) the blast cells
had a similar pattern of FSc/SSc, with most cells locating
within the myeloid blast region (54% to 91%; mean, 72.5%)
and spreading toward the differentiating myeloid cell region
(Fig le). Four markers showed a relative uniformity of this
patient group. The high proportions of CD34' cells were
accompanied by similarly high percentages of blasts positive
for class 11, CD13, and MP0 in all except one or two unusual
cases with each single marker studied (Table l ) . CD14 nega-
CD34-PE +
tivity, confirming the lack of monocytic differentiation, was
also uniform, whereas the expression of CD15 was heterogeneously positive in 39%of cases when tested witha VIMDSlike antibody reacting withthe nonsialated determinant of
the Lewis' antigen.' Among I8 cases of AML with t(8;2 I),
13 were studied for CD56 expression. Six of these (46%) had
more than 40% CD56' myeloblasts coexpressing CD34-a
leukemia-associated combination.'" A sign of asynchronous
development was the expression of nuclear TdT in myeloblasts that coexpressed membrane CD13 or cytoplasmic
MP0 in six of 14 cases tested(42%)-and
again, these
features are leukemia-associated.3'
Nevertheless, the most frequent aberration in AML M2
with t(8;21) was the overexpression of CD34 antigen, which
was formally quantified and found to be above the levels in
Fig 2. Incidence of CD34* leukemias among cases of AML. In this
study of 494 patients, cells with >5 x lo3 CD34 ABC per cell were
regarded as positive, and cases with >20% CD34' cells within the
blast gate werescored as positive. The numbers shownare percentage positive cases within each French-American-British category.
t(8;21) no t(8;Zll
Fig 3. Quantitation of ABC per cell in 53 casesof CD34' leukemia.
Mean ABC values in blast cell populations
are shown followingquantitation withCD34-FITC calibrated on the standard
AML and KG1 cell
line. The highest expression is observed in M2 with t(8;211. In six
cases, values are >250 x lo3 ABC per cell.
From www.bloodjournal.org by guest on November 14, 2014. For personal use only.
were still all in remission, and 32 CD34' patients hadan
excellent 62% 5-year survival rate (Fig 5); nevertheless, this
CD34- versus CD34' comparison, due to the small sample
size of CD34- patients with t(8;21), is uninterpretable. When
the non-t(8;21) group was analyzed separately, CD34- patients had a significantly higher CR rate (84%) and marginally better survival (36%) than the CD34' group (69% CR
and 3 1 % survival). However, the difference between CD34and CD34' subgroups is relatively small compared with the
greater impact of cytogenetics.
CD34 over-expression
or CD13
cumulative of these
% of cases tested
Fig 4. Aberrant and asynchronous staining combinations in AML
M2 with t(8;211. The relative frequency of CD34 overexpression, asynchronous combination with M P 0 and CD56, and aberrant combination of TdT staining with M P 0 andlor CD13 are shown. The cumulative data reveal that aberrant or asynchronous features are seen in
every case of AML with t(8;211, but these alterations are not fully
uniform (see Table 11.
The AML M2-associated t(8;21)(q22;q22.3) results in fusion of the AMLl gene on chromosome 21 with a gene
on chromosome 8 referred to as ET0.".33 These molecular
findings already provide insights into the altered function of
the AMLIETO protein in the processes normally regulated
by AMLl gene during myeloid differentiation.3 In this report, we characterize the complexity of these changes when
measured in these malignant AML blasts with quantitative
immunologic techniques. In all of these cases, irregular control of cellular differentiation can be demonstrated with simple combination staining for precursor cell features such as
CD34 or TdT versus signs of myeloid differentiation such
as MPO, CD1 3, and CD15 display. Furthermore, many of
these alterations appear to be characteristic for this particular
disorder. Further extended use of the quantitative techniques
used in this study on a larger cohort of patients willbe
required to see whether such changes occur at all in AML
without t(8;21). The most obvious of these changes is the
overexpression of CD34 antigen in MPO-positive malignant
myeloblasts. Nevertheless, these changes are not fully uniform (Fig 4), and16% of AML with t(8;21) are CD34-.
Thus, the complexity of the changes argues for a disturbed
regulatory control rather than for mere involvement of genes
that directly regulate these particular proteins.
One of the main features of our study has been the quantitation of CD34 antigen on normaland leukemic cells in
combination staining with other MAbs, eg, for cytoplasmic
MPO. This necessitated the use of CD34-FITC and CD34-
CD34' cells seen in normal and fetal BM in nine of I 1 cases
(81%; Fig 3). Furthermore, in eight of these samples (73%),
CD34 showed an asynchronous combination with MP0
above the levels seen in normal BM'8 (Fig l). A distinctive
live gating (R5 in Figs I f and h) clearly identified large
proportions of malignant blasts, indicating which parameters
should primarily be used when remission samples are analyzed from the same patient. The high levels of CD34 expression in combination with another marker of granulocytic
differentiation, CD15, were also recorded in seven of 18
cases analyzed (39%), although the double-stained blasts
were relatively few in two cases (Table l ) .
Finally, strong CD19 expression byCD34' myeloblasts
was detected in three cases, whereas in other samples the
CD19' peak, when studied with the RFB9 MAb used,
showed weak positivity but could notbe fully separated
from negative control; the CD19 ABC per cell waslow
(not shown). Cytoplasmic CD22 and CD79a, bona fide Blineage-specific antigens, were negative in all cases, including those that were TdT-positive. Taken together, these obTable 2. Clinical Outcome in 517 Patients Entered Into the MRC
servations demonstrate that aberrant andor asynchronous
AMLlO Trial-Effects of t(8;21) and CD34 Expression
combinations are identifiable in virtually every case of AML
M2 with t(8;21), although these changes are not fully uniform (Fig 4) and in 15% to 20% of AML M2 with t(8;21)
do not express strong CD34 positivity (Figs 2 and 3).
336 patients
No. of
Correlation with clinical outcome. Clinical outcome
CR rate (%l
38 patients with t(8;21) and 479 patients with other forms
Reason for failure (%l
of AML is shown in Table 2 and Fig 5 . All patients with
Induction death
t(8;21) obtained complete remissions ([CRs] 100%); in this
Resistant disease
group, significantly fewer relapses and deaths were seen
Survival (%) from entry at 5 years
(71% 5-year survival) than among patients who hadno
P < ,0005 for both CR rate (x2 test) and survival (log-rank test1
t(8;21) (34% survival, P < .0005;Table 2). Taken overall,
differences between t(8;211-positive and -negative patients.
the CR rate was significantly better in CD34- patients than
t P = .0003 by x* test.
in CD34' patients (84% v 75%, P = .008), but there was
P = .05 by log-rank test.
no survival difference (38% v 37% at 5 years). When the
5 P is not interpretabledue to the small number of CD34- patients
t(8;21) group was studied separately, six CD34- patients
in the t(8;211 group.
From www.bloodjournal.org by guest on November 14, 2014. For personal use only.
Fig 5. Kaplan-Meier survivalcurvesfor patients
from the time of entry onto the AMLlO trial according to the presence or absence of tl8;21) and
CD34 antigen on blasts at presentation.The criterion
for CD34 positivity was ~ 2 0 %
CD34' cells among
blasts. tl8;21)-, CD34- (-, n = 336); t18;21)-, CD34+
n = 143); t(8;21)+, CD"
I - . - . - , n = 6); and
t(8;21)+, CD34+ I - . -.
, n = 32).
( 9
. .,
-. -
PE reagents. If a single reagent is used in an investigation,
the molecules equivalent to soluble fluorochromes (MESF
units) are adequate to obtain comparative data.30 Inour case,
it was more appropriate to introduce standard CD34' cells
that expressed known values for ABC per cell established
by the QIFI test in indirect IF28,29,34
to secure the reproducibility of the performance of directly conjugated CD34 used in
diagnostic staining combinations (see the methods). In this
way, quantitative IF becomes a practical possibility. The
same methods have already been applied to document the
overexpression of CD10 antigen on 44% of cases with acute
lymphoblastic leukemia.34
Expression of CD34 in AML has previously been associated with the poorly differentiated French-American-British
subgroups MO and Ml,35.36and several studies have suggested that CD34 positivity may be related to a poor clinical
response in terms of CR and/or s u r v i ~ a l . ' ~In. ~this
~ -context,
the observations of Lanza et al,37 who also find that the
"bright" CD34 expression is associated with poor prognosis, are particularly interesting. This is because cytogenetic
analysis reveals that their AML M2 group lacks cases with
t(8;21) but primarily shows -5 and 5q- involvement, a known
poor prognostic sign.38Thus, our investigations, based on
large numbers of patients who were uniformly treated in the
AMLlO multicenter
may resolve the apparent discrepancies seen in the literature that appear to be due to different
proportions of patients with various chromosomal abnormalities in the individual studies. On one hand, when we excluded t(8;21) patients from the evaluation, CD34+ AML
cases were less likely to obtain CRand showed slightly
inferior 5-year survival as compared with the CD34- group
(Fig 5). These findings are reminiscent of the poor prognostic
significance of CD34 expression, save the t(8;21) group. On
the other hand, we confirm the survival advantage of patients
with t(8;21) despite the obviously bright CD34 expression
on most of these cases.1.23In conclusion, these findings indicate that individual markers such as CD34 positivity should
not be taken out of the context of the overall clinical and
cytogenetic presentation, but should be used for patient assessment as part of the diagnosis. Our studies also confirm
other observations that emphasize the relevance of chromosomal changes in interpreting the prognostic significance of
lymphoid marker expression such as CD19 and CD211.39,
well as CD7,40in the various forms of AML.
A clear example of the diagnostic advance inherent in our
investigation is the use of quantitative IF for rapid diagnosis,
as well as remission assessment, in t(8;21) AML. Sensitive
polymerase chain reaction techniques have uniformly revealed the persistence of AML1-ET0 fusion gene in patients
during remission, sometimes as long as 52 months after therapy.3,41-44
This is a clear step toward a better understanding
of a multistep process in malignancy, but it questions the
clinical utility of these molecular probes during the remission
phase. Residual-disease assessment with immunologic methods in these diseases should now include analysis of CD341
CD56,L4,16CD13/TdT?' CD19/CD13,'1,'2"4.39and CD34/
MPOLs(and this study), together with a quantitative analysis
of CD34 expression as delineated in this report. It will be
interesting to observe whether patients with identifiable molecular disease retain or lose signs of asynchrony in myeloid
development while remaining in an apparent clinical remission.
A.P.M. was on leave from the Department of Pathology, Karolinska Hospital, Stockholm, Sweden. We areparticularlyindebted to
the physicians andpatientswhoparticipatein
therapeutic trials for allowing us to investigate the presentation samples. We also thank DrL. Wong, MRCTissue Bank, Royal Postgraduate Medical School, London, for fetal BM samples, andVanessa
Lipton for secretarial assistance.
1. Swansbury GJ, Lawler
SD, Alimena G, Arthur D, Berger R,
Van den Berghe H, Bloomfield CD, de la Chappelle A, Dewald G ,
Garson OM: Long-term survival in acute myelogenous leukemia: A
second follow-up of the4thInternationalWorkshop
on Chromosomes in Leukemia. Cancer Genet Cytogenet 73:1, 1994
2. Walker H, Smith SJ, Betts DR: Cytogenetics in acute myeloid
leukaemia. Blood Rev 8:30, 1994
3. Nucifora G, Rowley JD: AML-l andthe (8;21) and (3;21)
translocations in acute and chronic myeloid leukemias. Blood 86: 1,
4. Biondi A, Rambaldi A, Pandolfi PP, Rossi V, Giudici G, Alcalay M, Lo-Coco F, Diverio D, Pogliani EM, Lanzi EM: Molecular
From www.bloodjournal.org by guest on November 14, 2014. For personal use only.
monitoring of the myl/retinoic acid receptor-alpha fusion gene in
acute promyelocytic leukemia by polymerase chain reaction. Blood
80:492, 1992
S. Adriaansen HJ, Boekhorst PA, Hagemeijer AM, van der Schoot
CE, Delwel HR. van-Dongen JJ: Acute myeloid leukemia M4 with
bone marrow eosinophilia (M4Eo) and inv(16)(p13q22) exhibits a
specific immunophenotype with CD2 expression. Blood 81 :3043,
6. Lanipa I, Acevedo S, Palau-Nagore MV, Bengio R, Mondini
N, Slavutsky I: Leukemic transformation in patients with SQ- and
additional abnormalities. Haematologica (Pavia) 76:363, 1991
7. Swirsky DM, Li YS, Matthews JG, Flemans RJ, Rees JK,
Hayhoe FG: 8;21 Translocation in acute granulocytic leukaemia:
Cytological, cytochemical and clinical features. Br J Haematol
56: 199, 1984
8. KnappW, Strobl H, Majidic 0: Flow cytometric analysis of
cell surface and intracellular antigens in leukemia diagnosis. Cytometry 18:187, 1994
9. Griffin JD, Mayer RJ, Weinstein HJ, Rosenthal DS, Coral FS,
Beveridge RP, Schlossman SF: Surface marker analysis of acute
myeloblastic leukemia: Identification of differentiation-associated
phenotypes. Blood 62:57, 1983
10. Terstappen LW, Safford M, LokenMR:Flow
analysis of human bone marrow. 111. Neutrophil maturation. Leukemia 4:6S7, 1990
11. Reading CL, Estey EH, Huh YO, Claxton DF, Sanchez G,
Terstappen LW, O’Brien MC, Baron S, Deisseroth AB: Expression
of unusual immunophenotype combinations in acute myelogenous
leukemia. Blood 81:3083, 1993
12. Touw I, Donath J, Pouwels K, vanBuitenen C, Schipper P,
Santini V, Hagemeijer A, Lowenberg B, Delwel R: Acute myeloid
leukemias with chromosomal abnormalities involving the 21q22 region identified by their in vitro responsiveness to interleukin-S. Leukemia S:687, 1991
13. Kita K, Nakase K, Miwa H, Masuya M, Nishii K, Morita N,
Takakura N, Otsuji A, Shirakawa S, Ueda T: Phenotypical characteristics of acute myelocytic leukemia associated with
t(8;21)(q22;q22) chromosomal abnormality: Frequent expression of
immature B-cell antigen CD19 together with stem cell antigen CD34.
Blood 80:470, 1992
14. Hunvitz CA, Raimondi SC, HeadD, Krance R, Mirro JJR,
Kalwinsky DK, Ayers CD, Behm FG: Distinctive immunophenotypic features of t(8;21)(q22;q22) acute myeloblastic leukemia in
children. Blood 80:3182, 1992
IS. Campos L, Guyotat D, Archimbaud E, Devaux Y, Treille D,
Larese A, Maupas J, Gentilhomme 0, Ehrsam A, Fiere D: Surface
marker expression in adult acute myeloid leukaemia: Correlations
with initial characteristics, morphology and response to therapy. Br
J Haematol 72:161, 1989
16. Coustan-Smith E, Behm FG, Hurwitz CA, Rivera GK, Campana D: N-CAM (CDS6) expression by CD34+ malignant myeloblasts has implications for minimal residual disease detection in
acute myeloid leukemia. Leukemia 7:8S3, 1993
17. DiGiusto D, Chen S, Combs J, Webb S, Namikawa R, Tsukamoto A, Che BP, Galy AH: Humanfetal bone marrow earlyprogenitors for T, B and myeloid cells are found exclusively in the population expressing high levels of CD34. Blood 84:421, 1994
18. Strobl H, Takimoto M, Majdic 0, Fritsch G, Scheinecker C,
Hocker P, Knapp W: Myeloperoxidase expression in CD34+ normal
human hematopoietic cells. Blood 82:2069, 1993
19. Vaughan WP, Civin CI, Weisenburger DD, Karp JE, Graham
ML, Sanger WC, Grierson HL, Joshi SS, Burke PJ: Acute leukemia
expressing thenormalhuman hematopoietic stem cell membrane
glycoprotein CD34 (MYlO). Leukemia 2:661, 1988
20. Geller RB, Zahurak M, Hunvitz CA, Burke PJ,Karp JE,
Piantadosi S, Civin CI: Prognostic importance of immunophenotyping in adults with acute myelocytic leukaemia: The significance of
the stem-cell glycoprotein CD34 (Mylo). Br J Haematol 76:340.
21. Solary E, Casasnovas RO, Campos L, Bene MC, Faure G,
Maingon P, Falkenrodt A, Lenormand B, Genetet N: Surface markers
in adult acute myeloblastic leukemia: Correlation of CD19+.
CD34+ and CD14+/DR-phenotypes with shorter survival. Groupe
d’Etude Immunologique des Leucemies (GEIL). Leukemia 6:393.
22. Myint H, Lucie NP: The prognostic significance of the CD34
antigen in acute myeloid leukemia. Leuk Lymph 7:42S, 1992
23. Marosi C, Koller U, Koller-Weber E, Schwarzinger I, Schneider B, Jager U, Vahls P, Nowotny H, Pirc-Danoewinata H, Steger
G: Prognostic impact of karyotype and immunologic phenotype in
125adult patients withdenovoAML.
Cancer Genet Cytogenet
61:14, 1992
24. Schouten HC, van-Putten WL, Hagemeijer A, Blijham GH,
Sonneveld P, Willemze R, Slater R, van Lely J, Verdonck LF. Lowenberg B : The prognostic significance of chromosomal findings in
patients with acute myeloid leukemia in a study comparing the efficacy of autologous and allogeneic bonemarrow transplantation.
Bone Marrow Transplant 8:377, 1991
25. Bennett JM, Catovsky D, Daniel MT, Flandrin G, Galton
DA, Gralnick HR, Sultan C: Proposed revised criteria
for the
classification of acute myeloid leukemia. A report of the FrenchAmerican-British Cooperative Group. Ann Intern Med 103:620,
26. Burnett AK, Goldstone AH, Hann IM, Stevens RF, Rees JKH,
Wheatley K, Gray RG: Evaluation of autologous bone marrow transplantation in acute myeloid leukaemia: A progress reportonthe
MRC Tenth AML Trial, in Dicke KA, Armitage JO (eds): Autologous Bone Marrow Transplantation, vol S. Omaha, NE, University
of Nebraska Medical Center, 1991, pp 3-7
27. Pizzolo G , Vincenzi C, Nadali G, VeneriD,Vinante
Chilosi M, Basso G, Connelly MC, Janossy G: Detection of membrane and intracellular antigens by flow cytometry following Ortho
Permeafix fixation. Leukemia 8:672, I994
28. Poncelet P, Carayon P: Cytofluorometric quantification of
cell-surface antigens by indirect immunofluorescence using monoclonal antibodies. J Immunol Methods 8S:6S, 1985
29. Poncelet P, Poinas G, Corbeau P, Devaux C, Tubiana N.
Muloko N, Tamalet C, Chermann JC, Kourilsky F,Sampol J: Surface
CD4 density remains constant on lymphocytes of HIV-infected patients in the progression of disease. Res Immunol 142:291, 1991
30. Vogt RF, Marti GE, Schwarz A: Quantitative calibration of
fluorescence intensity for clinical and research applications of immunophenotyping byflow
cytometry. Rev Biotechnol Bioengineer
3 1. Campana D, Coustan-Smith E, Janossy G: The immunological detection of minimal residual disease in acute leukemia. Blood
76: 163, 1990
32. Erickson PF, Robinson M, Owens G, Drabkin HA: The ET0
portion of acute myeloid leukemia t(8;21) fusion transcript encodes a
highly evolutionarily conserved, putative transcription factor. Cancer
Res S4:1782, 1994
33. Downing JR, Head DR, Curcio-Brint AM, Hulshof MC, Motroni
TA, Raimondi SC, Carroll AJ, Drabkin HA, Willman C, Theil KS: An
AMLIETO fusion transcript is consistently detected by RNA-baed
polymerase chain reaction in acute myelogenous leukemia containing
the (8;21)(q22;q22) translocation. Blood 81:2860, 1993
34. Lavabre-Bertrand T, Janossy G, Ivory K, Peters RE, SeckerWalker L, Ponvit-MacDonald A: Leukemia associated changes identified by quantitative flow cytometry. I. CD10 expression. Cytometry
18:209, 1994
From www.bloodjournal.org by guest on November 14, 2014. For personal use only.
35. Matutes E, Pombo De Ofiviera MS, Foroni L, Morilla R,
Catovsky D: The role of ultrastructural cytochemistry and monoclonal antibodies in clarifying the nature of undifferentiated cells in
acute leukaemia. Br J Haematol 69:205, 1988
36. Borowitz MJ, Gockerman JP, Moore JO, Civin CI, Page SO,
Robertson J, Bigner SH: Clinicopathologic and cytogenic features
of CD34 (My 10)-positive acute nonlymphocytic leukemia. Am J
Clin Pathol 91:265, 1989
37. Lanza F, Rigolin GM, Moretti S, Latorraca A, Castoldi G:
Prognostic value of immunophenotypic characteristics of blast cells
in acute myeloid leukemia. Leuk Lymph 13:81, 1994
38. Fagioli F, Cuneo A, Cali MG,BardiA,Piva N. Previati R,
Rigolin GM, Ferrari L, Spanedda R, Castoldi G Chromosome aberrations in CD34-positive acute myeloid leukemia. Correlation with clini119, 1993
copathologic features. Cancer Genet Cytogenet 71:
39. Ball ED, Davis RB, Griffin JD, Mayer RJ, Davey FR, Arthur
DC, Wurster Hill D, No11 W, Elghetany MT, Allen SL: Prognostic
value of lymphocyte surface markers in acute myeloid leukemia.
Blood 77:2242, 1991
40. Kornblau SM, Thall P, Huh YO, Estey E, Andreeff M: Analy-
sis of CD7 expression in acute myelogenous leukemia: Martingale
residual plots combined with optimal cutpoint analysis reveals absence of prognostic significance. Leukemia 9:1735, 1995
41. Kusec R, Laczika K, Knob1 P, Fried1 J, Greinix H, Kahls
P,Linkesch W, Schwarzinger I, Mitterbauer G, Purtscher B:
AMLlETO fusion mRNA can be detected in remission blood
samples of all patients with t(8;21) acute myeloid leukemia after
chemotherapy or autologous bone marrow transplantation. Leukemia 8:735, 1994
42. Chang KS, Fan YH, Stass SA, Estey EH, Wang G, Trujillo
JM, Erickson P, Drabkin H: Expression of AML1-ET0 fusion transcripts and detection of minimal residual disease in t(8;21)-positive
acute myeloid leukemia. Oncogene 8:983, 1993
43. Nucifora G, Larson RA, Rowley JD: Persistence of the 8;21
translocation in patients with acute myeloid leukemia type M2 in
long-term remission. Blood 82:712, 1993
44. Maruyama F, Stass SA, Estey EH, Cork A, Hirano M, Ino
T, Freireich El, Yang P, Chang KS: Detection of AMLIETO fusion
transcript as a tool for diagnosing t(8;21) positive acute myelogenous
leukemia. Leukemia 8:40, 1994