Transfusion Medicine | ORIGINAL ARTICLE
Hyperleukocytosis and leukocytapheresis in acute leukaemias:
experience from a single centre and review of the literature
of leukocytapheresis in acute myeloid leukaemia
Ø. Bruserud,1,2 K. Liseth,3 S. Stamnesfet,3 D. L. Cacic,3 G. Melve,3 E. Kristoffersen,3 T. Hervig3 & H. Reikvam1,2,3
1
Section of Haematology, Department of Clinical Science, University of Bergen, 2 Department of Medicine, Haukeland University Hospital,
and 3 Department of Immunology and Transfusion Medicine, Haukeland University Hospital, Bergen, Norway
Received 17 August 2012; accepted for publication 23 June 2013
SUMMARY
Background: Hyperleukocytosis is usually defined as leukocyte
count >100 × 109 L−1 and can be seen in newly diagnosed
leukaemias. Hyperleukocytic leukaemia is associated with a
risk of organ failure and early death secondary to leukostasis.
Mechanical removal of leukocytes by the apheresis technique,
leukocytapheresis, is a therapeutic option in these patients.
Methods: During a 16-year period, 16 patients were treated with
leukocytapheresis (35 apheresis procedures) for hyperleukocy-
tosis/leukostasis. We present our experience, and in addition we
review previous studies of hyperleukocytosis/leukocytapheresis
in patients with acute myeloid leukaemia (AML).
Results: We used a highly standardised approach for leuko-
cytapheresis in leukaemia patients with hyperleukocytosis. The
average leukocytapheresis number for each patient was 2·2
(range 1–6). Median leukocyte count before apheresis was
309 × 109 L−1 (range 104–935); the mean leukocyte count
reduction was 71%, corresponding to a mean absolute reduction
of 219 × 109 L−1 . No serious side effects were seen during or
immediately after apheresis.
Conclusions: The data suggest that our standardised technique
for leukocytapheresis effectively reduced the peripheral blood
leukaemia cell counts. Previous studies in AML also support
the conclusion that this is a safe and effective procedure for
the treatment of a potentially life-threatening complication, but
apheresis should always be combined with early chemotherapy.
Key words: acute myeloid leukaemia, hyperleukocytosis,
leukaemia, leukostasis.
Hyperleukocytosis is often defined as peripheral blood
leukocytes >50 × 109 L−1 or >100 × 109 L−1 (Dutcher et al.,
1987). The condition is associated with increased morbidity and
Correspondence: Prof. Øystein Bruserud, Section of Haematology,
Institute of Medicine, University of Bergen, Bergen, Norway.
Tel.: +55 97 50 00; fax: +55 97 29 50;
e-mail: Oystein.Bruserud@helse-bergen.no
© 2013 The Authors
Transfusion Medicine © 2013 British Blood Transfusion Society
mortality and can induce leukostasis, tumour lysis syndrome
and disseminated intravascular coagulopathy (DIC) especially
in patients with acute leukaemia. For some of the leukaemias,
hyperleukocytosis is more common in certain patient subsets,
e.g. in acute myeloid leukaemia (AML) hyperleukocytosis
seems to be associated with monocytic differentiation (FAB
classification M4–M5) (Cuttner et al., 1980, 1983).
The exact prevalence of hyperleukocytosis in acute leukaemia
is not known but it occurs probably in 5–10% of AML
patients and it is possibly somewhat higher (>20%) in
acute lymphoblastic leukaemia (ALL) (Cuttner et al., 1980;
Dutcher et al., 1987; Hoelzer et al., 1988). However, the
acute complications of hyperleukocytosis are closely linked
to leukostasis and seem to be most pronounced for cells
showing signs of myeloid differentiation (Bunin & Pui 1985).
The organs most commonly affected are the lungs and the
central nervous system (CNS), including the retina (Majhail
& Lichtin 2004). The main aim of the initial treatment of
hyperleukocytosis and/or leukostasis is therefore to reduce
the number of circulating leukaemia cells and thereby avoid
development of organ failure. The most common method for
rapid cytoreduction is leukocytapheresis that may be combined
with chemotherapy (Döhner et al., 2010). We present our
experience with leukocytapheresis in leukaemia patients with
a focus on apheresis technique and efficiency especially in
AML. We also give a detailed review of the previous studies of
leukocytapheresis in acute human AML.
MATERIALS AND METHODS
Patients included in the study
This retrospective study was approved by the Regional Ethics
Committee for Health Research. We included all leukaemia
patients receiving leukocytapheresis for hyperleukocytosis at
our institution during a 16-year period. The decision to perform
leukocytapheresis was taken by the responsible physician, and
these decisions were based on previously published guidelines
(Szczepiorkowski et al., 2010). The number of aphereses was
also decided by the treating physician.
doi: 10.1111/tme.12067
2 Ø. Bruserud et al.

Leukocytapheresis repeated daily until the WBC count was <100 × 109 L−1 , unless
the responsible clinician decided to end the leukocytapheresis
Our general guidelines for leukocytapheresis are described
treatment.
below. The leukocytapheresis procedure was performed with
a continuous-flow blood cell separator (COBE Spectra,
Gloucester, UK; software version 7.0) and cells were collected RESULTS
through a double lumen central vein apheresis catheter (Quinton
Mahurkar, Tyco/Kendall Healthcare, Wollerau, Switzerland). During a 16-year period, 35 leukapheresis procedures were
Anti-coagulation was administered according to the following performed in 16 patients with hyperleukocytosis (10 men and 6
women; median age 40 years, range 1–88 years). These patients
guidelines: acid citrate dextrose was added at a ratio of 1 : 12 for
had a variety of leukaemias and the most common diagnosis was
adults (body weight >30 kg) and for children (i.e. body weight
AML (Table 2). The other patients had ALL (five patients),
<30 kg) a ratio of 1 : 15 was used. Ionised calcium was measured
chronic myeloid leukaemia (CML; two patients), chronic
before and 2 h after start of leukocytapheresis, and prophylactic
myelomonocytic leukaemia (CMML; one patient) and chronic
calcium, 1 g of calcium gluconate orally, was administered to
lymphocytic leukaemia (CLL; one patient). Fourteen of the
all patients. The same dosage was given if patients experienced
patients had clinical symptoms, the most common manifestation
symptoms of hypocalcaemia, and this dose was repeated if
of hyperleukocytosis being dyspnoea with respiratory failure.
symptoms persisted. Hydroxyethyl starch or other sedimenting
All our patients fulfilled the criteria given by the American
agents were not used.
Society for Apheresis for performing leukocytapheresis in acute
Blood priming was used for all patients with body weight
leukaemia (Szczepiorkowski et al., 2010): (i) hyperleukocytosis
below 30 kg. This was done in order not to dramatically lower
associated with symptoms of leukostasis or (ii) sufficiently high
the haemoglobin (Hb) in smaller patients, since some Hb
leukocyte counts to consider prophylactic apheresis.
invariably will be lost in the leukapheresis procedure. We used
The average number of leukapheresis per patient was 2·2
blood priming to avoid further lowering the Hb value in smaller
(range 1–6), the median leukocyte count before the start of
patients. For these small patients, we believe that using blood
apheresis was 308·5 × 109 L−1 (range 104–935) and the median
priming is beneficial, since the risk of inducing symptomatic
Hb and platelet levels were 8·5 g dL−1 (range 4·2–10·6) and
anaemia is more important in this case than a possible increased
54 × 109 L−1 (range 12–582), respectively.
risk of leukostasis.
The mean reduction in the total leukocyte count when
Guidelines state that if possible red blood cells (RBC)
comparing the pretreatment level with the level immediately
transfusion should be avoided in patients with hyperleukocytosis
after the last apheresis was 71% and this corresponded to a
because it may increase the hyperviscosity. We therefore try to
mean absolute reduction of 219 × 109 L−1 (Table 2). It can
avoid or limit the number of erythrocyte transfusions to patients
be seen that leukocytapheresis caused a significant reduction
that are undergoing leukapheresis, since this may increase the
in peripheral blood leukocyte levels. All patients with myeloid
risk of leukostasis.
malignancies showed a reduction of total leukocyte counts to
The processed volume was always two times the patient’s
below the 100 × 109 L−1 that is commonly used as the definition
blood volume. The decision to process this volume was based on
of hyperleukocytosis; whereas ALL patients showed a reduction
a validated method from the producer of the leukocytapheresis
to below 400 × 109 L−1 ,i.e. used as the threshold for considering
equipment. (Cobe Spectra apheresis machine, Terumo BCT,
prophylactic leukocytapheresis (Szczepiorkowski et al., 2010).
Lakewood, CO, USA). The collect flow rate (CFR) was also
Further treatment was then based on chemotherapy alone, and
determined by a formula provided by the manufacturer and
leukapheresis treatment was ended to avoid interference with the
taking into consideration the patient’s leukocyte count and
chemotherapy. The results for the AML patients are presented in
blood volume. The CFR for patients above 30 kg is given in
detail in Table 3, and the following observations were then made:
Table 1. For patients below 30 kg, the CFR was based on the
formula: 0·0003 × inlet flow rate × leukocyte count. • The difference in peripheral blood leukocyte counts
Transfusion of platelets or RBC immediately before or during immediately before and after leukocytapheresis was mainly
the procedure was allowed if the Hb level was expected or caused by the effect of the apheresis procedures that
documented to fall below 7·0–8·0 g dL−1 or the platelet count last only for a few hours, even though the patients also
expected to fall below 30 × 109 L−1 during the procedure. After received chemotherapy. This shows that leukocytapheresis
reduction of the leukocyte counts, patients received transfusions is an effective therapeutic intervention to reduce leukocyte
of RBC when Hb levels were below 8·0–8·5 g dL−1 . levels. The mean reduction in leukocyte count per apheresis
Peripheral blood cell counts were measured (i) within 2 h for the AML patients was 48·3%.
after start of apheresis and (ii) within 1 h after the procedure • When patients receive chemotherapy together with
were ended. A white blood cell (WBC) count was also deter- leukocytapheresis, there is no or only a small increase in
mined in the apheresis product (i) approximately 2 h after the peripheral blood leukocyte counts during the first 18–24 h
start of the procedure and (ii) immediately after the procedure following the apheresis. The absolute increase until next
was completed. The leukocytapheresis procedures were usually apheresis did not exceed 25 × 109 L−1 for any procedure.

Transfusion Medicine, 2013 © 2013 The Authors

Transfusion Medicine © 2013 British Blood Transfusion Society
 Leukocytapheresis in leukaemia patients 3

Table 1. CFR and its dependency on total leukocyte counts in peripheral blood and inlet flow rate in leukaemia patients treated with
leukocytapheresis for hyperleukocytosis and leukostasis; guidelines used in this study for patients with a body weight above 30 kg

Collect flow rate (ml min−1 )

−1
9
Total peripheral blood leukocyte count (×10 L ) Inlet flow rate 40 (mL min−1 ) Inlet flow rate 60 (ml min−1 ) Inlet flow rate 80 (ml min−1 )

100–200 3 4 6
200–300 5 8 11
300–400 8 12 16
400–500 11 16 21
>500 13 20 27

Published by permission from TerumoBCT.

Table 2. Clinical characteristics of the patients and the effect of leukocytapheresis in leukaemia patients with hyperleukocytosis

WBC Leukocytes
before after last Absolute
Patient Age Number of apheresis apheresis reduction Percent Clinical course
number Gender (years) Diagnosis Clinical presentation apheresis (109 L−1 ) (109 L−1 ) (109 L−1 ) reduction after apheresis

1 F 66 AML Dyspnoea/lung failure 3 214 27·3 186·7 87 Death day 4

2 M 70 AML Dyspnoea/lung failure 3 181 34·7 146·3 81 Improvement
3 F 13 AML Prophylactic apheresis 2 246 53·1 192·9 78 No progression
4 M 5 T-ALL CNS and retinal affection 2 935 346 589 63 Improvement
5 F 88 CLL Dyspnoea/lung failure 1 278 78·4 199·6 72 Improvement
6 M 42 CML CNS affection and paresis 6 398 112 286 72 Improvement
7 F 19 AML Dyspnoea and CNS affection 2 104 33·1 70·9 68 Death day 17
8 M 3 B-ALL Priapism 2 394 75·0 319 81 Improvement
9 M 64 CMML Renal failure 2 117 60·5 56·5 48 Improvement
10 M 60 T-ALL Dyspnoea/lung failure 3 376 82·3 293·7 78 Improvement
11 M 1 ALL Dyspnoea/lung failure 1 820 185·2 634·8 77 Improvement
12 M 43 AML Dyspnoea/lung failure 2 339 53·5 285·5 84 Improvement
13 F 38 AML Dyspnoea/lung failure 2 182 57·0 125 69 Improvement
14 M 27 CML CNS affection and paresis 2 400 265 135 34 Improvement
15 F 1 ALL Prophylactic apheresis 1 528 132 396 75 No progression
16 M 42 AML Dyspnoea/lung failure 1 241 93·6 147·4 61 Improvement

AML, acute myeloid leukaemia; B-ALL, B-cell acute lymphoblastic leukaemia; CLL, chronic lymphocytic leukaemia; CML, chronic myeloid leukaemia;
CMML, chronic myelomonocytic leukaemia; CNS, central nervous system; F, female; M, male; T-ALL, T-cell acute lymphoblastic leukaemia; WBC,
white blood cell.
Improvement is defined as total regress of symptoms/clinical signs and objective improvement of affected organ functions (improved oxygenation and
reduction in serum creatinine) initially leading to the decision to perform leukocytapheresis.

• Even though the platelet counts are also reduced by the
apheresis, this decrease is small and severe post-apheresis
thrombocytopaenia was not observed. The same was true
for the Hb levels. It can be seen from Table 3 that
reduced post-apheresis platelet and erythrocyte levels were
detected for patients not receiving transfusions, but for
the other patients these apheresis-induced variations could
be controlled by transfusions and so safe levels could be
maintained.
These observations in AML patients are also representative
for the other patients with regard to leukocyte reduction
and avoidance of severe post-apheresis thrombocytopaenia or
anaemia. Thus, leukocytapheresis appears to be a safe procedure
(i.e. severe events during or immediately after the procedure are
uncommon) in leukaemia patients with high peripheral blood
leukocyte counts; it is also effective with regard to reduction of
leukocyte counts but it is not known whether leukocytapheresis
has a significant effect on early mortality in these patients.
The clinical course after leukocytapheresis varied. Two
patients who were critically ill at admission to hospital
died during the first week after leukaemia was diagnosed,
without signs of clinical improvement. Two patients receiving
prophylactic apheresis had a stable clinical situation following
apheresis. The majority of symptomatic patients had pulmonary
failure/dyspnoea. These patients generally had no immediate
clinical improvement during or after apheresis, but had a gradual
improvement during the following days after apheresis and start
of chemotherapy. However, it should be emphasised that optimal
evaluation of their pulmonary function with arterial blood gas
analyses was not possible due to the risk of haemorrhages after
arterial puncture.

© 2013 The Authors Transfusion Medicine, 2013

Transfusion Medicine © 2013 British Blood Transfusion Society
4 Ø. Bruserud et al.

Table 3. The effects of apheresis on peripheral blood cell counts in AML patients

Haemoglobin
Leukocyte count Platelet count level
Chemotherapy before or during apheresis (109 L−1 ) (109 L−1 ) (g 100 mL−1 )
FAB Effect of
Id class Chemotherapy regimen therapy Apheresis Before After Before After Before After

1 M4 Cytarabine 100 mg daily subcutaneously from Death 4 days after the start First 214 75·6 11 71 7·1 9.9
the first day of apheresis of treatment Second 75·6 36·8 71 41 1 9·9 8.71
Third 51·2 27·3 31 34 7·8 8.8
2 M4 From the same day as the first apheresis: Complete remission First 181 86·3 29 112 8·7 8.9
Idarubicin 12 mg m−2 days 1–3 Second 101 58·5 90 581 10·0 8.9
Cytarabine 200 mg m−2 days 1–7 Third 71·8 34·7 42 23 10·6 10.9
3 M0 The day before and on the day of the first Complete remission First 246 159 71 431 5·7 6.6
apheresis intravenous cytarabine 60 mg plus Second 116 53·1 29 41 7·9 6.31
oral thioguanin 60 mg; thereafter AIET
induction therapy Abrahamsson et al. (2011)
7 M2 From the same day as the first apheresis: Death in cytopenia First 104 71·5 43 60 8·3 9.1
Daunorubicin 45 mg m−2 days 1–3 Second 54·8 33·1 17 44 8·4 –
Cytarabine 200 mg m−2 days 1–7
12 M5 From the same day as the first apheresis: Complete remission First 339 184 46 75 8·6 9.2
Daunorubicin 45 mg m−2 days 1–3 Second 138 53·5 60 231 11·1 9.01
Cytarabine 200 mg m−2 days 1–7
13 M4 From the same day as the first apheresis: Complete remission First 182 – 53 – 5·8 –
Daunorubicin 45 mg m−2 days 1–3 Second – 57·0 – – – –
Cytarabine 200 mg m−2 days 1–7
16 M2 From the same day as the first apheresis: Complete remission First 241 93·6 70 73 10·0 8.71
Daunorubicin 45 mg m−2 days 1–3
Cytarabine 200 mg m2 days 1–7
Clofarabine 10 mg m−2 days 1–5

1 No platelet or erythrocyte transfusion given immediately before or during apheresis.

For patients presenting with priapism surgical intervention
should always be considered (Rodgers et al., 2012), which was
the case for patient 8 (Table 2), who underwent therapeutic
aspiration. His symptom regressed after leukocytapheresis and
surgical intervention.
DISCUSSION
The use of apheresis in the treatment of hyperleukocytosis
Hyperleukocytosis is usually defined as peripheral blood
leukocyte counts exceeding 50 or 100 × 109 L−1 and can be both
asymptomatic and symptomatic (referred to as leukostasis). In
this study, we describe our experience with leukocytapheresis
in 16 patients with various leukaemias, and we conclude
that leukocytapheresis is a safe procedure in these patients
with regard to toxicity during and immediately after the
procedure. Immediate complications due to the procedure
were not reported in any of our patients. One side effect
of leukocytapheresis is hypocalcaemia due to citrate toxicity.
All our patients were given calcium supplementation, serum
calcium levels were regularly controlled during the procedure,
and symptomatic hypocalcaemia was not observed. Excessive
erythrocyte loss was not seen in any patient either. Thus,
leukocytapheresis seems to be a safe procedure with regard
to immediate toxicity. To the best of our knowledge, no studies
have demonstrated an adverse effect of leukocytapheresis on the
long-term prognosis in leukaemia patients either, but it should
be emphasised that this question has not been addressed in
prospective clinical studies.
Leukocytapheresis removes leukocytes from the peripheral
circulation and the main aim is therefore to reduce or prevent
acute symptoms due to leukostasis. Another aim is to reduce
the risk of tumour lysis syndrome when chemotherapy is
initiated, although this is probably less important because
the majority of leukaemic cells are located in the bone
marrow or in peripheral lymphoid tissues. Patients treated
with leukocytapheresis also receive additional supportive care
and usually also cytoreductive chemotherapy. The supportive
treatment consists of prophylactic therapy of tumour lysis
syndrome with hydration, electrolyte corrections, allopurinol
and respiratory support. They will also receive transfusions with
RBC and platelets before or eventually during apheresis, as
described for our patients (see Materials and methods section).
All our 12 acute leukaemia patients fulfilled the criteria for
treatment with leukocytapheresis set by the current guidelines
from the American Society for Apheresis and they had either (i)

Transfusion Medicine, 2013 © 2013 The Authors

Transfusion Medicine © 2013 British Blood Transfusion Society

symptoms of leukostasis (14 patients) or (ii) so high peripheral
blood leukaemia cell counts that prophylactic leukocytapheresis
should be considered (two patients; Szczepiorkowski et al.,
2010). We used the AML criteria for the other patients with
myeloid malignancies (two CML and one CMML patient), and
the CLL patient was an exceptional patient. At least our acute
leukaemia patients are thus representative for acute leukaemia
patients that should receive this treatment, and this statement is
further supported by the frequency of acute leukaemia patients
treated with leukocytapheresis during the 16-year study period.
Our institution has been responsible for diagnosis and treatment
of acute leukaemia in the same geographic area during the
whole study period. The accurate number of acute leukaemia
patients treated at our institution during these 16 years is not
available, but in a previous study we documented that 98 adult
AML patients were admitted to our hospital during an 11 years
period (Bruserud et al., 2003), and the annual number of adult
ALL patients is lower (usually 2–5 per year). Even though an
accurate estimate is not possible, there is no doubt that our
adult acute leukaemia patients treated with leukocytapheresis
represent not more than 5–10% of our total acute leukaemia
patient population. This frequency is similar to other studies
(Cuttner et al., 1983; Dutcher et al., 1987; Hoelzer et al.,
1988; De Santis et al., 2011). Thus, our patient population is
representative because they fulfil the generally accepted criteria
for this treatment, and in addition they represent an expected
proportion of the total acute leukaemia population.
Although leukocytapheresis promptly and efficiently reduces
leukocyte count, the effects of leukocytapheresis on mortality
and morbidity are still controversial, no generally accepted
guidelines exist (Döhner et al., 2010), and most of the
available studies are small and retrospective. Leukostasis is
relatively common in AML compared with other haematologic
malignancies and leukocytapheresis has been tried in the
treatment of these conditions. These studies are reviewed in
detail below and summarised in Table 4 and the use of apheresis
in AML is discussed in detail below.
The apheresis did not have any immediate clinical effects
for the majority of our patients; they rather showed a gradual
improvement over the following days after apheresis and start
of chemotherapy. This is also consistent with observations and
recommendations from previous clinical studies emphasising
that leukocytapheresis alone is not sufficient and should not
delay start of chemotherapy (see below). For patients presenting
with priapism surgical intervention should always be considered
(Rodgers et al., 2012), which was also done for our patient with
this symptom.
Symptomatic hyperleukocytosis is relatively uncommon in
CLL (Baer et al., 1985), and this is also illustrated by our study of
an unselected patient population where only one of our patients
treated with leukocytapheresis had CLL. In the study of Baer
et al. (1985) leukocyte counts exceeding 500 × 109 L−1 were
seen in fewer than 10% of patients, and approximately one-
fifth of these patients had signs of hypervicosity. However, our
patient illustrates that symptomatic hyperleukocytosis can be
© 2013 The Authors
Transfusion Medicine © 2013 British Blood Transfusion Society
Leukocytapheresis in leukaemia patients 5
seen also at leukocyte counts below 500 × 109 L−1 . Symptomatic
hyperleukocytosis can also occur in CML, but symptoms usually
seem to occur at leukocytes exceeding 200 × 109 L−1 (Rowe &
Lichtman 1984). This was also true for our two CML patients
with symptomatic hyperleukocytosis.
Taken together our results show that leukocytapheresis in
leukaemia patients with hyperleukocytosis is a safe procedure,
i.e. severe or unexpected side effects during or immediately after
the procedure are uncommon. In our opinion, the major disease
compartment in these disorders is probably the extravascular
bone marrow and eventually lymphoid organs, however the
procedure is also effective with regard to reduction of leukocyte
levels and decreased levels are maintained without rapid
replacement of circulating cells from the bone marrow. Finally,
a recent report suggested that leukocytapheresis did not only
reduce leukocyte levels, the procedure also seems to normalise
the circulation because these patients may have an increased
blood volume that is restored and the reduction of the circulating
leukaemia cell burden may therefore be larger than reflected in
the reduction of the leukocyte count (Chekol et al., 2012).
In five of our AML patients intensive induction chemotherapy
was started the same day as the first apheresis, and repeated
apheresis were performed on the following day(s). Despite these
patients reached complete haematological remission. These case
reports thus suggest that apheresis started on the same day as
chemotherapy does not have any major impact on chemotherapy
efficiency, but further studies in additional patients are needed
to confirm this.
Hyperleukocytosis, leukostasis and leukapheresis in human
AML – a review of available studies
Relative few studies are available on the use of leukocytapheresis
in human AML. We identified all clinical studies on
leukocytapheresis in AML that were reported to the PubMed
database; these studies are summarised in Table 4. We did not
go through or include case reports.
Definitions. Hyperleukocytosis is defined as WBC exceeding
50 × 109 L−1 (Giles et al., 2001) or 100 × 109 L−1 (Porcu et al.,
1997; Thiebaut et al., 2000; Bug et al., 2007; Chang et al., 2007; De
Santis et al., 2011). Leukostasis is defined as hyperleukocytosis
combined with organ dysfunction due to hypoxia, tissue
infiltration or coagulopathy (Porcu et al., 1997).
Incidence and disease characteristics. Giles et al. (2001) reported
that during 7 years they observed WBC counts >50 × 109 L−1 in
146 (19%) of the 761 patients who received their first induction
treatment at their hospital. Chang et al. (2007) reported an
incidence of 29% among their AML patients.
As described by Giles et al. (2001), there is an association
between high WBC counts and inv(16) and a low count for
patients with t(8;21); the most common cytogenetic abnormality
among their patients was inv(16) and the frequencies of t(8;21)
or abnormalities on chromosomes 5 and/or 7 were low. A
Transfusion Medicine, 2013

Table 4. A summary of clinical studies of leukocytapheresis in AML patients with hyperleukocytosis
Patients
Porcu et al. (1997) n = 48 with no exclusion criteria
Median age 41·2 years (range
8–79 years
Transfusion Medicine, 2013
Median initial WBC count
201 × 109 L−1 , range
101–506 × 109 L−1
De Santis et al. (2011) n = 15
Out of 187 patients, 15 consecutive
AML patients (8%) had symptoms
of leukostasis
Median age 60 years (range
22–78 years)
All patients with WBC counts
>88 × 109 L−1 at admission
Bug et al. (2007) Two consecutive groups with
hyperleukocytosis treated either
without or with initial
leukocytapheresis
All patients had WBC counts
>100 × 109 L−1
Apheresis group: n = 25, median age
50 years with range 20–78 years
Transfusion Medicine © 2013 British Blood Transfusion Society
© 2013 The Authors
Apheresis
Daily apheresis with duration 3 h until
WBC count <100 × 109 L−1
Eighty-five apheresis in 48 patients:
Median reduction 56%
Median absolute reduction
118 × 109 L−1
Sixty-four per cent of patients had
WBC count <100 × 109 L−1 at the end
of procedure(s)
Apheresis begun within 12 h after Concomitant treatment with high doses of
admission
Two blood volumes were processed for
each session, each procedure lasting
2–3 h
Daily apheresis until clinical
improvement or WBC count
<100 × 109 L−1
Coagulation factors were transfused if
INR > 1·5, platelets were transfused if
platelet counts were <20 × 109 L−1
Apheresis initiated within 5·4 h, 14/25
patients received one apheresis
Median duration of apheresis 130 min
Chemotherapy
Hydroxyurea was started immediately after
the first apheresis
Induction chemotherapy begun within
48–72 h
hydroxyurea or conventional induction
chemotherapy with anthracycline plus
cytarabine
Cytoreductive hydroxyurea or cytarabine
administered within the first 24 h after
admission. Following apheresis patients
received age-adapted induction therapy
6
Observations – therapeutic effects
First-week mortality 27·1%
Survivors and non-survivors did not differ in
pretherapy WBC count, final WBC count,
Ø. Bruserud et al.
absolute and percent apheresis-induced
reduction of WBC
Adverse prognosis was associated with age,
presence of coagulopathy, respiratory
stress, renal failure and neurological
symptoms
First-week mortality was 7 out of 15 patients
The first apheresis reduced the mean WBC
count from 201 × 109 L−1 to
150 × 109 L−1
Six patients received a second apheresis; for
these patients the mean WBC count was
reduced from 279 to 198 × 109 L−1 during
the first apheresis and to
106 × 109 L−1 during the second apheresis
Apheresis did not have severe side effects
The median WBC count before apheresis
was 151 × 109 L−1 (range 103–399) and
after apheresis 80 × 109 L−1 (range
31–276); median reduction being 47%
Median platelet reduction from 54 to
45 × 109 L−1
Decreased early death rate for patients
treated with leukocytapheresis compared
with the control group (P = 0·015; early
death rate for all patients 26%)
 Table 4. Continued
© 2013 The Authors
Patients
Giles et al. (2001) One hundred and forty six patients with
WBC counts >50 × 109 L−1 were
diagnosed during a 7-year period
. Median WBC count before apheresis
Transfusion Medicine © 2013 British Blood Transfusion Society
Chang et al. (2007) Seventy-five patients with WBC counts
>100 × 109 L−1 ; 38 patients received
pretreatment with leukocytapheresis
and/or cranial irradiation
Thiebaut et al. (2000) Fifty-three patients (median age 59 years,
range 16–78 years)
Median WBC count 160 × 109 L−1 (range
100–480 × 109 L−1 )
WBC, white blood cell count.
Transfusion Medicine, 2013
Apheresis
Out of 146 patients, 71 underwent apheresis,
and 56 patients only one apheresis
117 × 109 L−1 and after apheresis
50 × 109 L−1
No adverse effects of apheresis
No significant effect of cranial irradiation on
the risk of intracranial haemorrhage
One to four apheresis for each patient
Effect of apheresis:
In 21 patients, there is no significant effect
Thirty-two patients achieved a WBC
count <100 × 109 L−1
Chemotherapy
Cytarabine-based (single doses >1 g m−2 )
intensive chemotherapy was initiated
within 24 h of the initial apheresis
Cytoreductive therapy with hydroxyurea
given to a minority of patients
Median WBC count at the start of therapy
was 85 × 109 L−1 (range
23–264 × 109 L−1 )
Observations – therapeutic effects
Two weeks early death rate was 13% for
patients receiving apheresis and 23%
for patients not receiving this
treatment (not significant)
The independent predictors of survival at
2 weeks were apheresis (P = 0·0056),
younger age (P = 0·014), lower WBC
count (P = 0·0003) and low
performance status (0·000006)
No statistically significant effect of
apheresis on long-term survival
More than two symptoms of leukostasis
was significantly associated with early
death; the early death rate for all
patients being 33%
Increased risk of early death was
associated with (i) more than two
symptoms of leukostasis, (ii) age
>65 years and (iii) start of
chemotherapy later than 48 h after
admission
Early death only 1%
Leukocytapheresis in leukaemia patients 7
8 Ø. Bruserud et al.
high frequency of monocytic variants among AML patients with
hyperleukocytosis has also been described (Thiebaut et al., 2000).
Pathophysiology of leukostasis. The reasons why hyperleukocy-
tosis at diagnosis is only observed for a subset of AML patients
and the pathophysiology behind leukostasis are incompletely
understood. Sissolak et al. (1993) described an altered adhe-
sion molecule pattern in peripheral blood blasts compared with
bone marrow blasts; this observation suggests that the molecule
adhesion pattern is important in the ability of blast to circu-
late or leukaemisation and contribute to the wide variation in
peripheral blood blast counts at diagnosis between patients.
Hyperleukocytosis with symptoms of organ failure is
probably caused by disturbed microcirculation due to increased
blood viscosity caused by the high concentration of large and
less deformable AML blast; this will interfere with the microcir-
culation and microembolisation may appear. Hyperleukocytosis
probably also causes interactions between leukaemic blasts
and endothelial cells leading to increased cytokine production
with altered local cytokine networks (Hatfield et al., 2010a),
endothelial cell damage and extravasation of AML cells. AML
cells show constitutive release of proteolytic enzymes, especially
matrix metolloproteases 9 (MMP-9) (Reikvam et al., 2010), that
may contribute to the tissue damage especially for AML cells
with monocytic differentiation (Paupert et al., 2008; Reikvam
et al., 2010; Hatfield et al., 2010b). Special mutations also seem
to influence these features, e.g. the common NPM-1 mutation
seems to be associated mainly with the FAB M4–M5 class
(Tsykunova et al., 2012).
Finally, symptoms and clinical signs of leukostasis have
also been reported in AML patients without hyperleukocytosis
(Soares et al., 1992), and this observation clearly shows that the
development of leukostasis depends not only on the level of
circulating blasts but also on the biological characteristics of the
AML cells. This hypothesis is also supported by the observation
by Bug et al. (2007) who described a significant association
between AML cell expression of the CD11c membrane molecule
and a high risk of early death in leukocytosis.
Symptoms and clinical signs. Porcu et al. (1997) performed
a study of unselected patients; a group of 48 consecutive
patients with WBC >100 × 109 L−1 was treated with initial
leukocytapheresis. The most common symptoms/signs in this
study were respiratory distress (39% of patients), neurological
symptoms (27%) and renal failure (14%); coagulopathy was also
common (31%). Similar frequencies have also been reported by
other investigators (Bug et al., 2007; Chang et al., 2007).
The prognostic impact of hyperleukocytosis, leukostasis and early
leukocytapheresis. The prognostic impact of hyperleukocytosis
has been evaluated both with regard to remission rate/disease-
free survival and with regard to early death. The presence
of leukocytosis (>50 × 109 L−1 ) seems to be associated with
a shorter long-term survival (Dutcher et al., 1987), but this
prognostic impact is weak compared with the most important
Transfusion Medicine, 2013
prognostic parameters that are genetic abnormalities and
response to the initial chemotherapy (Wheatley et al., 1999).
Hyperleukocytic AML is associated with an increased early
mortality (Greenwood et al., 2006; Marbello et al., 2008). Porcu
et al. (1997) described a first-week mortality of 27%, in the
small study by De Santis et al. (2011) the mortality was 47%;
early death rates within this range have also been described in
the studies by Bug et al. (2007) (n = 25, early death rate 38%)
and Chang et al. (2007) (n = 75, early death rate 33%). All these
studies included patients with WBC counts >100 × 109 L−1 ,
whereas Giles et al. (2001) included patients with WBC counts
>50 × 109 L−1 and a lower death rate of 13% was then observed
in patients receiving apheresis. The 2% early death rate in the
study of Thiebaut et al. (2000) seems to be exceptional.
The risk of early death during hyperleukocytosis is particularly
high in patients with clinical signs and symptoms of organ failure
due to the leukostasis, i.e. respiratory distress, neurological
symptoms, respiratory distress or coagulopathy (Porcu et al.,
1997; Bug et al., 2007; Chang et al., 2007). Age (Porcu et al.,
1997; Giles et al., 2001), very low performance status (Giles
et al., 2001; Bug et al., 2007), high C reactive protein (CRP)
levels (Bug et al., 2007) and high lactate dehydrogenase (LDH)
levels (Bug et al., 2007) have also been significantly associated
with prognosis. Bug et al. (2007) performed a multivariate
analysis and the factors associated with early death were then
dyspnoea, renal failure and high LDH levels.
Bug et al. (2007) described an association between AML cell
expression of CD11c and the risk of early death. This observation
supports the hypothesis that the development of leukostasis is
dependent not only on the degree of hyperleukocytosis but
also on the biological characteristics of the AML cells, and the
contribution of the biological characteristics can also explain
why the WBC count does not have a major prognostic impact
in AML patients with hyperleukocytosis.
Treatment of hyperleukocytosis – the combination of apheresis
and chemotherapy. De Santis et al. (2011) observed a mean
absolute WBC reduction of 50 × 109 L−1 by the first apheresis
whereas a second apheresis caused a mean absolute reduction of
100 × 109 L−1 . This difference may be due to a combined effect
of apheresis and chemotherapy during the second apheresis
because all their patients received concomitant chemotherapy
with conventional induction chemotherapy or high-dose
hydroxyurea. These observations illustrate the importance of
early chemotherapy also for patients receiving leukocytapheresis.
The importance of early chemotherapy was emphasised by Giles
et al. (2001); 45 of the 71 patients in their study started induction
chemotherapy on the same day as the apheresis.
The optimal number of apheresis is not known. The results
by Giles et al. (2001) suggested that apheresis improved
the prognosis whereas the number of apheresis was less
important. However, these authors included patients with
WBC counts >50 × 109 L−1 and the median count was low
(116 × 109 L−1 before apheresis, 51 × 109 L−1 after apheresis)
compared with the other studies including patients with WBC
© 2013 The Authors
Transfusion Medicine © 2013 British Blood Transfusion Society

counts >100 × 109 L−1 . The main conclusion from the study
by Giles et al. (2001) is that reduction of the WBC count
down to 50–100 × 109 L−1 may be an advantage at least if
intensive induction chemotherapy is not delayed. Finally, it is
also difficult to predict the optimal methodological approach for
leukocytapheresis in leukaemia patients with hyperleukocytosis
or leukostasis, given that detailed methodological descriptions
are usually not given in the previous studies. Our study seems to
be the first to give a detailed description of the leukocytapheresis
procedure.
The high early death rates in patients with hyperleukocytosis
(Table 4) clearly suggest that leukostasis by itself is associated
with increased early mortality, and the comparative study by Bug
et al. (2007) showed a reduced early mortality when leukostasis
was treated by leukocytapheresis. However, it is difficult to know
whether leukocytapheresis has an effect on early mortality or
long-term disease-free survival in AML as long as no randomised
clinical studies are available (Szczepiorkowski et al., 2010). Thus,
further clinical studies on leukocytapheresis in the treatment of
leukostasis should be encouraged. The first step should be to get
better documentation of efficiency and safety of leukocytaphere-
sis and to reach a general agreement on standardisation of the
leukocytapheresis procedure. How to combine leukocytaphere-
sis and chemotherapy also has to be standardised. The ideal
second step should then be a randomised clinical study. How-
ever, in the meantime the recommendation of leukocytapheresis
in patients with leukostasis has to be based on (i) the available
evidence showing increased early mortality in patients with
leukostasis compared with the lower early mortality rate in the
general AML patient population (usually being lower than 5%)
(Estey 2012), (ii) a comparative study (Bug et al., 2007) showing
reduced early mortality in patients receiving leukocytapheresis
and on the other hand (iii) the documented safety of the pro-
cedure in these patients and (iv) no evidence so far supporting
any effect of leukocytapheresis on the long-term disease-free
survival of patients receiving intensive chemotherapy.
Taken together the overall results from the available studies
from patients with hyperleukocytosis show that apheresis has an
acceptable safety in these patients and causes no severe imme-
diate side effects (De Santis et al., 2011). Both Bug et al. (2007)
and Giles et al. (2001) described a decreased early death rate
in patients receiving leukocytapheresis compared with a group
of historical controls. Even though Giles et al. (2001) observed
a decreased long-term survival of borderline significance for
patients receiving apheresis, they concluded that this observa-
tion may reflect the fact that patients chosen to have apheresis
were prognostically unfavourable, because they regarded a
direct effect of apheresis on long-term prognosis to be less
likely.
The importance of early cytoreductive chemotherapy is
emphasised by several authors, but the early chemotherapy
varies between the studies. Cytarabine or high-dose hydroxyurea
was administered within 24 h in one study (Bug et al., 2007),
in another study high-dose induction treatment was started on
the same day as the first apheresis (Giles et al., 2001) and in
© 2013 The Authors
Transfusion Medicine © 2013 British Blood Transfusion Society
Leukocytapheresis in leukaemia patients 9
a third study delayed chemotherapy for more than 48 h was
associated with increased early mortality (Chang et al., 2007).
Thiebaut et al. (2000) described a recruitment of bone marrow
cells into the S phase after apheresis; these results have to be
confirmed but this observation further supports that early start
of chemotherapy is important and if cell cycle-specific drugs are
used the apheresis-induced recruitment may even increase the
effect of chemotherapy.
Concluding comments
To conclude, leukocytapheresis is effective in reducing the
peripheral blood leukocyte count for various leukaemias and
it seems to be a safe procedure both with regard to immediate
complications and effects on long-term prognosis. However,
there is currently no consensus regarding indication for leukocy-
tapheresis in hyperleukocytosis. Larger prospective randomised
studies are missing. The aim of the treatment is to reduce early
mortality. Leukocytapheresis may not be necessary in all patients
with asymptomatic hyperleukocytosis and should not delay
the start of chemotherapy. Symptomatic leukostasis with organ
failure and special situations like serious respiratory failure,
CNS involvement (Szczepiorkowski et al., 2010) and priapism
(Rodgers et al., 2012) are currently regarded as indications for
leukocytapheresis. The current guidelines from the American
Society for Apheresis (ASFA) state that leukocytapheresis
should always be considered in AML patients with symptomatic
hyperleukocytosis; this is given a grade 1B recommendation
based on type II-2 evidence with the lack of randomised
clinical trials (Szczepiorkowski et al., 2010). In the case of
prophylactic leukocytapheresis the evidence is weaker, but it can
be considered in AML for patients with monocytic/monoblastic
subtypes and/or leukocyte count >100 × 109 L−1 , and for
ALL patients with leukocyte count >400 × 109 L−1 (grade 2C
recommendation) (Szczepiorkowski et al., 2010).
The possible effect of leukocytapheresis on early mortality and
long-time survival of leukaemia patients is not well documented.
Only a randomised comparison can give a final answer to the
question whether leukapheresis reduces early mortality without
affecting the long-term survival of these patients.
ACKNOWLEDGMENTS
We want to thank the staff at the Department of Immunology
and Transfusion Medicine, Haukeland University Hospital, with
performing the procedures and collecting the data for this study.
The study was supported by the Norwegian Cancer Society. Ø.
B., K. L. and T. H. treated the patients, performed the study and
wrote the manuscript. S. S. and G. M. collected the data. D. L. C.
wrote the manuscript. E. K. treated the patients. H. R. initiated
the study, collected the data and wrote the manuscript.
CONFLICT OF INTEREST
The authors have no competing interests.
Transfusion Medicine, 2013
10 Ø. Bruserud et al.
REFERENCES
Abrahamsson, J., Forestier, E., Heldrup, J. et al.
(2011) Response-guided induction therapy
in pediatric acute myeloid leukaemia with
excellent remission rate. Journal of Clinical
Oncology, 29, 310–315.
Baer, M.R., Stein, R.S. & Dessypris, E.N.
(1985) Chronic lymphocytic leukaemia
with hyperleukocytosis. The hyperviscosity
syndrome. Cancer, 56, 2865–2869.
Bruserud, O., Hovland, R., Wergeland, L.,
Huang, T.S. & Gjertsen, B.T. (2003)
Flt3-mediated signaling in human acute
myelogenous leukaemia (AML) blasts: a
functional characterization of Flt3-ligand
effects in AML cell populations with
and without genetic Flt3 abnormalities.
Haematologica, 88, 416–428.
Bug, G., Anargyrou, K., Tonn, T., Bialleck, H.,
Seifried, E., Hoelzer, D. & Ottmann, O.G.
(2007) Impact of leukocytapheresis on early
death rate in adult acute myeloid leukaemia
presenting with hyperleukocytosis. Transfu-
sion, 47, 1843–1850.
Bunin, N.J. & Pui, C.H. (1985) Differing com-
plications of hyperleukocytosis in children
with acute lymphoblastic or acute non-
lymphoblastic leukaemia. Journal of Clinical
Oncology, 3, 1590–1595.
Chang, M.C., Chen, T.Y., Tang, J.L., Lan, Y.J.,
Chao, T.Y., Chiu, C.F. & Ho, H.T. (2007)
Leukocytapheresis and cranial irradiation in
patients with hyperleukocytic acute myeloid
leukaemia: no impact on early mortality and
intracranial hemorrhage. American Journal
of Hematology, 82, 976–980.
Chekol, S.S., Bhatnagar, B., Gojo, I. & Hess,
J.R. (2012) Leukopheresis for profound
hyperleukocytosis. Transfusion and Aphere-
sis Science, 46, 29–31.
Cuttner, J., Conjalka, M.S., Reilly, M.,
Goldberg, J., Reisman, A., Meyer, R.J.
& Holland, J.F. (1980) Association of
monocytic leukaemia in patients with
extreme leukocytosis. The American Journal
of Medicine, 69, 555–558.
Cuttner, J., Holland, J.F., Norton, L.,
Ambinder, E., Button, G. & Meyer,
R.J. (1983) Therapeutic leukocytapheresis
for hyperleukocytosis in acute myelocytic
leukaemia. Medical and Pediatrical Oncol-
ogy, 11, 76–78.
De Santis, G.C., de, Oliveira, L.C., Romano,
L.G., Almeida Prado Bde, P. Jr., Simoes,
B.P., Rego, E.M., Covas, D.T. & Falcao,
R.P. (2011) Therapeutic leukocytapheresis
in patients with leukostasis secondary to
acute myelogenous leukaemia. Journal of
Clinical Apheresis, 26, 181–185.
Transfusion Medicine, 2013
Dutcher, J.P., Schiffer, C.A. & Wiernik,
P.H. (1987) Hyperleukocytosis in adult
acute nonlymphocytic leukaemia: impact
on remission rate and duration, and
survival. Journal of Clinical Oncology, 5,
1364–1372.
Döhner, H., Estey, E.H., Amadori, S. et al.
(2010) Diagnosis and management of acute
myeloid leukaemia in adults: recommenda-
tions from an international expert panel,
on behalf of the European LeukaemiaNet.
Blood, 115, 453–474.
Estey, E.H. (2012) Acute myeloid leukaemia:
2012 update on diagnosis, risk stratification,
and management. American Journal of
Hematology, 87, 89–99.
Giles, F.J., Shen, Y., Kantarjian, H.M. et al.
(2001) Leukocytapheresis reduces early
mortality in patients with acute myeloid
leukaemia with high white cell counts
but does not improve long-term survival.
Leukaemia & Lymphoma, 42, 67–73.
Greenwood, M.J., Seftel, M.D., Richardson, C.
et al. (2006) Leukocyte count as a predictor
of death during remission induction in
acute myeloid leukaemia. Leukaemia &
Lymphoma, 47, 1245–1252.
Hatfield, K.J., Bedringsaas, S.L., Ryningen,
A., Gjertsen, B.T. & Bruserud, O. (2010a)
Hypoxia increases HIF-1alpha expression
and constitutive cytokine release by primary
human acute myeloid leukaemia cells.
European Cytokine Network, 21, 154–164.
Hatfield, K.J., Reikvam, H. & Bruserud,
O. (2010b) The crosstalk between the
matrix metalloprotease system and the
chemokine network in acute myeloid
leukaemia. Current Medical Chemistry, 17,
4448–4461.
Hoelzer, D., Thiel, E., Loffler, H. et al. (1988)
Prognostic factors in a multicenter study for
treatment of acute lymphoblastic leukaemia
in adults. Blood, 71, 123–131.
Majhail, N.S. & Lichtin, A.E. (2004) Acute
leukaemia with a very high leukocyte count:
confronting a medical emergency. Cleveland
Clinic Journal of Medicine, 71, 633–637.
Marbello, L., Ricci, F., Nosari, A.M. et al.
(2008) Outcome of hyperleukocytic adult
acute myeloid leukaemia: a single-center
retrospective study and review of literature.
Leukaemia Research, 32, 1221–1227.
Paupert, J., Mansat-De Mas, V., Demur, C.,
Salles, B. & Muller, C. (2008) Cell-surface
MMP-9 regulates the invasive capacity
of leukaemia blast cells with monocytic
features. Cell Cycle, 7, 1047–1053.
Porcu, P., Danielson, C.F., Orazi, A., Heerema,
N.A., Gabig, T.G. & McCarthy, L.J. (1997)
Transfusion Medicine © 2013 British Blood Transfusion Society
Therapeutic Leukocytapheresis in hyper-
leucocytic leukaemias: lack of correlation
between degree of cytoreduction and early
mortality rate. British Journal of Haematol-
ogy, 98, 433–436.
Reikvam, H., Hatfield, K.J., Oyan, A.M.,
Kalland, K.H., Kittang, A.O. & Bruserud, O.
(2010) Primary human acute myelogenous
leukaemia cells release matrix metallopro-
teases and their inhibitors: release profile
and pharmacological modulation. European
Journal of Hematology, 84, 239–251.
Rodgers, R., Latif, Z. & Copland, M.
(2012) How I manage priapism in chronic
myeloid leukaemia patients. British Journal
of Haematology, 158, 155–164.
Rowe, J.M. & Lichtman, M.A. (1984) Hyper-
leukocytosis and leukostasis: common fea-
tures of childhood chronic myelogenous
leukaemia. Blood, 63, 1230–1234.
Sissolak, G., Hoffbrand, A.V., Mehta, A.B.
& Ganeshaguru, K. (1993) Interferon-
alpha inducible 2’-5’ oligoadenylate syn-
thetase transcripts in lymphoid and myeloid
leukaemias. Leukaemia, 7, 712–716.
Soares, F.A., Landell, G.A. & Cardoso,
M.C. (1992) Pulmonary leukostasis without
hyperleukocytosis: a clinicopathologic study
of 16 cases. American Journal of Hematology,
40, 28–32.
Szczepiorkowski, Z.M., Winters, J.L., Ban-
darenko, N. et al. (2010) Guidelines on
the use of therapeutic apheresis in clinical
practice – evidence-based approach from
the Apheresis Applications Committee of
the American Society for Apheresis. Journal
of Clinical Apheresis, 25, 83–177.
Thiebaut, A., Thomas, X., Belhabri, A.,
Anglaret, B. & Archimbaud, E. (2000)
Impact of pre-induction therapy leukocy-
tapheresis on treatment outcome in adult
acute myelogenous leukaemia presenting
with hyperleukocytosis. Annals of Hema-
tology, 79, 501–506.
Tsykunova, G., Reikvam, H., Hovland, R. &
Bruserud, O. (2012) The surface molecule
signature of primary human acute myeloid
leukaemia (AML) cells is highly associated
with NPM1 mutation status. Leukaemia, 26,
557–559.
Wheatley, K., Burnett, A.K., Goldstone,
A.H. et al. (1999) A simple, robust,
validated and highly predictive index for
the determination of risk-directed therapy
in acute myeloid leukaemia derived from
the MRC AML 10 trial. United Kingdom
Medical Research Council’s Adult and
Childhood Leukaemia Working Parties.
British Journal of Haematology, 107, 69–79.
© 2013 The Authors