The most common forms of venous thromboembolic (VTE) diseases encountered during pregnancy and the postpartum period include deep venous thrombosis (DVT), pulmonary embolism (PE), septic pelvic thrombophlebitis (SPT), and ovarian vein thrombosis (OVT). Thromboembolic disease is a major contributor to both perinatal and maternal morbidity and mortality worldwide. The World Health Organization (WHO) estimates that VTE accounted for 3.2% of maternal deaths between 2003 and 2012.¹ A large part of the strategy for reducing maternal morbidity and mortality from VTE lies in a systematic approach to thromboprophylaxis, and there is evidence that progress is being made.

Data from the 1990s suggested that VTE had surpassed hemorrhage and hypertension as leading causes of maternal mortality in developed countries.² In 2004, the Royal College of Obstetricians and Gynaecologists recommended thromboprophylaxis during pregnancy, labor, and after normal vaginal delivery for women in the United Kingdom (UK).³ The time period from 2006-2008 was the first time in the prior two decades that VTE was not the leading direct cause of death in the UK.⁴ In 2011, the American College of Obstetricians and Gynecologists (ACOG) updated the Practice Bulletin on Thromboembolism in Pregnancy to include recommendations for thromboprophylaxis for all patients undergoing cesarean delivery including the use of pneumatic compression devices or pharmacologic prophylaxis.⁵ In the United States, between 2011 and 2013, 9.2% of maternal mortality was attributable to thrombotic PE making it the sixth most common cause of maternal mortality.⁶

The most recent data from the National Inpatient Sample indicate that the rates of DVT and PE are currently 0.4 and 0.2 per 1000 delivery hospitalizations. Interestingly, since 1994, the rate of DVT among US women has decreased by about one third, while the rate of PE has doubled.⁷

The cornerstones of VTE management lie in ongoing systems-based strategies for prevention, a high index of suspicion, and prompt treatment. This chapter reviews the etiology, diagnosis, treatment, and prevention of VTE.

A practical and user-friendly version of the complex regulatory pathways of hemostasis and fibrinolysis is presented in Figure 7-1.

Pregnancy

Important physiologic changes to this system occur during pregnancy: the procoagulant Factors I (fibrinogen), VII, VIII, and X, Von Willeband factor, and plasminogen activator inhibitor-1 and -2 are increased, while the anticoagulant-free protein S is decreased.⁵ These changes begin with conception and may persist for up to 12 weeks postpartum. As a result of physiologic changes in pregnancy, VTE occurs at a rate that is fourfold higher compared to the nonpregnant state.⁸ The postpartum period is even more thrombogenic, with VTE twice as likely as a given 6-week period during pregnancy. That VTE does not occur more often is remarkable, given the paradoxical challenges presented to the hemostatic system during pregnancy.

During early placentation, syncytiotrophoblasts penetrate maternal uterine vessels to establish the primordial uteroplacental circulation. Subsequently, endovascular extravillous cytotrophoblasts invade decidual and superficial myometrial spiral arteries, orchestrating a morphologic conversion of these vessels to achieve high-volume, low-resistance blood flow into the intervillous space. To ensure maternal survival, decidual hemorrhage must be avoided throughout pregnancy. The most profound hemostatic challenge is faced by mothers during the third stage of labor. Following separation of the placenta from the uterine wall after delivery of the infant, hemostasis must be rapidly achieved in 140 remodeled spiral arteries to avoid potentially catastrophic hemorrhage. While local factors such as high decidual tissue factor (aka thromboplastin) expression contribute to this placental site hemostasis, dramatic changes in the mother’s expression of clotting and anticlotting factors are also required to meet this hemostatic challenge.

In addition to an innate hypercoagulability, venous stasis and vascular trauma complete Virchow classic triad⁹ (Fig. 7-2). Venous stasis is present as a result of mechanical impedance of the lower extremity vasculature by the gravid uterus, and estrogen-mediated vascular dilation.¹⁰ Endothelial damage is often present during the puerperium, especially with operative delivery, hypertensive disease, tobacco use, and infections.

Risk Factors

In a review of International Classification of Disease-9 (ICD-9) codes from over 9 million pregnancy admissions and over 73,000 postpartum admissions in the United States National Inpatient Sample (NIS), a list of medical and obstetric risk factors for VTE was identified. The most important risk factors for VTE are personal history of thrombosis and the presence of an inherited or acquired thrombophilia. See Tables 7-1 and 7-2 for a listing of common risk factors in the development of VTE.

Thrombophilia

As many as half of VTE events during pregnancy occur in women with a preexisting thrombophilia.¹¹ These thrombophilic states can be divided into inheritable mutations and acquired disorders.

Inheritable Thrombophilia

The most common significant inherited thrombophilias include heterozygosity for factor V Leiden (FVL), and prothrombin G20210A (PGM) gene mutations. Rarer causes of inherited thrombophilias include antithrombin (AT) deficiency, protein S deficiency, and protein C deficiency. See Table 7-3 for a summary of these thrombophilias, their respective inheritance patterns, and risks of thromboses.¹²

Acquired Thrombophilia

The most common acquired thrombophilia is antiphospholipid syndrome (APS). Approximately 2% of APS patients will experience a VTE in pregnancy, accounting for approximately 14% of VTE events in pregnancy. Diagnosis of APS requires one clinical criterion and one laboratory criterion, as defined at the international consensus conference in 2006¹³ (see Table 7-4). APS is a thrombogenic disorder that arises from autoimmune targeting of proteins binding to exteriorized anionic phospholipids on endothelial cell membranes, such as cardiolipin and phosphatidylserine. In more than half of APS patients, the responsible antibodies arise as a result of underlying disorders such as systemic lupus erythematosus (SLE). The type and concentration of antiphospholipid antibody predict its pathogenicity. Low-positive anticardiolipin IgG and IgM are seldom associated with medical complications. Medium or high titers of anticardiolipin and presence of lupus anticoagulant are associated with fourfold higher rates of thrombosis.

Other examples of acquired thrombophilias include nephrotic syndrome, heparin-induced thrombocytopenia (HIT), and malignancy. The etiology of hypercoagulability in patients with nephrotic range proteinuria is incompletely understood, but may be explained by decreased levels of antithrombin and protein S, increased blood viscosity, or abnormal platelet function. The mechanism of thrombosis in HIT is related to autoantibodies against platelet factors in patients exposed to heparin. Tissue factor and cancer procoagulant have been implicated in the hypercoagulable state observed in patients with malignancy.

Screening for Thrombophilias in Pregnancy

The American College of Obstetricians and Gynecologists currently recommend screening for inheritable thrombophilias in cases where results will affect management decisions such as duration and dosage of treatment.¹¹ Screening may be considered in the following clinical settings:

• 
  

  A personal history of VTE that was associated with a nonrecurrent risk factor (eg, fractures, surgery, prolonged immobilization)

  
  
• 
  

  A first-degree relative (eg, parent or sibling) with known high-risk thrombophilia

Routine screening for inheritable thrombophilias is not recommended for patients with a history of recurrent fetal loss and placental abruption because there is a lack of evidence that anticoagulation reduces the risk of recurrence.

The American College of Obstetricians and Gynecologists recommends testing for antiphospholipid antibodies for the following indications:¹⁴

• 
  

  A prior unexplained or new arterial or venous thromboembolism

  
  
• 
  

  A history of one fetal loss

  
  
• 
  

  A history of three or more recurrent embryonic or fetal losses

Severe preeclampsia and placental insufficiency diagnosed before 34 weeks of gestation are included in the criteria for diagnosis of APS; however, there is insufficient evidence that screening and treating women with these conditions improves subsequent pregnancy outcomes.

Most cases of VTE in pregnancy occur in the lower extremities, with predisposition for the left lower extremity, secondary to the anatomic compression of the left iliac vein by the right iliac, and ovarian arteries.^15,16

Diagnosis

Clinicians caring for pregnant patients must maintain a high index of suspicion for DVT. Disease prevalence and pretest probability are often invoked to guide screening and diagnostic approaches in the general population. The clinical diagnosis of DVT in pregnancy is complicated by the increased risk of VTE during pregnancy and the overlap of symptoms with expected physiologic changes, such as lower extremity edema.

Clinical Signs and Symptoms

• 
  

  Unilateral extremity pain and swelling (80%)

  
  
• 
  

  Erythema, warmth, tenderness

  
  
• 
  

  Lower abdominal, flank, buttock, or back pain

  
  
• 
  

  Homans sign (pain in calf with dorsiflexion of foot)

  
  
• 
  

  Reflex arterial spasm, with cool, pale extremity and decreased pulses (“phlegmasia alba dolens”)

Clinical Decision Aids

• 
  

  The Wells and modified Wells score are the most commonly used risk scoring systems in nonpregnant populations; however, they have not been validated in a pregnant population and may not be helpful for risk stratification¹⁷ (Table 7-5). The LEFt clinical prediction rule is a risk assessment tool that has been validated in pregnancy. LEFt uses the presence of three variables to predict DVT: left leg symptoms, calf circumference difference of more than or equal to 2 cm, and first trimester presentation.¹⁸ Studies of test performance showed that the negative predictive value was good for a population with a low prevalence of DVT; however, as the authors pointed out, the rule would not be well suited to ruling out VTE as a stand alone test during pregnancy.

Laboratory Studies

• 
  

  D-dimer: D-dimer is the breakdown product of cross-linked fibrin and elevations may be detected by enzyme-linked immunosorbent assay (ELISA) in acute thrombotic events outside of pregnancy. In normal pregnancy, however, physiologic elevations of D-dimer are found in a gestational-age dependent fashion, with 84% of women in the first trimester with a normal D-dimer, 33% in the second trimester, and 1% in the third trimester, reducing the test’s specificity.¹⁹ D-dimer levels further peak at the time of delivery and early puerperium. Obstetric complications such as placental abruption, preeclampsia, and sepsis can also elevate D-dimer levels. Currently available D-dimer assays are more sensitive than previous generations of tests and therefore may generate more false positive results in pregnant patients. Higher cut-points may improve specificity; however, studies to validate the performance of this test have yet to be done.²⁰ The use of D-dimer as an initial screening tool in pregnancy is therefore not recommended.⁵ However, the American College of Chest Physicians (ACCP) guidelines suggest that D-dimer may have a role in further risk-stratifying patients with negative compression ultrasound results.²¹

Imaging

• 
  

  Compression Doppler ultrasonography: The primary diagnostic tool for DVT is compression ultrasonography (CUS) of the proximal veins. Sonographic signs of thrombus may include visualization of a clot, noncompressibility of a venous segment, and absence of Doppler flow. In nonpregnant patients, compression color Doppler ultrasound is both highly sensitive (92%) and specific (98%) for popliteal and femoral vein thrombosis, but slightly less effective for evaluating calf vein thrombosis with a sensitivity of only 50% to 70% and specificity of 60%.²² These results have been extrapolated to pregnant women with relatively limited information available regarding test performance of CUS in this relatively higher risk cohort. One small study utilizing a protocol of CUS at presentation and for follow-up imaging at the clinician’s discretion reported a false negative rate of 0.7% during pregnancy.²³

  
  
• 
  

  Both ACOG and ACCP currently recommend that when pregnant patients have signs or symptoms of DVT, CUS should be the initial diagnostic test.^5,21 Depending on the results, surveillance, treatment, or additional imaging may be considered (Fig. 7-3). Of note, ACCP suggests that follow-up proximal CUS (day 3 and day 7) be performed for patients with a negative initial CUS study in whom DVT is suspected. For patients who have symptoms of iliac vein thrombosis and negative proximal CUS, Doppler US of the iliac vein or consideration of alternative imaging is recommended.

  
  
• 
  

  CT scan venography: Contrast venography was previously considered the gold standard for diagnosing DVT in pregnancy, with a negative predictive value of 98%. However, given its invasive nature and high rate of complication, contrast venography has fallen out of favor. Contrast agents are injected into lower extremity veins and the venous system of the leg and pelvis are evaluated radiographically. Although, when used with an abdominal lead shield, it exposes the fetus to very low levels of radiation (0.0005 Gy), well below that associated with childhood cancers and teratogenicity, the availability of other imaging modalities with a more favorable risk profile has made venography an infrequently used modality.

  
  
• 
  

  Magnetic resonance imaging (MRI): MRI is useful for detecting thigh and pelvic vein thrombosis. The accuracy of MR venography has been reported to be similar to CT venography in nonpregnant patients.²⁴ The performance of MR venography in pregnancy has not been studied. Although the safety of MRI in pregnant women is yet to be proven, no adverse effects have been noted. Administration of gadolinium is not necessary for MR venography.

In summary, compression Doppler ultrasound is recommended as the initial test in pregnant women with suspected DVT. If this study proves positive for a DVT, treatment should be initiated. With equivocal test results, additional confirmatory imaging, such as with venous MRI, should be performed as detailed in Figure 7-3. Serial imaging with proximal CUS or iliac Dopplers may also be appropriate.

General DVT Management Principles

• 
  

  Heparin anticoagulation (see separate section for details).

  
  
• 
  

  Therapeutic anticoagulation should be for 12 to 20 weeks following diagnosis.

  
  
• 
  

  Prophylactic anticoagulation should be initiated after initial treatment, for 6 to 12 weeks and until the patient reaches 6 weeks postpartum.

  
  
• 
  

  For complicated DVTs, including those involving the iliofemoral vessels, prophylaxis is recommended for 4 to 6 months.

  
  
• 
  

  Conversion to oral warfarin may be considered in the postpartum period, if the patient is compliant with drug level monitoring.

  
  
• 
  

  Warm moist heat packs and leg elevation may swelling and provide symptomatic relief.

  
  
• 
  

  Sequential compression devices should be avoided for a limb known to have a DVT.

Pulmonary embolism complicates approximately 1 in 2500 pregnancies. A thrombotic obstruction in the pulmonary vascular tree results in obstruction to pulmonary arterial blood flow, vasoconstriction of small arterial vessels, and progressive loss of alveolar surfactant. In pregnancy, thromboemboli most commonly originate in the iliac vessels. In nonpregnant populations, it is estimated that up to one third of patients with DVT have a concomitant PE, although less than half of patients with PE will have evidence of DVT on CUS.²⁵

Diagnosis

Clinical Signs and Symptoms

• 
  

  Dyspnea

  
  
• 
  

  Tachycardia

  
  
• 
  

  Hemoptysis

  
  
• 
  

  Pleuritic chest pain

  
  
• 
  

  Low pulse oximetry reading

  
  
• 
  

  Unilateral leg swelling and calf tenderness

  
  
• 
  

  Low end-tidal CO₂ (<30 mm Hg)

  
  
• 
  

  Syncope

Clinical Decision Aids

Clinical decision aids for working up suspected PE include Wells score for PE, the Pulmonary Embolism Rule-out Criteria (PERC), and the Simplified (Tables 7-6, 7-7, and 7-8).^26-28 These aids were designed primarily for use in the Emergency Department and have very good negative predictive values when the disease prevalence is low.²⁹ Therefore they may not be applicable to populations in which the prevalence of VTE may be higher, for example during pregnancy, among hospital inpatients, and in patients with other medical comorbidities. None have been validated in pregnancy.