INTRODUCTION

Venous thromboembolism (VTE) is one of the leading causes of death in pregnancy and the peripartum period in the developed world.^1,2 Overall, 75% to 80% of pregnancy-related thrombotic events are venous and only 20% to 25% are arterial.³ Though undiagnosed VTE can cause significant morbidity and mortality, there are safe and effective ways of diagnosing and treating VTE. All clinicians looking after women in pregnancy or in the months following delivery must therefore be aware of the clinical features of VTE, and appropriate diagnostic and treatment strategies.

INCIDENCE

Venous thromboembolism is found in approximately 0.2% of pregnancies.⁴ This is evenly divided between women with deep vein thromboses (DVT) and pulmonary emboli. Overall there is a fivefold increased incidence of VTE during pregnancy compared to the general population. The risk of thromboembolism increases throughout pregnancy, but the highest risk is in the 3 months post-partum; up to 50% of cases of thromboembolism are in the 6 weeks post-partum, where the risk is 60-times greater than in the general population. This increased risk is present following all deliveries, including term births, first-trimester miscarriages, and terminations of pregnancy.^1,4,5,6

Arterial thromboembolism is rare in the child-bearing age group but may occur in the context of antiphospholipid syndrome (APLS) or cerebral vasculitides, e.g., Takayasu’s, Kawasaki’s disease. Arterial thrombosis may occur in peripheral vessels, coronary, or cerebral vessels. The risk of myocardial infarction in pregnancy is increased three- to four-fold compared to non-pregnant women.⁷ Most recently, severe Covid-19 infection in pregnancy has been associated with both venous and arterial thromboembolism. A Dutch study of 184 patients with COVID-19 pneumonia found that 31% had venous or arterial thromboembolism [acute pulmonary embolism (PE), ischemic stroke, deep vein thrombosis, or myocardial infarction].⁸

PATHOPHYSIOLOGY

Virchow’s triad describes the three components that lead to clot formation: hypercoagulability, venous stasis, and endothelial damage. Pregnancy affects all three of these areas.

Hypercoagulability

There is a steady increase in production of coagulation factors VII, VIII, and IX as well as fibrinogen through pregnancy due to increased estrogen production, while levels of endogenous anticoagulant peptides such as protein S fall. Thrombi that form also dissolve slower, as there is increased plasminogen activator inhibitor type 1 and 2 activity and reduced tissue plasminogen activator activity.^9,10 These changes are likely an evolutionary response to limit hemorrhage in the peri- and post-partum periods.

Venous stasis

Venous stasis in the lower limbs occurs due to compression of the lower limb veins by the gravid uterus, plus progesterone-induced venodilation. Deep vein thromboses are more likely to occur in the left leg than the right in pregnancy (compared to equal distribution in the non-pregnant state) as the gravid uterus causes the right iliac artery to press on the left iliac vein.^12,13 The gravid uterus also increases the incidence of isolated pelvic vein thrombosis compared to the non-pregnant population.^14,15

Endothelial dysfunction

In pregnancy there are changes in the vascular endothelium such as increased microparticle production, which may cause a procoagulant state.¹⁴ There can also be direct injury to the pelvic veins during delivery, particularly in instrumental vaginal or cesarean delivery.¹⁵

RISK FACTORS (SEE CHAPTER: THROMBOEMBOLISM AND THROMBOPHILIAS IN PREGNANCY)

Risk factors for venous thromboembolism include both pre-pregnancy features and factors that develop during pregnancy.^{16,17,17,18,19,20} These include the following:

Pre-pregnancy factors.

• Obesity.
• Heritable thrombophilia.
• Assisted-reproductive technology.
• Medical co-morbidities such as inflammatory bowel disease, particularly if active inflammation.
• Smoking.
• Personal or family history of venous thromboembolism.
• May–Thurner anatomy – when the left iliac vein is compressed by the right iliac artery.

Pregnancy-specific factors.

• Gestation (increased risk in third trimester).
• Post-partum period.
• Hospitalization during pregnancy.
• Pre-eclampsia.
• Stillbirth.
• Multi-fetal pregnancies.
• Post-partum hemorrhage.
• Pre-term delivery.
• Cesarean section, particularly emergency cesarean section.
• Hyperemesis.
• Infection.

Risk factors for arterial thrombus.

• Medical conditions such as Takyasu’s vasculitis, peripartum cardiomyopathy, and anti-phospholipid syndrome.
• Paradoxical emboli.
• Assisted-reproductive technology.
• Heritable thrombophilias.
• Essential thrombocythemia.

CLINICAL FEATURES

The clinical features of venous thromboembolism are not altered by pregnancy but it can be difficult in some cases to distinguish the symptoms of VTE from the normal physiological changes associated with pregnancy, particularly in the third trimester.

Deep vein thrombosis

Deep vein thromboses of the lower limbs typically present with leg swelling, pain, and tenderness, which may affect the calf or the whole limb. These features can also be seen in infection, lymphoedema, and in normal pregnancy. Chan et al. found that women with calf circumference difference of at least 2 cm were 18 times more likely to receive a diagnosis of DVT in the symptomatic leg.²¹ Deep vein thromboses are more commonly seen in the left than the right in pregnancy.¹³ Isolated pelvic vein thromboses are also more common in pregnancy than outwith pregnancy; they can often be an incidental finding during a routine pregnancy ultrasound, but may present with abdominal pain, fever, nausea, and vomiting, and are an important differential in a pregnant woman with these symptoms and no other cause found.^22,23 Cerebral sinus and venous thrombosis is a rare site of DVT in pregnancy and is discussed further in the chapter on Stroke in pregnancy.

Pulmonary embolism

The presentation of pulmonary embolism is the same both in and out of pregnancy. The most common symptoms of pulmonary embolism are dyspnoea, pleuritic chest pain, hemoptysis, pre-syncope, and syncope.²⁴ Sinus tachycardia is often the only sign on clinical examination, though pulmonary embolism can also be associated with complete circulatory collapse. There may be desaturation, particularly on exertion, but this is not always a feature of pulmonary emboli. ECG changes may be present but have low sensitivity. They include sinus tachycardia, S1Q3T3, p-pulmonale, right-axis deviation, as well as right-bundle branch block and atrial dysrhythmias, though commonly ECGs show no abnormalities.²⁵

Unfortunately none of these signs or symptoms are specific to pulmonary embolism, and do overlap with features of a normal pregnancy. Given, however, the serious impact a PE may have, it is perhaps unsurprising that many women undergo further investigations to look for PE, and yet only 2–7% of those investigated are confirmed to have a venous thromboembolus.^2,26,27

INVESTIGATIONS

Those women who are clinically stable with only a suspected deep vein thrombosis and no other obvious cause for symptoms should, after clinical review, undergo Doppler imaging (discussed below). However, those women with suspected pulmonary embolism should also have baseline investigations to exclude other potential causes of their symptoms. Depending on local availability, these should include the following:

• Arterial blood gas analysis, if SpO₂ is <94%.
• Full blood count.
• Renal function.
• Electrolytes.
• Chest radiograph.

Unlike in the non-pregnant state, measuring D-dimer levels is unhelpful to distinguish women who may avoid pulmonary imaging. Average D-dimer levels increase through pregnancy, but cases of venous thromboembolism have been found even with low D-dimer levels, rendering the test inappropriate for use in pregnancy.^10,28

Thrombophilia testing is not recommended as part of the work up for a woman with VTE in pregnancy. The acute management of newly diagnosed VTE is the same regardless of the presence or absence of a recognized inherited thrombophilia, but their presence may impact on the risk of recurrence and recommendations regarding duration of anticoagulation. This is discussed further in the chapter on Prevention of thromboembolic disease and thrombophilias in pregnancy.

IMAGING

Lower-limb imaging

Duplex ultrasonography of the lower limb is recommended as the first-line investigation for deep vein thromboses by several national societies.^11,29,30,31 In those women where the initial duplex is negative but the clinical suspicion of deep vein thrombosis is high, then no treatment should be given but imaging should be repeated in a week. If an iliac vein thrombosis is suspected but the initial duplex is negative then further imaging with magnetic resonance venography (MRV) scanning is appropriate.

In cases of suspected pulmonary emboli, most international societies also suggest a duplex ultrasound of the lower limbs as the first-line investigation, and if a deep vein thrombosis is found then treatment of the venous thromboembolism can be started without further imaging.^11,29,30,31 However, women with no clinical features of deep vein thrombosis rarely have a duplex scan prior to imaging of their chest, and duplex is rarely positive in those cases with no clinical features of DVT.^32,33 Therefore, specific imaging of the chest is required in most cases of suspected pulmonary embolism.

Chest imaging (see chapter on Radiation in pregnancy)

Both ventilation/perfusion (V/Q) scanning and computerized tomography of pulmonary arteries (CTPA) can be used to diagnose pulmonary emboli.

In the non-pregnant population, CTPA scans are in general preferred to VQ scans. CT scans are more readily available and convenient for the patient, as well as having slightly better sensitivity and specificity: CTPA has an estimated sensitivity and specificity of 80–100% and 78–100%, respectively, whilst VQ scanning has estimated sensitivity of 41–100% and specificity of 72–97%.^34,35

In pregnancy, however, additional consideration has to be given to radiation exposure to the fetus and to the maternal breast, as the changes in breast tissue may increase the risk of breast cancer in comparison to the non-pregnant state. In both cases, fetal exposure is very low and the increased risk of childhood cancer is negligible.^36,37

Ventilation/perfusion scan

Ventilation/perfusion scanning consists of an inhalation of a radio-isotope such as Tc-99 m DTPA (ventilation scan) plus injection of a radioisotope such as Tc-99 microaggregated albumin (perfusion scan). Fetal exposure with VQ scanning is 0.02–0.9 mGy, whilst maternal breast tissue exposure is 0.16–1.2 mGy. After V/Q scanning, there is considerable variability in milk radioactivity, and close contact with an infant may result in additional exposure. A conservative recommendation suggest expressing and storing breast milk for at least 12 h after which it may be used safely.^38,39

Computerized tomography of pulmonary arteries

CTPA scans are increasingly readily available in developed countries, which makes them appealing in many busy healthcare settings. Although the fetal radiation exposure is lower with CTPA than VQ scans (0.05–0.5 mGy), breast exposure is higher, with average exposures on newer CT scans of 3–10 mGy. In the non-pregnant woman, previous studies suggested a single exposure to 20 mGy radiation increased the risk of breast cancer by 1/2000, on a background rate where approximately 1 : 8 women develop breast cancer in their lifetime; however in the pregnant woman this is assumed to be increased, and lifetime rates of up to 13% have been predicted.^29,36 VQ scans had previously been recommended as first-line investigation for pulmonary emboli in pregnancy because of the breast cancer risk.^11,29,30 However, as modern techniques reduce radiation exposure, and the expected life-time risk is only increased by 1.0003–1.0007, both VQ scans and CTPA scans can be considered as first-line imaging for pulmonary emboli, and CTPA should not be avoided because of risks of maternal breast cancer.³¹ A recent study of almost 6000 women exposed to CTPA in pregnancy found no increased rate of breast cancer in the short term (5 years) compared to the background population.⁴⁰

Echocardiography

Acute pulmonary emboli may cause right ventricular pressure overload, which can be detected on echocardiography. With the increasing prevalence of bed-side echocardiography in many emergency departments, this can be a useful tool in the diagnosis of acute pulmonary emboli before even a CTPA is available. Echocardiography is not sufficient to confirm a diagnosis of pulmonary embolism; however, it can also help eliminate other diagnoses in the acute setting, such as acute cardiac valvular dysfunction.

CLINICAL DECISION TOOLS¹¹

Given the non-specific nature of the signs and symptoms of venous thromboembolism, many clinical decision tools have been developed to calculate the pre-test probability of a VTE, and thus limit the number of patients who undergo radiological investigations for VTE. These tests include the Wells and modified-Wells score, the Geneva score, the PERC criteria, and the YEARS score.^{41,42,43,44,45} Versions of these scoring systems modified to be appropriate for pregnancy have been suggested, and indeed the European Society for Cardiology suggests that D-dimer measurement and clinical prediction rules should be considered to rule out PE during pregnancy or the post-partum period. However, none have been consistently validated across different pregnant populations, and therefore at the present time none of these rules should be used to omit imaging in woman with suspected venous thromboembolism.^{46,28,31,47,48}

TREATMENT

The mainstay of treatment for women with venous thromboembolism who are hemodynamically stable is anticoagulation with heparin, most commonly with low molecular weight formulations. Unlike coumarins (warfarin) and the newer direct oral anticoagulants, heparin does not cross the placenta. Therefore, there is no risk of fetal hemorrhage or teratogenicity. All women started on anticoagulation should have a baseline platelet level measured.

Low molecular weight heparin

Low molecular weight heparins are the most commonly used anticoagulant in pregnancy. Treatment-dose LMWH is weight dependent, and should be dosed based on a woman’s pre-pregnancy or current weight if the former is not known. A variety of different low molecular weight heparins are available, and local guidelines should be followed as to the type and dose of LMWH; however, suggested regimens are listed below. Therapeutic dose LWMH can be given as either a single daily dose or split in two daily doses, depending on local guidelines and which LMWH is used.⁴⁹ Once-daily versus twice-daily dosing with enoxaparin, nadroparin, or dalteparin has not been compared directly but observational studies in pregnant women with acute PE have not demonstrated any increase in the risk of recurrence with a once-daily regimen compared with twice-daily schedules.⁵⁰

Most women who have been diagnosed with deep vein thromboses can be managed as outpatients. Women with newly diagnosed pulmonary emboli can be managed as outpatients if they are hemodynamically stable and after review by and discussion with senior clinicians. However, they will require regular review in the first few weeks, so for some women it may be more appropriate for them to be treated in hospital initially. The physiological changes of pregnancy (such as a fall in blood pressure and a physiological tachycardia) mean that clinical risk scores derived from the non-pregnant population, such as PESI score, should not be used in pregnancy.⁴⁹

Therapeutic LMWH should be continued for the remainder of the pregnancy plus 6 weeks post-partum, and until at least 3 months of treatment have been given.²⁹ LMWH and warfarin can be used in the post-partum period and in breastfeeding; the DOACs are found in breast milk, and though no adverse effects have been found in the infants, they are avoided in breastfeeding mothers.⁵¹

Anti Xa levels

The anti-Xa assay directly measures the efficacy of LMWH at inhibiting the clotting factor Xa, and therefore at interrupting the clotting cascade. It is therefore a useful measure of efficacy of anticoagulation and adherence with therapy. Anti-Xa levels do not need to be routinely measured in all women on therapeutic LMWH; however, they are recommended in women at the extremes of body weight, those with renal impairment and those with recurrent venous thromboembolism. Clinicians should aim for anti-Xa levels of 0.6–1.0 units/ml 4 h after last injection in twice-daily regimens. Slightly higher target levels are used for those on a once-daily regimen.^11,29

Low molecular weight heparin around the time of delivery

There is a risk of bleeding and spinal epidural hematoma when using regional anesthesia in women receiving low molecular weight heparin. The Society for Obstetric Anesthesia and Perinatology Consensus statement suggests there should be a delay of at least 24 h between therapeutic/high-dose LWMH administration and neuraxial administration.

All obstetric units should have local guidelines and the woman informed about if and when she should pause LMWH administration in the peripartum period. All women should also be counseled that, if an operative or instrumental delivery is required but insufficient time has passed since her last LMWH dose, she may require a general anesthetic.

Post-partum, therapeutic LMWH can be recommenced at least 24 h after neuraxial anesthetic procedure and at least 4 h after spinal catheter removal.⁵² Prophylactic dose LMWH can be restarted at 4 h after vaginal birth and 6–12 h after cesarean birth, providing there is full neurological recovery and no evidence of active bleeding or coagulopathy. The decision to recommence therapeutic or intermediate dose anticoagulation should be made by a senior obstetrician and obstetric anesthetist. The risk of bleeding is greater with intermediate and therapeutic than lower-dose LMWH.

Unfractionated heparin

Unfractionated heparin (UFH) is as safe for the fetus as low molecular weight heparin. It has a much shorter half-life of only 1–2 h, which makes it ideal in cases where rapid reversal of anticoagulation is required – for example, when delivery, surgery, or reperfusion therapy due to hemodynamic instability may be required. However, achieving therapeutic levels requires regular monitoring of aPTT and therefore UFH is only suitable in the inpatient setting.

Warfarin, direct thrombin inhibitors, factor Xa inhibitors and the direct oral anticoagulants

Warfarin is a vitamin K antagonist, which was previously the mainstay of long-term anticoagulation outside of pregnancy. However, use in the first trimester is associated with warfarin embryopathy, a syndrome which includes facial dysmorphism, epiphyseal stipling, choanal atresia, central nervous system, and cardiac abnormalities.⁵³ There is also an increased risk of hemorrhage in women on warfarin around the time of delivery. Therefore, warfarin is not recommended for women with venous thromboembolism in pregnancy.^11,29,30

In recent years, direct oral anticoagulants (DOACs), such as rivaroxaban, apixaban, and the oral thrombin inhibitor dabigatran, have been increasingly used in the management of VTE in the non-pregnant population. However, ex vivo studies have shown these drugs cross the placenta,^54,55 and systematic reviews have shown the DOACs to be associated with miscarriage and congenital abnormalities.^55,56 None of these drugs are recommended for use in pregnancy.

Fondaparinux is a synthetic factor Xa inhibitor. This does not cross the placenta, and appears to be tolerated well in pregnancy, with no clear link to increased congenital malformations.^57,58,59 Fondaparinux is therefore the preferred option for anticoagulation in women who cannot tolerate LMWH because of the development of allergic skin reactions (approximately 2%) or heparin-induced thrombocytopenia in pregnancy.

MANAGEMENT OF WOMEN WHO ARE HEMODYNAMICALLY COMPROMISED BY VENOUS THROMBOEMBOLISM

Venous thromboembolism causes an estimated 10–30 deaths per 100,000 in the general global population.⁶⁰ In the UK in 2015–2017, 16% of all maternal deaths were due to venous thromboembolism.⁶¹ Right ventricular failure due to pressure overload is the presumed cause of death in most cases.³¹ Deaths occur because the pulmonary embolism causes significant cardiovascular instability, and there is often a very rapid deterioration in the woman’s clinical state. Features include tachycardia, significant hypotension, end-organ hypoperfusion with altered mental status, clammy skin, oliguria and increased serum lactate, and cardiopulmonary arrest.

The first treatment for women with hemodynamic instability due to massive pulmonary emboli is oxygen therapy, through high-flow oxygen, non-invasive ventilation, and if required intubation and mechanical ventilation. Cardiovascular support through cautious use of crystalloid fluids, vasopressors, and inotropes may also be necessary. In severe cases, ECMO or extracorporeal life support may be appropriate.³¹ However, these treatments are usually only to support the woman while she undergoes reperfusion therapy to eliminate the clots.

Reperfusion strategies that can be used in pregnancy include systemic thrombolysis, percutaneous catheter-directed thrombolysis, and surgical embolectomy. Systemic thrombolysis is the most commonly used treatment for massive/high-risk PE outside of pregnancy, and is associated with improved mortality but at the cost of increased hemorrhage.⁶² Though there is a paucity of data on deaths from pulmonary embolism, one meta-analysis showed maternal survival for women in pregnancy and the post-partum with massive/high-risk PEs was good – 92% survival in those who underwent thrombolysis (both systemic and catheter directed), and 85% in those who had surgical embolectomy.⁶³In total, 28% of women who underwent thrombolysis had a major bleed, including one fatality due to intra-cranial hemorrhage; 20% of women who had thrombectomy also had a major bleeding episode. Bleeding was more common if the diagnosis and treatment occurred in the post-partum period compared to antenatally (58% and 18%, respectively). Fetal outcomes varied, with 88% and 80% survival in cases of non-fatal PE treated with thrombolysis and thrombectomy, respectively.⁶³ Thrombolysis post-partum can result in uncontrolled bleeding and should be reserved for life-threatening events.

Decisions for management of all pregnant/post-partum women who have a massive/high/intermediate risk of pulmonary embolism should be taken by a multidisciplinary team involving staff experienced in the management of pulmonary embolism in pregnancy, such as obstetricians, physicians, intensivists, and radiologists.³⁶ Catheter-directed thrombolysis (CDT) if available allows lower doses than systemic thrombolysis as the tPA is delivered directly to the clot and therefore a lower bleeding risk. While awaiting decisions regarding lysis, an echocardiogram, CTPA, and troponin measurement should be performed and intravenous unfractionated heparin commenced.

INFERIOR VENA CAVA FILTER

Inferior vena cava (IVC) filters are netting or mesh devices, which are placed radiologically in the inferior vena cava. They can be permanent or removable. Indications for their use include a contraindication to anticoagulation, complications necessitating cessation of anticoagulant use, and failure of anticoagulation to prevent further venous thromboembolism.⁶⁴ Typically removable IVC filters are used in pregnancy. Indications for their use are similar to those out of pregnancy – and most commonly they are used in those with recurrent venous thromboembolism despite anticoagulation, and in those with iliac vein thromboses to prevent the occurrence of pulmonary emboli.²⁹

Management of pregnancy in women with a personal or family history of thromboembolism is discussed in the chapter on Prevention of thromboembolic disease and thrombophilias in pregnancy.

CONCLUSIONS

Venous thromboemboli occur more commonly in pregnancy and are a potential cause of significant maternal morbidity and mortality. Venous thromboembolism presents similarly in and out of pregnancy, though clinical decision rules and D-dimer should not be used in pregnancy. Anticoagulation with low molecular weight heparin is the mainstay of treatment, but in those women who are hemodynamically unstable following massive/high-risk/intermediate-risk pulmonary emboli, all available techniques, including thrombolysis and thrombectomy, should be deployed to help with maternal survival.

PRACTICE RECOMMENDATIONS