ORIGINAL ARTICLE

Early Complications of Hyperleukocytosis and Leukapheresis

in Childhood Acute Leukemias
Oussama Abla, MD,* Paola Angelini, MD,*
Giancarlo Di Giuseppe, BSc (Hons),* Mohamed F. Kanani, MD,w
Wendy Lau, MBBS, FRCPC,* Johann Hitzler, MD, FRCP (C), FAAP,*
Lillian Sung, MD, PhD,* and
Ahmed Naqvi, MBBS, DCH, MCPS, MRCP, FRCPCH*

microcirculation. The release of proinflammatory cytokines

Summary: Hyperleukocytosis in children with acute lymphoblastic (eg, tumor necrosis factor-a, interleukin-2, and interleukin-
leukemia (ALL) and acute myeloid leukemia (AML) is associated 8) by leukemic cell lysis can also cause life-threatening car-
with early morbidity and mortality. The benefit from leukapheresis diopulmonary complications in acute myeloid leukemia
is controversial, and its complications are not well defined. We
analyzed the frequency of early complications in children with ALL
(AML), especially M4/M5 subtypes.4 To avoid these com-
and AML presenting with white blood cell (WBC) count plications, a reduction of leukemic blasts can be accom-
>100109/L, and the type and frequency of complications related plished with hyperhydration, early start of induction che-
to leukapheresis. During a 12-year period, 84 of 634 (13%) ALL motherapy including hydroxyurea and leukapheresis.
and 18 of 143 (12.5%) AML patients presented with hyper- Leukapheresis is commonly used in patients with acute
leukocytosis. Leukapheresis was performed in 18 ALL and 12 lymphoblastic leukemia (ALL) and initial WBC count of
AML patients. The median initial WBC was 474 109/L in the >400 109/L and in those with AML presenting with
leukapheresis group compared with 175 109/L in the non- WBC > 100109/L.5 Discrepancy of survival benefit for
leukapheresis group. Neurological leukostasis occurred in 6 ALL leukapheresis has been shown in adult AML studies,6 and
(7.1%) and 4 AML (22.2%) patients. Pulmonary leukostasis
occurred in 16 ALL (19%) and 4 AML patients (22.2%). Neuro-
guidelines on when and for how long to perform this pro-
logical symptoms improved in few patients after leukapheresis, cedure are lacking. Leukapheresis is a resource-intensive
except in patients with very high WBC (> 650 109/L in ALL and intervention that requires adequate venous access, experi-
>400109/L in AML). Leukapheresis improved respiratory enced personnel, and specialized equipment. Furthermore,
symptoms in some patients but caused worsening symptoms in leukapheresis is not recommended for patients with acute
others. Early death was associated with neurological complications, promyelocytic leukemia presenting with elevated WBC as it
AML diagnosis, and coagulopathy. Leukapheresis did not delay may exacerbate the coagulopathy.7 Early complications of
initiation of chemotherapy, nor did it impact early response to leukapheresis have not been well characterized in leukemic
chemotherapy or long-term survival. Complications included children presenting with hyperleukocytosis. We analyzed the
femoral vein thrombosis, electrolyte imbalances, and hemodynamic
instability, which were all reversible. The role of leukapheresis as a
frequency and patterns of early complications in children
cytoreductive procedure in childhood hyperleukocytic leukemia with ALL and AML presenting with WBC > 100109/L,
remains to be well defined. as well as the frequency and type of complications related to
leukapheresis. In addition, we aimed to determine whether
Key Words: hyperleukocytosis, leukapheresis, children, acute leukapheresis was beneficial in preventing severe complica-
leukemias tions of hyperleukocytosis, including early death.
(J Pediatr Hematol Oncol 2016;38:111–117)
MATERIALS AND METHODS
Research ethics approval was obtained for the study at
H yperleukocytosis, defined as a white blood cell (WBC)
count >100109/L, is a complication seen in 5% to
20% of patients with acute leukemia.1,2 It can be associated
The Hospital for Sick Children, Toronto, Canada. The
medical records of patients with ALL and AML diagnosed
between January 1992 and July 2005 and with an initial
with early morbidity and mortality because of pulmonary, WBC > 100109/L were reviewed in detail to determine age
neurological, and metabolic complications.3 These can be at diagnosis, presence, or absence of a mediastinal mass, sex,
due to severe leukostasis caused by the hyperviscosity in the use of packed red blood cell (PRBC), platelet or fresh frozen
plasma (FFP) transfusions at presentation, use of urate
Received for publication May 4, 2015; accepted December 17, 2015. oxidase (rasburicase), use of dialysis, and initial method used
From the *Leukemia/Lymphoma Section, Department of Pediatrics, for cytoreduction. Complete blood count (CBC) data at
Division of Haematology/Oncology, The Hospital for Sick Chil- diagnosis including WBC, hemoglobin concentration and
dren, University of Toronto, Toronto, ON, Canada; and wPediatric
Department, Pediatric Oncology/Hematology Division, Tawam
platelet count, as well as coagulation studies were recorded.
Hospital, Al Ain, United Arab Emirates. Routine blood chemistry values (serum electrolytes, crea-
The authors declare no conflict of interest. tinine, blood urea nitrogen, uric acid, lactate dehydrogenase,
Reprints: Oussama Abla, MD, Leukemia/Lymphoma Section, calcium, and phosphorus) were documented, as were the
Department of Pediatrics, Division of Haematology/Oncology, The
Hospital for Sick Children, 555 University Avenue, Toronto, ON,
peak levels of potassium, creatinine, phosphorus, and cal-
Canada M5G 1X8 (e-mail: oussama.abla@sickkids.ca). cium during the first 10 days. Diagnostic bone marrow (BM)
Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved. morphology and immunophenotyping, cerebrospinal fluid

J Pediatr Hematol Oncol  Volume 38, Number 2, March 2016 www.jpho-online.com | 111
Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
Abla et al J Pediatr Hematol Oncol  Volume 38, Number 2, March 2016

status at diagnosis and findings from available neuroimaging

tests were also recorded. In addition, we collected data on TABLE 1. Clinical Features of Children Presenting With Acute
Leukemia and Hyperleukocytosis
end of induction BM morphologic remission status and
survival. n (%)
Characteristics ALL (n = 84) AML (n = 18)
Definition of Complications
Early morbidity and mortality were defined as adverse Sex
Male 56 (66.7) 11 (61.1)
events that occurred during the first 14 days after pre-
Female 28 (33.3) 7 (38.9)
sentation. Cause of death was determined by clinical findings Age (y)
or autopsy. A neurological event was defined as any neu- Median 5.3 5.1
rological complication that met the National Cancer Insti- Range 0.1-17.0 0.3-16.4
tute Common Terminology Criteria for Adverse Events WBC ( 109/L)
(CTCAE, version 3.0) grade 3 or 4. Pulmonary leukostasis Median 207.9 221.3
syndrome was defined by the presence of a triad: infiltrates Range 101.0-1117.2 106.2-718.3
on chest roentgenogram, tachypnea, and hypoxia.8 Patients Category
100-199 38 (45.2) 5 (27.8)
with a concurrent mediastinal mass on presentation were 200-299 16 (19.0) 6 (33.3)
excluded from the pulmonary leukostasis analysis. Hyper- 300-399 8 (9.5) 2 (11.1
uricemia, hyperkalemia, hyperphosphatemia, and hypo- > 400 22 (26.2) 5 (27.8)
calcemia were defined if they were outside the normal range Hemoglobin (g/L)
for age according to our institutional values. A patient was Median 75 78
considered to have renal dysfunction if the peak serum cre- Range 32-159 49-115
atinine was greater than twice the upper limit of normal for Platelet count ( 109/L)
age and sex (CTCAE, version 3.0). Coagulopathy was Median 22 27
Range 1-474 4-349
defined as having an international normalized ratio, and/or Immunophenotype
prothrombine time, and/or partial thromboplastin time, ALL B-lineage 51 (60.7)
and/or fibrinogen levels outside the normal range for age ALL T-lineage 33 (39.3)
according to our institutional values. AML-M1 6 (33.3)
AML-M2 3 (16.7)
Treatment of Hyperleukocytosis AML-M4 5 (27.8)
The methods used for the reduction of WBC were AML-M5 4 (22.2)
recorded for each patient presenting with hyperleukocytosis, CNS status
CNS 1 63 (75.0) 11 (61.1)
that is, continuous intravenous hyperhydration, chemo- CNS 2 8 (9.5) 1 (5.6)
therapy, and leukapheresis. Cytoreduction with leuka- CNS 3 9 (10.7) 4 (22.2)
pheresis was performed at the discretion of the treating Unknown 4 (4.8) 2 (11.1)
physician. For patients undergoing leukapheresis, number of Coagulation studies*
sessions, type of circuit priming (with whole blood, normal Abnormal INR 27 (54.0) 9 (50.0)
saline, or FFP), clinical complications, and laboratory Prolonged PT 6 (17.6) (n = 34) 1 (16.7) (n = 6)
abnormalities (CBC, coagulation studies, calcium, phos- Prolonged PTT 10 (12.2) 2 (11.1)
phorus, potassium, uric acid, creatinine, and urea) were Low fibrinogen 16 (41.0) (n = 39) 5 (41.7) (n = 12)
Coagulopathy 41 (48.8 10 (55.6%)
recorded immediately before and after leukapheresis. In Metabolic complications
addition, the time interval between presentation to the hos- Hyperkalemia 3 (3.6) 0
pital and administration of the first dose of chemotherapy, Hyperphosphatemia 6 (7.1) 1 (5.6)
duration of admission to intensive care unit, and total length Hypocalcemia 14 (16.7) 4 (22.2)
of hospital admission were calculated for the leukapheresis Hyperuricemia 51 (60.7) 9 (50.0)
and nonleukapheresis groups. Received rasburicasew 16 (51.6) 5 (50)
Renal dysfunction 5 (6.0) 2 (11.1)
Statistical Analysis Dialysis 4 (4.8) 1 (5.6)
Initial LDH
Statistical analysis was performed using Fisher exact Median (n = 21, 11) 4962 4249
test, to test for association between categorical factors. The Range 550-20688 489-7324
Wilcoxon rank-sum test was used to compare continuous Transfusions
variables. All P values reflect 2-sided tests. Statistical FFP 34 (40.5) (n = 83) 6 (33.3)
analysis was performed using SPSS, version 17.0. PRBC 41 (50.0) (n = 82) 11 (61.1)
Platelet 76 (91.6) (n = 83) 16 (88.9)
Received leukapheresis 18 (21.4) 12 (66.7)
RESULTS Double-blood volume 13 (15.5) 10 (55.6)
Double-blood volume 2 4 (4.8) 1 (5.6)
Patients and Management Double-blood volume 3 1 (1.2) 1 (5.6)
During the study period, 634 children were diagnosed
with ALL and 143 were diagnosed with AML. Eighty-four *There were 82 ALL pateints who had coagulation information avail-
able at diagnosis: 50 had INR, whereas 34 had PT only (2 had both INR and
of 634 (13%) ALL patients and 18 of 143 (12.5%) AML PT checked).
patients presented with hyperleukocytosis. The presenting w31 ALL and 10 AML patients were diagnosed after 2000.
features of patients with ALL and AML are shown AML indicates acute myeloid leukemia; ALL, acute lymphoblastic
in Table 1. All patients received continuous intravenous leukemia; CNS, central nervous system; FFP, fresh frozen plasma, INR,
hyperhydration with either allopurinol or rasburicase to international normalized ratio; LDH, lactate dehydrogenase; PRBC, packed
red blood cell; PT, prothrombine time; PTT, partial thromboplastin time;
reduce serum uric acid concentration. Cytoreduction with WBC, white blood cell.
leukapheresis was performed in 30 patients, 18 had ALL

112 | www.jpho-online.com Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
J Pediatr Hematol Oncol  Volume 38, Number 2, March 2016 Leukapheresis in Pediatric Leukemias

and 12 had AML. More than 1 leukapheresis was per- present in 1 patient already at the time of admission but
formed in 7 patients, with a total of 39 procedures. Twenty- developed within days 1 to 10 in all others. Four patients
two of these were primed with reconstituted whole blood required dialysis (3 for abnormal creatinine and 1 for
(FFP added to RBC), 16 were primed with normal saline, hyperphosphatemia), of whom 2 had been treated with
and 1 with FFP. The median initial WBC count was urate oxidase and 1 had undergone leukapheresis. Six
474 109/L (range, 167.0 to 1117.2 109/L) in the leuka- patients (7.1%) had a neurological complication at pre-
pheresis group compared with 175 109/L (range, 101.1 to sentation. A total of 5 patients underwent leukapheresis, 3
710 109/L) in the nonleukapheresis group (P < 0.001). of whom had T-ALL and presented with seizures because
Leukapheresis resulted in an average WBC count reduction of CNS hemorrhage (all had an initial WBC > 650 109/L
of 52.8% at first procedure with no differences between and an initial platelet count ranging from 4 to 33 109/L), 1
ALL and AML (54% vs. 48.6%, P = 0.49), 45.6% further patient had B-precursor ALL and presented with seizures
reduction at second leukapheresis (46.3% vs. 38.6%, and a WBC of 969.1 109/L and another patient had T-
P = 0.26), and 55.4% further reduction at third leukaphe- ALL and presented with confusion and a presenting WBC
resis (69% vs. 41.8%). Patients who underwent the proce- of 460109/L. There was 1 death from CNS hemorrhage
dure were more often AML (P < 0.001), had central nerv- and 1 nonfatal ischemic stroke among these 5 patients who
ous system (CNS) involvement (P = 0.006), coagulopathy underwent leukapheresis. One patient with T-ALL who
(P = 0.004), and neurological complications (P < 0.001) at presented with seizures and a WBC of 314.6 109/L did not
first presentation, and more often received rasburicase undergo leukapheresis, and eventually recovered. Pulmo-
(P = 0.015), FFP (P = 0.008), and PRBC transfusion nary leukostasis syndrome was present in 16 (19%) patients
(P = 0.005) before leukapheresis. The median time between with ALL at presentation; 9 of these received chemotherapy
hospital admission and the start of chemotherapy was 38.5 (median initial WBC of 156 109/L; range, 105 to 326) and
hours (range, 14 to 202 h) in the leukapheresis group and 36 recovered thereafter, whereas the remaining 7 patients
hours (range, 12 to 118 h) in the nonleukapheresis group (median initial WBC 726109/L; range, 273 to 969)
(P = 0.289). The median duration of intensive care unit underwent leukapheresis. Three of these 7 patients required
admission was 2.25 days (range, 0 to 15 d) in the leuka- mechanical ventilation and all had initial WBCZ650 109/
pheresis group versus 0 days (range, 0 to 10 d) in the non- L. One of these patients died because of CNS hemorrhage.
leukapheresis one (P < 0.001), whereas the median length of One patient who was hypoxic preleukapheresis had wor-
hospital admission was 26.5 days (range, 1 to 153 d) in the sening of his pulmonary symptoms after the procedure but
leukapheresis group versus 16 days (range, 7 to 127 d) in the recovered later. The remaining 5 ALL patients had an
nonleukapheresis one (P = 0.028). Where data were avail- improvement of their symptoms after leukapheresis.
able, there was no significant difference in end of induction Overall, CNS hemorrhage and severe pulmonary leu-
BM complete remission (CR) rates between the 2 groups, kostasis (requiring mechanical ventilation) were more fre-
with an 88% (22 of 25 patients who survived to end of quent in ALL patients with initial WBCZ650109/L, and
induction) CR rate in the leukapheresis group versus 94% the incidence of early death was 1.2% (1/84) which was
(65/69 patients who survived to end of induction) CR rate in secondary to CNS hemorrhage.
the nonleukapheresis group (P = 0.378).
Hyperleukocytosis Complications in AML
Hyperleukocytosis Complications in ALL Table 2 depicts complications stratified by initial WBC
Early complications of hyperleukocytosis in ALL in children with AML. Two of 18 (11%) patients developed
according to initial WBC are listed in Table 2. Renal dys- renal dysfunction, which was present at admission in 1
function developed in 5 of 84 ALL patients (6%), and was patient. This patient had received urate oxidase. Renal

TABLE 2. Early Complications in Children With Acute Leukemia and Hyperleukocytosis

No. Pulmonary Renal Dysfunction Neurological Early Death
Diagnosis WBC Patients Leukostasis* (Dialysis) Complications (Cause)
ALL 100-199 38 6 — — —
200-299 16 3 2 (1) — —
300-399 8 1 2 (1) 1 —
Z400 22 6 1 (2)w 5 One (multiple parenchymal
and
brain stem hemorrhages)
AML 100-199 5 – – – –
200-299 6 1 1 (0) 1 One (pulmonary leukostasis
and sepsis)
300-399 2 2 — 1 —
Z400 5 1 1 (1) 2 Two (intracranial hemorrhage,
pulmonary hemorrhage
leading to ARDS)
Total 102 20 (19.6%) 7 (5) (6.9%) 10 (9.8%) 4 (3.9%)
*Respiratory events data available for 83 ALL patients.
wOne patient received dialysis to treat hyperphosphatemia (PO4 = 3.0 mmol/L), the other 3 were dialysed because of abnormal creatinine.
ARDS indicates acute respiratory distress syndrome; AML, acute myeloid leukemia; ALL, acute lymphoblastic leukemia; WBC, white blood cell.

Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved. www.jpho-online.com | 113
Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
Abla et al J Pediatr Hematol Oncol  Volume 38, Number 2, March 2016

dysfunction in the other patient, who did not receive urate 590 109/L, respectively), who were initially asymptomatic;
oxidase, developed within 10 days. One of the above 2 both of them recovered. Respiratory distress worsened in
patients underwent dialysis, and both had leukapheresis. 10 patients after leukapheresis, 3 had ALL (2 T-ALL, 1
Neurological complications were present in 4 patients B-precursor ALL, WBCZ350109/L) and 7 had AML.
(22.2%) at presentation. One patient with AML-M2 and One patient with AML-M5 who presented with WBC of
with an initial WBC = 417  109/L presented with paralysis, 450 109/L and chest infiltrates developed hypoxia and
ataxia, confusion, and CNS hemorrhage. This patient had needed to be intubated during the procedure, but sub-
received PRBC transfusion before and during leukaphe- sequently recovered; however, the patient suffered relapsed
resis, experienced tentorial herniation because of increased disease 6 months later and died. Another patient with
intracranial pressure, and died within 24 hours from pre- AML-M5 and a WBC of 201 109/L, with preexisting chest
sentation. The remaining 3 patients presented with seizures infiltrates developed severe hypoxia after the leukapheresis
(WBC = 360.8  109/L), dizziness, headache (WBC = and after starting chemotherapy with cytarabine, etoposide,
718.3  109/L), and confusion (WBC = 201.4 109/L). All and doxorubicin. He died after 72 hours from presentation
were leukapheresed and the latter patient died within 72 due to concurrent pulmonary leukostasis/hemorrhage and
hours from presentation because of pulmonary leukostasis, sepsis.
multiple organ dysfunction, and septic shock. Four patients New-onset coagulopathy (prolonged international
(22.2%) presented with pulmonary leukostasis, 1 recovered normalized ratio) occurred in 5 patients (12.8%) after
after chemotherapy only, whereas 3 patients were leuka- cytoreductive procedure. Other complications likely related
pheresed. One of these 3 patients died due to cerebral to leukapheresis included hypokalemia (n = 14), hypo-
complications, whereas 2 had worsening of respiratory calcemia (n = 5), hypomagnesemia (n = 5), hypertension
symptoms (1 was intubated) that subsequently improved. (n = 5), hypotension (n = 2), cardiac arrest (n = 1), brady-
Overall, 3 AML patients experienced early death, 1 cardia (n = 4), tachycardia (n = 3), femoral vein thrombosis
died due to pulmonary leukostasis/hemorrhage and sepsis (n = 4), hypothermia (n = 1), and rigors (n = 1).
after leukapheresis (initial WBC was 201  109/L), whereas
the other 2 died due to CNS hemorrhage (both with initial Long-term Outcome
WBC > 400109/L) one of whom had also pulmonary Where data were available, there was no significant
hemorrhage. All of these 3 patients underwent leukaphe- difference in the relapse rate (RR) between the leukaphe-
resis, however, the death in these 3 patients can be attrib- resis group (RR = 38.5%, or 10/26) and the non-
uted to the severe pulmonary and CNS toxicity related to leukapheresis group (RR = 34.3%, or 23/67) (P = 0.810).
hyperleukocytosis rather than to the leukapheresis proce- Late deaths occurred in 13/26 (50.0%) patients from the
dure itself. leukapheresis group compared with 21/72 (29.2%) in the
Early death due to hyperleukocytosis was associated nonleukapheresis group (P = 0.091), and the median time
with neurological complications (P < 0.001), with an AML from diagnosis to death was 521 days (range, 22 to 3254 d)
diagnosis (P = 0.017) and with initial coagulopathy in both groups.
(P = 0.015).

Complications After Leukapheresis DISCUSSION

Six leukapheresis procedures had no associated com- Because of the increased risk of severe neurological
plications, whereas complications occurred after the and pulmonary complications, children with hyper-
remaining 33 procedures (85%) in 26 patients. Among all leukocytic acute leukemias have a higher early mortality
children with acute leukemia, neurological complications rate than those with nonhyperleukocytic leukemias.8–11
occurred in 6 patients (15.4%) after cytoreduction. Two Children with AML and hyperleukocytosis have even worse
patients (T-ALL with initial WBC of 350 109/L, and prognosis with an early mortality rate between 9% and
AML with initial WBC of 800109/L) had seizures in the 16.9%1,9compared with 4% in children with ALL and
presence of mildly abnormal electrolytes (calcium = 0.88, hyperleukocytosis.8 Our study showed that, despite
magnesium = 0.59, potassium = 2.9) and normal neuro- aggressive supportive care measures, including leukaphe-
imaging; both of them recovered. One AML patient resis, hyperleukocytosis in pediatric ALL and AML con-
developed brain herniation due to increased intracranial tinues to be associated with severe CNS toxicities, such as
pressure during the procedure and subsequently died. Three hemorrhage/ischemia as well as pulmonary leukostasis/
had an ischemic stroke: 1 patient with AML and initial hemorrhage, with increasing risk in those with a very high
WBC count 467.5 109/L, was neurologically normal pre- initial WBC count. In this study, among children with
leukapheresis but was coagulopathic, developed ischemic ALL, CNS hemorrhage occurred in 3/84 patients (3.5%)
stroke with seizures, and pulmonary hemorrhage after the and was more common in those with initial
second apheresis and died within 72 hours from pre- WBCZ650109/L, 1 case was fatal. This is similar to the
sentation. Two more ischemic strokes occurred in 1 coa- report by Maurer et al12 where the incidence of CNS
gulopathic B-cell ALL and 1 AML-M4 with initial WBC hemorrhage in children with ALL was 3% (4/124) and all
counts of 672.1 and 297.6, respectively; in both cases, the occurred in those with a WBC > 600 109/L. In contrast,
leukapheresis circuit was primed with RBCs and a PRBC the report by Lowe and colleagues showed a CNS hemor-
transfusion was administered after the procedure. The ALL rhage rate of 2% (4/178 with initial WBC > 200 109/L)
patient presented also with pulmonary leukostasis which and all 4 had WBC > 400 109/L. In our study, however,
got worse after leukapheresis but eventually recovered later, minor neurological symptoms like seizures and confusion
however, he relapsed and died at 1 year from diagnosis, were present in 2 ALL patients with initial WBC of 314.6
whereas the AML patient recovered fully and is currently in and 460109/L, respectively. Among children with AML,
CR. New respiratory distress after cytoreduction occurred CNS hemorrhage occurred at a lower initial WBC
in 2 patients with T-ALL, (WBC = 350 109/L and (400 109/L) than ALL cases, with 2 early deaths due to

114 | www.jpho-online.com Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
J Pediatr Hematol Oncol  Volume 38, Number 2, March 2016 Leukapheresis in Pediatric Leukemias

CNS hemorrhage, one of these had also pulmonary bleed- included 89 children from whom PHIS data were available.
ing. One of the contributing factors to the death of ALL Sixteen (18%) of them underwent leukapheresis and one of
and AML patients with CNS hemorrhage is that those 3 them (6.3%) experienced early death compared with 3 of 73
patients, in addition to thrombocytopenia and coagulop- (4.3%) for those who did not receive luekapheresis, sug-
athy, received PRBC transfusion at presentation for a gesting that leukapheresis did not reduce induction mor-
hemoglobin level of <60 g/L. In fact, Lichtman and tality.1 A more recent meta-analysis by Oberoi et al20 that
Rowe13 showed a correlation between a higher blood vis- evaluated early deaths in patients with AML and hyper-
cosity and an increased risk of adverse event. Further, red leukocytosis, did not show any difference in early mortality
cell transfusions and high hemoglobin concentrations have rate based on approach to leukapheresis or use of
been associated with increased morbidity and mortal- hydroxyurea and/or low-dose chemotherapy. A study by
ity.12,14 In addition, severe thrombocytopenia is a well- Lowe et al8 showed that neurological complications were
known risk factor for CNS hemorrhage in patients with more frequent in children with ALL and initial
hyperleukocytosis.15 WBC > 400109/L (17.9% vs. 3.6%) and that cytor-
Although pulmonary hemorrhage, edema, or infection eduction may be considered for these patients or those who
may be contributing factors to the respiratory symptoms, have complications at presentation. Our study showed that
pulmonary leukostasis worsened with the increase in WBC severe pulmonary and neurological complications were more
count. An initial WBC cut-off of 650 109/L was a risk frequent in children with ALL and WBC > 650109/L; in
factor for severe pulmonary leukostasis requiring mech- AML cases pulmonary leukostasis was more common with
anical ventilation in children with ALL; whereas this com- initial WBC of >200 109/L, whereas CNS hemorrhage was
plication occurred at an earlier initial WBC (201 109/L) in more common with a WBC of >400 109/L. It is hard to
AML cases. The 2 patients with AML who developed judge whether leukapheresis in our patients had any impact
pulmonary leukostasis requiring intubation were both on early death. Four of the 5 patients with ALL who pre-
AML-M5. The higher incidence of leukostasis among chil- sented with neurological event and underwent leukapheresis
dren with AML compared with ALL despite their lower recovered after the procedure (1 had initial WBC > 400
presenting WBC counts could be related to the larger mean 109/L, developed ischemic stroke but improved after leuka-
cell volume of myeloblasts (especially in AML-M5/M4), the pheresis), whereas 1 died due to CNS hemorrhage. Among
presence of adhesion molecules on the surface of these cells the leukapheresed ALL and AML patients who developed a
and their chemotactic response to cytokines.4 Furthermore, stroke (1 ALL and 2 AML), it is difficult to determine
respiratory distress can develop after starting chemotherapy whether the stroke was triggered by a preceding cerebral
because of blast cell lysis, leading to the release of enzymes leukostasis or whether it was an adverse effect of the proce-
that can cause alveolar damage and edema.4 dure itself. In at least one of the patients who developed an
Leukapheresis is a procedure where WBCs are ischemic stroke after leukapheresis, the stroke may have been
removed from the blood, whereas the plasma, RBCs, and precipitated by the use of RBCs (hematocrit = 0.7) in the
platelets are returned to the patient. Although leukaphe- apheresis priming process, combined with a PRBC trans-
resis has been used to treat leukostasis associated with fusion for a hemoglobin level of 69 g/L after second leuka-
hyperluekocytosis, primarily in AML and ALL, no pheresis. Although leukapheresis was helpful in improving
randomized trials evaluating its benefits have been pub- respiratory complications in 18 patients, it caused a worsen-
lished. One of the rationales behind the use of leukapheresis ing of respiratory symptoms in at least 10 patients and new
is to rapidly remove a large number of leukemic cells to respiratory distress in 2 patients. The report by Lowe et al8
avoid leukostasis; however, immediate start of chemo- showed that 1 neurological and 5 respiratory complications
therapy may be as effective and less expensive. Another occurred after cytoreduction in children with ALL and
historical reason to use leukapheresis is to control the hyperleukocytosis.
metabolic complications associated with hyperleukocytosis, Leukapheresis has a few disadvantages as well. First, it
however, as the discovery of rasburicase for the manage- involves the insertion of a temporary large central venous
ment of hyperuricemia in patients at a high risk for tumor line that may require anesthesia in children and can pre-
lysis syndrome,16 this rationale no longer exists. In addi- dispose to bleeding, or thrombosis at the catheter site. The
tion, a Children’s Oncology Group study showed that incidence of central venous line–related femoral vein
metabolic complications are no longer the primary concern thrombosis was 10.3% in our cohort. Second, citrate is an
in children with AML and hyperleukocytosis.1 Leukaphe- anticoagulant that is usually required to prevent clotting of
resis can also increase the fraction of leukemic cells in the S- the apheresis circuit. However, citrate binds calcium and
phase and therefore may increase the efficacy of cytar- can lead to symptoms of hypocalcemia that may necessitate
abine.17 Finally, priming the leukapheresis circuit with FFP calcium administration. This by itself may cause calcium-
may reduce the risk of hemorrhage.17 Several retrospective phosphate precipitation and worsen tumor lysis syn-
studies have looked at the role of leukapheresis in reducing drome.21 The incidence of asymptomatic hypocalcemia in
early mortality and morbidity in patients with hyper- our study was 12.8%. Third, leukapheresis can decrease the
leukocytic leukemias, but they have all yielded con- platelet count by as much as 44%, therefore a after pro-
troversial results.6,18–20 Two retrospective studies of adults cedure CBC is required to determine the need for platelet
with AML and hyperleukocytosis, 50% of whom under- transfusion. Fourth, blood loss is a common side effect
went leukapheresis, showed that early mortality was sig- after leukapheresis as RBCs are collected with WBCs.
nificantly reduced in the leukapheresis group but long-term However, transfusion of RBC is not recommended as this
survival was similar in the 2 groups.6,18 A recent study may worsen blood hyperviscosity. Fifth, leukapheresis may
showed no improvement of early mortality in 52 patients cause fragmentation of WBC with subsequent intravascular
with hyperleukocytic leukemia who underwent leukaphe- coagulation that may increase the risk of early death.17
resis.19 A large Children’s Oncology Group study of 256 Further, leukapheresis has been reported to cause a delay in
children with AML and an initial WBCZ100109/L starting induction chemotherapy,1,8,20 however, our study

Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved. www.jpho-online.com | 115
Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
Abla et al J Pediatr Hematol Oncol  Volume 38, Number 2, March 2016

WBC > 100109/L (especially with M4/M5 subtypes) or

TABLE 3. Incidence of New Complications After 39 in ALL with WBC > 400 109/L.28 Furthermore, the
Leukapheresis Procedures in 30 Patients
guidelines stated that undiluted PRBCs should be avoided
Complications n (%) (N = 39) in small children with hyperviscosity.
Neurological (n = 6) 6 (15.4) We conclude that leukemic children presenting with
Ischemic stroke 3 (7.7) very high WBC counts are still at major risk of severe CNS
Brain herniation 1 (2.6) and pulmonary injuries in spite of all current supportive
Seizures 2 (5.1) measures. Leukapheresis may be of benefit in reversing
Respiratory (n = 2) 2 (5.1) leukostasis syndrome in some patients, but it is unknown
Coagulopathy 5 (12.8) whether it can help in reducing early mortality in those with
Metabolic (n = 26) 26 (67) very high WBC counts. We observed a high incidence of
Hypokalemia 14 (35.9.5)
Hyperkalemia 1 (2.5)
side effects because of the procedure, but most of them were
Hypomagnesemia 5 (12.8) reversible. We could not find evidence that leukapheresis
Hypocalcemia 5 (12.8) can delay the start of chemotherapy, nor that it can affect
Hyperuricemia 1 (2.6) end of induction CR rates or long-term survival. Therefore,
Hemodynamic instability (n = 15) 15 (38) we recommend that rasburicase and aggressive correction
Cardiac arrest 1 (2.6) of thrombocytopenia/coagulopathy and absolute avoidance
Hypertension 5 (12.8) of PRBC transfusions should be immediately implemented
Hypotension 2 (5.1) in leukemic patients presenting with hyperleukocytosis.
Tachycardia 3 (7.7) After these aggressive measures and hemodynamic stability
Bradycardia 4 (10.3)
of the patient, chemotherapy should be started as soon as
Other
Femoral vein thrombosis 4 (10.3) possible and at a pediatric tertiary care center with avail-
Hypothermia 1 (2.6) able intensive care and apheresis teams. Leukapheresis
Rigors 1 (2.6) should be performed only by experienced personnel with
Early death 4 (10.3) adequate monitoring and preventive measures, and should
not delay the start of chemotherapy. Although there might
be ethical concerns, a large prospective randomized trial
would be needed to define the role of leukapheresis with
did not find any significant delay in starting chemotherapy chemotherapy versus chemotherapy alone.
between the leukapheresis and nonleukapheresis groups. In
addition, our study showed that the 39 leukapheresis pro-
cedures were complicated by electrolyte imbalances (67%) REFERENCES
and hemodynamic instability (38%), which were all rever-
sible (Table 3). In contrast, leukapheresis complications 1. Sung L, Aplenc R, Alonzo TA, et al. Predictors and short-term
outcomes of hyperleukocytosis in children with acute myeloid
have not been well characterized before, but it appeared to leukemia: a report from the Children’s Oncology Group.
be well tolerated in other pediatric reports of childhood Haematologica. 2012;97:1770–1773.
hyperleukocytic leukemias8,22–25 except from mild hypo- 2. Porcu P, Cripe LD, Ng EW, et al. Hyperleukocytic leukemias
calcemia26 and minor clinical problems, such as tachycardia and leukostasis: a review of pathophysiology, clinical presen-
and hypotension.23 tation and management. Leuk Lymphoma. 2000;39:1–18.
A new clinical grading scale was recently proposed by 3. Dearth J, Salter M, Wilson E, et al. Early death in acute
Novotny and colleagues, to determine the risk of leuko- leukemia in children. Med Pediatr Oncol. 1983;11:
stasis in patients with hyperleukocytic leukemias. Patients 225–228.
were divided into 4 groups as risk of leukostasis: not 4. Hijiya N, Metger ML, Pounds S, et al. Severe cardiopulmo-
nary complications consistent with systemic inflammatory
present, possible, probable, and highly probable. This was response syndrome caused by leukemia cell lysis in childhood
based on the severity of pulmonary, neurological, and other acute myelomonocytic leukemia. Pediatr Blood Cancer.
symptoms.27 This tool may assist in the standardization of 2005;44:63–69.
approaches to patients with hyperleukocytosis, but has not 5. Aqui N, O’Doherty U. Leukocytapheresis for the treatment of
been widely accepted yet. Furthermore, it is difficult to hyperleukocytosis secondary to acute leukemia. Hematology
apply this grading scale to pediatric patients with acute Am Soc Hematol Educ Program. 2014;2014:457–460.
leukemias, particularly very young children. 6. Giles Fl, Shen Y, Kantarjian HM, et al. Leukapheresis reduces
The 2013 Guidelines on the use of Therapeutic early mortality in patients with acute myeloid leukemia with
Apheresis in Clinical Practice was published by the high white cell counts but does not improve long-term survival.
Leuk Lymphoma. 2001;42:67–73.
American Society of Apheresis: hyperleukocytosis secon- 7. Sanz MA, Grimwade D, Tallman MS, et al. Management of
dary to AML and ALL is a category I recommendation acute promyelocytic leukemia: recommendation from an
(defined as a disorder for which apheresis is accepted as expert panel on behalf of the European Leukemia Net. Blood.
first-line therapy, either as a primary stand-alone treatment 2009;113:1875–1891.
or in conjunction with other modes of treatment) and grade 8. Lowe EJ, Pui CH, Hancock ML, et al. Early complications in
1B (strong recommendation, moderate-quality evidence, children with acute lymphoblastic leukemia presenting with
can apply to most patients in most circumstances without hyperleukocytosis. Pediatr Blood Cancer. 2005;45:10–15.
reservation) for the treatment of leukostasis and a category 9. Creutzig U, Zimmermann M, Reinhardt D, et al. Early deaths
III recommendation (optimum role of apheresis therapy is and treatment-related mortality in children undergoing therapy
for acute myeloid leukemia: analysis of the multicenter clinical
not established; decision making should be individualized) trials AML-BFM 93 and AML-BFM 98. J Clin Oncol. 2004;
and grade 2C (very weak recommendation, based on 22:4384–4393.
observational studies or case series; other alternatives may 10. Inaba H, Fan Y, Pounds S, et al. Clinical and biologic features
be equally reasonable) for prophylaxis in AML with and treatment outcome of children with newly diagnosed acute

116 | www.jpho-online.com Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
J Pediatr Hematol Oncol  Volume 38, Number 2, March 2016 Leukapheresis in Pediatric Leukemias

myeloid leukemia and hyperleukocytosis. Cancer. 2008;113: 20. Oberoi S, Lehrnbecher T, Phillips B, et al. Leukapheresis and
522–529. low-dose chemotherapy do not reduce early mortality in acute
11. Bunin NJ, Pui CH. Differing complications of hyperleukocy- myeloid leukemia hyperleukocytosis: a systematic review and
tosis in children with acute lymphoblastic or acute non- meta-analysis. Leuk Res. 2014;38:460–468.
lymphoblastic leukemia. J Clin Oncol. 1985;3:1590–1595. 21. Howard SC, Jones DP, Pui CH. The tumor lysis syndrome.
12. Maurer HS, Steinherz PG, Gaynon PS, et al. The effect of N Engl J Med. 2011;364:1844–1854.
initial management of hyperleukocytosis on early complica- 22. Haase R, Merkel N, Diwan O, et al. Leukapheresis and
tions and outcome of children with acute lymphoblastic exchange transfusion in children with acute leukemia and
leukemia. J Clin Oncol. 1988;6:1425–1432. hyperleukocytosis. A single center experience. Klin Padiatr.
13. Lichtman MA, Rowe JM. Hyperleukocytic leukemias: rheo- 2009;221:374–378.
logical, clinical, and therapeutic considerations. Blood. 1982; 23. Greze V, Chambon F, Merlin E, et al. Leukapheresis in
60:279–283. management of hyperleukocytosis in children’s leukemias.
14. Harris AL. Leukostasis associated with blood transfusion in J Pediatr Hematol Oncol. 2014;36:513–517.
acute myeloid leukaemia. Br Med J. 1978;1:1169–1171. 24. Veljkovic D, Kuzmanovic M, Micic D, et al. Leukapheresis in
15. Nowacki P, Zdziarska B, Fryze C, et al. Co-existence of management hyperlecocytosis induced complications in two
thrombocytopenia and hyperleukocytosis (‘critical period’) as a pediatric patients with chronic myelogenous leukemia. Trans-
risk factor of haemorrhage into the central nervous system in fus Apher Sci. 2012;46:263–267.
patients with acute leukaemias. Haematologia. 2002;31:347–355. 25. Woloskie S, Armelagos H, Meade JM, et al. Leukodepletion
16. Kikuchi A, Kigasawa H, Tsurusawa M, et al. A study of for acute lymphocytic leukemia in a three-week-old infant. J
rasburicase for the management of hyperuricemia in pediatric Clin Apher. 2001;16:31–32.
patients with newly diagnosed hematologic malignancies at 26. Yilmaz D, Karapinar B, Karadas N, et al. Leukapheresis in
high risk for tumor lysis syndrome. Int J Hematol. 2009;90: childhood acute leukemias: single-center experience. Pediatr
492–500. Hematol Oncol. 2014;31:318–326.
17. Thiebaut A, Thomas X, Belhabri A, et al. impact of pre- 27. Novotny JR, Muller-Beissenhirtz H, Hergt-Rosenthal S, et al.
induction therapy leukapheresis on treatment outcome in adult Grading of symptoms in hyperleukocytic leukaemia: a clinical
acute myelogenous leukemia presenting with hyperleukocyto- model for the role of different blast types and promyelocytes in
sis. Ann Hematol. 2000;79:501–506. the development of leukostasis syndrome. Eur J Haematol.
18. Bug G, Anargyrou K, Tonn T, et al. Impact of leukapheresis 2005;74:501–510.
on early death rate in adult acute myeloid leukemia presenting 28. Schwartz J, Winters JL, Padmanabhan A, et al. Guidelines on
with hyperleukocytosis. Transfusion. 2007;47:1843–1850. the use of therapeutic apheresis in clinical practice-evidence-
19. Pastore F, Pastore A, Wittmann G, et al. The role of based approach from the Writing Committee of the American
therapeutic leukapheresis in hyperleukocytic AML. PLoS Society for Apheresis: the sixth special issue. J Clin Apher.
One. 2014;9:e95062. 2013;28:145–284.

Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved. www.jpho-online.com | 117
Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.