Thrombosis Research 135 (2015) 846–851

Contents lists available at ScienceDirect

Thrombosis Research
journal homepage: www.elsevier.com/locate/thromres

Regular Article

Prediction of thrombotic and hemorrhagic events during polycythemia

vera or essential thrombocythemia based on leukocyte burden☆,☆☆
Yoojoo Lim 1, Jeong-Ok Lee 1, Se Hyun Kim, Jin Won Kim, Yu Jung Kim, Keun-Wook Lee,
Jong Seok Lee, Soo-Mee Bang ⁎
Department of Internal Medicine, Seoul National University College of Medicine, Seoul National University Bundang Hospital, Seongnam, Republic of Korea

a r t i c l e i n f o a b s t r a c t

Article history: Background: Evidences suggest an association between leukocytosis and thrombotic or hemorrhagic complica-
Received 6 December 2014 tion in polycythemia vera (PV) and essential thrombocythemia (ET), but clinical implication is not well known.
Received in revised form 21 January 2015 Objective: To evaluate whether leukocyte burden during follow-up is related to thrombotic or hemorrhagic
Accepted 18 February 2015 events in PV and ET.
Available online 26 February 2015
Patients/Methods: We retrospectively analyzed patients with PV or ET treated at Seoul National University
Bundang Hospital, Korea. Time-weighted averages of leukocytes during the follow-up period were defined as
Keywords:
Polycythemia vera
leukocyte burden and were calculated for each patient and compared between patient subgroups. In each patient
Essential thrombocythemia with events, leukocyte burden for the 3-month period before the event was compared with that for the entire
Predictor follow-up period.
Thrombosis Results: In 102 patients with PV or ET, 35 events (16 thrombotic, 19 hemorrhagic) occurred in 29 patients
Hemorrhage (median follow-up, 54 months). Leukocyte burden were significantly higher in patients with events than in
Time-weighted averages event-free patients (12,015 × 103 /μL vs. 9,567 × 103/μL, P = 0.003). The difference was more prominent in ET
patients than in PV patients, and in patients with hemorrhagic events than in those with thrombotic events. In
patients with events, the leukocyte burden in the pre-event period was higher than in the entire follow-up period
(16,767 × 103/μL vs. 12,015 × 103/μL, P = 0.002). In all patients, leukocyte burden during entire follow-up period
of 11,000 × 103/μL or higher was an independent risk factor for vascular events.
Conclusion: In PV or ET patients, leukocyte burden during disease course is related to increased incidence of
thrombotic or hemorrhagic events.
© 2015 Elsevier Ltd. All rights reserved.

Introduction There is a paucity of understanding concerning occurrence of

Polycythemia vera (PV) and essential thrombocythemia (ET) are
myeloproliferative neoplasms (MPN) characterized by increased
production of mature red blood cells (RBC) or platelets, respectively.
Since PV and ET both lead to an increased risk of thrombotic or
hemorrhagic complications and one of the most important current
treatment goals in these MPNs is to minimize the risk of these vascular
complications.
☆ Financial Support: This study was supported by a grant from Seoul National
University Bundang Hospital Research Fund (grant no. 11–2012-025)
☆☆ Role of the Funding Source: The funding source had no involvement in the study
design; the collection, analysis and interpretation of data; writing of the report; nor the
decision to submit the article for publication.
⁎ Corresponding author at: Department of Internal Medicine, Seoul National University
College of Medicine, Seoul National University Bundang Hospital, 82 Gumi-Ro, 173 Beon-
Gil, Bundang-Gu, Seongnam, Gyeonggi-Do, 463–707, Republic of Korea. Tel.: +82 31 787
7038.
E-mail address: smbang7@snu.ac.kr (S.-M. Bang).
1
Contributed equally to this work.
thrombotic or hemorrhagic complications, thus finding predictive
biomarkers of these complications is an area of current interest in
PV and ET. Growing evidence suggests that erythrocytosis or
thrombocytosis per se is not always predictive of vascular complica-
tions, except in cases of extreme thrombocytosis, and recent clinical
and preclinical findings suggest that leukocytosis is associated with
thrombotic or hemorrhagic complications [1–5]. Most published
clinical data suggest that higher leukocyte count at diagnosis may
be related to a higher risk of vascular complications. However, it is
uncertain whether this risk could be reduced with a tighter control
on leukocyte count during the course of the disease, or if high initial
leukocyte count is essentially an uncontrollable biomarker
reflecting poor prognosis. Besides providing insight into the
disease, it would also be useful to determine if there is a threshold
leukocyte count value below which the occurrence of vascular
complications may be avoided.
This study was designed to determine whether the leukocyte
burden, defined on the area under the curve (AUC) of the leukocyte-
time course, during the course of PV/ET is a risk factor associated with

http://dx.doi.org/10.1016/j.thromres.2015.02.023
0049-3848/© 2015 Elsevier Ltd. All rights reserved.
 Y. Lim et al. / Thrombosis Research 135 (2015) 846–851
thrombotic/hemorrhagic complications, and to enable practical
guidance in management of PV and ET patients. We retrospectively
analyzed the blood cell counts over a long period in patients with
PV or ET, and calculated the leukocyte burden for each patient. We
then sought to determine possible correlation between the calculat-
ed burden and the occurrence of thrombotic or hemorrhagic
complications.
Methods
Patients
The retrospective, consecutive 2004–2012 database of patients
treated for PV or ET at our hospital by 2 hematologists (J.O.L and
S.M.B) was analyzed by a third hematologist (Y.L). All the patients had
been diagnosed in accordance with the WHO criteria [6], and were
managed according to traditional risk factor groups. Most of the patients
were given antiplatelets, unless there were absolute contraindications.
Phlebotomy was used for hematocrit control in PV patients with
low-risk of thrombosis. Patients at higher risk of thrombosis (age
older than 60 years or prior thrombosis), with extreme thrombocytosis
or leukocytosis, or with disease-related symptoms were given
cytoreductive therapy with hydroxyurea or anagrelide. Patients were
excluded if: (1) their disease had evolved to primary myelofibrosis or
chronic or acute myeloid leukemia during the follow-up period,
(2) the erythrocytosis or thrombocytosis was considered to be second-
ary, or if 3) they had developed any solid cancer during the follow-up
period or had had cancer treatment within 5 years before the diagnosis
of PV or ET.
Definition of Events
We defined thrombotic events as either arterial or venous thrombo-
sis, including ischemic stroke, cerebral transient ischemic attacks, acute
thrombotic myocardial infarction, peripheral arterial thrombosis, and
venous thromboembolism. Major or clinically relevant non-major
hemorrhagic events were considered in evaluating the hemorrhagic
events. Major bleeding was defined as either life threatening or
requiring at least 1 unit of RBC transfusion, or that resulting in a 2 g/dL
or greater decrease in hemoglobin (Hb) count without transfusion, or
that requiring intervention for hemostasis. Clinically relevant
non-major bleeding was defined as overt bleeding that did not meet
the above criteria but was associated with the need for medical interven-
tion, contact with a physician, or interruption of medication or with
discomfort or impairment of activities of daily life. The events described
above that were documented to have happened either at or after the
diagnosis of PV or ET were recorded as relevant events. If multiple events
were documented in 1 patient, the event occurring first was considered
to be the relevant event for the comparison and analysis of blood cell
burden.
Study Design
We aimed to prove the assumption that a higher leukocyte
burden during PV or ET, in terms of either severity or duration, or
both, would lead to a higher incidence of thrombotic and/or hemor-
rhagic complications. To predict the leukocyte burden during the
entire follow-up period, we calculated the time-weighted averages
(i.e. the chronologically-related AUC) of all leukocyte counts during
each follow-up period. The detailed calculation of time-weighted
averages are as follows. For each patient, the average of all pairs of
consecutive measurements was multiplied by the number of days
between the measurements to obtain a product, and then the sum
of all such products for a given period was divided by the total
number of days in that period (Fig. 1). In cases of patients with
events, time-weighted averages for the 3-month period before the
847
Fig. 1. Calculation of blood cell burdens. The figure and the equation given below describes
the method of estimating blood cell burdens by calculation of the time-weighted averages
of a blood count over a follow-up period for a given patient. Each period (A, B, or C) is the
number of days between consecutive blood counts (a and b, b and c, or c and d, respective-
ly). The schematic is not drawn to scale. Abbreviations: PV, polycythemia vera; ET, essen-
tial thrombocythemia.
event (pre-event period) were also determined, and compared
with those for the entire follow-up period. Same calculations and
analysis were also done for Hb and platelets.
Medical records of eligible patients were reviewed retrospectively
for collecting data regarding clinical parameters, events, and laboratory
values. Any and all data for complete blood cell counts for the 3-months
prior to diagnosis of PV or ET were also collected if available, from our
and other hospitals’ laboratories. The determination of JAK2 V617F
and exon 12 mutations were based on polymerase chain reaction
(PCR) and DNA sequencing. Body mass index (BMI) was calculated as
weight (Kgs) divided by the square of height (meters) and the
overweight cutoff was at BMI ≥ 25 kg/m2 for cardiovascular risk factor
analysis, in accordance to WHO BMI cut-off points for overweight [7].
Statistical Analysis
Categorical variables were compared using chi-square test or
Fisher’s exact test as applicable, and continuous variables were
compared using Student’s t-test or nonparametric tests. Paired t-test
or Wilcoxon signed-rank test were used when comparing continuous
variables of different time periods within each patient. Odds ratios for
vascular events according to the leukocyte burden were fitted with a
restricted cubic spline basis to visualize the correlation between vascu-
lar event risk and leukocyte burden. Logistic regression was used for
multivariable analysis and to determine the cut-off value of time-
weighted leukocyte count above which a risk of vascular event is
predictable. P-values of b 0.05 were considered to indicate a statistically
significant difference. Statistical analyses were performed with IBM
SPSS version 20.0 (IBM Corp., Armonk, NY, USA).
The study protocol was approved by the Institutional Review Board
(IRB) of Seoul National University Bundang Hospital (IRB approval
number: B-1201-143-107) and was conducted in accordance with the
principles of the Declaration of Helsinki.
Results
Patient Characteristics
A total of 102 patients were included in this analysis, of which PV
had been diagnosed in 33 patients (32.4%), and ET in 69 patients
(67.6%). The median age was 64 years (range, 24–87 years), and 52.9%
of the patients were men (Table 1). Data regarding the JAK2 mutation
848 Y. Lim et al. / Thrombosis Research 135 (2015) 846–851

Table 1 Table 2
Patient Characteristics. Calculated burdens of leukocytes, hemoglobin, and platelets of the entire follow-up period,
according to event occurrence.
Patient Characteristics Total (N = 102) PV (N = 33) ET (N = 69)
Event (−) (N = 73) Event (+) (N = 29) P-value
Gender
Male 54 (52.9%) 17 (51.5%) 37 (53.6%) Total (N = 102)
Female 48 (47.1%) 15 (48.5%) 32 (46.4%) Leukocytes (×103/μL) 9,567 12,015 0.003
Age, years Hemoglobin (g/dL) 14.1 14.4 0.480
3
Median (range) 64 (24–87) 61 (18–82) 66 (24–87) Platelets (×10 /μL) 594 641 0.354
JAK2 mutation statusa, n (%) Polycythemia vera (N = 33)
No mutation 39 (38.2%) 10 (30.3%) 29 (42.0%) Leukocytes (×103/μL) 12,019 12,498 0.767
Mutation (+) 55 (53.9%) 22 (66.7%) 33 (47.7%) Hemoglobin, (g/dL) 16.3 15.8 0.382
Heterozygotic 50 (49.0%) 19 (57.6%) 31 (44.9%) Platelets (×103/μL) 390 472 0.195
Homozygotic 3 (2.9%) 2 (6.1%) 1 (1.4%) Essential thrombocythemia (N = 69)
Cardiovascular risk factors, n (%) Leukocytes (×103/μL) 8,577 11,674 0.007
Diabetes mellitus 23 (22.5%) 7 (21.2%) 16 (23.2%) Hemoglobin (g/dL) 13.2 13.5 0.562
3
Hypertension 61 (59.8%) 26 (78.8%) 35 (50.7%) Platelets (×10 /μL) 677 761 0.133
a
Hyperlipidemia 15 (14.7%) 5 (15.2%) 10 (14.5%) Thrombosis (N = 16)
Obesity (BMI ≥ 25 30 (29.4%) 14 (42.4%) 16 (23.2%) Leukocytes (×103/μL) 9,567 11,358 0.076
kg/cm2) Hemoglobin (g/dL) 14.1 15.1 0.045
History of cardiovascular disease, n (%) Platelets (×103/μL) 594 579 0.808
Any 17 (16.7%) 7 (21.2%) 10 (14.5%) Hemorrhage (N = 13) a
Angina 7 2 5 Leukocytes (×103/μL) 9,567 12,823 0.003
Myocardial infarct 2b 2b 0 Hemoglobin (g/dL) 14.1 13.6 0.512
3
Stroke 7b 3b 4 Platelets (×10 /μL) 594 718 0.079
Transient ischemic 1 0 1 a
In patients with multiple events, the event occurring first was considered to be the
attack relevant event for categorization.
Smoking, n (%)
Current or 22 (21.6%) 8 (24.4%) 14 (20.3%)
ex-smoker, gastrointestinal tract bleeding; 4 of musculoskeletal bleeding; and 1 of
Treatment, n (%) retinal bleeding. Multiple events were documented in 5 patients
Phlebotomy 35 (34.3%) 19 (57.6%) 16 (23.2%)
(2 events each in 4 patients, 3 events in 1 patient). Three of the patients
Antiplatelet drugs 93 (91.2%) 28 (84.8%) 59 (85.5%)
Hydroxyurea 82 (80.4%) 28 (84.8%) 54 (78.3%) had a history of thrombotic or hemorrhagic events before the diagnosis
Anagrelide 19 (18.6%) 2 (6.1%) 17 (24.6%) of PV or ET. The diagnoses of MPN in the 29 patients with events
No drugs 2 (2.0%) 1 (3.0%) 1 (1.4%) were PV in 12 (41.4%) and ET in 17 (58.6%) patients. Incidence rates of
Counts at diagnosis, median (range) thrombosis and hemorrhage were 4.9%/patient-year and 3.5%/patient-
Leukocyte 11,435 15,240 10,700
(×103/μL) (4,790–42,630) (6,990–35,230) (4,790–42,630)
year in PV patients, and 4.2%/patient-year and 3.7%/patient-year in ET
Hemoglobin (g/dL) 15 (8–24) 19 (16–24) 14 (8–18) patients, respectively. When clinical parameters were compared
Platelet (×103/μL) 752 (133–2,125) 415 (133–975) 1,004 between the groups with or without events, no statistically signifi-
(537–2,215) cant differences were found except for the number of patients with
Time-weighted average counts during the follow-up period, median (range)
hypertension. Significantly higher number of patients with hyper-
Leukocyte 9,341 12,332 8,752
(×103/μL) (4,840–21,670) (5,470–21,670) (4,840–19,940) tension was in the event group compared to the event-free group
Hemoglobin (g/dL) 14 (9–19) 16 (14–19) 13 (9–17) (52.1% vs. 20.6%, P = 0.011).
Platelet (×103/ μL) 601 (150–1,410) 399 (150–836) 680 (312–1,410)

Abbreviations: PV, polycythemia vera; ET, essential thrombocythemia; BMI, body mass
index.
a
JAK2 mutation status data were available in 94 patients (92.2%).
b
History of both myocardial infarct and cerebrovascular event in 1 PV patient.
status were available for 94 patients (92.2%), of which 55 patients were
positive for mutation, all carrying the V617F mutation. Overall, 35
patients (34.3%) had received at least 1 phlebotomy during follow-up.
Pharmacologic treatments recorded during the follow-up period were
as follows: 93 patients (91.2%) received antiplatelet drugs; 82 patients
(80.4%), hydroxyurea; and 19 patients (18.6%), anagrelide. Two patients
did not take any drugs.
The median follow-up period was 54 months (range, 2–106
months). The median values of the blood counts at diagnosis and the
medians of the time-weighted averages calculated for the entire
follow-up period are presented in Table 2.
Thrombotic and Hemorrhagic Events
A total of 35 thrombotic/hemorrhagic events were recorded in 29
patients who had been diagnosed with PV (12 patients) or ET (17
patients). The 35 events comprised of 19 thrombotic complications in
18 patients, and 16 hemorrhagic complications in 15 patients, which
were recorded either at (14 cases) or after the diagnosis (21 cases) of
PV or ET. The 19 thrombotic events comprised of 8 cases of transient is-
chemic attacks; 7, of stroke; and 4, of myocardial infarction. The 16
hemorrhagic events comprised of 6 cases of intracranial bleeding; 5 of
Blood Cell Burden and Thrombotic/hemorrhagic Events
The leukocyte burden for the entire follow-up period was compared
in subgroups based on the occurrence of thrombotic or hemorrhagic
events. The leukocyte burden was significantly higher in the
event group than in the event-free group (12,015 × 103/μL vs.
9,567 × 103/μL, P = 0.003). However, no significant differences were
found between the two groups when comparing Hb or platelet burden
(Table 2). When the odds ratios for thrombotic/hemorrhagic events
were fitted according to leukocyte burden with a restricted cubic spline
basis, the increased risk for the events correlated with increasing leuko-
cyte burden (Fig. 2). In patients with events, the blood cell burdens in
the pre-event period were significantly higher than in the follow-up pe-
riod for leukocytes (16,767 × 103 μL vs. 12,015 × 103 μL, P = 0.002) and
platelets (786 × 103 μL vs. 641 × 103 μL, P = 0.002) (Table 3). Same
analysis in all patients excluding those who presented with a throm-
botic or hemorrhagic event at the time of diagnosis of PV/ET still
showed significantly higher leukocyte burden in event-group com-
pared to event-free group (13,297 × 10 3 μL vs. 9,567 × 10 3 μL,
P b 0.001) and in the pre-event period compared to entire period
(17,029 × 103 μL vs. 13,297 × 103 μL, P = 0.017).
There were no statistically significant differences between the
groups with or without events in terms of the initial leukocyte count
(P = 0.078), Hb levels (P = 0.875), or platelet counts (P = 0.910),
although higher tendency of initial leukocyte count was observed in
the event group compared to the event-free group (15,235x103/μL vs.
12,020x103/μL).
 Y. Lim et al. / Thrombosis Research 135 (2015) 846–851 849

Analysis of Blood Cell Burden According to Event Types

In both the thrombosis and the hemorrhage groups, the leukocyte

burdens were higher in patients with events than that in those without
events, but statistically significant difference was evident only in the
hemorrhage group (Table 2). Leukocyte and platelet burdens were
higher in the 3-month period before hemorrhage, while Hb and platelet
burdens were higher in the 3-month period before thrombosis, each
compared to the entire period (Table 3).

Mutivariable Analysis of Classical Risk Factors of Thrombosis/hemorrhage

Univariable analysis of risk of events on the basis of the classical risk

Fig. 2. Relationship between leukocyte burden and the risk of vascular events. The odds
ratio of the occurrence of a vascular event was plotted against the leukocyte burden.
factors of thrombosis or hemorrhage showed that no classical risk
factors other than hypertension significantly contributed to the inci-
dence of thrombosis or hemorrhage in the study population. Apart
from hypertension, the leukocyte burden was the only other statistically
significant risk factor identified. At multivariable analysis, these two risk
factors remained meaningful contributors to thrombotic/hemorrhagic
complications. Age over 60 years and history of thrombosis, which are
recognized as classical risk factors, were not proven significant. Applica-
tion of logistic regression with different cut-off values of leukocyte
burden for the entire follow-up period showed that leukocyte burden
higher than 11,000 × 103/ μL can predict higher risk of a vascular
event, with a hazard ratio of 3.20 (Table 4).

Analysis of Blood Cell Burden According to Diagnosis Discussion

When the analysis was done separately according to the type of In all patients with PV or ET managed according to current treatment
diagnosis of MPN, the difference in leukocyte burden of the whole guidelines, we determined the time-weighted averages of blood cell
follow-up period between the event and event-free groups were more counts to estimate the burden of the blood cells over a generally long
distinct in the ET group. This difference was not maintained in the PV follow-up period, and analyzed their impacts on risk of thrombotic or
group. However, in the group with any thrombotic or hemorrhagic hemorrhagic events. Patients who experienced any thrombotic or
event, the leukocyte burden in the pre-event 3 months was significantly hemorrhagic events had higher leukocyte burden than those who did
higher compared to that of the entire period in both PV and ET groups. not. In patients with events, a higher leukocyte burden seemed
The burden of pre-event 3 month Hb was significantly higher than that to contribute significantly to the risk of such events, as shown by
of the entire period in the PV group, and the burden of pre-event comparison of the pre-event leukocyte burden to that of the entire
3 month platelets was higher than that of the entire period in the ET period. Risks of events were especially higher at leukocyte burden
group (Table 3). higher than 11,000 × 103/μL. The influence of elevated Hb or platelet
was particularly maintained only in the immediate pre-event period.
Thrombosis and hemorrhage are classical and important compli-
cations of PV and ET. As the risks of such complications are over 10
Table 3
Comparison of the blood cell burdens for the entire follow-up period with those for the times higher than those in the general population, it is important
pre-event 3-month period in PV and ET patients who had vascular events of thrombosis to predict and manage the PV/ET patients of the complications. The
or hemorrhage. reported incidence rate of thrombosis ranges from 1.1 to 4.9%/
Events (+) (N = 29) Entire period Pre-event period P-value patient-year in PV patients [8–11], and from 1.3 to 6.6%/patient-
Total (N = 102)
Leukocytes (×103/ μL) 12,015 16,767 0.002
Hemoglobin (g/dL) 14.4 15.2 0.092 Table 4
Platelets (×103/μL) 641 786 0.002 Multivariable analysis of thrombosis/hemorrhage risk.
PV (N = 33)
Leukocytes (×103/μL) 12,498 16,358 0.023 Parameter N (total = 102) P-value Hazard ratio
Hemoglobin (g/dL) 15.8 17.6 0.015 Univariable analysis
Platelets (×103/μL) 472 504 0.217 Comorbid diabetes mellitus 23 0.777 0.859
ET (N = 69) Comorbid hypertension 61 0.011 3.531
Leukocytes (×103/μL) 11,674 17,074 0.013 Comorbid hyperlipidemia 15 0.354 1.855
Hemoglobin (g/dL) 13.5 13.4 0.903 Current smoker or history of smoking 22 0.425 1.605
Platelets (×103/μL) 761 999 0.005 Male gender 54 0.828 0.882
Thrombosis (N = 16) a Obesity (body mass index ≥ 25 kg/m2) 24 0.956 0.972
Leukocytes (×103/μL) 11,358 13,119 0.054 History of cardiovascular disease 17 0.922 1.059
Hemoglobin (g/dL) 15.1 16.6 0.010 JAK2 V617F mutation (+) 55 0.516 1.368
Platelets (×103/μL) 579 670 0.038 Age ≥ 60 years 61 0.234 1.734
Hemorrhage (N = 13) a Leukocyte burden ≥ 11,000 × 103 μL 29 0.008 3.609
Leukocytes (×103/μL) 12,823 20,976 0.005
Hemoglobin (g/dL) 13.6 17.2 0.911 Multivariable analysis
Platelets (×103/μL) 718 1,164 0.011 Comorbid hypertension 61 0.031 3.118
Leukocyte burden ≥ 11,000 × 103 μL 29 0.017 3.195
Abbreviations: PV, polycythemia vera; ET, essential thrombocythemia.
a Age ≥ 60 years 61 0.649 1.275
In patients with multiple events, the event occurring first was considered to be the
History of cardiovascular disease 17 0.751 0.819
relevant event for categorization.
850 Y. Lim et al. / Thrombosis Research 135 (2015) 846–851
year in ET patients [8,11,12]. The incidence rate of hemorrhage
ranges from 0.9 to 6.2%/patient-year in PV patients [9,10,13–15],
and from 0.3 to 5.9%/patient-year in ET patients [12,14–16].
Classical risk factors in PV and ET have been age over 60 years and
history of thrombosis for thrombotic events, and platelet counts over
1,500 × 103/μL for hemorrhagic events [17]. Recently, leukocytosis has
been suggested as a predictive biomarker of thrombotic/hemorrhagic
events. In PV patients enrolled in the European Collaboration in
Low-dose Aspirin in Polycythemia Vera (ECLAP) study, increased risk
of myocardial infarction was seen in patients with leukocyte count
above 15,000 × 103/ μL [4]. Similar results of increased thrombotic/
hemorrhagic complications were seen in ET patients in several different
studies [1,2,5,18,19]. In a more recent study analyzing ET patients
enrolled in the prospective PT1 cohort, Campbell et al. showed a
U-shaped pattern of risk curve for vascular complications and leukocyte
counts during the follow-up period, by using different statistical
methods [20].
Concerning the association between leukocytosis and thrombotic/
hemorrhagic events, there have been reports of polymorphonuclear
leukocyte activation in PV and ET, showing in vivo leukocyte activation
that is associated with laboratory signs of endothelium and coagulation
system activation [2,21]. There have been other studies suggesting
increased circulating leukocyte-platelet aggregates in these patients,
which may provoke prothrombotic phenomena [22]. To explain the
association between leukocytosis and hemorrhage, increased secretion
of proinflammatory mediators of tissue destruction by neutrophils has
been suggested [23], but it has not yet been proven in MPN patients.
As of now, it is still not clear whether leukocytosis is a marker of
vascular complications or whether it actually contributes to the causing
the complications.
The results of our study are in agreement with those of recent
studies that indicated that leukocytosis is more closely related with
vascular complications than erythrocytosis or thrombocytosis in PV
and ET patients. Also, the incidence rate of 4.9%/patient-year in PV and
4.2%/patient-year in ET for thrombosis and 3.5%/of patients/year in PV
and 3.7% of patients/year in ET for hemorrhage for the patients in this
analysis is comparable to those in previously reported studies. However,
most of earlier studies seeking associations between leukocytosis and
thrombotic/hemorrhagic complications in MPNs focused on leukocyte
count only at the time of diagnosis of MPNs. Some studies have looked
into the effects of leukocytosis during disease course, but the studies
have analyzed the effect of spot leukocyte counts [2,20].
To the best of our knowledge, this study is the first to assess the
effect on vascular events of exposure to leukocytosis, both in terms of
its strength and duration rather than spot leukocytosis values, by
designing a calculation method to quantify the burden of exposure to
leukocytosis. The rationale for this was that if leukocytosis actually
contributed to vascular complications, then evaluating the strength
and length of time under exposure would be more physiologically
relevant than evaluation based on spot values. We assumed that
confirming the association between the leukocyte burden during the
disease course and the risk of vascular complications would provide a
hint to understanding the pathophysiology. Also, this apparently is the
first study to compare the blood cell burdens of the whole period to
the pre-event period, and to establish the higher burden in the pre-
event period in patients with thrombotic/hemorrhagic complications.
We would like to increase the awareness of vascular complications
and encourage clinicians to control persistent leukocytosis in PV or ET
patients. We already know from the results of randomized prospective
studies in ET patients that hydroxyurea plus low-dose aspirin was
superior to anagrelide in reducing rate of arterial thrombosis and
serious hemorrhage [24]. The findings from our study may be associated
with the reported lower complication rate with hydroxyurea as
hydroxyurea is superior to anagrelide in reducing leukocyte burden.
Of further interest in our study is that hemorrhagic events were more
closely related with leukocytosis than thrombotic events. Current
explanations of hemorrhagic events in PV or ET mostly depend on
acquired von Willebrand syndrome related to extreme thrombocytosis,
and possibly on proinflammatory mediators released by leukocytes
[19,25]. The mechanism underlying the close relationship between
the leukocyte burden and higher risk of hemorrhage needs to be
further elucidated.
Our study has limitation of being a single institution study,
restricting the analysis to a limited number of patients. Moreover,
since it was a retrospective study, unrecorded relevant events and/or
important blood cell count values could be missing. Third, the time
intervals between blood cell counts were of different durations among
the patients, which might have affected the blood-cell burden calcula-
tion. Nevertheless, all the patients were treated and followed-up inde-
pendently by 2 hematologists according to the same protocol for the
standard of care. In addition, our institution is the main university
hospital of Seongnam-si, where most of the study patients live; hence,
the likelihood of patients with events visiting other clinics is very low.
In conclusion, the results of this study confirm that leukocyte burden
is a risk factor for thrombotic/hemorrhagic complications in PV and ET
patients. The notable findings in this study are as follows: (i) the
risk of thrombotic/hemorrhagic complication is correlated with high
leukocyte burden, especially when the average is above the threshold
of 11,000 × 103 μL, and (ii) the leukocyte burden is higher in the period
immediately before a thrombotic/hemorrhagic event than in the entire
follow-up period. These two findings suggest that leukocyte burden is
more likely a direct causative factor in vascular complications among
PV and ET patients than simply a marker of poor prognosis. The results
suggest that leukocytosis should be monitored and treated in PV and ET
patients. However, further prospective clinical studies in a larger
number of patients as well as fundamental studies are warranted to
provide confirmation and further explanation of these findings.
Addendum
Y. Lim, J.O. Lee, and S.M. Bang were responsible for conception and
design of the study. J.O. Lee and S.M. Bang were responsible for grant
funding. All authors were responsible for data acquisition. Y. Lim and
J.O. Lee were responsible for data analysis, Y. Lim, J.O Lee and S.M.
Bang were responsible for data interpretation. Y. Lim was responsible
for drafting of the article. All authors were responsible for critical
revision and final approval of the manuscript.
Conflict of Interest Statement
There are no potential conflicts of interest to disclose for any author.
Acknowledgements
Assistance in statistical analysis was provided by the Medical
Research Collaborating Center (MRCC) of Seoul National University
Bundang Hospital.
The authors are indebted for J. Patrick Barron, Professor Emeritus of
Tokyo Medical University and Adjunct Professor of Seoul National
University Bundang Hospital for his thoughtful editing of this
manuscript.
References
[1] Carobbio A, Finazzi G, Antonioli E, Guglielmelli P, Vannucchi AM, Delaini F, et al.
Thrombocytosis and leukocytosis interaction in vascular complications of essential
thrombocythemia. Blood 2008;112:3135–7.
[2] Carobbio A, Finazzi G, Guerini V, Spinelli O, Delaini F, Marchioli R, et al. Leukocytosis
is a risk factor for thrombosis in essential thrombocythemia: interaction with treat-
ment, standard risk factors, and Jak2 mutation status. Blood 2007;109:2310–3.
[3] Falanga A, Marchetti M, Vignoli A, Balducci D, Barbui T. Leukocyte-platelet interac-
tion in patients with essential thrombocythemia and polycythemia vera. Exp
Hematol 2005;33:523–30.
 Y. Lim et al. / Thrombosis Research 135 (2015) 846–851
[4] Landolfi R, Di Gennaro L, Barbui T, De Stefano V, Finazzi G, Marfisi R, et al. Leuko-
cytosis as a major thrombotic risk factor in patients with polycythemia vera.
Blood 2007;109:2446–52.
[5] Carobbio A, Antonioli E, Guglielmelli P, Vannucchi AM, Delaini F, Guerini V, et al.
Leukocytosis and risk stratification assessment in essential thrombocythemia. J
Clin Oncol 2008;26:2732–6.
[6] Tefferi A, Vardiman JW. Classification and diagnosis of myeloproliferative
neoplasms: the 2008 World Health Organization criteria and point-of-care diagnos-
tic algorithms. Leukemia 2008;22:14–22.
[7] WHO Expert Consultation. Appropriate body-mass index for Asian populations and
its implications for policy and intervention strategies. Lancet 2004;363:157–63.
[8] Passamonti F, Rumi E, Pungolino E, Malabarba L, Bertazzoni P, Valentini M, et al. Life
expectancy and prognostic factors for survival in patients with polycythemia vera
and essential thrombocythemia. Am J Med 2004;117:755–61.
[9] Marchioli R, Finazzi G, Landolfi R, Kutti J, Gisslinger H, Patrono C, et al. Vascular and
neoplastic risk in a large cohort of patients with polycythemia vera. J Clin Oncol
2005;23:2224–32.
[10] Landolfi R, Marchioli R, Kutti J, Gisslinger H, Tognoni G, Patrono C, et al. Efficacy and
safety of low-dose aspirin in polycythemia vera. N Engl J Med 2004;350:114–24.
[11] Patrono C, Rocca B, De Stefano V. Platelet activation and inhibition in polycythemia
vera and essential thrombocythemia. Blood 2013;121:1701–11.
[12] Cortelazzo S, Viero P, Finazzi G, D'Emilio A, Rodeghiero F, Barbui T. Incidence and risk
factors for thrombotic complications in a historical cohort of 100 patients with
essential thrombocythemia. J Clin Oncol 1990;8:556–62.
[13] Passamonti F, Brusamolino E, Lazzarino M, Barate C, Klersy C, Orlandi E, et al. Efficacy
of pipobroman in the treatment of polycythemia vera: long-term results in 163
patients. Haematologica 2000;85:1011–8.
[14] Chou YS, Gau JP, Yu YB, Pai JT, Hsiao LT, Liu JH, et al. Leukocytosis in polycythemia
vera and splenomegaly in essential thrombocythemia are independent risk factors
for hemorrhage. Eur J Haematol 2013;90:228–36.
[15] De Stefano V, Za T, Rossi E, Vannucchi AM, Ruggeri M, Elli E, et al. Recurrent
thrombosis in patients with polycythemia vera and essential thrombocythemia:
incidence, risk factors, and effect of treatments. Haematologica 2008;93:372–80.
851
[16] Palandri F, Polverelli N, Catani L, Sollazzo D, Ottaviani E, Parisi S, et al. Bleeding in
essential thrombocythaemia: a retrospective analysis on 565 patients. Br J Haematol
2012;156:281–4.
[17] Cervantes F, Passamonti F, Barosi G. Life expectancy and prognostic factors in the
classic BCR/ABL-negative myeloproliferative disorders. Leukemia 2008;22:
905–14.
[18] Tefferi A, Gangat N, Wolanskyj A. The interaction between leukocytosis and
other risk factors for thrombosis in essential thrombocythemia. Blood 2007;
109:4105.
[19] Finazzi G, Carobbio A, Thiele J, Passamonti F, Rumi E, Ruggeri M, et al. Incidence and
risk factors for bleeding in 1104 patients with essential thrombocythemia or
prefibrotic myelofibrosis diagnosed according to the 2008 WHO criteria. Leukemia
2012;26:716–9.
[20] Campbell PJ, MacLean C, Beer PA, Buck G, Wheatley K, Kiladjian JJ, et al. Correlation
of blood counts with vascular complications in essential thrombocythemia: analysis
of the prospective PT1 cohort. Blood 2012;120:1409–11.
[21] Falanga A, Marchetti M, Evangelista V, Vignoli A, Licini M, Balicco M, et al. Polymor-
phonuclear leukocyte activation and hemostasis in patients with essential
thrombocythemia and polycythemia vera. Blood 2000;96:4261–6.
[22] Arellano-Rodrigo E, Alvarez-Larrán A, Reverter JC, Villamor N, Colomer D, Cervantes
F. Increased platelet and leukocyte activation as contributing mechanisms for
thrombosis in essential thrombocythemia and correlation with the JAK2 mutational
status. Haematologica 2006;91:169–75.
[23] Moxon-Emre I, Schlichter LC. Neutrophil depletion reduces blood–brain barrier
breakdown, axon injury, and inflammation after intracerebral hemorrhage. J
Neuropathol Exp Neurol 2011;70:218–35.
[24] Harrison CN, Campbell PJ, Buck G, Wheatley K, East CL, Bareford D, et al.
Hydroxyurea compared with anagrelide in high-risk essential thrombocythemia. N
Engl J Med 2005;353:33–45.
[25] Elliott MA, Tefferi A. Thrombosis and haemorrhage in polycythaemia vera and
essential thrombocythaemia. Br J Haematol 2005;128:275–90.