Keywords:
anticoagulation, cancer, deep venous thrombosis, malignancy, pulmonary embolism, thromboprophylaxis
Abbreviation list
CT: computed tomography
DVT: deep venous thrombosis
PE: pulmonary embolism
RAM: risk assessment models
VTE: venous thromboembolism
Take-home messages
Patient-oriented messages
Though cancer is associated with increased incidence of venous thromboembolism (VTE; including deep vein thrombosis and pulmonary embolism), awareness of symptoms, accurate and timely diagnosis and appropriate therapy prevents poor outcomes. Moreover, screening of asymptomatic patients for the identification of patients at high risk for future VTE is another measure for prevention.
Impact on Practice statement
The increased awareness of venous thromboembolism (VTE) in cancer patients along with the ever-increasing availability of non-invasive imaging tests will hopefully encourage clinicians to suspect and initiate a diagnostic workup for DVT and/or pulmonary embolism more frequently than in the past. Moreover, high-risk patients for future VTE may be identified by dedicated risk assessment models (RAMs) and benefit from receiving anticoagulant therapy for primary prevention.
Introduction
Thromboembolic events that occur during cancer and their therapies encompass venous thromboembolism (VTE) and arterial events and are referred to as cancer-associated thrombosis [1, 2]. Venous thromboembolism, including deep vein thrombosis (DVT) and pulmonary embolism (PE) can occur at any time during the course of cancer or even be the first presenting sign of the disease [2-4]. Despite substantial improvements in cancer treatment, the risk of VTE has increased in recent years and is the second-leading cause of death in patients with malignancies [5]. Furthermore, VTE can complicate the management of patients with active cancer (surgery, hospitalisation, and systemic therapy), lead to increased, health-care costs, and have an impact on the psychologic burden of patients [1].
Pathogenesis of VTE in cancer patients
Historically, the association between cancer and a clinical hypercoagulable state was first described in 1823 by the French physician Jean-Baptiste Bouillaud. In 1865, another French physician, Armand Trousseau, further highlighted this association [6]. Mechanisms promoting VTE in cancer patients have been extensively studied ever since. Cancer-related thrombosis seems to be determined by the cancer itself (type, stage, time since diagnosis), anticancer therapy and patient-related risk factors (including demographics, genetic predisposition, and comorbidities) [1, 7].
In a classical sense, cancer patients are generally in a hypercoagulable or prothrombotic state, due to cancer’s ability to affect all components of Virchow’s triad. The famous Virchow’s triad that was formulated to explain the pathogenesis of VTE, includes stasis of venous flow, endothelial injury, and hypercoagulability [6]. Disruption of each component of Virchow’s triad by the factors associated with malignancy leads to the thrombophilia observed in these patients (Central Illustration). Potential factors that increase the risk of VTE in patients with cancer are:
- Stasis of blood flow due to mechanical compression of malignant tumours as well as reduced performance status and patient immobility (>3 days) post-surgery.
- Endothelial injury due to cancer therapy and placement of central venous catheters (commonly used to administer cancer therapy).
- Hypercoagulability itself due to the release of coagulant factors by cancer cells (tissue factor and inflammatory cytokines), which affect haemostasis, platelet functions and clotting cascade.
CENTRAL ILLUSTRATION. Factors Contributing to Increased Thrombotic Risks in Cancer.
From a modern perspective, the aetiology of dysregulated haemostasis in cancer has been linked to other contributing factors such changes due to tumour biology, extent of disease, coagulation activation, local and systemic inflammation, cancer therapeutics, and patient-related factors including genetics.
Prediction of VTE in cancer patients (primary thromboprophylaxis)
Recent European Society of Cardiology (ESC) guidelines recommend primary thromboprophylaxis for hospitalised with active cancer throughout hospitalisation and post-surgery patients (Class I, Level of Evidence [LoE] B) [1]. However, primary thromboprophylaxis in the outpatient setting, may only be considered for high-risk patients provided there are no contraindications (Class IIb, LoE B) [1]. High-risk patients may be identified by dedicated risk assessment models (RAMs) that combine the additive predictive power of clinical risk variables and increase the ability to stratify cancer patients according to their risk for future VTE events.
Based on current guidelines, VTE risk should be periodically assessed and individually determined based on proposed validated scores such as the Khorana or the COMPASS-CAT (prospective COmparison of Methods for thromboembolic risk assessment with clinical Perceptions and AwareneSS in real-life patients—Cancer Associated Thrombosis) score [1, 8, 9]. Over the last decades, additional clinical RAMs have emerged, which vary with respect to variables included, predictive performance, and current stage of clinical validation (Table 1) [10].
Table 1. Comparison of risk assessment models.
| Score | Variables | Points | Risk categories |
|---|---|---|---|
| Khorana |
Pancreatic or gastric cancer (very high-risk tumours) Lung, gynaecologic, lymphoma, bladder, or testicular (high-risk tumours) Pre-chemotherapy Hb of <10 g/dl or erythropoietin-stimulating agents Pre-chemotherapy white blood cell count of >1 x 109/l Pre-chemotherapy platelet count of >350 x 109/l Body mass index of >35 kg/m2 |
+2 +1 +1 +1 +1 +1 |
≥3 high risk of VTE |
| Vienna CATS |
Pancreatic or gastric cancer (very high-risk tumours) Lung, gynaecologic, lymphoma, bladder, or testicular (high-risk tumours) Pre-chemotherapy Hb of <10 g/dl or erythropoietin-stimulating agents Pre-chemotherapy white blood cell count of >1 x 109/l Pre-chemotherapy platelet count of >350 x 109/l Body mass index of >35 kg/m2 D-dimer of >1.44 mg/l Soluble P-selectin of >53.1 ng/l |
+2 +1 +1 +1 +1 +1 +1 +1 |
≥5 high risk of VTE |
| PROTECHT |
Pancreatic or gastric cancer (very high-risk tumours) Lung, gynecologic, lymphoma, bladder, or testicular (high-risk tumours) Pre-chemotherapy Hb of <10 g/dl or erythropoietin-stimulating agents Pre-chemotherapy white blood cell count of >1 x 109/l Pre-chemotherapy platelet count of >350 x 109/l Body mass index of >35 kg/m2 Platinum-based chemotherapy Gemcitabine chemotherapy |
+2 +1 +1 +1 +1 +1 +1 +1 |
≥3 high risk of VTE |
| CONKO |
Pancreatic or gastric cancer (very high-risk tumours) Lung, gynaecologic, lymphoma, bladder, or testicular (high-risk tumours) Pre-chemotherapy Hb of <10 g/dl or erythropoietin-stimulating agents Pre-chemotherapy white blood cell count of >1 x 109/l Pre-chemotherapy platelet count of >350 x 109/l WHO performance status ≥2 |
+2 +1 +1 +1 +1 +1 |
≥3 high risk of VTE |
| ONKOTEV |
Khorana score >2 Metastatic disease Previous VTE Vascular/lymphatic macroscopic compression |
+1 +1 +1 +1 |
≥2 high risk of VTE |
| COMPASS-CAT |
Anti-hormonal or anthracycline therapy Time since cancer diagnosis ≤6 months Central venous catheter Advanced stage of cancer Cardiovascular risk factors Recent hospitalisation Personal history of VTE Platelet count ≥350 × 109 /L |
+6 +4 +3 +2 +5 +5 +1 +2 |
≥7 high risk of VTE |
| TiC-Onco |
Pancreatic or gastric cancer (very-high-risk tumours) Lung, gynaecologic, lymphoma, bladder, or testicular (high-risk tumours) Pre-chemotherapy Hb of <10 g/dl or erythropoietin-stimulating agents Pre-chemotherapy white blood cell count of >1 x 109/l Pre-chemotherapy platelet count of >350 x 109/l Body mass index of >35 kg/m2 |
+2 +1 +1 +1 +1 +1 |
≥3 high risk of VTE |
Hb: haemoglobin; VTE: venous thromboembolism; WHO: World Health Organisation
The Khorana score was the first risk prediction model for VTE in ambulatory cancer patients prior to the initiation of chemotherapy [8]. It is a simple, point-based risk score that relies on 5 variables (type of cancer, complete blood count components [haemoglobin, platelet, and white blood cells], and body mass index). Each variable is assigned 1 point, except for the variable of very high-risk tumours, which counts for 2. Although the Khorana score is widely recommended by current guidelines, a high proportion of VTE events seem to occur in patients categorised as low/intermediate risk (0–1 points). This limitation seems to be linked to its modest sensitivity and the inconsistent performance across cancer types.
Several prediction models were designed subsequently as modifications to the Khorana score with a view to improve its discriminatory performance (Vienna CAT, PROTECHT-, CONKO-, ONKOTEV). The Vienna CAT score adds D-dimer and soluble P selectin measurements to the aforementioned 5 variables, improving the positive predictive value, though it has not yet been validated externally [11]. In the PROTECHT score, platinum-based and/or gemcitabine chemotherapy (each adding +1 point) are included as predictive variables on top of items of the Khorana score [12]. In the CONKO-score, the risk items of the Khorana score are used, only substituting BMI with the World Health Organization (WHO) performance status (+1 point for a WHO performance status of ≥2) [13]. Finally, the ONKOTEV score allocates 1 point each to patients with a Khorana score ≥3, metastatic disease, macroscopic compression of vascular or lymphatic structures by the tumour, and a prior history of VTE [14]. Taken together, these scores showed only limited benefit in comparison to the original Khorana score.
In 2017, the COMPASS-CAT score was proposed as novel clinical RAM. The variables included in the score comprise various clinical characteristics and comorbidities, treatment types, and platelet counts. External validation in a large cohort of ambulatory cancer patients indicated moderate model discrimination [9]. In 2018, the Tic-Onco risk score, where clinical variables are incorporated in addition to a genetic risk score, was proposed [15]. This model achieved superior discriminatory ability compared to the Khorana score, however, no external validation of the score is available to date. Finally, novel models have been developed based on tumour type with a view to improve the identification of ambulatory patients at high-risk for VTE thrombosis (Table 2) [10].
Table 2. Tumour type specific risk assessment models.
| Cancer type | Variables | Points | Risk Categories | |
| ROADMAP-CAT | Lung Adenocarcinoma |
Mean rate index of thrombin generation <125 nM/min Procoagulant phospholipid-dependent clotting time <44 s |
1 1 |
- |
| THROMBOGYN | Gynaecologic cancer |
Haemoglobin <11.5 BMI >30 |
1 1 |
≥3 high risk of VTE |
| THROLY | Lymphoma |
Previous VTE and/or arterial events ECOG 2-4 BMI >30 Extra-nodal localisation Mediastinum involvement Development of neutropenia during therapy Haemoglobin <10 g/dL |
2 1 2 1 2 1 1 |
≥3 high risk of VTE |
| SAVED | Multiple myeloma |
Prior surgery Asian ethnicity VTE history Age ≥80 years Dexamethasone standard dose (120-160 mg) Dexamethasone high dose (>160 mg) |
2 -3 3 1 1 2 |
≥2 high risk of VTE |
| IMPEDE VTE | Multiple myeloma |
Immunomodulatory agent BMI ≥25 kg/m2 Pelvic, hip or femur fracture Erythropoietin stimulating agent Doxorubicin Dexamethasone low-dose Dexamethasone high-dose Asian ethnicity VTE history Central venous catheter Existing thromboprophylaxis (therapeutic dose) Existing thromboprophylaxis (prophylactic dose) |
4 1 4 1 3 2 4 -3 5 2 -4 -3 |
≥8 high risk of VTE |
BMI: body mass index; ECOG: Eastern Cooperative Oncology Group score; VTE: venous thromboembolism;
Diagnosis of VTE in cancer patients (secondary prevention)
Thromboembolic events can either occur in patients with a known history of malignancy or can be the first manifestation of the disease [1, 2]. Accurate and timely diagnosis of cancer-associated VTE is particularly important, taking into consideration the poor outcomes observed in patients not receiving appropriate anticoagulant therapy. Fortunately, the increased awareness of VTE in cancer patients and the ever-increasing availability of non-invasive imaging tests have generated a tendency for clinicians to suspect and initiate a diagnostic workup for DVT and/or PE more frequently than in the past.
Taking a thorough history and performing a complete physical exam is essential. Patients with symptoms or signs suggestive of VTE, should be screened according to the 2019 ESC Guidelines for the diagnosis of PE and the second consensus document on diagnosis and management of acute DVT. Diagnosis of symptomatic patients requires clinical assessment, evaluation of pre-test probability, and imaging testing (for DVT and/or PE). Still, the proposed validated diagnostic pathways in the guidelines involving clinical pre-test probability assessment tools followed by laboratory testing (D-dimer assay) and/or diagnostic imaging may not be as accurate in the cancer patient population.
Pulmonary embolism
Clinical presentation
The clinical signs and symptoms of acute PE in cancer patients may be non-specific [1, 2, 4]. Moreover, classic clinical symptoms of PE (dyspnoea, chest pain, presyncope, syncope, hemodynamic instability and/or haemoptysis) quite often mimic symptoms of cancer itself in oncologic patients.
Pulmonary embolism may present with a wide spectrum of symptoms and signs ranging from pleuritic pain or dyspnoea to haemodynamic instability and/or cardiac arrest. Clinical presentation is strongly influenced by the extent of PE, the pulmonary pressure increment and eventually, the right ventricular systolic dysfunction due to right ventricular overload. In some cases, however, PE may be asymptomatic or found incidentally. Hypoxaemia and hypocapnia are frequent, but not always present. A chest X-ray is frequently abnormal and, although its findings are usually non-specific in PE, it may be useful for excluding other causes of dyspnoea or chest pain. Electrocardiographic changes indicative of right ventricular strain (incomplete or complete right bundle branch block, T wave inversion in leads V1-V4, QR pattern in V1, and/or S1Q3T3 pattern) are often common in more critical cases of PE.
Assessment of clinical (pre-test) probability
In addition to symptoms, knowledge of the predisposing factors for VTE is important in determining the clinical probability. The combination of symptoms and clinical findings with the presence of predisposing factors for VTE allows patients with suspected PE to be classified into distinct categories of clinical or pre-test probability, which correspond to an increasing actual prevalence of confirmed PE. This pre-test assessment can be done either by implicit (empirical) clinical judgement or by using prediction rules. The value of clinical judgement has been confirmed in several large series but lacks standardisation. As a consequence, current ESC guidelines suggest the use of clinical prediction rules. Among others, the most frequently used are the revised Geneva score and the Wells rule. A score of ≤4.0 by Wells criteria makes PE unlikely (PE risk <5%). Although cancer itself is included as a risk factor in both prediction models, it is unclear whether this scoring system is valid in cancer patients because only a minority of patients with cancer were included in models of this score. Thus, the diagnosis of acute PE remains still challenging.
D-dimer testing
D-dimer levels are elevated in the presence of acute thrombosis because of simultaneous activation of coagulation and fibrinolysis. The negative predictive value of D-dimer testing is high, and a normal D-dimer level renders acute PE unlikely. On the other hand, the positive predictive value of elevated D-dimer levels is low and D-dimer testing is not useful for confirmation of PE. In cancer patients, D-dimer levels may be elevated in the absence of PE thereby lowering its specificity and utility. Indeed, because >50% of cancer patients have elevated D-dimer levels, the D-dimer test has lower specificity in this group of patients compared with the general population. Consequently, the utility of D-dimer testing resides in its ability to exclude PE when the levels are normal.
Diagnostic imaging
The use of current diagnostic algorithms followed by D-dimer testing and/or diagnostic imaging) validated for non-cancer patients seem not to be as safe and accurate in cancer patients. In most cancer patients with suspected PE, it seems rational to proceed directly to diagnostic imaging to confirm or exclude PE.
The diagnostic imaging tests for suspected PE may vary, depending on the availability of and expertise in various hospitals and clinical settings. Current ESC guidelines recommend screening of patients with symptoms or signs suggestive of PE with computed tomographic pulmonary angiography (CTPA). Among other imaging tests for the diagnosis of PE (ventilation/perfusion scintigraphy, pulmonary angiography, magnetic resonance angiography) CTPA is the predominant modality for imaging the pulmonary vasculature in patients with suspected PE. It allows adequate visualisation of the pulmonary arteries down to the subsegmental level. The available data suggest that a negative CTPA is an adequate exclusion criterion for PE in patients with low or intermediate clinical probability of PE. On the other hand, it remains controversial whether patients with a negative CTPA and a high clinical probability should be further investigated.
In particular, for patients presenting with haemodynamic instability, current ESC guidelines suggest bedside echocardiography. In cancer patients with haemodynamic instability, the differential diagnosis includes cardiac tamponade, acute heart failure decompensation or hypovolaemia. Transthoracic echocardiography will yield evidence of acute right ventricular dysfunction in the setting of PE with advanced right ventricular pressure overload. Typical echocardiographic findings among patients with acute PE include: 1) right ventricular dilation and depressed contractility of the right ventricular free wall compared to the apex (McConnell sign), 2) flattened intraventricular septum, 3) a distended inferior vena cava with diminished inspiratory collapsibility, 4) a 60/60 sign (coexistence of acceleration time of pulmonary ejection <60ms and a mid-systolic “notch” with mildly elevated (<60 mmHg) peak systolic gradient at the tricuspid), 5) right heart mobile thrombus detected in right heart cavities, 6) decreased tricuspid annular plane systolic excursion (TAPSE) by M-Mode (<16 mm) and 7) decreased peak systolic (S’) velocity of tricuspid annulus (<9.5 cm/s).
Ancillary bedside imaging testing includes transoesophageal echocardiography which may allow direct visualisation of thrombi in the pulmonary artery and its main branches and compression ultrasonography that may detect DVT. In a highly unstable patient, echocardiographic evidence of right ventricular dysfunction is sufficient to prompt immediate reperfusion without further testing especially in cases when there is visualisation of emboli in right heart chambers or/and pulmonary artery and its main branches.
The aforementioned echocardiographic findings have low sensitivity as standalone findings, as they were reported to be normal in haemodynamically stable patients despite the presence of PE. Although transthoracic echocardiography has poor sensitivity in diagnosis, it seems to be crucial for risk stratification in patients with proven acute PE. Thus, as soon as the patient is stabilised using supportive treatment, final confirmation of the diagnosis by CT angiography should be sought.
Deep vein thrombosis
Clinical presentation
Management of DVT has similarities to that of PE, however, many diagnostic features present particularities [2, 3, 7]. Clinical signs and symptoms of lower DVT remain the cornerstone of diagnostic strategy and include pain, swelling, increased skin veins visibility, erythema, and cyanosis accompanied by unexplained fever. Still, many cancer patients with VTE may show no signs and symptoms. More often, their signs might be masked by the underlying malignancy.
Assessment of clinical (pre-test) probability
Wells’ score has been widely validated and can be applied in non-cancer patients for DVT screening, however, it seems of limited efficiency in patients with cancer.
D-dimer testing
Normal D-dimers render DVT unlikely, however, D-dimers have low specificity even in non-cancer patients. Taken together, D-dimers allow ruling out DVT in patients with ‘unlikely’ DVT. However, in patients with ‘likely’ DVT, D-Dimer testing is not necessary and imaging is required.
Diagnostic imaging
Venous ultrasound is recommended as first-line imaging method for DVT diagnosis. It is useful for screening and to confirm/rule out DVT in both symptomatic and asymptomatic patients. Venous duplex ultrasound combines vein compression (B-mode imaging) and pulsed Doppler spectrum analysis with and without colour. Rarely, alternative imaging modalities may be considered for selected patients and may include computed tomography, magnetic resonance imaging, and contrast venography (when the evaluation by venous ultrasound is inconclusive).
Conclusions
Venous thromboembolism (VTE), including deep vein thrombosis and pulmonary embolism is the second-leading cause of death in patients with malignancies. Accurate and timely diagnosis according to current guidelines is particularly important for initiation of appropriate therapy (secondary thromboprophylaxis). Current guidelines also suggest the consideration of anticoagulant therapy (primary thromboprophylaxis) for patients identified as high-risk for VTE based on validated risk assessment models. In all cases, a multidisciplinary approach is recommended with a view to individualised treatment.
Our mission: To reduce the burden of cardiovascular disease.