RESEARCH ARTICLE

A JH

Thrombosis and survival in essential thrombocythemia:

A regional study of 1,144 patients
Marco Montanaro,1* Roberto Latagliata,2 Michele Cedrone,3 Antonio Spadea,4 Angela Rago,5 Jonny Di Giandomenico,6
Francesca Spirito,7 Raffaele Porrini,8 Marianna De Muro,9 Sabrina Crescenzi Leonetti,10 Nicoletta Villiva,11
Cinzia De Gregoris,1 Massimo Breccia,2 Enrico Montefusco,8 Cristina Santoro,2 Giuseppe Cimino,5 Ignazio Majolino,7
Maria Gabriella Mazzucconi,2 Giuliana Alimena,2 and Alesssandro Andriani11
To identify prognostic factors affecting thrombosis-free survival (TFS) and overall survival (OS), we report the
experience of a Regional cooperative group in a real-life cohort of 1,144 patients with essential
thrombocythemia (ET) diagnosed from January 1979 to December 2010. There were 107 thrombotic events
(9.4%) during follow-up [60 (5.3%) arterial and 47 (4.1%) venous thromboses]. At univariate analysis, risk factors
for a shorter TFS were: age >60 years (P < 0.0054, 95% CI 1.182.6), previous thrombosis (P < 0.0001, 95% CI
1.584.52) and the presence of at least one cardiovascular risk factor (P 5 0.036, 95% CI 1.153.13). Patients with
a previous thrombosis occurred 24 months before ET diagnosis had a shorter TFS compared to patients with
a previous thrombosis occurred <24 months (P 5 0.0029, 95% CI 1.56.1); furthermore, patients with previous
thrombosis occurred <24 months did not show a shorter TFS compared with patients without previous
thrombosis (P 5 0.303, 95% CI 0.643.21). At multivariate analysis for TFS, only the occurrence of a previous
thrombosis maintained its prognostic impact (P 5 0.0004, 95% CI 1.483.79, RR 2.36). The 10-year OS was 89.9%
(95% CI 87.392.5): at multivariate analysis for OS, age >60 years (P < 0.0001), anemia (P < 0.0001), male gender
(P 5 0.0019), previous thromboses (P 5 0.0344), and white blood cell >15 3 109/l (P 5 0.0370) were independent
risk factors. Previous thrombotic events in ET patients are crucial for TFS but their importance seems related
not to the occurrence per se but mainly to the interval between the event and the diagnosis.
C 2014 Wiley Periodicals, Inc.
Am. J. Hematol. 89:542546, 2014. V

Introduction
Essential thrombocythemia (ET) is a myeloproliferative neoplasm characterized by abnormal bone marrow megakariocyte proliferation and an
increased platelet (PLT) count [1]. Life expectancy in ET patients, in the first 20 years from diagnosis, is generally considered similar to that of
normal population of the same age [2]; however, afterword the median survival becomes shorter.
The main causes of death are thrombotic events and evolution into myelofibrotic phase or secondary myelodisplasia/acute leukemia [3]; furthermore, the most frequent complications occurring in ET patients are either arterial or venous thromboses [4]. The treatment of ET patients with
cytoreductive drugs is effective in reducing the incidence of these thrombotic events [5] but could be associated with an higher rate of leukemic
evolution. Thus, a crucial point in ET treatment is to evaluate which patients could really benefit from a cytoreductive therapy [6].
To date, many clinical studies focused on factors predicting thrombotic complications and survival [713]; whereas a general agreement exists
on the prognostic role of age superior to 60 years and previous thrombotic events, some other clinical (white blood cell [WBC] and PLT counts,
cardiovascular [CV] risk factors) and biological factors (JAK2 V617F mutation, allele burden) have been evaluated with conflicting results.
In our retrospective study, we report the experience of a Regional cooperative group in a large real-life cohort of patients with ET, to identify
reliable prognostic factors affecting thrombosis-free survival (TFS) and overall survival (OS).

Methods
Study population. In the present retrospective analysis, data from 1,144 patients affected by ET diagnosed from January 1979 to December 2010 in 11 hematological Centers
(five universitary Institutes and six community-based Hospitals) were collected in the database of our cooperative group.
The diagnosis was made according to Polycythemia Vera Study Group (PVSG) criteria or WHO 2001 and 2008 criteria, depending on the date of diagnosis. Permission was
obtained by all patients alive at the time of data collection.
The following parameters were taken into consideration to find prognostic risk factors for thrombosis in univariate and multivariate models: age, sex, WBC count, hemoglobin level, PLT count, spleen size, JAK2 V617F mutation and allele burden, and history of previous thrombosis (before and at diagnosis). In addition, we considered as CV risk
factors at diagnosis the presence of arterial hypertension, diabetes, smoking attitude, and hypercholesterolemia.
1

Department of Hematology, Belcolle Hospital, Viterbo, Italy; 2Department of Cellular Biotechnologies and Hematology, University La Sapienza,, Rome, Italy; 3Department of Hematology, San Giovanni Hospital,, Rome, Italy; 4Unit of Hematology, Regina Elena National Cancer Institute, Rome, Italy; 5Department of Hematology, Polo
Universitario Pontino, Latina, Italy; 6Department of Hematology, University Tor Vergata,, Rome; 7Department of Hematology, San Camillo Hospital, Rome, Italy;
8
Department of Hematology, SantAndrea Hospital, Rome, Italy; 9Department of Hematology, University Campus Biomedico,, Rome, Italy; 10Department of Hematology, Sandro Pertini Hospital, Rome, Italy; 11Department of Hematology, Nuovo Regina Margherita Hospital, Rome, Italy

Conflict of interest: Nothing to report.

*Correspondence to: Marco Montanaro, Department of Hematology, Belcolle Hospital, Strada Sammartinese Viterbo, Italy. E-mail: mmontanaro51@gmail.com
Received for publication: 14 January 2014; Revised: 26 January 2014; Accepted: 27 January 2014
Am. J. Hematol. 89:542546, 2014.
Published online: 31 January 2014 in Wiley Online Library (wileyonlinelibrary.com).
DOI: 10.1002/ajh.23685
C 2014 Wiley Periodicals, Inc.
V

542

American Journal of Hematology, Vol. 89, No. 5, May 2014

doi:10.1002/ajh.23685

RESEARCH ARTICLE

Role of previous thrombotic events in Essential Thrombocythemia

Definition of thrombotic events. In the present analysis, the following events

were considered as thrombotic complications:
 Arterial thrombosis: acute myocardial infarction, angina pectoris, ischemic
stroke, transient ischemic attack, and peripheral arterial occlusions.
 Venous thrombosis: pulmonary embolism, peripheral or splanchnic venous
thrombosis, and cerebral venous sinus thrombosis.
Statistical analysis. Data were expressed as mean 6 standard deviation (SD)
(normally distributed data), median, and interquartile range (IR; nonnormally distributed data), or as percentage frequencies, and within-patient comparisons were
made by paired t test and v2 test, as appropriate, at significance levels of P < 0.05.
TFS was calculated from the date of diagnosis to the occurrence of any event
above defined as thrombotic complication. OS was calculated from the date of diagnosis to death due to any cause.
The KaplanMeier product-limit method was used to estimate univariate survival curves, and the log-rank test was adopted to compare the survival curves. Cox

TABLE I. Main Epidemiological and Clinical Features of Patients at

Diagnosis
Pts number

1144

Median age (years) (IR)

F/M number (%)
Median WBC (3109/L) (IR)
Median Hb (g/dL) (IR)
Median PLT (3109/L) (IR)
JAK2 V617F mutated/all performed (%)
Median JAK2 V617F allele burden (%) (IR)
Splenomegaly (%)

62.4 (46.872.2)
727/417 (63.5/36.5)
8.9 (7.211.0)
14.0 (13.015.0)
815 (6701014)
423/716 (59.1)
19.9 (8.238.8)
213/1097 (19.4)

TABLE II. Different Types of Arterial and Venous Thrombotic Events

Arterial thromboses
Cardiac
Cerebral
Peripheral
Intraabdominal
Others
Venous thromboses
Peripheral (leg/arm)
Abdominal
Cerebral
Others

Previous
thrombotic
events

Thrombotic
events
during
follow-up

162
60
68
25
3
6
47
34
6
7
/

60
20
34
3
1
2
47
37
7
3
/

proportional hazards regression was used to carry out multivariate survival analyses.
All analyses were performed using Statistica 7.1 (Statsoft, Tulsa, OK), Stata 9 (StataCorp LP, College Station, TX) software, and Microsoft Excel.

Results
Baseline characteristics and ET treatment
The main epidemiological and clinical features of all patients are
reported in Table I. Female gender was more frequent than male (2/
1) and the median age was 62.4 years (IR 46.871.2). Splenomegaly,
referred as a spleen enlargement below costal margin, was present at
diagnosis in about 20% of the 1,097 patients in whom this data was
available.
The assessment of JAK2 V617F mutational status was performed
in 716 patients and resulted positive in 423 (59.1% of the evaluated
patients). In addition, the allele burden was measured in 263 patients,
with a median value of 19.9% (IR 8.238.8).
Previous thromboses, considering events occurred before or at the
time of ET diagnosis, were reported in 209 patients (18.3%): these
events were arterial in 162 patients (77.5% of patients with previous
thrombotic events and 14.2% of all patients) and venous in 47 (22.5%
of patients with previous thrombotic events and 4.1% of all patients;
Table II). According to the interval from the thrombotic episode to
the diagnosis of ET, previous thromboses in 186 cases with a known
date of occurrence were also divided into thromboses occurred <24
months before diagnosis of ET (95 cases, 51.1% of all previous
thrombotic events with a known date of occurrence) and thromboses
occurred 24 months before diagnosis of ET (91 cases, 48.9% of all
previous thrombotic events with a known date of occurrence). The
clinical features at diagnosis of these two groups of patients are
reported and compared in Table III: the only difference was a significantly higher median age in patients with previous thrombosis
occurred 24 months before diagnosis of ET.
According to conventional thrombotic risk assessment [6], 695
patients (61.0%) were high risk patients because of an age >60 years
and/or previous thrombosis, whereas 449 patients (39.0%) were low
risk because of an age <60 years and no previous thrombosis.
After ET diagnosis, 265 patients (23.1%) did not receive any cytoreductive therapy because of low thrombotic risk or physician judgment and were managed with antiplatelet agents only. Among the
879 patients (76.9%) who received cytoreductive therapy because of
high thrombotic risk or physician judgment, 629 were treated with
hydroxyurea (HU), 31 with alpha-interferon (IFN), 18 with anagrelide
(ANA), and 19 with Pipobroman (PY); the remaining 182 patients
were sequentially treated with more than one of the previous drugs

TABLE III. Clinical Features at Diagnosis of Patients with Previous Thrombotic Episodes According to Interval from Thrombosis to ET Diagnosis

Gender (M/F)
Median age (yrs) (Interquartile range) (IR)
Hb median (g/dl) (IR)
PLT median (3109/l) (IR)
WBC median (3109/l) (IR)
Median interval diagnosisCHT (mos) (IR)
*
Cardiovascular (CV) risk factors (no./%):
0
1
2
Type of previous thrombosis (n /%):
Arterial
Venous

< 24 months before

ET diagnosis

 24 months before
ET diagnosis

40/55
64.1 (52.771.8)
13.9 (12.514.7)
800 (6691066)
9.2 (7.811.3)
0.9 (07.0)

42/49
70.9 (61.078.0)
14.2 (13.015.2)
778 (652926)
8.6 (7.110.8)
1.7 (0.45.6)

20 (21.0)
42 (44.2)
33 (34.8)

17 (18.7)
44 (48.4)
30 (32.9)

78 (82.1)
17 (17.9)

67 (73.6)
24 (26.4)

P
0.636
0.001
0.136
0.453
0.121
0.194
0.454

0.873

Bold font indicates the significance of p values.

doi:10.1002/ajh.23685

American Journal of Hematology, Vol. 89, No. 5, May 2014

543

RESEARCH ARTICLE

Montanaro et al.
TABLE IV. Prognostic Factors for Global TFS at Univariate Analysis

Global previous thromboses

Previous thromboses < 24 months
Previous thromboses  24 months
Age > 60 years
WBC > 8.9 3 109/l
WBC > 15.0 3 109/l
PLTS > 861 3 109/l
PLTS > 1.000 3 109/l
PLTS > 1.500 3 109/l
Gender
V617F JAK-2 mutation
Cardio-vascular factors (1 vs. 0)
Cardio-vascular factors (2 vs. 0)
Hb < normal median value
(F < 12.5/M < 13.5 g/dl)

(HU and IFN in 40 patients, HU and ANA in 40 patients, HU and

PY in 50 patients, other sequential therapies in 52 patients).

95%CI

HR

<0.0001
0.303
<0.0001
0.0054
0.32
0.34
0.0012
0.23
0.13
0.14
0.71
0.036
0.0017
0.32

1.584.52
0.643.22
2.009.93
1.182.60
0.541.22
0.613.29
0.350.78
0.481.17
0.724.01
0.892.02
0.491.63
1.153.13
1.328.08
0.752.18

2.67
1.44
4.46
1.74
0.81
1.42
0.52
0.75
1.69
1.34
0.89
1.90
3.26
1.28

Thrombotic events during follow-up

On the whole, there were 107 thrombotic events (9.4%) during
follow-up: the arterial thrombotic events were 60 (5.3%), the venous
thrombotic events were 47 (4.1%). The rate of thrombosis was 1.4%/
year. The frequency of different arterial and venous thrombotic events
during follow-up is shown in Table II.

Prognostic factors for TFS

On univariate analysis (Table IV), the risk factors at diagnosis that
resulted statistically significant for a shorter global TFS were: age >60
year (P < 0.0054, 95% CI 1.182.6), previous thrombosis (P < 0.0001,
95% CI 1.584.52) and the presence of at least 1 CV risk factor
(P 5 0.036, 95% CI 1.153.13). It is worth of note that patients with a
previous thrombotic event occurred 24 months before ET diagnosis
had a significantly shorter TFS compared to patients with a previous
thrombotic event occurred in the 24 months before ET diagnosis
(P 5 0.0029, 95% CI 1.56.1); furthermore, patients with previous
thrombotic events occurred <24 months did not show a significant
shorter TFS when compared with patients without previous thrombotic events (P 5 0.303, 95% CI 0.643.21; Figure 1). As shown in
Table IV, the prognostic significance of CV risk factors was higher
when the presence of at least two of them was considered
(P 5 0.0017, 95% CI 1.328.08). The PLT count above the median
value of 816 3 109/L or > 1000 3 109/L was associated with a significantly better TFS (P 5 0.0012, 95% CI 0.350.78).
On the contrary, WBC count (either above the median value of 8.9
3 109/L, or >15 3 109/L), Hb (<14 g/dL), JAK2 V617F mutation,
JAK2 V617F allele burden >50% did not reach the cut-off value of
significance (Table IV).
The same features were also evaluated in univariate analysis for
arterial TFS and venous TFS: the corresponding results are shown in
Table V. The presence of a previous thrombotic event on the whole
retained the negative role for both types of thrombosis during followup: however, previous arterial or venous thromboses were predictive
only for the same type of thrombotic event during follow-up A PLT
count above the median value retained the protective role for both
types of thrombosis during follow-up: in addition, age >60 years and
presence of at least 1 CV risk factor had a negative impact for arterial

Figure 1. Thrombosis-free survival according to previous thrombotic

events. [Color figure can be viewed in the online issue, which is available
at wileyonlinelibrary.com.]

TABLE V. Prognostic Factors for Arterial and Venous TFS at Univariate Analysis

Global previous thromboses

Previous thromboses < 24 months
Previous thromboses  24 months
Previous arterial thromboses
Previous venous thromboses
Age > 60 years
WBC > 8.9 3 109/l
WBC > 15.0 3 109/l
PLTS > 861 3 109/l
PLTS > 1.000 3 109/l
PLTS > 1.500 3 109/l
Gender
V617F JAK-2 mutation
Cardio-vascular factors (1 vs. 0)
Cardio-vascular factors (2 vs. 0)
Hb < normal median value (F < 12.5/M < 13.5 g/dl)

Arterial thromboses during follow-up

Venous thromboses during follow-up

95%CI

HR

95%CI

HR

0.0003
0.524
<0.0001
0.0005
0.13
0.0028
0.549
0.376
0.001
0.0087
0.47
0.20
0.69
0.014
0.0072
0.88

1.315.17
0.473.86
1.5112.5
1.235.88
0.519.03
1.333.71
0.501.43
0.514.51
0.300.86
0.321.01
0.191.83
0.812.38
0.391.85
1.585.62
1.1415.69
0.532.12

2.6
1.35
4.35
2.69
2.15
2.23
0.85
1.50
0.51
0.57
0.59
1.39
0.86
2.98
4.24
1.05

0.0003
0.398
<0.0001
0.195
<0.0001
0.442
0.399
0.672
0.047
0.845
0.001
0.45
0.93
0.78
0.08
0.19

1.246.27
0.455.50
1.3515.7
0.634.65
1.2426.7
0.692.31
0.411.42
0.344.79
0.281.04
0.542.13
0.9313.66
0.672.38
0.362.55
0.502.54
0.739.05
0.713.66

2.78
1.57
4.61
1.72
5.75
1.26
0.76
1.29
0.53
1.07
3.56
1.26
0.95
1.13
2.58
1.16

Bold font indicates the significance of p values.

544

American Journal of Hematology, Vol. 89, No. 5, May 2014

doi:10.1002/ajh.23685

RESEARCH ARTICLE

Role of previous thrombotic events in Essential Thrombocythemia

TFS only, while a PLT count >1.500 3 109/l had a negative impact
for venous TFS only (Table V).
At the multivariate analysis for global TFS, only the occurrence of
a previous thrombosis maintained its prognostic impact (P 5 0.0004,
95% CI 1.483.79, RR 2.36), while age >60 years had only a trend of
significance (P 5 0.058, 95% CI 0.992.56, RR 1.59). For arterial TFS,
the occurrence of a previous thrombosis 24 months before the diagnosis of ET (P 5 0.0046, 95% CI 1.385.76, RR 2.82) and the presence
of at least 1 CV risk factor (P 5 0.037, 95% CI 1.078.32, RR 2.98)
were independent prognostic factors. For venous TFS, only the occurrence of previous thrombosis was an independent prognostic factor
(P 5 0.002, 95% CI 1.475.45, RR 2.83).

Prognostic factors for OS

After a median follow-up of 5.4 years (IR 2.310.8), 93 patients
(8.1%) have died, 173 (15.1%) were lost at follow up, and 878
(76.7%) are still living at the time of the last evaluation.
The global OS at 5, 10, and 15 years was 95.5% (95% CI 94.1
96.9), 89.9% (95% CI 87.392.5), and 82.2% (95% CI 78.186.3),
respectively.
On univariate analysis, the risk factors at diagnosis that resulted
statistically significant for a shorter global OS were: age >60 years
(P < 0.0001, 95% CI 3.058.29), Hb levels below normal values
(P < 0.0001, 95% CI 1.805.56), male gender (P 5 0.0008, 95% CI
1.293.03), previous thromboses (P 5 0.0012, 95% CI 1.193.56) and
WBC  15 3 109/l (P 5 0.035, 95% CI 0.803.30). On the contrary,
PLT count, JAK2 V617F mutation, JAK2 V617F allele burden >50%,
spleen enlargement and the presence of either 1 or 2 CV risk factors did not reach the cut-off value of significance. As shown in Table
VI, all the factors with prognostic value at univariate analysis retained
their independent role at the multivariate analysis.

Discussion
Thrombotic episodes constitute the most frequent and clinically
significant events affecting morbidity and mortality in patients with
ET. Thus, they have been widely examined, aiming to prevent them
with the identification of prognostic factors at disease onset: up to
now, there is a worldwide agreement on the predictive role of older
age and previous thromboses, and these two factors are generally
TABLE VI. Prognostic Factors for OS at Multivariate Analysis

Age  60 years
Hb < normal median value
(F < 12.5/M < 13.5 g/dl)
Male gender
Previous thromboses
WBC  15 3 109/l

95%CI

HR

<0.0001
<0.0001

3.6012.16
2.024.97

6.61
3.17

0.0019
0.0344
0.0370

1.303.16
1.042.82
1.054.25

2.03
1.71
2.11

used to define the thrombotic risk of each patient at diagnosis [713].

On the other hand, as shown in Table VII, there were conflicting
results when other factors potentially affecting the thrombotic risk
were examined.
In the present study, we reported retrospective data on a large
cohort of ET patients collected by different hematological Centers
from a single and well-defined geographic area: CV risk factors at
baseline were considered together with many other disease-related
factors and all these features were evaluated for global but also for
distinct arterial and venous thrombotic risk in the follow-up. In our
cohort of patients, before 2008 the diagnosis was made according to
PVSG criteria and it is likely that some cases of prefibrotic IMF were
included. However, it has been recently shown that there was no difference between true ET and prefibrotic Idiopatic Myelofibrosis
(IMF) as to the incidence of thrombotic complications during followup [14].
As expected, the occurrence of a prior thrombotic event had a
high statistical significance in predicting all types of thrombosis during follow-up, in agreement with all previous reports (Table VII).
However, the most important finding in our cohort of patients was
the relevance of the interval from previous thrombosis to ET diagnosis in determining the risk of new thrombotic events in the follow-up:
only the occurrence of less recent previous thromboses (24 months
before ET diagnosis) had a prognostic role for TFS, whereas more
recent previous thrombotic episodes (<24 months before ET diagnosis) did not add any risk compared to patients without previous
thrombosis. The relevance of the interval time resulted considering all
thrombotic events in the follow-up, but it was also maintained when
arterial and venous thromboses in the follow-up were considered separately (Table V).
This finding deserves some comments: in our opinion, it could be
due to the different weight of ET-independent risk factors in the genesis of previous thrombotic events that occurred 24 months before
diagnosis of ET. In other words, in patients with less recent previous
thrombotic events, CV or other patient-related and ET-independent
factors could have played a more important role than ET-related factors and could have continued to act subsequently, adding their
effects to ET-related factors after ET diagnosis; on the contrary, it is
conceivable that in patients with more recent previous thrombotic
events, ET was already present when thrombosis occurred and thus a
more relevant role could have been played by ET-related factors,
which are generally controlled by treatment during follow-up. It is
noteworthy, however, that the distribution of CV risk factors was
similar in both groups and this finding is not in accordance with our
hypothesis: in addition, the role of higher median age in patients with
thrombosis occurred 24 months before diagnosis of ET should be
considered, even if the absence of a significant role of age in the multivariate analysis makes its importance questionable. As a matter of
fact, a confirmation of this finding is warranted in other independent
cohorts of patients.

TABLE VII. Comparison of Prognostic Factors for TFS at Multivariate Analysis in Different Studies

Author (years)

Tefferi
(2007)

Passamonti
(2008)

Carobbio
(2008)

Girodon
(2010)

Palandri
(2011)

Carobbio
(2011)

Barbui
(2012)

Present
study

No. patients
Median follow-up (yrs)
Age > 60 yrs
Previous thromboses
Cardiovascular risk factors
WBC >11.109/L
JAK2 V617F mutation
JAK2 V617F allele burden
Gender

605
13.6
YES
YES (A)
NR
YES
NR
NR
NR

605
5.6
YES
YES
NR
NR
NR
NR
NR

657
4.5
YES
YES
NR
YES
NO
NO
NO

311
9.5
YES
YES
NR
NO
NO
NR
YES

532
7.6
YES
YES
NR
YES
NO
NO
NO

891
6.2
YES
YES (A)
YES (A)
YES (A)
YES (A)
NR
YES

1220
NR
YES
YES
YES
NO
YES
NR
NO

1144
5.4
NO
YES
YES (A)
NO
NO
NO
NO

doi:10.1002/ajh.23685

American Journal of Hematology, Vol. 89, No. 5, May 2014

545

RESEARCH ARTICLE

Montanaro et al.

At univariate analysis, the age >60 years had a significant prognostic

role for global TFS and arterial TFS (but not for venous TFS). However,
in contrast with all other reported studies, in our experience older age
lost its prognostic value at multivariate analysis, with only a trend of
significance (P 5 0.058) for global TFS. A possible explanation for this
unexpected result could be the presence as covariate of CV risk factors,
which were significant in our multivariate analysis, while were not considered in many of other previous studies or were considered with a different definition (for example, Carobbio et al [12] and Barbui et al [13]
did not mention hypercholesterolemia among them); these CV risk factors are strictly linked to the increased age and thus could be responsible for this otherwise unexplained finding.
The presence of JAK-2 V617F mutation was evaluated for its prognostic significance on thrombotic risk in several studies and was considered in the recent International Prognostic Score - Essential
Thrombocythemia (IPSET) risk score [13] as the most powerful predicting factor. However, data from previous studies were largely controversial and also in our experience the presence of this mutation as
well as the allele burden did not show any prognostic role.
The role of high WBC count as potential risk factor for TFS was
addressed in several studies, but with conflicting results. Our data did
not show any prognostic impact of WBC count at baseline for either

References

1. Tefferi A. Polycythemia vera and essential

thrombocythemia: 2012 update on diagnosis,
risk stratification, and management. Am J Hematol 2012;87:285293.
2. Wolanskyj A, Schwager S, McClure R, et al.
Essential Thrombocythemia beyond the first
decade: life expectancy, long-term complication
rates and prognostic factors. Mayo Clin Proc
2008;81:159166.
3. Elliott MA, Tefferi A. Thrombosis and haemorrhage in polycythaemia vera and essential
thrombocythaemia. Br J Haematol 2005;128:
275290.
4. Bleeker J, Hogan W. Thrombocytosis: Diagnostic evaluation, thrombotic risk stratification and
risk-based management strategies. Thrombosis
2011:536062. doi: 10.1155/2011/536062.
5. Cortelazzo S, Finazzi G, Ruggeri M, et al.
Hydroxyurea for patients with essential thrombocythemia and a high risk of thrombosis. N
Engl J Med 1995;332:11321136.
6. Finazzi G. How to manage essential thrombocythemia. Leukemia 2012;26:875882.

546

American Journal of Hematology, Vol. 89, No. 5, May 2014

global TFS and arterial or venous TFS: this finding is also in agreement with data from PT-1 trial, recently presented by Campbell et al
[15].
Five prognostic factors for OS at univariate and multivariate analysis were identified in our cohort of patients. Comparing our results
with the IPSET model recently reported by Passamonti et al [16], the
role of older age, previous thrombotic episodes and high WBC count
was confirmed in both studies: in the IPSET model, however, lower
Hb levels were not significant as in our study. This latter finding is
probably due to the presence among our patients of some subjects
with prefibrotic IMF, who are expected to have lower Hb levels and a
shorter survival when compared with true ET.
In conclusion, the worldwide recognized prognostic role of previous thrombotic events and older age in ET patients probably deserves
some deeper insight. The importance of previous thromboses seems
to be related not to the occurrence per se of the event but mainly to
the interval between the event and the diagnosis of ET, while the role
of the older age could be counteracted when considered together with
CV risk factors at multivariate analysis. Furthermore, both JAK-2
mutation and WBC count in our experience do not seem to have a
significant impact on the thrombotic risk. A prospective observational
trial is warranted to give a conclusive answer to these hypotheses.

7. Tefferi A, Gangat N, Wolanskyj A. The interaction between leukocytosis and other risk factors
for thrombosis in essential thrombocythemia.
Blood 2007;109:4105.
8. Passamonti F, Rumi E, Arcaini L, et al. Prognostic
factors for thrombosis, myelofibrosis, and leukemia in essential thrombocythemia: A study of 605
patients. Haematologica 2008;93:16451651.
9. Carobbio A, Antonioli E, Guglielmelli P, et al.
Leukocytosis and risk stratification assessment
in essential thrombocythemia. J Clin Oncol
2008;26:27322736.
10. Girodon F, Dutrillaux F, Broseus J, et al. Leukocytosis is associated with poor survival but not
with increased risk of thrombosis in essential
thrombocythemia: A population-based study of
311 patients. Leukemia 2010;24:900903.
11. Palandri F, Polverelli N, Catani L, et al. Impact of
leukocytosis on thrombotic risk and survival in 532
patients with essential thrombocythemia: A retrospective study. Ann Hematol 2011;90:933938.
12. Carobbio A, Thiele J, Passamonti F, et al. Risk
factors for arterial and venous thrombosis in
WHO-defined essential thrombocythemia: An

13.

14.

15.

16.

international study of 891 patients. Blood 2011;

117:58575859.
Barbui T, Finazzi G, Carobbio A, et al. Development and validation of an International Prognostic Score of thrombosis in World Health
Organizationessential
thrombocythemia
(IPSET-thrombosis). Blood 2012;120:51285133.
Barbui T, Thiele J, Passamonti F, et al. Survival
and disease progression in essential thrombocythemia are significantly influenced by accurate
morphologic diagnosis: An international study. J
Clin Oncol 2011;29:31793184.
Campbell PJ, MacLean C, Beer PA, et al. Correlation of blood counts with vascular complications
in essential thrombocythemia: Analysis of the prospective PT1 cohort. Blood 2012;120:14091411.
Passamonti F, Thiele J, Girodon F, et al. A prognostic model to predict survival in 867 World
Health Organization-defined essential thrombocythemia at diagnosis: A study by the International
Working Group on Myelofibrosis Research and
Treatment. Blood 2012;120:11971201.

doi:10.1002/ajh.23685