Thrombosis Research 135, Suppl.

1 (2015) S8S11

The mechanisms of cancer-associated thrombosis

Anna Falanga *, Marina Marchetti, Laura Russo
Department of Immunohematology and Transfusion Medicine, Hospital Papa Giovanni XXIII, Bergamo, Italy

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

Keywords: Patients with cancer may display many types of hemostatic disorders that signicantly contribute to mor-
Thrombosis bidity and mortality in this disease. A complex coagulopathy develops in parallel with malignancy and is
Cancer characterized by activation of clotting mechanisms to different extent in different patients and in different
Hemorrhage types of tumor. The pathogenesis of hemostatic alterations in cancer is multifactorial; however, the tumor
Hypercoagulable state
tissue capacity to interact with and activate the host hemostatic system plays an important role. New molec-
Tissue factor
ular pathways of regulation of these properties have been recently demonstrated. Intervention strategies to
prevent and treat venous thromboembolism (VTE) in cancer patients have been addressed by large RCTs and
guidelines for VTE management have been updated.
In this review, we will present an updated overview of the complex coagulopathy associated to malignancy
and of recent advances in the thrombotic risk assessment of cancer patients.
2015 Elsevier Ltd. All rights reserved.

Introduction
Patients with cancer may present many types of hemostatic
disorders with increased frequency of related complications. A tight
relationship exists between malignant disease, the occurrence of
coagulation abnormalities and thrombosis. The relationship relies
on the evidence that cancer induces a prothrombotic switch of
the host hemostatic system, and vice versa, that blood clotting
activation, positively stimulates tumor growth and dissemination.
Abnormalities in circulating thrombotic marker levels is a common
laboratory nding in these patients, also in the absence of overt
thrombo-hemorragic disease, and demonstrates that activation of
blood coagulation and brinolysis parallels the development and
spread of cancer [1].
The clinical manifestations of hemostasis deregulation may vary
from a subclinical asymptomatic hypercoagulable state to manifest
thrombosis of the large vessels or, further, to systemic disseminated
intravascular coagulation with severe bleeding. Therefore, prevent-
ing these complications is clinically relevant because signicantly
contribute to the morbidity and mortality of these patients.
The pathogenesis of cancer-associated coagulopathy is complex
and involves various mechanisms. Most importantly, tumor cells
gain the capacity to activate the host hemostatic system, and this
phenomenon is driven by the same oncogenes responsible for the
neoplastic transformation. Indeed, by this process, cancer tissues
become capable to express different procoagulant proteins (i.e
Tissue Factor [TF], Cancer Procoagulant, Factor VII), which contribute
to the occurrence of the overt symptomatic coagulopathy in vivo
[2]. The shedding of procoagulant microparticles is also regulated
by oncogenic events and further adds to the pathogenesis of the
cancer-associated hypercoagulable state. Finally, the changes in the
stromal cells of the tumor niche induced by tumor TF shade new
lights on the hemostasis/cancer interaction.
In addition, many clinical factors inuence the thrombotic risk
of cancer patients, who represent a high-risk category of subjects.
Clinical factors include general cardiovascular risk factors, cancer-
specic risk factors, and anti-cancer therapies (see Table 1). Recent
data conrm that the risk of developing venous thromboembolism
(VTE) is increased up to seven-fold in these patients as compared
to the general population [3]. To help clinicians in the prevention
and management of thrombotic events in cancer patients, a number
of guidelines have been released from national and international
scientic societies [4,5].
This paper wishes to provide the most recent discoveries on the
coagulopathy of cancer patients, the mechanisms underlying this
phenomenon, and nally the risk assessment to predict thrombosis
in each single patient.
Pathogenesis of coagulation in cancer
The pathogenesis of the coagulation in cancer is complex and
involves several clinical and biological factors

Clinical factors
* Correspondence to: Anna Falanga, MD, Department of Immunohematology and
Transfusion Medicine & the Hemostasis and Thrombosis Center, Hospital Papa
Multiple clinical risk factors (i.e. patients-, cancer-, and
Giovanni XXIII, Piazza OMS, I-24127, Bergamo, Italy. Tel.: +39 035 2673663; fax:
+39 035 2674832. treatment-related) concur to blood clotting activation and to
E-mail address: annafalanga@yahoo.com (A. Falanga). thrombotic manifestations in cancer [6,7]. Regarding cancer-related

0049-3848/$ see front matter 2015 Elsevier Ltd. All rights reserved.
 A. Falanga et al. / Thrombosis Research 135, Suppl. 1 (2015) S8S11 S9

features, VTE risk varies greatly depending on the primary cancer
site and the stage of the disease. Patients with pancreatic, stomach,
lung cancer, and hematologic malignancies, including lymphomas
and myeloma, have the strongest association with VTE. Additionally,
advanced, metastatic cancers are associated with an increased risk
of VTE compared to localized tumors. To the some guise, active
anticancer treatments, including chemotherapy, hormonal therapy,
anti-angiogenic agents, combination regimens, and surgery, sig-
nicantly increase the thrombotic risk. In particular, the use of
systemic chemotherapy seems to increase the risk of VTE beyond to
the expected from the disease itself, and is associated with up to
6-fold increase in VTE risk.
Biological factors
Away from clinical factors, biological pathways likely play an
important role in the pathogenesis of hemostatic alterations in
cancer. The principal mechanisms of thrombosis include the expres-
sion of hemostatic proteins by tumor cells (i.e. TF), the production
by tumor and/or host cells of microparticles (MP), inammatory
cytokines (i.e. TNF-, IL-1), and proangiogenic factors (VEGF,
bFGF), and the tumor cell expression of adhesion molecules to
bind platelets, endothelial cells and leucocytes (Fig. 1). The same
properties also contribute to tumor progression [8]. TF, constitu-
tively expressed on the malignant cell surface, plays a fundamental
role in thrombin generation in cancer, but also contributes to
tumor progression by directly inducing VEGF expression by both
malignant and host vascular cells. This property regulates tumor
neo-vascularization and provides an important link between coagu-
lation, inammation, thrombosis and tumor growth and metastasis
[8]. Another emerging mechanism of tumor-promoted clotting ac-
tivation is the heparanase activity, which associated with a high
metastatic potential of tumor cells. Heparanase, besides degrading
the extracellular matrix, upregulates the expression of TF by inter-
acting with the tissue factor pathway inhibitor (TFPI) on the surface
of endothelial and tumor cells, thereby causing the dissociation of
TFPI with a subsequent increase in TF activity [9]. Among tumor
tissue procoagulant activities, much relevance is gaining the role of
the tumor-derived MP. These small membrane vescicles carry high
concentrations of procoagulant phosphatidylserine and TF [10].
In the last decade, molecular studies of experimental models
of human cancers (including hepatoma, brain and colon cancer)
have demonstrated that oncogene and repressor gene-mediated
neoplastic transformation (i.e. activation of Met, loss of PTEN,
induction of K-ras and loss of p53) activate clotting as an integral
feature of neoplastic transformation. Moreover, a mutation of EGFR
gene renders cancer cells hypersensitive to the action of coagulation
proteins, such as TF. As a result, a microenvironment promoting
tumor growth is generated [2]. These data conrm that a reciprocal
cancer-thrombosis connection exists, by which cancer cells support
clot formation, and clotting proteins support cancer growth and
dissemination. In the development of human cancers, cycles of
microenvironmental and genetic changes are increasingly manifest.
Indolent or dormant phenotype of cells lacking tumor initiating
capacity may be altered in the presence of procoagulant TF, which
induces several changes in the tumor microenvironment, capable to
affect the fate of these cells from a dormant to a fully malignant
phenotype [11].
Finally, much relevant to the occurrence of the prothrombotic
state of cancer is the contribution of host endothelial cells, platelets
and leukocytes. Recent data support the role of platelets in the
maintenance of the vascular integrity within tumors, acting as
guardian of tumor vasculature; therefore platelets may be a target
for the specic destabilization of tumor vessels [12]. Extensive
experimental evidence also shows the presence of large quantities
of neutrophils in the tumor where they release extracellular DNA
traps (NETs) and affect tumor growth and neoangiogenesis [13].
Assessment of the thrombotic risk in cancer patients
The role of most of clinical risk factors (Table 1) in cancer-
associated thrombosis is well known. Differently, considerable
effort has been made and is still ongoing to identify the predictive
value of VTE events of several circulating thrombotic biomarkers.

Fig. 1. Cancer-hemostatic system interactions. Tumor cells can activate the hemostatic system in multiple ways. Tumor cells release procoagulant activities, and
microparticles (MP), by which activate the coagulation cascade. Tumor cells also activate the host haemostatic cells (endothelial cells, leukocytes, and platelets), by either
release of soluble factors or by direct adhesion contact, thus eliciting the expression of a procoagulant phenotype of these cells. In addition, the neutrophils can release
neutrophil extracellular trap (NETs) and the adhesion of a large quantity of NETs to the vasculature may initiate thrombosis by providing a scaffold for platelet adhesion,
activation and thrombin generation.
S10 A. Falanga et al. / Thrombosis Research 135, Suppl. 1 (2015) S8S11
Table 1
Thrombotic risk factors in cancer patients.
General factors Cancer-specic factors Treatment-related factors
Older age Site of cancer: brain, Surgery
Gender pancreas, kidney, Chemo- and
Immobility stomach, lung, bladder, hormone-therapy
Previous VTE gynecologic, hematologic Anti-angiogenic therapy
Infections malignancies Erythropoiesis
Stage of cancer:
Hypertension stimulating agents
Hyperlipidemia advanced stage Blood transfusions
Initial period of Central venous lines
diagnosis
These include thrombinantithrombin complex [TAT], prothrombin
fragment F1+2 [F1+2], brinopeptide A and B, plasminogen activator
inhibitor 1, D-dimer, and others [14]. Data from the Vienna
Cancer and Thrombosis (CAT) prospective study show a signicant
association with increased VTE risk of elevated P-selectin [15], or
elevated D-dimer plus F1+2 levels [16]. High MP concentration
have been found to be increased in both solid and hematological
malignancies and have been utilized to select patients to entry
into a randomized clinical trial of thromboprophylaxis during
chemotherapy. The results of this small, but very original trial are
promising [17]. In this context, the thrombin generation assay, a
global test of coagulation activation, also is under evaluation to
predict VTE in cancer [18].
The interaction and relative effects of the several risk factors
associated with VTE in cancer patients is highly complex, making
the pretreatment risk assessment dicult. The updated ASCO
VTE Guidelines Panel recommends against the use of a single
risk factor or a single biomarkers to identify high-risk patients,
and instead recommends the use of a Risk Score models that
incorporate multiple variables to identify the high-risk patients [5].
The combination of some of the clinical factors with laboratory
biomarkers allowed the development of the Khorana VTE risk
assessment model (RAM) specic for cancer patients undergoing
chemotherapy [19]. This RAM is based on 5 predictive variables,
including cancer site, platelet count, hemoglobin level or the use of
erythropoiesis-stimulating agents, leukocyte count, and body mass
index [19]. Subsequently, the Vienna CAT Study group proposed
an expansion of this score by including two additional biomarkers
(i.e. P-selectin and D-Dimer), which considerably improved the
predictive value of VTE in the study population [14]. More recently,
another modi?ed Khoranas RAM, the Protecht score, was designed
by adding platinum- or gemcitabine-based chemotherapy to the 5
original predictive variables to increase the high-risk cancer patients
identication [20]. In addition, in the setting of patients with
multiple myeloma, Palumbo et al. published a RAM based on expert
recommendations, specic for the denition of patients at risk
for thalidomide or lenalidomide treatment-associated thrombosis
[21]. Finally, the Ottawa Score is capable to distinguish between
cancer patients with thrombosis at high or low risk of recurrent
VTE [22,23]. The ASCO panel recommends the use of RAMs to
periodically assess the thrombotic risk of cancer patients during the
different disease phases.
The search for thrombotic markers capable to predict survival
in cancer patients is currently active. Results of a large cohort
of individuals free of clinically evident cardiovascular and cancer
disease (MOLI-SANI project) show that elevated plasma D-dimer
levels are independently associated with increased risk of death for
any cause [24]. Furthermore, in patients with ovarian masses, the
preoperative D-dimer levels, alone or in combination with CA-125,
signicantly predict for malignant versus benign ovarian lesions
[25].
Conclusions
Cancer patients are at high risk of developing thrombosis. The
pathogenesis of blood coagulation activation in cancer is complex
and involves both clinical and biological factors. The thrombotic
risk varies according to the type and stage of malignancy, and is
increased by concomitant patient-related general thrombotic risk
factors and anti-cancer therapies. However, the tumor cell-specic
prothrombotic properties and the host cell inammatory response
dramatically contribute to thrombosis in these patients. New path-
ways of regulation of these properties have been recently identied.
This may lead to the denition of target-specic treatments to hit
both cancer and thrombosis. For the time being, the development
of risk assessment models to predict thrombosis in cancer is impor-
tant for identifying high-risk patients and predispose the adequate
preventive measures.
Acknowledgements
The authors wish to thank Associazione Italiana per la Ricerca
sul Cancro (A.I.R.C., grants IG10558 and 5 per mille 12237) for its
support.
Conict of interest statement
The authors have no conicts of interest to declare
References
[1] Falanga A, Russo L. Epidemiology, risk and outcomes of venous thromboem-
bolism in cancer. Hamostaseologie 2012;32(2):11525.
[2] Magnus N, DAsti E, Meehan B, Garnier D, Rak J. Oncogenes and the
coagulation system forces that modulate dormant and aggressive states in
cancer. Thromb Res 2014;133(Suppl 2):S19.
[3] Timp JF, et al. Epidemiology of cancer-associated venous thrombosis. Blood
2013;122(10):171223.
[4] Farge D, et al. International clinical practice guidelines for the treatment and
prophylaxis of venous thromboembolism in patients with cancer. J Thromb
Haemost 2013;11(1):5670.
[5] Lyman GH, et al. Venous thromboembolism prophylaxis and treatment in
patients with cancer: American Society of Clinical Oncology clinical practice
guideline update. J Clin Oncol 2013;31(17):2189204.
[6] Falanga A, Russo L, Verzeroli C, Mechanisms of thrombosis in cancer. Thromb
Res 2013;131(Suppl 1):S5962.
[7] Khorana AA, McCrae KR. Risk stratication strategies for cancer-associated
thrombosis: an update. Tromb Res 2014;133(S2):S358.
[8] Falanga A, Marchetti M, Vignoli A. Coagulation and cancer: biological and
clinical aspects. J Thromb Haemost 2013;11(2):22333.
[9] Nadir Y, Brenner B. Heparanase multiple effects in cancer. Thromb Res
2014;133(Suppl 2):S904.
[10] Falanga A, Tartari CJ, Marchetti M. Microparticles in tumor progression.
Thromb Res 2012;129(Suppl 1):S1326.
[11] Magnus N, et al. Tissue factor expression provokes escape from tumor
dormancy and leads to genomic alterations. Proc Natl Acad Sci U S A
2014;111(9):35449.
[12] Stegner D, Dutting S, Nieswandt B. Mechanistic explanation for platelet
contribution to cancer metastasis. Thromb Res 2014;133(Suppl 2):S14957.
[13] Demers M, Wagner DD. Neutrophil extracellular traps: A new link to cancer-
associated thrombosis and potential implications for tumor progression.
Oncoimmunology, 2013;2(2):e22946.
[14] Pabinger I, Thaler J, Ay C. Biomarkers for prediction of venous thromboem-
bolism in cancer. Blood 2013;122(12):20118.
[15] Ay C, et al. High plasma levels of soluble P-selectin are predictive of venous
thromboembolism in cancer patients: results from the Vienna Cancer and
Thrombosis Study (CATS). Blood 2008;112(7):27038.
[16] Ay C, 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(25):41249.
[17] Zwicker JI, et al. Prediction and prevention of thromboembolic events with
enoxaparin in cancer patients with elevated tissue factor-bearing micropar-
ticles: a randomized-controlled phase II trial (the Microtec study). Br J
Haematol 2013;160(4):5307.
[18] Ay C, et al. Prediction of venous thromboembolism in patients with cancer
 A. Falanga et al. / Thrombosis Research 135, Suppl. 1 (2015) S8S11
by measuring thrombin generation: results from the Vienna Cancer and
Thrombosis Study. J Clin Oncol 2011;29(15):2099103.
[19] Khorana AA, et al. Development and validation of a predictive model for
chemotherapyassociated thrombosis. Blood 2008;111(10):49027.
[20] Verso M, et al. A modied Khorana risk assessment score for venous
thromboembolism in cancer patients receiving chemotherapy: the Protecht
score. Intern Emerg Med 2012;7(3):2912.
[21] Palumbo A, et al. Prevention of thalidomide- and lenalidomide-associated
thrombosis in myeloma. Leukemia 2008;22(2):41423.
[22] Louzada ML, et al., Development of a clinical prediction rule for risk stratica-
tion of recurrent venous thromboembolism in patients with cancer-associated
venous thromboembolism. Circulation 2012;126(4):44854.
S11
[23] den Exter PL, Kooiman J, Huisman MV. Validation of the Ottawa prognostic
score for the prediction of recurrent venous thromboembolism in patients
with cancer-associated thrombosis. J Thromb Haemost 2013;11(5):9981000.
[24] Di Castelnuovo A, et al. Association of D-dimer levels with all-cause mor-
tality in a healthy adult population: ndings from the MOLI-SANI study.
Haematologica 2013;98(9):147680.
[25] Amirkhosravi A, et al. Blood clotting activation analysis for preoperative
differentiation of benign versus malignant ovarian masses. Blood Coagul
Fibrinolysis 2013;24(5):5107.