Cytometry Part B (Clinical Cytometry) 94B:82–93 (2018)

Original Article
AIEOP-BFM Consensus Guidelines 2016 for
Flow Cytometric Immunophenotyping of
Pediatric Acute Lymphoblastic Leukemia
Michael N. Dworzak,1* Barbara Buldini,2 Giuseppe Gaipa,3 Richard Ratei,4
Ondrej Hrusak,5 Drorit Luria,6 Eti Rosenthal,7 Jean-Pierre Bourquin,8 Mary Sartor,9
Angela Schumich,1 Leonid Karawajew,10 Ester Mejstrikova,5 Oscar Maglia,3
Georg Mann,1 Wolf-Dieter Ludwig,4 Andrea Biondi,3 Martin Schrappe,11 and
Giuseppe Basso,2 on behalf of the International-BFM-FLOW-network
1
Children’s Cancer Research Institute and St. Anna Children’s Hospital, Department of Pediatrics,
Medical University of Vienna, Vienna, Austria
2
Laboratory of Pediatric Onco-Hematology, Women and Child Department,
University of Padova, Padova, Italy
3
Tettamanti Research Center and Department of Pediatrics, Ospedale San Gerardo,
University of Milano-Bicocca, Monza, Italy
4
Clinic for Oncology and Tumor Immunology, HELIOS Klinikum Berlin-Buch, Berlin, Germany
5
Department of Pediatric Hematology and Oncology, 2nd Faculty of Medicine, Charles University and
University Hospital Motol, Prague, Czech Republic
6
Department of Pediatric Hematology and Oncology, Schneider’s Children’s Medical Center,
Petach-Tikva, Israel
7
Cancer Research Center and the Hematology Laboratory, Jeffrey Modell Foundation (JMF) Center,
Edmond and Lily Safra Children’s Hospital, Sheba Medical Center, Tel Hashomer, Israel
8
Division of Oncology and Children’s Research Center, University Children’s Hospital,
University of Zurich, Zurich, Switzerland
9
Flow Cytometry Unit, Institute of Clinical Pathology and Medical Research,
Westmead Hospital, Sydney, Australia
10
Department of Pediatric Oncology/Hematology, Charite Universit€atsmedizin, Berlin, Germany
11
Department of Pediatrics, University Medical Center Schleswig-Holstein, Kiel, Germany

Immunophenotyping by flow cytometry (FCM) is a worldwide mainstay in leukemia diagnostics. For con-
cordant multicentric application, however, a gap exists between available classification systems, techno-
logic standardization, and clinical needs. The AIEOP-BFM consortium induced an extensive standardiza-
tion and validation effort between its nine national reference laboratories collaborating in
immunophenotyping of pediatric acute lymphoblastic leukemia (ALL). We elaborated common guidelines
which take advantage of the possibilities of multi-color FCM: marker panel requirements, immunological
blast gating, in-sample controls, tri-partite antigen expression rating (negative vs. weak or strong posi-
tive) with capturing of blast cell heterogeneities and subclone formation, refined ALL subclassification,
and a dominant lineage assignment algorithm able to distinguish “simple” from bilineal/“complex” mixed
phenotype acute leukemia (MPAL) cases, which is essential for choice of treatment. These guidelines

Additional Supporting Information may be found in the online ver- Grant sponsor: The Fondazione Tettamanti, Fondazione Cariplo, Asso-
sion of this article. ciazione Italiana per la Ricerca sul Cancro, Consiglio Nazionale delle
*Correspondence to: M. N. Dworzak, M.D., Associate Professor of Ricerche (G. G.).
Pediatrics, Children’s Cancer Research Institute (CCRI) and St. Anna Grant sponsor: The Wilhelm Sander Stiftung 2004.072.1 (R. R.).
Kinderspital, Zimmermannplatz 10, A-1090 Vienna, Austria. Email: Grant sponsor: The Czech Health Research Council; Grant numbers:
[email protected] NV15-28525A and NPU I nr. LO1604 (O. H. and E. M.).
Grant sponsor: The EU-funded network-project ENCCA; Grant number: Received 3 June 2016; Revised 16 January 2017; Accepted 6
EU-FP7, GA HEALTH-F2-2011–261474 (M. N. D.). February 2017
Grant sponsor: The Fondazione Citta della Speranza, Consiglio Nazio- Published online 10 February 2017 in Wiley Online Library
nale delle Ricerche, Ministero dell’Universita a Ricerca Scientifica a (wileyonlinelibrary.com).
Tecnologica, Associazione Italiana per la Ricerca sul Cancro (G. B.). DOI: 10.1002/cyto.b.21518

C 2017 International Clinical Cytometry Society

V
 IBFM-FLOW STANDARDS FOR IMMUNOPHENOTYPING OF PEDIATRIC ALL 83

are a first step toward necessary inter-laboratory standardization of pediatric leukemia immunophenotyp-
ing for a concordant multicentric application. V
C 2017 International Clinical Cytometry Society

Key terms: standardization; flow cytometry; leukemia; pediatric; immunophenotyping:

How to cite this article: Dworzak MN, Buldini B, Gaipa G, Ratei R, Hrusak O, Luria D, Rosenthal E, Bourquin
J-P, Sartor M, Schumich A, Karawajew L, Mejstrikova E, Maglia O, Mann G, Ludwig W-D, Biondi A, Schrappe M,
and Basso G. AIEOP-BFM Consensus Guidelines 2016 for Flow Cytometric Immunophenotyping of Pediatric
Acute Lymphoblastic Leukemia. Cytometry Part B 2018; 94B: 82–93.

INTRODUCTION the diagnostic mainstay for treatment selection and trial

Immunophenotyping by flow cytometry (FCM) is a inclusion. In particular, the topic of mixed phenotype
cornerstone of the diagnostic work-up of leukemia. acute leukemia (MPAL; according to WHO 2008/2016),
Recent years have rendered important improvements previously captured somewhat differently as bipheno-
and fine-tuning of the diagnostic pathways used to typic acute leukemia (BAL) by the European Group for
assess leukemia, the technical armamentarium of FCM, the Immunological Characterization of Leukemias
and the sub-classification based on detailed biological (EGIL), has perturbed clinicians significantly with
insights—as summarized in the 2008 and 2016 revisions respect to best therapeutic options (9,16–24).
of the World Health Organization (WHO) classification The standardization project debuted in July 2009 and
of myeloid neoplasms and acute leukemia (1–4). Howev- comprised several task group meetings (literature search
er, general implementation of these advancements into and topic discussions, joint case reviews, preliminary
daily practice in a comprehensive, standardized and concordance assessments of local case cohorts, guide-
quality controlled way is lagging behind (5–7). line drafting) and culminated in the internal release of a
In the run-up to the multinational treatment trial for first consensus guideline in December 2012, which is
pediatric acute lymphoblastic leukemia (ALL), jointly delineated herein in its most recent 2016 version.
designed by the Associazone Italiana Ematologia Oncolo-
CONSENSUS GUIDELINES
gia Pediatrica (AIEOP) and the Berlin-Frankfurt-M€ unster
(BFM) group, AIEOP-BFM ALL 2009, the nine involved Antibody Panel
national FCM reference labs—providing service to an The consensus panel (Table 1) is designed to fulfil
annual recruitment of approximately 1000 children with EGIL and WHO 2008/2016 requirements for ALL-
ALL—were induced by the trial authorities to elaborate subtyping, MPAL (16), BAL (9,10), as well as early T-cell
a common immunophenotyping standard-operating- precursor (ETP) leukemia discrimination (12,25,26).
procedure (SOP) as an up-date to previously used stand- A multicolor staining approach is compulsory (i.e.,
ards (8–10). Several issues of technology and classifica- 4 colors; 6-color approach recommended) and should
tion were put forward under the premises that it is rea- be based on a CD45-backbone plus markers against
sonable to conserve local practices—based on ample (both) lymphoid lineages in each tube to pull out leuke-
experience—but with some adaptations to meet corpo- mic and potential control cells. The choice of
rate requirements. First, a methodological transition was fluorochrome-conjugated antibody clones, vendors, and
specified to respect the advent and chances of multi- tube combinations is largely left to individual lab’s
color FCM toward (i) improved immunological blast cell expertise with the exception of a strong recommenda-
gating with the ability to explore smaller and/or more tion regarding prominent lineage-defining markers like
heterogeneous populations than previously possible, (ii) the intracellular myeloperoxidase antibody (iMPO, clone
advanced resolution and sensitivity in determining cellu- 8E6), the antibodies against iCD22 and iLysozyme
lar antigen expressions, (iii) inclusion of novel markers (clones RFB4 and LZ-2, respectively), as well as iCD3
or marker combinations with relevance for subclassifica- (clones UCHT1 or SK7). The necessity of panel standard-
tion and/or outcome prognostication, and (iv) introduc- ization regarding certain antigens has been investigated
tion of a detailed, semiquantitative capturing of expres- in a still ongoing validation process in which we have
sion types (1,2,11–15). Accordingly, it was decided to been comparing the individual antibody panels of sever-
introduce the rating of weak and strong expression— al centers against an antibody panel completely stan-
deliberately applied by the WHO 2008 guidelines—into dardized between laboratories (M.N. Dworzak et al.
AIEOP-BFM practice (16). Toward this end, we needed manuscript in preparation). In case of applying a
to clarify uncertainties regarding clear-cut thresholds for sequential staining approach (using a screening tube as
a concordant application throughout many labs. Sec- well as a narrowed, lineage-specific secondary panel),
ond—and related to this previous aspect—we needed to the screening must include the markers CD19, CD10,
specify hematopoietic lineage assignment which remains iCD79a, iCD22, iCD3/CD3, CD7, iMPO, and CD45 to

Cytometry Part B: Clinical Cytometry

84 DWORZAK ET AL.

warrant dominant lymphoid-lineage determination (see Table 1

below) in all cases. The AIEOP-BFM consensus antibody panel for pediatric ALL
Optional/recommended markers add to drawing a Mandatory and optional markers (each combined with CD45)
comprehensive picture of the individual leukemic phe-
Intracellulara,b iCD3, iCD22, iCD79a, iIgM (l-chain),
notypes and fulfil two alternative purposes: First, anti- iLysozyme, iMPO
a
gens like CD11a, CD38, CD58, CD99, and CD123 have a Surface CD2c, CD3, CD5, CD7; CD10, CD19,
strong rationale toward leukemia follow-up and minimal CD20; CD11c, CD11b, CD13, CD14,
residual disease (MRD) assessment (27–31), which is a CD15, CD33, CD64, CD65d, CD117;
CD34, (CD45), CD56, HLA-DR
decisive task for immunophenotyping in AIEOP-BFM tri- if T-ALL: CD1a, CD4, CD8, TCRab, TCRgd
als. Second, they are associated with genetic lesions or if B-IV suspected: j-chain, k-chain
novel clinical subtypes potentially impacting on out- (surface staining after pre-washing or
come. Genotype-associated markers are for example intracellular)
Optional / all cases: NG2e, CD371c,f
CD123 (hyperdiploidy), CD66c (hyperdiploidy, BCR/ Recommended if BCP-ALL: CD11ac, CD22, CD24,
ABL), NG2 (MLL-rearrangements), CRLF2 (CRLF2-rear- CD38, CD44, CD58, CD66c,
rangements), or lack of CD44 positivity (TEL/AML1; also CD123c, CRLF2c,g
in mature B-ALL with MYC-translocation) (1,15,31–39). if T-ALL: CD99, iTdT
if BAL according to general panel:
The consortium, however, considers the fact that per CD24, iTdT
AIEOP-BFM trial practice immunophenotyping is general-
a
ly not used as gateway for genetic assessments. More Mandatory markers for WHO, EGIL, ETP classifications.
b
important than genotype surrogates, the consensus anti- Prefix “i” stands for intracellular staining.
c
Phycoerythrin-conjugate (PE) recommended.
body panel allows capturing novel clinical ALL subtypes d
Available only labelled with fluorescein isothiocyanate
for which specific genetic lesions have not been (FITC).
e
defined. This includes ETP and the newly described B- Clone 7.1.
f
cell precursor (BCP)-ALL entity with a propensity to an Clone 50C1.
g
Clone 1D3.
early switch to the monocytic lineage (40). Such switch-
ALL, apart from being CD2-positive in many cases, is
most efficiently determined at diagnosis based on
CD371 positivity (M.N. Dworzak et al. manuscript in setup and tracking beads according to manufacturer’s
preparation). recommendations are described elsewhere (27,44,45).
Sample Priority, Preparation, Staining and Acquisition Analysis and Gating Strategy
Bone marrow (BM) is preferred over peripheral blood Immunological blast gating is applied using a CD45/
for assessing the immunophenotype at diagnosis. The SSC backbone strategy for delineating the blast cell
latter may be used in case BM is not available or of poor region in the CD45-low/SSC-low area (5,46), plus
quantity/quality or in case of massive peripheral blasto- lineage-defined gating as appropriate. Doublets exclu-
sis. In principle, a stain-lyse procedure is used (27). sion of events with increased FSC-Width is mandatory
Additional staining with a live-cell-permeant nucleated- because such events frequently exhibit false expressions
cell dye is recommended to allow distinguishing non- (Supporting Information Fig. 1; Ref. 44). Assessment of
nucleated events (erythrocytes, debris) from nucleated individual antigen expression is done preferably using
cells (NC) (27). This is particularly relevant to compen- histograms, contour plots, or pseudocolor density plots,
sate for diminished sample quality after prolonged trans- rather than with ordinary dot plots or parameter band
portation to central reference laboratories (by excluding diagrams. Usage of plots with biexponential or logical
residual non-nucleated events from blast cell assessment display is specifically advantageous for investigating low
and enumeration among NC) and specifically in ALL intensities and negative values (44,47). A negative region
cases with a very low CD45 expression on blasts (which is then defined for each antigen by in-sample cell con-
fall into the CD45 area where non-nucleated events are trols (i.e., mature, cross-lineage lymphocytes gated in
found). For cellular permeabilization and intracellular plots with CD45 and/or cross-lineage markers) which do
staining, we use commercial reagents according to the not express the antigen. This approach is currently seen
manufacturers’ recommendations (27,41,42). as one of the most appropriate control measures
Sample acquisition on flow-cytometers should mea- because of the exposure of all populations to identical
sure a minimum of 30,000 NC per tube. This warrants conditions (44,48,49). Secondly, we address autofluores-
that—at a stipulated sensitivity of 1%—also minor blast cence. To this end, we use unstained control tubes (one
subpopulations as well as residual normal lymphocytes permeabilized for comparison with intracellular stainings
(to be used as internal control, see below) can be immu- and one non-permeabilized for surface antigens) (48,49).
nophenotyped with sufficient precision and narrow CV However, to respect that different cell types within a
by acquiring several hundreds of such events (43,44). sample may exhibit different background fluorescence
Necessary procedures for quality control of the instru- (Supporting Information Fig. 2; Refs. 48 and 49) we also
mental setup and over-time performance by recording recommend the use of further two control tubes (one

Cytometry Part B: Clinical Cytometry

 IBFM-FLOW STANDARDS FOR IMMUNOPHENOTYPING OF PEDIATRIC ALL 85

FIG. 1. Overview scheme and definitions of antigen expression rating according to the AIEOP-BFM consensus guidelines. Note that, for the tripartite
major expression capturing (negative, weak, strong positive) as well as for the Bethesda-style fine-rating, cross-lineage lymphocytes are used as
negative-control populations, with the exception of a few broadly expressed antigens like CD45, which receive adapted fine-rating. For CD45 we rec-
ommend assessment of its expression on the blast subset (red) versus residual granulocytes (blue; high SSC) and mature lymphocytes (blue; low
SSC).

Cytometry Part B: Clinical Cytometry

86 DWORZAK ET AL.

Table 2 categories (weak and strong, see Fig. 1). Only in border-
The AIEOP-BFM dominant lineage assignmenta line situations the prevalence of non-overlap of blasts
Lineage Criteria Antigens
with the negative control population is quantified
(instead of using, for example, the standard deviation of
BCP-ALL 2 positive of: CD19b; CD10,
(i)CD22, iCD79a
median fluorescence intensity; Refs. 1,50). A threshold
T-ALL all 3 of: (i)CD3pos,c CD7pos; marker is then placed to delimitate the negative region
iMPO negative or weak as defined above and to assess the proportion of blasts
AML 2 positive of: iMPO, CD13, CD33, outside this region (Fig. 1; Supporting Information Fig.
CD64, CD65, CD117 3). A threshold 10% of gated blasts is used for all anti-
and: BCP-/T-ALL criteria
not met gens (independently from surface or intracellular loca-
tion) to consider an antigen as “positive.” Hence, posi-
a
Of note, these markers are relevant for dominant lineage tivity is only accepted once a fluorescence shift is large
assignment, but are insufficient for a thorough description of enough to exceed potential background variability
leukemic immunophenotypes.
b
BCP-ALL needs strong positivity in 2 of the four antigens
(1,50–52). However, threshold markers are not used for
– in the rare case of CD19-negativity, specifically CD10 must partitioning of otherwise single blast populations for
be strong positive. Mind that rare cases of MLL-rearranged reporting of proportions of cells aside from the marker.
BCP-ALL may drop out of this scheme due to biology-inherent In full concordance with the WHO-style tri-partite
lack of CD10, as well as weak (i)CD22 and iCD79a expression consensus rating (negative-weak-strong), a more detailed
(CD19 is then usually strong positive).
c
For T-ALL, iCD3 positivity must be either strong, or if rated further description of antigen distribution and intensity
weak, CD2 and/or CD5 should be any positive in addition. is recommended (e.g., Bethesda-style rating; Fig. 1).
Surface CD3 expression needs to be tested in addition. According to the latter, in particular immunophenotypic
subclones of leukemic blasts can be delineated upon
permeabilized and one non-permeabilized) stained only description of only partially expressed markers. This
with the CD45 reagent. With these tubes we discrimi- may occur in both, the weak and the strong positive
nate blasts from other cell populations according to dif- expression categories. Such fine-rating makes use of a
ferential CD45 expression and then estimate the poten- descriptive and semi-quantitative scale rather than an
tially different autofluorescence of blasts and control exact enumeration procedure (2). The focus of such a
cells in the unstained channels. The higher level of the more detailed interpretation is more on robustness in
two background regions (from in-sample cell control or daily routine and less on biologic accuracy. Accordingly,
autofluorescence) should be regarded as more relevant. an exception exists with respect to broadly expressed
With paucicellular specimens, where we need to priori- pan-leukocyte markers such as CD45 and CD11a, as well
tize tubes in order to run the full antigen panel, we pre- as CD38, CD58, and CD99. With those, the system of in-
fer the control tubes containing only the CD45 reagent sample negative-control lymphocyte subsets cannot easi-
because giving most information except for CD45 back- ly be applied because lymphocytes of all lineages are
positive with these markers. In case of CD45 and
ground itself. It should be emphasized that we do not
CD11a, expression on blasts is compared to the
recommend the usage of isotype controls as has been
unstained channel control (negative) as well as to residu-
established before by several authoritative publications
al mature neutrophils (SSChigh and/or CD15bright) in the
(43,44,47–49). This omission relies on the use of well-
sample which are positive (Fig. 1). In case of CD38,
described internal control populations available now
CD58, and CD99, blasts should be compared to those
that whole blood or BM samples are used for FCM. residual SSClow lymphocyte subsets with a higher
However, in case of totally missing cross-lineage control expression of the marker: if lower than those but not
cells as well as in particular in situations of suspected negative, rating is dim positive; if similar in expression,
artifactual results, fluorescence minus one (FMO) tests the rating is medium; if stronger, the rating is bright
are pertinent. These are also advised for learning expres- (e.g., this is the type of CD38 expression in plasma
sion patterns with new combinations of antibodies (44). cells); if broadly overlapping, heterogeneous. Mind that
Definition of Antigen Expression the latter rating is related to a certain (broad) pattern of
expression of a single marker. This has to be distin-
The consortium follows in its guidelines recent devel- guished from the term blast population heterogeneity
opments in the field with respect to (i) increased techni- (see below), which is used to describe immunopheno-
cal sensitivity which renders previously applied higher typic subset formation in a given case of leukemia.
cut-offs for positivity (e.g., 20%) obsolete (44), and (ii)
the rejection of reporting antigen expression solely by Blast Percentage Recording
percentages (1,2,5,44,50). Antigen expression of the gat- The percentage of blasts among total NC (see above;
ed blasts is determined by assessing the fluorescence Ref. 27) of the sample is recorded. Notably, formal diag-
shift and distribution pattern of the leukemic population nosis of leukemia and treatment indication usually
against appropriate control as delineated before depends on morphologic blast enumeration respecting
(1,2,5,50). The amount of deviation from negative is fur- conventional thresholds (e.g., 25% as per AIEOP-BFM
ther used to differentiate positivity into two major ALL 2009 protocol, or 20% as per WHO 2008

Cytometry Part B: Clinical Cytometry

 IBFM-FLOW STANDARDS FOR IMMUNOPHENOTYPING OF PEDIATRIC ALL 87

Table 3
The AIEOP-BFM subclassification of ALLa
Subtype Discriminators Remarks
B-I (pro-B) CD10neg BCP-ALL lineage criteria fulfilled
B-II (common) CD10pos /
B-III (pre-B) iIgMpos CD10neg or weak pos may occurb
B-IV (mature B) j- or k-chainpos may occur with FAB L1/L2 morphologyc
T-I (pro-T)d only iCD3pos and CD7pos T-ALL lineage criteria fulfilled
T-II (pre-T) 1 of CD2pos, CD5pos, CD8pos surface (s) CD3weak pos allowede
T-III (cortical T) CD1apos sCD3weak may occure
neg
T-IV (mature T) CD1a and sCD3pos sCD3 strong
, or sCD3weak pos with TCRpos
ETP (only additive to CD1aneg, CD8neg if CD5strong pos: 2pos of HLADR,
T-I or T-II) usually CD5neg or weak pos and 1pos of CD11b,13,33,34,65,117;
HLADR, CD11b,13,33,34,65,117 sCD3weak pos may occure
a
Adapted from Refs. 8 and 9.
b
CD10neg/weak B-III is frequently associated with MLL-rearrangements (12).
c
Light-chainpos cases without FAB L3-morphology and without MYC-translocation are eligible for conventional ALL treatment, and
thus must be separated from Burkitt-type mature B-ALL (40–43).
d
T-I is very rare and can be reported together with T-II (as T-I/II).
e
Dim or even more frequently partial surface positivity with CD3 (e.g., in a minor blast subpopulation) occurs when sensitive
methodology is used and should not mislead to diagnose mature T-ALL in the absence of TCR expression.

definition). This threshold requirement is, however, qualify for the label “heterogeneous blast population,”
independent of the process of immunophenotyping of such cases must either fulfil the MPAL/BAL- or ETP-
the leukemia. Hence, the leukemic immunophenotype criteria by inter-lineage clonal heterogeneity (see Sup-
can be determined as long as the blast population can porting Information Fig. 3), or exhibit intra-lineage clon-
be clearly distinguished from normal background cells al heterogeneity that would lead to different ALL sub-
regardless of numerical blast-percentage cut-off (53). We type classifications (see Supporting Information Fig. 4A
consider the fact of nearly systematic hemodilution of and B) when assessing blast subsets separately. To this
BM aspirations collected for immunophenotyping, end, the following antigens are of interest in ALL when
which indeed hampers fulfilling criteria derived from only partially positive: CD10, (i)IgM, j-chain, k-chain,
microscopic traditions. In such case, with a malignant CD1a, CD3, CD5, TCRab, TCRgd.
blast percentage <25% by FCM, however, the immuno- In addition, very rare situations of interlineage hetero-
phenotypic report should not anticipate formal diagnosis geneity exist where both, clearly separate blast subpopu-
of ALL but state, for example, “malignant lymphoblasts lations of different lineages as well as further subsets
suggestive of ALL with a low blast percentage.” interconnecting the former occur (see Supporting Infor-
mation Fig. 4C), which can best be summarized as (bilin-
Reporting of Blast Immunophenotype – Blast Population eal) MPAL cases with “complex immunophenotype.”
Heterogeneity Possible are also situations where an antigen is rated
The leukemic cell population can exhibit heterogene- clearly negative in the great majority of blasts, but a
ities. Therefore, it must strictly be avoided that immuno- minute (<10%) subset is unambiguously positive. This is
phenotypic reporting lumps together features of imma- typical for CRLF2, NG2, and other antigens (see Support-
ture blasts and maturing cells (not an issue in ALL, but ing Information Fig. 5). Such an antigen expression on
typical for myeloid leukemias with maturation), as well very minor subsets of blasts should not be rated partially
as of heterogeneous blast populations by reporting positive according to our definition, but deserves report-
many antigens as partially positive. Rather, samples must ing because of potential biologic relevance (e.g., by
vigilantly be assessed for signs of more than one immu- “blast heterogeneity” and by a description of that sub-
nophenotypic blast “clone.” Such cases should be classi- population in the diagnostic summary).
fied by an extra label as “heterogeneous blast pop-
ulation.” Two major situations correspond to this: First, Dominant Lineage Assignment
leukemia with separate, immunophenotypically largely Inclusion into AIEOP-BFM ALL treatment trials is
non-interconnected blast clones (typically occurring as based on morphologically and genetically defined lym-
bilineal). In such case, the detailed phenotype should be phoblastic leukemia with a dominant ALL-
reported for each blast clone separately (including per- immunophenotype according to the AIEOP-BFM lineage
centage among NC, lineage and subtype assignment, assignment criteria (Table 2) adapted from Mejstrikova
antigen expression pattern, for example, each in the et al. (20). This is irrespective of whether or not a case
Flow Diagnostics Essential Code [FDE-code] format; fulfils also criteria of MPAL (WHO 2008/2016, Refs. 3,4,
Ref. 53). Second, leukemia with a branched, interrelated and 16; Supporting Information Table 1) or BAL (EGIL,
blast population, in which such heterogeneity is cap- Refs. 9 and 10; Supporting Information Table 2)—even
tured by reporting antigens as partially positive. To in case of MPO positivity—in a single blast population.

Cytometry Part B: Clinical Cytometry

88 DWORZAK ET AL.

FIG. 2. Diagnostic algorithm based on the AIEOP-BFM consensus guidelines. This algorithm is used to define the leukemic immunophenotype as pre-
requisite for clinical trial inclusion to AIEOP-BFM ALL trials. Note that dominant lineage assignment can lead to a diagnosis of ALL according to the
AIEOP-BFM consensus despite fulfilled MPAL criteria according to WHO 2008/2016. LIN refers to usage of a lineage marker for blast cell gating as
appropriate. A prefix in parenthesis indicates that the respective marker needs to be tested both on surface as well as intracellular.

Nevertheless, apart from assigning the dominant lineage blasts is found, such case is designated per AIEOP-BFM
of a single or simply branched leukemic clone (e.g., by rules as ALL fulfilling also MPAL-criteria and may enter
partial positivity with MPO), a finding of fulfilled MPAL/ the ALL protocol for reasons of clinical pragmatism.
BAL-criteria needs to be reported as secondary detail. However, the admixture and type of aberrant myelo-
Thus, dominant lineage assignment overrules French- blasts—which adds the MPAL label to this case—must
American-British (FAB)-classification as well as MPAL/ be described in the summary report. Notably, before
BAL designations for clinical decision making in AIEOP- accepting the MPAL label based on such suspicious cell
BFM with the—important—exception of cases with subsets it has to be ruled out by careful phenotypic
more than one (separated or complex) blast population investigation that these cells just represent remaining
including a non-lymphoblastic component (see Support- normal hematopoiesis.
ing Information Fig. 6). In such case, dominant lineage Of note, originally the EGIL did not consider B/T-BAL
assignment is not applicable as far as both separate com- (9), however, newer observations notify the existence of
ponents comprise each 10% of cells of the sample. In such rare occurrences including also cases of tri-lineage
other words, a (probably MLL-rearranged) MPAL case differentiation (16,17,23). Hence, also such cases should
with a lymphoblast subset of 10% and 10% leukemic be recorded as BAL using the EGIL scoring system. In B/
monoblasts (of total cells) is reported as bilineal leuke- T-BAL, the higher EGIL-rank will favor dominant lineage
mia (neither as AML or ALL, therefore not to be includ- assignment for AIEOP-BFM ALL trial treatment arm (B or
ed into AIEOP-BFM ALL trials). By contrast, if in a case T) selection.
of leukemia with >90% typical lymphoblasts an addi- Cases not fitting into the ALL category will be classi-
tional small subset of aberrant myeloid differentiated fied as either AML, AUL (acute undifferentiated

Cytometry Part B: Clinical Cytometry

 IBFM-FLOW STANDARDS FOR IMMUNOPHENOTYPING OF PEDIATRIC ALL 89

Table 4 a subset of blasts (in 40–50% of B-III cases; see Support-

Contents of immunophenotyping reports in AIEOP-BFM trials ing Information Fig. 4A), hence all observed differentia-
Item Contents
tion stages (in populations of blasts 10%) should be
reported in the subtype classification (e.g., B-II/III). In a
Identification Name, code number, date of birth,
question/suspicion, type of sample,
minor subgroup of cases with iIgM expression, CD10 is
dates of sampling, processing/ negative or only weakly positive. This does not alter des-
analysis, reporting, name/ ignation as B-III, but is recognized as frequently herald-
signature(s) of technical analyst ing MLL-gene rearrangements (usually NG2 positive, but
and medical reporter not MLL/AF4; Ref. 13). If iIgM is only partially
Antigen expression “WHO-style” rating: negative, positive
weak, positive strong expressed, such cases might even be classified B-I/III.
“Bethesda-style” rating: as add-on, at Rare cases expressing immunoglobulin light-chains (sur-
least partially positive face or cytoplasmic) occur without further features of
Dominant lineagea B-, T-, Myeloid-lineage independently mature Burkitt-type lymphocytic leukemia (like FAB L3-
of MPAL/BAL-status; not applicable
if bilineal leukemia with
morphology, MYC-translocation, and a high proliferation
non-lymphoblastic component, rate/high S-phase; Refs. 56–58). Such unusual cases (or
AUL, NK/myeloid precursor- or blast subpopulations) should also be given the B-IV des-
DC-leukemia ignation as appropriate (isolated or as part of the classifi-
Subtypeb B-I to B-IV, T-I to T-IV; report ETP as cation, e.g., B-III/IV), and they are eligible for inclusion
additive result
Heterogeneities WHO 2008 MPAL, adapted EGIL BAL into conventional ALL treatment trials similar to B-I to B-
criteria (also B/T and B/T/My) III. They frequently exhibit either MLL-rearrangements
(Intra-lineage) blast clone or other precursor ALL-type genetic lesions (56,58,59).
heterogeneity In T-ALL, similar to physiologic thymocyte differentia-
Quantities Blast percentage among NC (if
applicable, separate by subclones)
tion (60), CD3 starts to be expressed on the cell surface
Conclusions Diagnosis as relevant for treatment already in rather early maturation stages and not only in
choice; additional information of stage T-IV. Surface expression of (s)CD3 alone therefore
potential clinical interest (e.g., poorly discriminates T-ALL subtypes unless strongly posi-
toward underlying genetic lesions) tive or co-expressed together with T-cell receptors.
a
Bilineal leukemia with a lymphoblastic and a separate non- Hence, it should be emphasized that CD1a expression,
lymphoblastic blast subset, as well as AUL, NK/myeloid pre- irrelevant of the presence (indeed already occurring) of
cursor- and DC-leukemia are excluded from AIEOP-BFM ALL sCD3 defines T-III ALL in the EGIL acceptance. More-
trials. over, T-ALL very frequently exhibits blast population het-
b
Combinations of subtype labels are appropriate in case of
blast subset formation at different maturation levels; B-IV with
erogeneities due to immunophenotypic subclone forma-
FAB-L3 morphology and MYC-rearrangements (“Burkitt-type”) tion, and this typically involves sCD3 (see Supporting
are excluded from AIEOP-BFM ALL trials, whereas B-IV cases Information Fig. 4A and B; in about 50% of T-ALLs). Like-
with L1/L2-morphology and other genetic lesions are included. wise, CD3 surface expression (without TCR ab or gd
expression) does not exclude ETP (12). This prognosti-
leukemia), truly bilineal/“complex” immunophenotype cally relevant new subtype is reported as additional
acute leukemia, or very rare entities like NK/myeloid result along with the classical T-I to T-IV subtyping.
precursor or dendritic cell (DC)-derived leukemia (3,16). AIEOP-BFM criteria for ETP definition are stringent as
As mentioned before, these categories are not consid- per references 12,25: negative for CD1a and CD8, usual-
ered apt for inclusion into AIEOP-BFM ALL trials. Nota- ly CD5 negative or weak positive, as well as strong posi-
bly, it should be made sure to exclude also AML FAB M7 tive for at least one of CD34, CD117, HLADR, CD13,
(positive for CD41, CD61, CD42b, and CD36; typically CD33, CD11b, and CD65. In case of strong positivity of
also negative for CD11a; Ref. 54). CD5, at least two of the latter must be strong positive
(25,61). There is also novel evidence that the myeloid
Immunophenotypic Subtype Classification of ALL marker CD371 may be positive in ETP (compare Sup-
porting Information Fig. 3).
The AIEOP-BFM consortium has a long tradition in
using the EGIL-based subclassification for ALL subtyping Clinical Reporting
and description of patient cohorts (8–10,55). Based on The report should allow grasping a comprehensive
this frame, we designed an AIEOP-BFM ALL subtype clas- picture of the leukemic case as derived from immuno-
sification which includes also ETP as an additional vari- phenotyping according to the AIEOP-BFM algorithm (see
able (Table 3). These criteria assign the leukemic sub- Fig. 2). This includes case identification items, the data
type within a given dominant ALL lineage (Table 3) on individual marker expression patterns, and an inte-
using the described AIEOP-BFM antigen expression defi- grated classification of primary (dominant) lineage, as is
nitions (Table 2). necessary for trial inclusion and treatment selection, pat-
The following leukemic phenotype peculiarities are terns and heterogeneities of differentiation with aspects
recognized along the major lines of subclassification: of subset formation, as well as the quantitative aspects
Intracellular IgM expression is frequently found in only of the blast population(s) in the sample (Table 4). A

Cytometry Part B: Clinical Cytometry

90 DWORZAK ET AL.

summary or conclusion should state the overriding diag- exhibiting a differentiation drift from a subset with a
nosis and additional information of potential clinical or lower maturation level toward one with a higher. We
diagnostic interest (for example a probable genotype considered that describing rather the comprehensive
association). Results should be reported in a standard- subtype picture of a given ALL (like B-II/B-III) might be
ized format which may make use of the FDE code (for more useful than simplification (like quoting only B-III)
an example see Supporting Information Fig. 6) which in particular since subtype simplification is not neces-
has been designed in particular also for database captur- sary for any current clinical choices. Analyses of larger
ing of immunophenotype information (53). case cohorts are needed to clinically validate this more
detailed subtyping.
DISCUSSION The importance we put on discriminating blast cell
In elaborating the guidelines presented herein, we heterogeneities in the sense of subset formation directly
considered that technical standardization alone is insuffi- led to the additional nomenclature we adopted with
cient if not devoted to answer questions pertinent to respect to antigen distribution, as originally stated in the
established classification systems or clinical needs, 2006 Bethesda recommendations (2). From this nomen-
whereas classification devoid of practical standards fos- clature, the annotation “partially expressed” has the
ters solecism. Hence, we considered taking advantage of most relevant impact, whereas the distinction between
the possibilities of multicolor FCM by embedding it into weak and strong expression, as suggested in the 2008
the context of clinical trial applications. Consequently, WHO guidelines, appears sufficient for most other ques-
in order to keep up with existing classification systems tions asked in leukemia immunophenotyping apart from
we needed to regulate the impact of the increased sensi- scientific analyses (3,16). It needs to be determined
tivity of current multi-color FCM on the reporting. Not whether more exact, automated, and even compound
each minute fluorescence shift or miniscule blast subset expression evaluation, bar-coding, or vector-based statis-
expressing a marker should lead to results like “antigen tical pattern transformation are of additive value (6). It
positive” in representing the leukemic case in total. We will be most important for all such automation or statis-
therefore needed to choose a still artificial, but pragmat- tical approaches not to disregard and appropriately deal
ic threshold as discriminator of positive expression. It with blast heterogeneities and sub-clone formation.
also means that in certain situations we may rate an anti- For a “WHO-style” negative/weak/strong-distinction,
gen as negative in our report, oriented on the majority we realized that elaborate rules are necessary for con-
of blasts, but positive in a separate blast subclone of cordant application in a consortium. Our validation data
<10% which should not be overlooked and still be (Supporting Information Fig. 7) underscore that the
reported (see Supporting Information Fig. 5). More gen- approach of quantifying non-overlap of blasts with nega-
erally speaking and based on our experience with MRD tive control renders synonymous results regarding CD19
assessment, we believe that also at initial immunopheno- and iCD3 expression grading compared to overlap
typing exists no simple cut-off regarding percentage of assessment with residual antigen-positive lymphocytes as
cells in- or excluding leukemia (compare Ref. 24). Rath- per WHO criteria. Based on this, we agreed to use the
er, we support reporting phenotypically aberrant popu- “non-overlap” approach for our grading. This is broadly
lations suggestive of ALL even when blast percentages applicable to many different antigens and even in sam-
are too small for a formal diagnosis of overt leukemia. ples lacking positive normal populations, either mature
Based on the higher resolution ability of multicolor or immature. Immature counterpart populations have
FCM, we could introduce the specification “blast clone been proposed by others as gauges for rating leukemic
heterogeneity” which relates to such situations where expressions (2,44). However, often such normal cells
leukemic subsets with different maturation levels (or of are not found in leukemic samples at diagnosis, which
different lineage) are seen. Commonly such phenotypic hinders their use as direct in-sample comparators.
subclone formation occurs with antigens like m-chains Concordant rating of antigen expressions as weak or
(iIgM), CD3, and TCR on the cell surface (Supporting strong is most important also for making a diagnosis of
Information Fig. 4), which can lead to perturbation of MPAL according to WHO (3,4,16). “Simple” MPAL cases
leukemia subclassification. As a consequence, we inten- (mostly BCP-ALL with a single blast clone and some
sified group training and awareness on this matter to MPO-positivity) are now included into the AIEOP-BFM
eradicate respective solecism for the future. We also ALL treatment strategy because it was noted that pediat-
specified our EGIL-based subclassification to avoid perti- ric MPAL (or BAL) often does not herald a dismal out-
nent flaws in T-II vs.T-IV distinction by including TCR come when it occurs together with ALL features (includ-
expression and CD3 grading into the diagnostic interpre- ing findings of microscopy and genetics like ETV6/
tation (Table 3). We also adapted the classification to RUNX1- or other ALL-specific rearrangements)(17–22).
allow for combinations of differentiation subtype-labels Hence, we put all efforts into tracing and exploring
where appropriate (Supporting Information Fig. 4). For MPAL cases fulfilling criteria of bilineal or “complex”
example, this means that we rather do not assign homo- immunophenotypes (see Supporting Information Fig. 6).
geneous leukemic subtype labels based on either highest Only such more extreme MPALs most probably need
differentiation level reached or on the larger proportion special therapy in children, which, however, is yet not
of blasts when there is a heterogeneous blast population well established. Along this line, the AIEOP-BFM

Cytometry Part B: Clinical Cytometry

 IBFM-FLOW STANDARDS FOR IMMUNOPHENOTYPING OF PEDIATRIC ALL 91

consortium puts emphasis on a diagnostic net effect, 2. Wood BL, Arroz M, Barnett D, DiGiuseppe J, Greig B, Kussick SJ,
respecting all information from microscopy, genotyping, Oldaker T, Shenkin M, Stone E, Wallace P. 2006 Bethesda Interna-
tional Consensus recommendations on the immunophenotypic anal-
and immunophenotyping, thus reducing the relevance ysis of hematolymphoid neoplasia by flow cytometry: Optimal
of MPO assessment (by FCM and/or cytochemistry) and reagents and reporting for the flow cytometric diagnosis of hemato-
poietic neoplasia. Cytom B Clin Cytom 2007;72 Suppl 1:S14–S22.
antigen scoring (9,10) for guiding treatment choice. This
3. Vardiman JW, Thiele J, Arber DA, Brunning RD, Borowitz MJ, Porwit
is also in line with the recent 2016 update of the WHO A, Harris NL, Le Beau MM, Hellstr€ om-Lindberg E, Tefferi A, et al.
classification where the meaning of some MPO positivity The 2008 revision of the World Health Organization (WHO) classifi-
in an otherwise indisputable ALL has been diminished cation of myeloid neoplasms and acute leukemia: Rationale and
important changes. Blood 2009;114:937–951.
(4). 4. Arber DA, Orazi A, Hasserjian R, Thiele J, Borowitz MJ, Le Beau
We emphasize that also technical issues—as shown in MM, Bloomfield CD, Cazzola M, Vardiman JW. The 2016 revision to
Supporting Information Figs. 1 and 2—are relevant the World Health Organization classification of myeloid neoplasms
and acute leukemia. Blood 2016;127:2391–2405.
causes of flaws and pitfalls in immunophenotyping (e.g., 5. Bene MC, Nebe T, Bettelheim P, Buldini B, Bumbea H, Kern W,
sample quality, permeabilization procedure, antibody Lacombe F, Lemez P, Marinov I, Matutes E, et al. Immunophenotyp-
clones, compensation, and negative control strategy). ing of acute leukemia and lymphoproliferative disorders: A consen-
sus proposal of the European LeukemiaNet Work Package 10. Leu-
For example, we learned to do better with iCD22 assess- kemia 2011;25:567–574.
ment, which we found quite sensitive to choice of per- 6. van Dongen JJ, Lhermitte L, B€ ottcher S, Almeida J, van der Velden
meabilization and antibody (clone RFB4 giving best VH, Flores-Montero J, Rawstron A, Asnafi V, L ecrevisse Q, Lucio P,
et al. EuroFlow antibody panels for standardized n-dimensional flow
results, data not shown). Whether there is a need for cytometric immunophenotyping of normal, reactive and malignant
more standardization (e.g., regarding antibody choice in leukocytes. Leukemia 2012;26:1908–1975.
general and combination set-up) as fostered by the Euro- 7. Kalina T, Flores-Montero J, van der Velden VH, Martin-Ayuso M,
B€ottcher S, Ritgen M, Almeida J, Lhermitte L, Asnafi V, Mendonça A,
Flow consortium (6), will depend on outcome of our et al. EuroFlow standardization of flow cytometer instrument set-
ongoing panel validation process. Instead of a full unifi- tings and immunophenotyping protocols. Leukemia 2012;26:1986–
cation, it might be relevant for a consortium like ours to 2010.
8. van der Does-van den Berg A, Bartram CR, Basso G, Benoit YC,
consider a set of common reagents against certain anti- Biondi A, Debatin KM, Haas OA, Harbott J, Kamps WA, K€ oller U,
gens of very high diagnostic importance or with a rele- et al. Minimal requirements for the diagnosis, classification, and
vant degree of divergence when tested with different evaluation of the treatment of childhood acute lymphoblastic leuke-
mia (ALL) in the “BFM Family” Cooperative Group. Med Pediatr
antibodies. Certain less well investigated antibodies Oncol 1992;20:497–505.
against novel or research markers might also need a 9. Bene MC, Castoldi G, Knapp W, Ludwig WD, Matutes E, Orfao A,
higher degree of standardization. Along this line, we are van’t Veer MB. Proposals for the immunological classification of
currently evaluating the impact of operator interpreta- acute leukemias. European Group for the Immunological Characteri-
zation of Leukemias (EGIL). Leukemia 1995;9:1783–1786.
tion on results as well as of using a fully standardized 10. Bene MC, Bernier M, Casasnovas RO, Castoldi G, Knapp W, Lanza F,
panel as compared to local practice (manuscript in prep- Ludwig WD, Matutes E, Orfao A, Sperling C, et al. The reliability
aration). Eventually, we may also be able to establish and specificity of c-kit for the diagnosis of acute myeloid leukemias
and undifferentiated leukemias. The European Group for the Immu-
which markers are essential for accurate classification nological Classification of Leukemias (EGIL). Blood 1998;92:596–
according to our designated consensus and which could 599.
be left aside (2,62). 11. Ratei R, Karawajew L, Lacombe F, Jagoda K, Del Poeta G, Kraan J,
De Santiago M, Kappelmayer J, Bj€ orklund E, Ludwig WD, et al. Nor-
In conclusion, we see our guidelines presented herein mal lymphocytes from leukemic samples as an internal quality con-
as a first step of necessary inter-laboratory standardiza- trol for fluorescence intensity in immunophenotyping of acute leu-
tion in order to serve patients to the best we can by kemias. Cytom B Clin Cytom 2006;70:1–9.
12. Coustan-Smith E, Mullighan CG, Onciu M, Behm FG, Raimondi SC,
contemporary immunophenotyping. Pei D, Cheng C, Su X, Rubnitz JE, Basso G, et al. Early T-cell precur-
sor leukaemia: A subtype of very high-risk acute lymphoblastic leu-
ACKNOWLEDGEMENTS kaemia. Lancet Oncol 2009;10:147–156.
13. Attarbaschi A, Mann G, K€ onig M, Steiner M, Strehl S,
This study was performed under the auspices of the Schreiberhuber A, Schneider B, Meyer C, Marschalek R, Borkhardt
International-BFM Study Group. The authors also thank A, et al. Mixed lineage leukemia-rearranged childhood pro-B and
CD10-negative pre-B acute lymphoblastic leukemia constitute a dis-
Andishe Attarbaschi, Grit Czerwony, Marianne Dunken, tinct clinical entity. Clin Cancer Res 2006;12:2988–2994.
Claudia M€ oller, Simona Sala, Rosa Kornm€ uller and Sus- 14. Mullighan CG, Collins-Underwood JR, Phillips LA, Loudin MG, Liu
anne Suhendra for their excellent contribution or assis- W, Zhang J, Ma J, Coustan-Smith E, Harvey RC, Willman CL, et al.
Rearrangement of CRLF2 in B-progenitor- and Down syndrome-
tance in data analysis. We are grateful to Renate Panzer- associated acute lymphoblastic leukemia. Nat Genet 2009;41:1243–
Gr€umayer and co-workers, as well as to Margit K€ onig for 1246.
their valuable contributions to the full characterization 15. Zangrando A, Intini F, te Kronnie G, Basso G. Validation of NG2 anti-
of certain exemplary leukemias described in the manu- gen in identifying BP-ALL patients with MLL rearrangements using
qualitative and quantitative flow cytometry: A prospective study.
script. And we thank all doctors, nurses, documenta- Leukemia 2008;22:858–861.
rists, and laboratory technicians of the involved clinical 16. Borowitz MJ, Bene M-C, Harris NL, Porwit A, Matutes E. Acute leu-
centers for their close collaboration. kaemias of ambiguous lineage. In: Swerdlow S, Campo E, Harris NL,
et al, editors. WHO Classification of Tumours of Haematopoietic
and Lymphoid Tissues. Lyon: IARC, 2008; pp 149–155.
LITERATURE CITED 17. Rubnitz JE, Onciu M, Pounds S, Shurtleff S, Cao X, Raimondi SC,
1. Basso G, Buldini B, De Zen L, Orfao A. New methodologic Behm FG, Campana D, Razzouk BI, Ribeiro RC, et al. Acute mixed
approaches for immunophenotyping acute leukemias. Haematolog- lineage leukemia in children: The experience of St Jude Children’s
ica 2001;86:675–692. Research Hospital. Blood 2009;113:5083–5089.

Cytometry Part B: Clinical Cytometry

92 DWORZAK ET AL.

18. Al-Seraihy AS, Owaidah TM, Ayas M, El-Solh H, Al-Mahr M, Al- role of P2RY8-CRLF2-harboring clones as main relapse-driving force
Ahmari A, Belgaumi AF. Clinical characteristics and outcome of chil- in childhood ALL. Blood 2012;120:5134–5142.
dren with biphenotypic acute leukemia. Haematologica 2009;94: 37. Attarbaschi A, Morak M, Cario G, Cazzaniga G, Ensor HM, te
1682–1690. Kronnie T, Bradtke J, Mann G, Vendramini E, Palmi C, et al. Treat-
19. Gerr H, Zimmermann M, Schrappe M, Dworzak M, Ludwig WD, ment outcome of CRLF2-rearranged childhood acute lymphoblastic
Bradtke J, Moericke A, Schabath R, Creutzig U, Reinhardt D. Acute leukaemia: A comparative analysis of the AIEOP-BFM and UK NCRI-
leukaemias of ambiguous lineage in children: Characterization, prog- CCLG study groups. Br J Haematol 2012;158:772–777.
nosis and therapy recommendations. Br J Haematol 2010;149:84– 38. Vaskova M, Mejstrikova E, Kalina T, Martinkova P, Omelka M, Trka J,
92. Stary J, Hrusak O. Transfer of genomics information to flow cytome-
20. Mejstrikova E, Volejnikova J, Fronkova E, Zdrahalova K, Kalina T, try: Expression of CD27 and CD44 discriminates subtypes of acute
Sterba J, Jabali Y, Mihal V, Blazek B, Cerna Z, et al. Prognosis of chil- lymphoblastic leukemia. Leukemia 2005;19:876–878.
dren with mixed phenotype acute leukemia treated on the basis of 39. Attarbaschi A, Mann G, Schumich A, K€ onig M, Pickl WF, Haas OA,
consistent immunophenotypic criteria. Haematologica 2010;95:928– Gadner H, Dworzak MN. CD44 deficiency is a consistent finding in
935. childhood Burkitt’s lymphoma and leukemia. Leukemia 2007;21:
21. Steiner M, Attarbaschi A, Dworzak M, Strobl H, Pickl W, Kornm€ uller 1110–1113.
R, Haas O, Gadner H, Mann G, Austrian Berlin-Frankfurt-M€ unster 40. Slamova L, Starkova J, Fronkova E, Zaliova M, Reznickova L, van
Study Group. Cytochemically myeloperoxidase positive childhood Delft FW, Vodickova E, Volejnikova J, Zemanova Z, Polgarova K,
acute leukemia with lymphoblastic morphology treated as lympho- et al. CD2-positive B-cell precursor acute lymphoblastic leukemia
blastic leukemia. J Pediatr Hematol Oncol 2010;32:e4–e7. with an early switch to the monocytic lineage. Leukemia 2014;28:
22. Matutes E, Pickl WF, Van’t Veer M, Morilla R, Swansbury J, Strobl H, 609–620.
Attarbaschi A, Hopfinger G, Ashley S, Bene MC, et al. Mixed-pheno- 41. Lanza F, Latorraca A, Moretti S, Castagnari B, Ferrari L, Castoldi G.
type acute leukemia: Clinical and laboratory features and outcome Comparative analysis of different permeabilization methods for the
in 100 patients defined according to the WHO 2008 classification. flow cytometry measurement of cytoplasmic myeloperoxidase and
Blood 2011;117:3163–3171. lysozyme in normal and leukemic cells. Cytometry 1997;30:134–
23. Weinberg OK, Arber DA. Mixed phenotype acute leukemia: Histori- 144.
cal overview and a new definition. Leukemia 2010;24:1844–1851. 42. Kappelmayer J, Gratama JW, Karaszi E, Menendez P, Ciudad J, Rivas
24. Borowitz MJ. Mixed phenotype acute leukemia. Cytometry B Clin R, Orfao A. Flow cytometric detection of intracellular myeloperoxi-
Cytom 2014;86:152–153. dase, CD3 and CD79a: Interaction between monoclonal antibody
25. Inukai T, Kiyokawa N, Campana D, Coustan-Smith E, Kikuchi A, clones, fluorochromes and sample preparation protocols.
Kobayashi M, Takahashi H, Koh K, Manabe A, Kumagai M, et al. J Immunol Methods 2000;242:53–65.
Clinical significance of early T-cell precursor acute lymphoblastic 43. Wood B, Jevremovic D, Bene MC, Yan M, Jacobs P, Litwin V, ICSH/
leukaemia: Results of the Tokyo Children’s Cancer Study Group ICCS Working Group. Validation of cell-based fluorescence assays:
Study L99-15. Br J Haematol 2012;156:358–365. Practice guidelines from the ICSH and ICCS - part V - assay perfor-
26. Zhang J, Ding L, Holmfeldt L, Wu G, Heatley SL, Payne-Turner D, mance criteria. Cytom B Clin Cytom 2013;84:315–323.
Easton J, Chen X, Wang J, Rusch M, et al. The genetic basis of early 44. Johansson U, Bloxham D, Couzens S, Jesson J, Morilla R, Erber W,
T-cell precursor acute lymphoblastic leukaemia. Nature 2012;481: Macey M. Guidelines on the use of multicolour flow cytometry in
157–163. the diagnosis of haematological neoplasms. British Committee for
27. Dworzak MN, Gaipa G, Ratei R, Veltroni M, Schumich A, Maglia O, Standards in Haematology. Br J Haematol 2014;165:455–488.
Karawajew L, Benetello A, P€ otschger U, Husak Z, et al. Standardiza- 45. Owens MA, Vall HG, Hurley AA, Wormsley SB. Validation and quali-
tion of flow cytometric minimal residual disease evaluation in acute ty control of immunophenotyping in clinical flow cytometry.
lymphoblastic leukemia: Multicentric assessment is feasible. Cytom- J Immunol Methods 2000;243:33–50.
etry B Clin Cytom 2008;74:331–340. 46. Arnoulet C, Ben e MC, Durrieu F, Feuillard J, Fossat C, Husson B,
28. Karawajew L, Dworzak M, Ratei R, Rhein P, Gaipa G, Buldini B, Jouault H, Maynadie M, Lacombe F. Four- and five-color flow cytom-
Basso G, Hrusak O, Ludwig WD, Henze G, et al. Minimal residual etry analysis of leukocyte differentiation pathways in normal bone
disease analysis by eight-color flow cytometry in relapsed childhood marrow: A reference document based on a systematic approach by
acute lymphoblastic leukemia. Haematologica 2015;100:935–944. the GTLLF and GEIL. Cytom B Clin Cytom 2010;78:4–10.
29. Veltroni M, De Zen L, Sanzari MC, Maglia O, Dworzak MN, Ratei R, 47. Mahnke YD, Roederer M. Optimizing a multicolor immunopheno-
Biondi A, Basso G, Gaipa G. Expression of CD58 in normal, regener- typing assay. Clin Lab Med 2007;27:469–485.
ating and leukemic bone marrow B cells: Implications for the detec- 48. Hulspas R, O’Gorman MR, Wood BL, Gratama JW, Sutherland DR.
tion of minimal residual disease in acute lymphocytic leukemia. Considerations for the control of background fluorescence in clini-
Haematologica 2003;11:1245–1252. cal flow cytometry. Cytom B Clin Cytom 2009;76:355–364.
30. Dworzak MN, Fr€ oschl G, Printz D, De Zen L, Gaipa G, Ratei R, 49. Tanqri S, Vall H, Kaplan D, Hoffman B, Purvis N, Porwit A,
Basso G, Biondi A, Ludwig WD, Gadner H. CD99 expression in T- Hunsberger B, Shankey TV, ICSH/ICCS Working Group. Validation
lineage ALL: Implications for flow cytometric detection of minimal of cell-based fluorescence assays: Practice guidelines from the ICSH
residual disease. Leukemia 2004;18:703–708. and ICCS - part III - analytical issues. Cytom B Clin Cytom 2013;84:
31. Djokic M, Bj€ orklund E, Blennow E, Mazur J, S€ oderh€all S, Porwit A. 291–308.
Overexpression of CD123 correlates with the hyperdiploid geno- 50. Del Vecchio L, Brando B, Lanza F, Ortolani C, Pizzolo G, Semenzato
type in acute lymphoblastic leukemia. Haematologica 2009;94: G, Basso G. Recommended reporting format for flow cytometry
1016–1019. diagnosis of acute leukemia. Haematologica 2004;89:594–598.
32. Kalina T, Vaskova M, Mejstrikova E, Madzo J, Trka J, Stary J, Hrusak 51. Gratama JW, D’hautcourt JL, Mandy F, Rothe G, Barnett D, Janossy
O. Myeloid antigens in childhood lymphoblastic leukemia: Clinical G, Papa S, Schmitz G, Lenkei R. Flow cytometric quantitation of
data point to regulation of CD66c distinct from other myeloid anti- immunofluorescence intensity: Problems and perspectives. Europe-
gens. BMC Cancer 2005;5:38. an Working Group on Clinical Cell Analysis. Cytometry 1998;33:
33. Hrusak O, Porwit-MacDonald A. Antigen expression patterns reflect- 166–178.
ing genotype of acute leukemias. Leukemia 2002;16:1233–1258. 52. Lenkei R, Gratama JW, Rothe G, Schmitz G, D’hautcourt JL,
34. De Zen L, Orfao A, Cazzaniga G, Masiero L, Cocito MG, Spinelli M, Arekrans A, Mandy F, Marti G. Performance of calibration standards
Rivolta A, Biondi A, Zanesco L, Basso G. Quantitative multiparamet- for antigen quantitation with flow cytometry. Cytometry 1998;33:
ric immunophenotyping in acute lymphoblastic leukemia: Correla- 188–196.
tion with specific genotype. I. ETV6/AML1 ALLs identification. Leu- 53. Hrusak O, Basso G, Ratei R, Gaipa G, Luria D, Mejstrıkova E,
kemia 2000;14:1225–1231. Karawajew L, Buldini B, Rozenthal E, Bourquin JP, et al. Flow diag-
35. Bugarin C, Sarno J, Palmi C, Savino AM, te Kronnie G, Dworzak M, nostics essential code: A simple and brief format for the summary
Schumich A, Buldini B, Maglia O, Sala S, et al. Fine tuning of surface of leukemia phenotyping. Cytom Part B (Clin Cytom) 2014;86B:
CRLF2 expression and its associated signaling profile in childhood 288–291.
B-cell precursor acute lymphoblastic leukemia. Haematologica 2015; 54. Boztug H, Schumich A, P€ otschger U, M€ uhlegger N, Kolenova A,
100:e229–e232. Reinhardt K, Dworzak M. Blast cell deficiency of CD11a as a marker
36. Morak M, Attarbaschi A, Fischer S, Nassimbeni C, Grausenburger R, of acute megakaryoblastic leukemia and transient myeloproliferative
Bastelberger S, Krentz S, Cario G, Kasper D, Schmitt K, et al. Small disease in children with and without Down syndrome. Cytometry B
sizes and indolent evolutionary dynamics challenge the potential Clin Cytom 2013;84:370–378.

Cytometry Part B: Clinical Cytometry

 IBFM-FLOW STANDARDS FOR IMMUNOPHENOTYPING OF PEDIATRIC ALL 93

55. Ratei R, Schabath R, Karawajew L, Zimmermann M, M€ oricke A, lymphoblastic leukemia with MLL rearrangement: An uncommon
Schrappe M, Ludwig WD. Lineage classification of childhood acute and distinct subset of childhood acute leukemia. Leukemia 2008;22:
lymphoblastic leukemia according to the EGIL recommendations: 1056–1059.
Results of the ALL-BFM 2000 trial. Klin Padiatr 2013;225Suppl1: 60. Dworzak MN, Fritsch G, Buchinger P, Fleischer C, Printz D, Zellner
S34–S39. A, Sch€ollhammer A, Steiner G, Ambros PF, Gadner H. Flow cytomet-
56. Kansal R, Deeb G, Barcos M, Wetzler M, Brecher ML, Block AW, ric assessment of human MIC2 expression in bone marrow, thymus,
Stewart CC. Precursor B lymphoblastic leukemia with surface light and peripheral blood. Blood 1994;83:415–425.
chain immunoglobulin restriction: A report of 15 patients. Am J 61. Zuurbier L, Gutierrez A, Mullighan CG, Cante-Barrett K, Gevaert
Clin Pathol 2004;121:512–525. AO, de Rooi J, Li Y, Smits WK, Buijs-Gladdines JG, Sonneveld E,
57. Li S, Lew G. Is B-lineage acute lymphoblastic leukemia with a et al. Immature MEF2C-dysregulated T-cell leukemia patients have
mature phenotype and L1 morphology a precursor B-lymphoblastic an early T-cell precursor acute lymphoblastic leukemia gene signa-
leukemia/lymphoma or Burkitt leukemia/lymphoma?. Arch Pathol ture and typically have non-rearranged T-cell receptors. Haematolog-
Lab Med 2003;127:1340–1344. ica 2014;99:94–102.
58. Tsao L, Draoua HY, Osunkwo I, Nandula SV, Murty VV, Mansukhani 62. Ratei R, Karawajew L, Lacombe F, Jagoda K, Del Poeta G, Kraan J,
M, Bhagat G, Alobeid B. Mature B-cell acute lymphoblastic leukemia De Santiago M, Kappelmayer J, Bj€ orklund E, Ludwig WD, et al. Dis-
with t(9;11) translocation: A distinct subset of B-cell acute lympho- criminant function analysis as decision support system for the diag-
blastic leukemia. Mod Pathol 2004;17:832–839. nosis of acute leukemia with a minimal four color screening panel
59. Blin N, Mechinaud F, Talmant P, Garand R, Boutard P, Dastugue N, and multiparameter flow cytometry immunophenotyping. Leukemia
McIntyre EA, Harousseau JL, Avet-Loiseau H. Mature B-cell 2007;21:1204–1211.

Cytometry Part B: Clinical Cytometry