G Model

JJCC-1895; No. of Pages 5

Journal of Cardiology xxx (2019) xxx–xxx

Contents lists available at ScienceDirect

Journal of Cardiology
journal homepage: www.elsevier.com/locate/jjcc

Original article

Identification of risk factors for venous thromboembolism and

evaluation of Khorana venous thromboembolism risk assessment
in Japanese lung cancer patients
Makoto Hiraide (MSc)a,b, Taro Shiga (MD)c,*, Yuichi Minowa (BS)a, Yasuhiro Nakano (MSc)a,
Hidenori Yoshioka (MSc)a, Kenichi Suzuki (PhD)a, Chikao Yasuda (MD)d,
Harumi Takahashi (PhD)b, Toshihiro Hama (PhD)a
a
Department of Pharmacy, The Cancer Institute Hospital, Japanese Foundation for Cancer Research, Tokyo, Japan
b
Department of Biopharmaceutics, Meiji Pharmaceutical University, Kiyose City, Tokyo, Japan
c
Department of Onco-Cardiology/Cardiovascular Medicine, The Cancer Institute Hospital, Japanese Foundation for Cancer Research, Tokyo, Japan
d
Department of Medical Safety Management, The Cancer Institute Hospital, Japanese Foundation for Cancer Research, Tokyo, Japan

A R T I C L E I N F O A B S T R A C T

Article history: Background: The reported incidence of venous thromboembolism (VTE) in cancer patients is 4–20%. The
Received 26 February 2019 Khorana VTE risk score (KRS) and the Vienna VTE risk score (VRS) have been proposed as scoring models
Received in revised form 30 May 2019 for evaluation of cancer-associated VTE. However, the risk factors of VTE in Japanese lung cancer patients
Accepted 14 June 2019
have not been clarified.
Available online xxx
Methods: This retrospective study included 682 hospitalized Japanese patients with newly diagnosed
lung cancer who were examined for VTE on admission between January 2014 and December 2016.
Keywords:
Results: Seventy-one (10.4%) of the 682 patients were diagnosed with VTE. Multivariate logistic
Body mass index
D-dimer
regression analysis showed that body mass index (BMI) 25 kg/m2 (OR, 2.02; 95% CI, 1.06–3.72), white
Khorana risk score blood cell (WBC) count >11 109/L (OR, 2.31; 95% CI, 1.11–4.61), pre-chemotherapy serum D-dimer
Lung cancer concentration 1.44 mg/mL (OR, 2.73; 95% CI, 1.49–4.99), and non-small cell lung cancer (OR, 3.13; 95% CI,
Venous thromboembolism 1.32–9.23) were significantly associated with VTE in these patients. The cut-off values for BMI, WBC
White blood cell count, and D-dimer concentration determined using receiver operating characteristic curves were
25.4 kg/m2, 11.2  109/L, and 1.95 mg/mL, respectively.
Conclusions: In this study, we were able to identify four independent risk factors for cancer-associated
VTE in Japanese lung cancer patients for the first time. Moreover, we showed that a cut-off level of
25 kg/m2 for BMI was a risk factor for VTE in this cohort.
© 2019 Published by Elsevier Ltd on behalf of Japanese College of Cardiology.

Introduction The development of thrombosis involves a complex interaction

The reported incidence of venous thromboembolism (VTE) in
cancer patients is 4–20% [1], and thromboembolic disease is the
second most common cause of death following cancer progression;
therefore, adequate treatment of thromboembolism in cancer
patients is important [2]. In fact, the number of patients with
cancer-associated thromboembolism has been constantly increas-
ing worldwide [1,3].
* Corresponding author.
E-mail address: taro.shiga@jfcr.or.jp (T. Shiga).
of factors such as plasminogen activator inhibitor-1 (PAI-1)
production, tissue factor release, and cytokine production by
tumor cells [4,5]. These complex mechanisms may lead to not only
venous thrombosis but also arterial thrombosis [6]. Previous
studies have revealed several significant risk factors for cancer-
associated VTE, including the primary site of the cancer, presence
of metastatic disease, application of antineoplastic therapies
including chemotherapy and hormonal therapy, surgery, distur-
bances to venous flow caused by extravascular compression by the
tumor, intravenous tumor invasion, long-term bedridden status,
and intravenous catheterization [1,7,8]. A population-based study
has shown that, in comparison with non-cancer patients, cancer
patients had a 4.1-fold greater risk of thrombosis, and that

https://doi.org/10.1016/j.jjcc.2019.06.013
0914-5087/© 2019 Published by Elsevier Ltd on behalf of Japanese College of Cardiology.

Please cite this article in press as: Hiraide M, et al. Identification of risk factors for venous thromboembolism and evaluation of Khorana
venous thromboembolism risk assessment in Japanese lung cancer patients. J Cardiol (2019), https://doi.org/10.1016/j.jjcc.2019.06.013
G Model
JJCC-1895; No. of Pages 5
2 M. Hiraide et al. / Journal of Cardiology xxx (2019) xxx–xxx
chemotherapy increased the risk to 6.5-fold [9]. Thus, management
of VTE in cancer patients is an issue of vital importance.
The American Society of Clinical Oncology has categorized risk
factors for cancer-associated VTE into cancer-related factors,
treatment-related factors, patient-related factors, and biomarkers
[10]. Previous studies have demonstrated that certain laboratory
parameters are associated with an increased risk of cancer-
associated VTE, including a high platelet (PLT) count [11,12] and a
high leukocyte count [13]. A risk-scoring model incorporating
5 clinical and laboratory parameters – cancer site, PLT count,
hemoglobin (Hb) and/or use of erythropoiesis-stimulating agents,
white blood cell (WBC) count, and body mass index (BMI) – was
developed by Khorana et al. to predict cancer-associated throm-
bosis in ambulatory cancer patients [13].
Biomarkers reflecting activation of blood coagulation and
fibrinolysis, such as high levels of D-dimer, were also indepen-
dently predictive of VTE occurrence in cancer patients [14–16]. In
the Vienna VTE risk score (VRS) model, the risk of developing VTE
has been reported to increase in patients with D-dimer values of
1.44 mg/mL [17]. The Khorana VTE risk score (KRS) and VRS have
been proposed as scoring models for evaluation of cancer-
associated VTE [13,17].
Although the KRS and VRS models can be used in clinical
practice for evaluating the risk of VTE in cancer patients, these
models were created employing clinical data from cancer patients
in Western countries. Risk factors for VTE such as WBC count have
also been reported in Asian countries [18,19], but the risk factors in
Japanese cancer patients are currently unknown. In addition, in
Japanese cancer patients, the incidence of VTE in each cancer type
differs from that in Western countries. Moreover, we are unaware
of the cut-off levels for BMI in Japanese cancer patients that might
indicate the risk of cancer-associated VTE. The BMI cut-off values
for Western patients would differ from those for Japanese patients,
who have a lower BMI on average [20]. Thus, a risk assessment
model that can effectively evaluate the clinical risk of VTE in
Japanese patients is needed. In the present study, therefore, we
attempted to clarify the risk factors for VTE in Japanese cancer
patients. Compared with other cancer types, patients with lung
cancer are known to have a higher VTE incidence rate (13.2%) [21],
but few reports on the incidence rate of VTE in Japanese lung
cancer patients are available. Accordingly, we investigated the
incidence of VTE and risk factors related to cancer-associated VTE
in Japanese lung cancer patients and evaluated the applicability of
KRS and VRS to these patients.
Methods
Subjects and study design
In this retrospective study, newly diagnosed lung cancer
patients hospitalized between January 2014 and December
2016 were selected at the Cancer Institute Hospital, Japanese
Foundation for Cancer Research. Patients hospitalized for the first
time for treatment within the study period were included. The
exclusion criteria were a history of VTE, continuous anticoagula-
tion treatment, and insufficient clinical data. This study was
approved by the Institutional Review Board of the Cancer Institute
Hospital, Japanese Foundation for Cancer Research. The require-
ment for informed consent was waived due to the retrospective
nature of the study.
Using electronic medical records, we collected the following
clinical information: age, sex, height, body weight, BMI, body
surface area, performance status (PS), smoking history, anamnestic
history, complications (cardiovascular disease, diabetes mellitus,
hypertension, dyslipidemia, and chronic pulmonary disease),
clinical staging, primary site, genetic mutation, histologic type,
Please cite this article in press as: Hiraide M, et al. Identification of risk factors for venous thromboembolism and evaluation of Khorana
venous thromboembolism risk assessment in Japanese lung cancer patients. J Cardiol (2019), https://doi.org/10.1016/j.jjcc.2019.06.013
metastasis, recurrence, history of surgery, date of thromboembo-
lism diagnosis, site of thrombosis, laboratory data (WBC, PLT count,
Hb, and D-dimer concentration), and concomitant drug use. Both
symptomatic and asymptomatic deep vein thrombosis (DVT) or
pulmonary embolism (PE) were defined as VTE in this study. DVT
events were confirmed by venous ultrasonography or contrast-
enhanced computed tomography (CT). PE events were confirmed
by CT pulmonary angiography.
Risk assessment for VTE
In the KRS model, VTE risks are evaluated on the basis of clinical
and laboratory parameters [13], including the following five
clinical items: tumor type - stomach and pancreatic cancers,
classified as “very-high-risk”; lymphoma and lung, kidney,
gynecologic, and genitourinary cancers excluding prostate cancer,
classified as “high-risk”, a pre-chemotherapy PLT count of 350
 109/L, a pre-chemotherapy Hb concentration of <10 g/dL, and/or
use of erythropoiesis-stimulating agents, a pre-chemotherapy
WBC count of >11 109/L, and a BMI of 35 kg/m2. The VRS model
additionally includes a D-dimer level of 1.44 mg/mL. In order to
evaluate the applicability of this risk assessment to our patients,
the subjects were classified into two groups based on VTE event:
those with VTE (VTE group) and those without VTE (non-VTE
group). The following factors were selected for our risk analysis:
age, sex, body weight, BMI, performance status (PS), complications
(cardiovascular disease, diabetes mellitus, hypertension, dyslipi-
demia, and chronic pulmonary disease), smoking history, clinical
staging, primary site, genetic mutation, histologic types, metasta-
sis, history of surgery, and laboratory data (WBC, PLT count, Hb and
D-dimer concentration). All variables included in the KRS and VRS
models were collected and we used the same cut-off values of WBC
and D-dimer concentration employed in those scores, except for
BMI  25 kg/m2, which is the diagnostic criterion for obesity in
Japan [22]. We used clinical information and laboratory data
collected at the time of initial admission after lung cancer
diagnosis.
Statistical analysis
Data are presented as medians with ranges (minimum value–
maximum value) or means  standard deviation (SD) where
appropriate. Differences between two groups were analyzed using
Student’s t test, Mann–Whitney U-test or Fisher’s exact test.
Correlations were tested using Spearman’s rank correlation test.
Any variable with a p-value of 0.20 in the univariate analysis was
considered a potential independent variable and included in
multivariate analysis after examining whether there was an
association between the extracted variables using univariate
analysis [23]. Multivariate logistic regression analysis using the
stepwise method was then performed for extracting the variables
that were significant risk factor(s) for VTE. We estimated cut-off
values by receiver operating characteristic (ROC) curve analysis for
each continuous variable selected in the multivariate analysis. A
two tailed p-value of <0.05 was considered statistically significant
for all analyses. All statistical analyses were performed using JMP
PRO version 12 (SAS Institute Japan, Tokyo, Japan).
Results
Patient characteristics and VTE prevalence
A total of 811 patients were newly diagnosed with lung cancer
and admitted to our hospital for initial treatment during the study
period. Of these patients, 729 were hospitalized for treatment.
However, 38 patients were excluded because they were on
G Model
JJCC-1895; No. of Pages 5

M. Hiraide et al. / Journal of Cardiology xxx (2019) xxx–xxx 3

Table 1
Baseline characteristics of the total study population (n = 682).

Characteristics

Age (years)b 67.0 (29–87)

Sex (male)a 449 (65.8)
Weight (kg)c 59.0  11.9
BMI (kg/m2)c 22.3  3.3
Comorbiditiesa
Cardiovascular disease 51 (7.5)
Diabetes mellitus 129 (18.9)
Hypertension 235 (34.5)
Dyslipidemia 97 (14.2)
Chronic pulmonary disease 105 (15.4)
Smokinga 484 (71.0)
Surgerya 132 (19.4)
ECOG performance statusa
0–1 579 (84.9)
Fig. 1. Selection of the study cohort. VTE, venous thromboembolism. 2–4 103 (15.1)
Cancer typea
Small cell lung cancer 116 (17.0)
Non-small cell lung cancer 566 (83.0)
anticoagulants prior to cancer treatment at the time of admission. Adenocarcinoma 428 (62.8)
A further 9 patients were also excluded because of insufficient Squamous cell carcinoma 102 (15.0)
data. Finally, 682 eligible and consecutive patients were included Other NSCLCs 36 (5.2)
Genetic mutationa
(Fig. 1). Table 1 shows the baseline patient characteristics. The EGFR mutation positive 187 (27.4)
682 included patients with lung cancer had a median age of ALK positive 28 (4.1)
67.0 years, and 66.8% (n = 449) of them were men. In terms of Stagea
tumor histology, 62.8% (n = 428) of the tumors were adenocarci- I-II 20 (2.9)
III–IV 662 (97.1)
nomas, 15.0% (n = 102) were squamous cell carcinomas, 17.0%
Cancer metastasisa
(n = 116) were small cell lung cancers (SCLC), and 5.2% (n = 36) were Lymph node metastasis 96 (14.1)
other non-small cell lung cancers (NSCLC). Distant metastasis was Distant metastasis 381 (55.9)
found in 381 patients (55.9%). The mean BMI was 22.3  3.3 kg/m2 Baseline laboratory valuesc
. Mean WBC count and D-dimer were 7.7  3.7  109/L and WBC count  109 (/L) 7.72  3.74
Hemoglobin (g/dL) 13.0  1.6
1.8  4.9 mg/mL, respectively. Patients with a high KRS (3 points) Platelet count  109 (/L) 265.4  97.6
were larger in the VTE group than in the non-VTE group [14.1% (10/ D-dimer (mg/mL) 1.77 4.88
71) vs 7.2% (44/611); p = 0.055]. Overall, VTE events occurred in VTEa 71 (10.4)
71 of the 682 patients (10.4%). Forty-eight patients (7.0%) DVT alone 48 (7.0)
PE alone 9 (1.3)
developed DVT alone, 9 (1.3%) developed PE alone, and 14 (2.1%)
PE + DVT 14 (2.1)
developed both DVT and PE. The average time (range) between KRSa
cancer detection and VTE development in these patients was 217 1 point 536 (78.6)
(1–1040) days. 2 points 92 (13.5)
3 points 45 (6.6)
4 points 9 (1.3)
Factors associated with VTE 5 points 0 (0.0)

Characteristics of patients with or without VTE are given in
Table 2. Univariate analysis demonstrated significant differences
(p < 0.2) in the seven parameters: BMI 25 kg/m2 (p = 0.163),
diabetes mellitus (p = 0.107), PS 2 (p = 0.158), NSCLC (p = 0.002),
adenocarcinoma (p = 0.019), WBC count >11 109 /L (p < 0.001),
and D-dimer concentration 1.44 mg/mL (p < 0.001). The cut-off
values for BMI, WBC count, and D-dimer concentration employing
ROC analysis were 25.4 kg/m2 [area under the curve (AUC), 0.63],
11.2  109/L (AUC, 0.62), and 1.95 mg/mL (AUC, 0.67), respectively.
As adenocarcinoma is included in NSCLC, the six significant
parameters except for adenocarcinoma were further used for
multivariate logistic analysis to identify significant risk factors for
VTE. We performed multivariate logistic analysis, using each cut-
off value in Japanese lung cancer patients (Table 3). As a result, BMI
[odds ratio (OR), 2.21; 95% confidence interval (CI), 1.13–4.15;
p = 0.021], WBC count (OR, 2.49; 95% CI, 1.23–4.88; p = 0.012), D-
dimer concentration (OR, 4.84; 95% CI, 2.62–8.91; p < 0.001), and
NSCLC (OR, 3.65; 95% CI, 1.54–10.80; p = 0.002) were significantly
associated with VTE. We also performed multivariate logistic
analysis, using the same cut-off values employed in KRS and VRS.
As a result, BMI (OR, 2.02; 95% CI, 1.06–3.72; p = 0.032), WBC count
(OR, 2.31; 95% CI, 1.11–4.61; p = 0.026), D-dimer concentration (OR,
2.73; 95% CI, 1.49–4.99; p = 0.001), and NSCLC (OR, 3.13; 95% CI,
1.32–9.23; p = 0.007) were significantly associated with VTE
(Table 4).
ALK, anaplastic lymphoma kinase; BMI, body mass index; DVT, deep vein
thrombosis; ECOG, Eastern Cooperative Oncology Group; EGFR, epidermal
growth factor receptor; KRS, Khorana risk score; NSCLC, non-small cell lung
cancer; PE, pulmonary embolism; VTE, venous thromboembolism; WBC, white
blood cell.
a
b
n (%).
Median (range).
c
Mean  SD.
Discussion
In this study, the prevalence of total VTE was 10.4% in patients
newly diagnosed with lung cancer before initiation of cancer
therapy. Zhang et al. reported previously that the incidences of
VTE, DVT, PE, and PE + DVT were 13.2%, 6.2%, 4.9%, and 2.1%,
respectively [21], in 672 pre-treatment Chinese patients with lung
cancer, and our present results (incidence rates for VTE, DVT, PE,
and PE + DVT: 10.4%, 7.0%, 1.3%, and 2.1%, respectively) were
compatible with those.
We identified four clinical and laboratory parameters that were
independently associated with VTE in Japanese lung cancer
patients: BMI 25 kg/m2, WBC count >11 109 /L, D-dimer
concentration 1.44 mg/mL, and NSCLC. In this study, we used
ROC analysis to calculate the cut-off value of the VTE risk factor in
Japanese lung cancer patients. The cut-off values for BMI, WBC

Please cite this article in press as: Hiraide M, et al. Identification of risk factors for venous thromboembolism and evaluation of Khorana
venous thromboembolism risk assessment in Japanese lung cancer patients. J Cardiol (2019), https://doi.org/10.1016/j.jjcc.2019.06.013
G Model
JJCC-1895; No. of Pages 5

4 M. Hiraide et al. / Journal of Cardiology xxx (2019) xxx–xxx

Table 2 Table 4
Comparison of clinical characteristics between patients with and without VTE. Multivariate analysis of the risk of the venous thromboembolism in Japanese
patients with lung cancer using the same cut-off values employed in KRS and VRS.
Parameter VTE non-VTE p-Value
(n = 71) (n = 611) Parameter Odds ratio 95% Confidence p-Value
interval
Age (years)b 64.5  10.8 66.0  10.1 0.252
65 yearsa 42 (59.2) 375 (61.4) 0.897 BMI 25 (kg/m2) 2.02 1.06–3.72 0.032
Sex (male)a 43 (60.6) 406 (66.4) 0.426 Diabetes mellitus 0.47 0.17–1.05 0.068
BMI (kg/m2)b 22.5  3.5 22.2  3.3 0.672 ECOG performance status 2 1.01 0.46–2.07 0.984
25 (kg/m2)a 20 (28.2) 121 (19.8) 0.163 Non-small cell lung cancer 3.13 1.32–9.23 0.007
35 (kg/m2)a 0 (0.0) 1 (0.002) 1.000 WBC count >11 109 (/L) 2.31 1.11–4.61 0.026
Comorbiditiesa D-dimer 1.44 (mg/mL) 2.73 1.49–4.99 0.001
Cardiovascular disease 3 (4.2) 48 (7.9) 0.346
BMI, body mass index; ECOG, Eastern Cooperative Oncology Group; KRS,
Diabetes mellitus 8 (11.3) 121 (19.8) 0.107
Khorana risk score; VRS, Vienna risk score; WBC, white blood cell.
Hypertension 30 (42.3) 205 (33.6) 0.232
Dyslipidemia 12 (16.9) 85 (13.9) 0.470
Chronic pulmonary disease 11 (15.5) 94 (15.4) 1.000
Smokinga 45 (63.4) 439 (71.8) 0.211
Surgerya 16 (22.5) 116 (19.0) 0.427 present study, only 1 of the 682 patients (0.1%) had a BMI of
ECOG performance statusa
35 kg/m2. Accordingly, we considered that it might be appro-
0–1 56 (78.9) 522 (85.4) 0.158
2–4 15 (21.1) 89 (14.6) 0.158
priate to set the cut-off BMI value as 25 kg/m2 in a Japanese
Cancer typea clinical setting. Moreover, we were able to show that the OR for
Non-small cell lung cancer 69 (97.2) 497 (81.3) 0.002 BMI 25 kg/m2 for VTE was 2.02 (p = 0.032), using multivariate
Adenocarcinoma 54 (76.1) 374 (61.2) 0.019 analysis in the present study. From this result, we considered it
Squamous cell carcinoma 10 (14.1) 92 (15.1) 0.863
reasonable to use BMI 25 kg/m2 for assessment of VTE risk in
Genetic mutationa
EGFR mutation positive 25 (35.2) 162 (26.5) 0.203 Japanese cancer patients. The occurrence of thrombosis has been
ALK positive 1 (1.4) 27 (4.4) 0.346 reported to be associated with a decrease in endothelial function,
Stagea a decrease in adiponectin, and an increase in PAI-1 secondary to
I–II 3 (4.2) 17 (2.8) 0.462
obesity [24].
III–V 68 (95.8) 594 (97.2) 0.462
Cancer metastasisa
Similarly to the KRS model, the present study showed that an
Lymph node metastasis 6 (8.5) 90 (14.7) 0.204 increase in the WBC count (>11 109/L) was significantly
Distant metastasis 39 (54.9) 342 (56.0) 0.800 associated with the risk of developing VTE. The association
Baseline laboratory valuesa between the increase in WBCs and VTE indicates that cancer cells
WBC count >11 109 (/L) 18 (25.4) 65 (10.6) 0.001
may increase the blood concentrations of inflammatory cytokines
Hemoglobin <10 (g/dL) 4 (5.6) 21 (3.4) 0.312
Platelet count 350  109 (/L) 13 (18.3) 87 (14.2) 0.371 such as interleukin-6 and exert coagulant-promoting action
D-dimer 1.44 (mg/mL) 27 (38.0) 112 (18.3) <0.0001 through activation of neutrophils and monocytes [25].
ALK, anaplastic lymphoma kinase; BMI, body mass index; ECOG, Eastern
The D-dimer concentration is known to be a useful biomarker
Cooperative Oncology Group; EGFR, epidermal growth factor receptor; VTE, for predicting VTE [14,15]. The risk of developing VTE increases in
venous thromboembolism; WBC, white blood cell. patients with a D-dimer concentration of 1.44 mg/mL [17]. In this
a
n (%). study, we obtained similar results, suggesting that patients with D-
b
Mean  SD Values determined by Student’s t-test, Fisher’s exact test, or
dimer concentrations of 1.44 mg/mL (OR 2.73, p = 0.001) were at
Mann–Whitney U-test.
higher risk of developing VTE in comparison with those with a D-
dimer concentration of <1.44 mg/mL. Regarding the D-dimer cut-
Table 3
off value, our results revealed a higher cut-off level for D-dimer
Multivariate analysis of the risk of the venous thromboembolism in Japanese than that in the VRS [17]. However, we consider that VTE should
patients with lung cancer not be excluded in patients with D-dimer levels of <1.95 mg/mL.
Complications related to VTE in cancer patients are not rare even
Parameter Odds ratio 95% Confidence p-Value
interval when D-dimer levels are lower than the cut-off value. VTE can
sometimes be detected even in patients with D-dimer levels lower
BMI 25.4 (kg/m2) 2.21 1.13–4.15 0.021
Diabetes mellitus 0.53 0.22–1.12 0.098
than 1.0 mg/mL. We consider that patients with D-dimer levels of
ECOG performance status 2 0.79 0.37–1.60 0.530 1.95 mg/mL may have a high likelihood of VTE. Furthermore, on
Non-small cell lung cancer 3.65 1.54–10.8 0.002 this basis, we believe that it is important to simultaneously
WBC count >11.2  109 (/L) 2.49 1.23–4.88 0.012 examine patient symptoms, clinical background, and other factors
D-dimer 1.95 (mg/mL) 4.84 2.62–8.91 <0.001
that can cause VTE.
BMI, body mass index; ECOG, Eastern Cooperative Oncology Group; WBC, white In this study, NSCLC was extracted as an independent risk factor
blood cell. associated with the incidence of VTE. Patients with NSCLC,
especially adenocarcinoma, have been reported to have a higher
incidence of VTE [26,27].
count, and D-dimer concentration employing ROC analysis were Our study had several limitations. First, it was a single-center,
25.4 kg/m2 (AUC, 0.63), 11.2  109/L (AUC, 0.62), and 1.95 mg/mL retrospective study.Therefore, a prospective study would definitely
(AUC, 0.67), respectively. There have been few reports on the cut- be necessary to evaluate our established model and the extracted
off values for VTE risk factors in Japanese patients. Therefore, the risk factors for VTE. Second, we were unable to investigate the
cut-off value obtained in this study would be useful for VTE relationship between biomarkers other than D-dimer, such as
screening in Japanese patients with lung cancer. tissue factor activity and soluble P-selectin [11] and VTE. Finally, we
BMI 35 kg/m2 is included as one of the risk factors for cancer- were unable to investigate the influence of chemotherapy on the
associated VTE in the KRS [9]. In a Japanese clinical setting, the incidence of VTE because all our data including patient character-
diagnostic criterion for obesity is considered to be a BMI istics and clinical laboratory data were collected at the first
exceeding 25 kg/m2 [22], and the proportion of adults with a hospital admission after diagnosis of lung cancer, which was before
BMI of 35 kg/m2 has been reported to be 0.2-0.3% [22]. In the chemotherapy.

Please cite this article in press as: Hiraide M, et al. Identification of risk factors for venous thromboembolism and evaluation of Khorana
venous thromboembolism risk assessment in Japanese lung cancer patients. J Cardiol (2019), https://doi.org/10.1016/j.jjcc.2019.06.013
G Model
JJCC-1895; No. of Pages 5
M. Hiraide et al. / Journal of Cardiology xxx (2019) xxx–xxx
Conclusions
Various mechanisms are involved in the onset of VTE in cancer
patients, and thrombosis can be a fatal factor when VTE develops.
In clinical practice, it is important to comprehensively judge the
risks of individual patients from an array of clinical information
and choose an appropriate treatment. In this study, we identified
for the first time four independent risk factors for cancer-
associated VTE in Japanese lung cancer patients. Furthermore,
we showed that BMI  25 kg/m2 in Japanese lung cancer patients is
a risk factor for VTE, in comparison with BMI  35 kg/m2 in
Western countries, while other clinical parameters of KRS/VRS we
investigated had similar cut-off values. We also found that most
risk factors used in the KRS and VRS models could be useful for
clinical assessment of Japanese lung cancer patients.
Funding
This research received no grant from any funding agency in the
public, commercial, or not-for-profit sectors.
Conflicts of interest
The authors declare that there is no conflict of interest.
References
[1] Lyman GH, Khorana AA, Falanga A, Clarke-Pearson D, Flowers C, Jahanzeb M,
et al. American Society of Clinical Oncology guideline: recommendations for
venous thromboembolism prophylaxis and treatment in patients with cancer. J
Clin Oncol 2007;25:5490–505.
[2] Khorana AA, Francis CW, Culakova E, Kuderer NM, Lyman GH. Thromboembo-
lism is a leading cause of death in cancer patients receiving outpatient
chemotherapy. J Thromb Haemost 2007;5:632–4.
[3] Ikeda S, Koga S, Yamagata Y, Eguchi M, Sato D, Muroya T, et al. Comparison of the
effects of edoxaban, an oral direct factor Xa inhibitor, on venous thromboembo-
lism between patients with and without cancer. J Cardiol 2018;72:120–7.
[4] Mukai M, Oka T. Mechanism and management of cancer-associated thrombo-
sis. J Cardiol 2018;72:89–93.
[5] Hisada Y, Mackman N. Cancer-associated pathways and biomarkers of venous
thrombosis. Blood 2017;130:1499–506.
[6] Varki A. Trousseau’s syndrome: multiple definitions and multiple mecha-
nisms. Blood 2007;110:1723–9.
[7] Horsted F, West J, Grainge MJ. Risk of venous thromboembolism in patients with
cancer: a systematic review and meta-analysis. PLoS Med 2012;9e1001275.
[8] Khorana AA, Francis CW, Culakova E, Kuderer NM, Lyman GH. Frequency, risk
factors, and trends for venous thromboembolism among hospitalized cancer
patients. Cancer 2007;110:2339–46.
Please cite this article in press as: Hiraide M, et al. Identification of risk factors for venous thromboembolism and evaluation of Khorana
venous thromboembolism risk assessment in Japanese lung cancer patients. J Cardiol (2019), https://doi.org/10.1016/j.jjcc.2019.06.013
5
[9] Heit JA, Silverstein MD, Mohr DN, Petterson TM, O’Fallon WM, Melton 3rd LJ.
Risk factors for deep vein thrombosis and pulmonary embolism: a population-
based case-control study. Arch Intern Med 2000;160:809–15.
[10] Khorana AA, Connolly GC. Assessing risk of venous thromboembolism in the
patient with cancer. J Clin Oncol 2009;27:4839–47.
[11] Khorana AA, Francis CW, Culakova E, Lyman GH. Risk factors for chemothera-
py-associated venous thromboembolism in a prospective observational study.
Cancer 2005;104:2822–9.
[12] Simanek R, Vormittag R, Ay C, Alguel G, Dunkler D, Schwarzinger I, et al. High
platelet count associated with venous thromboembolism in cancer patients:
results from the Vienna Cancer and Thrombosis Study (CATS). J Thromb
Haemost 2010;8:114–20.
[13] Khorana AA, Kuderer NM, Culakova E, Lyman GH, Francis CW. Development
and validation of a predictive model for chemotherapy-associated thrombosis.
Blood 2008;111:4902–7.
[14] Ay C, Vormittag R, Dunkler D, Simanek R, Chiriac AL, Drach J, et al. D-dimer and
prothrombin fragment 1 + 2 predict venous thromboembolism in patients
with cancer: results from the Vienna cancer and thrombosis study. J Clin Oncol
2009;27:4124–9.
[15] Pabinger I, Thaler J, Ay C. Biomarkers for prediction of venous thromboembo-
lism in cancer. Blood 2013;122:2011–8.
[16] Arpaia G, Carpenedo M, Verga M, Mastrogiacomo O, Fagnani D, Lanfredini M,
et al. D-dimer before chemotherapy might predict venous thromboembolism.
Blood Coagul Fibrinolysis 2009;20:170–5.
[17] Ay C, Dunkler D, Marosi C, Chiriac AL, Vormittag R, Simanek R, et al. Prediction
of venous thromboembolism in cancer patients. Blood 2010;116:5377–82.
[18] Chen JS, Hung CY, Chang H, Liu CT, Chen YY, Lu CH, et al. Venous thromboem-
bolism in Asian patients with pancreatic cancer following palliative chemo-
therapy: low incidence but a negative prognosticator for those with early
onset. Cancers (Basel) 2018;10:e501.
[19] Wang KL, Yap ES, Goto S, Zhang S, Siu CW, Chiang CE. The diagnosis and
treatment of venous thromboembolism in Asian patients. Thrombosis J
2018;16:4. http://dx.doi.org/10.1186/s12959-017-0155-z.
[20] WHO Expert Consultation. Appropriate body-mass index for Asian popula-
tions and its implications for policy and intervention strategies. Lancet
2004;363:157–63.
[21] Zhang Y, Yang Y, Chen W, Guo L, Liang L, Zhai Z, et al. Prevalence and
associations of VTE in patients with newly diagnosed lung cancer. Chest
2014;146:650–8.
[22] Japan Society for the Study of Obesity. Guidelines for the management of
obesity disease 2016 editors. Tokyo: Life Science; 2016.
[23] Kawaguchi T, Kumabe T, Kanamori M, Saito R, Yamashita Y, Sonoda Y, et al. Risk
assessment for venous thromboembolism in patients with neuroepithelial
tumors: pretreatment score to identify high risk patients. Neurol Med Chir
2013;53:467–73.
[24] Kato H, Kashiwagi H, Shiraga M, Tadokoro S, Kamae T, Ujiie H, et al. Adipo-
nectin acts as an endogenous antithrombotic factor. Arterioscler Thromb Vasc
Biol 2006;26:224–30.
[25] Ay C, Pabinger I, Cohen AT. Cancer-associated venous thromboembolism:
burden, mechanisms, and management. Thromb Haemost 2017;117:219–30.
[26] Vitale C, D’Amato M, Calabrò P, Stanziola AA, Mormile M, Molino A. Venous
thromboembolism and lung cancer: a review. Multidiscip Respir Med
2015;10:28.
[27] Blom JW, Osanto S, Rosendaal FR. The risk of a venous thrombotic event in lung
cancer patients: higher risk for adenocarcinoma than squamous cell carcino-
ma. J Thromb Haemost 2004;2:1760–5.