A. Introduction. Approximately 600,000 cases and 50,000 deaths from pulmonary embolism (PE) occur annually in the United States. Prompt treatment can significantly reduce the mortality rate.
B. Causes of Pulmonary Embolism.
a. Most pulmonary emboli (95%) arise from deep venous thrombosis (DVT) of the lower extremities. The highest risk for PE occurs with proximal, lower extremity DVT (i.e., thrombosis of the popliteal, superficial femoral, or common femoral vein).
b. PE can also originate from the upper extremity, central venous catheters or pelvic venous thrombosis and right atrial thrombi. In particular, peripherally inserted central catheters (PICCs) have been shown to be strongly associated with upper extremity DVT.
c. Virchow’s triad describes the major risk factors for venous thrombosis (and PE).
i. Venous stasis (e.g., an immobile patient in the hospital).
ii. Endothelial damage (e.g., hip or knee surgery).
iii. Hypercoagulable state (e.g., malignancy, antiphospholipid antibody syndrome).
d. Rare causes of nonthrombotic PE include fat embolism, amniotic fluid embolism, septic embolism, and tumor embolism.
C. Clinical Manifestations of Pulmonary Embolism. The symptoms and signs of PE are nonspecific.
a. Symptoms. Pleuritic chest pain and dyspnea are the most common symptoms. A feeling of apprehension or impending doom is frequently described.
b. Signs. Tachypnea is present in most patients with PE. Tachycardia, low-grade fever, and evidence of lower extremity swelling or tenderness are other important signs.
D. Approach to the Patient. PE can pose a substantial diagnostic challenge. The differential diagnosis includes other cardiac and pulmonary disorders that cause dyspnea, chest pain, and hypoxemia. Common competing diagnoses include pneumonia, congestive heart failure (CHF), asthma, chronic obstructive pulmonary disease (COPD), pneumothorax, myocardial infarction, and aortic dissection.
a. The history and physical examination are essential to the diagnosis. A sudden onset of chest pain and dyspnea is typical. However, symptoms of new-onset syncope or an unexplained exacerbation of COPD are increasingly recognized presentations. The physical examination may reveal an edematous, erythematous, and tender upper or lower extremity suggesting DVT. However, remember that physical examination is often unreliable and/or unremarkable in the diagnosis of DVT or PE.
i. Patients with PE have a normal chest radiograph about 20% of the time. Common radiographic findings among patients with PE are nonspecific and include atelectasis, parenchymal opacities, pleural effusion, and cardiomegaly.
ii. Some classic roentgenographic signs of PE exist:
1. Westermark’s sign is a region of oligemia (decreased vascularity) visible as a focal radiolucent area/
2. Hampton’s hump is a peripheral wedge-shaped density that occurs with pulmonary infarction.
3. Subsegmental atelectasis, a small pleural effusion, an enlarged central pulmonary artery, and an elevated hemidiaphragm may be found.
iii. These classic radiographic signs of PE are nonspecific and cannot definitively establish the diagnosis. Therefore, the main utility of a chest radiograph is in identifying alternative diagnoses (e.g., pneumonia, pneumothorax, pulmonary edema).
c. Electrocardiogram (EKG). Most patients with PE have nonspecific EKG findings:
i. The most common finding is sinus tachycardia; other arrhythmias are rare. It is not uncommon for the EKG to be normal.
ii. The classic findings of acute right-sided heart strain are seen in approximately 25% of patients and include:
1. Right bundle branch block.
2. P pulmonale.
3. Right axis deviation.
4. S1Q3T3 (a large S wave in lead I, a large Q wave in lead III, and an inverted T wave in lead III). A handy mnemonic is “SiQ from Thrombus!”
5. T wave inversions in V1–V4 occur if PE leads to significant right ventricular wall strain.
d. Arterial blood gases. Patients with PE may have alveolar hyperventilation (low PaCO2), hypoxemia (low PaO2), a widened alveolar-to-arterial (A-a) gradient, or a combination of the three. Most patients will have a widened A-a gradient. Patients can have normal oxygen saturation because hyperventilation can increase PaO2. Do not rule out PE based on normal oxygen saturation.
e. D-dimer assay. While this test has a poor specificity, the highly sensitive enzyme-linked immunosorbent assay (ELISA) has >99% negative predictive value and can be used to rule out PE in patients with an “unlikely” pretest probability (see below).
f. Biomarkers. Brain natriuretic peptide (BNP) has a poor sensitivity and specificity for diagnosis of PE. However, elevated levels (>90 pg/mL) are associated with worse outcomes. Similarly, troponin has no diagnostic role in acute PE but correlates physiologically with right strain and worse outcomes.
g. Ventilation-perfusion (V/Q) scan. The V/Q scan is used in patients who cannot tolerate intravenous contrast (e.g., renal insufficiency and contrast allergy). The V/Q scan has improved sensitivity when evaluating for chronic thromboembolic disease. If perfusion defects in areas of normal ventilation are observed, PE is probable. V/Q scans are interpreted either as normal or low, intermediate, or high probability of PE based on standardized criteria.
i. Using the V/Q scan result and the pretest probability, a clinical likelihood of PE can be calculated. Unfortunately, only certain combinations yield acceptable clinical certainty.
1. A low pretest probability combined with a low-probability V/Q scan rules out PE in 96% of cases.
2. A combination of a high pretest probability and a high-probability V/Q scan diagnoses PE in 95% of cases.
h. Noninvasive lower extremity testing. Most pulmonary emboli arise from proximal lower extremity DVT. Because the treatments of DVT and PE are generally the same, documentation of DVT is adequate reason to stop diagnostic testing and initiate anticoagulation therapy.
i. Compression Doppler ultrasonography has excellent sensitivity and specificity in patients with clinically suspected symptomatic DVT (i.e., leg swelling or pain). In asymptomatic patients, the sensitivity of compression ultrasonography is much lower (in the 70% range), perhaps owing to inability to detect thrombus in the more proximal iliac and femoral veins.
ii. Complete lower extremity ultrasound has been shown to have much better sensitivity but is operator dependent and has not been prospectively studied in patients with suspected PE.
i. Computed tomographic (CT) angiography is the gold standard for diagnosis of PE. It can also aid in diagnosis of other causes of dyspnea if no PE is found (e.g., pneumothorax, pneumonia, heart failure, rib fractures). Although sensitivity depends on the experience of the reader and the quality of the study, it approaches 90%. The addition of lower extremity CT venography (scanning the lower extremities as the dye moves through the venous circulation) adds sensitivity to the test because it can help detect thrombosis in these vessels. The CT angiogram is a highly specific test for PE.
j. Pulmonary angiography. Pulmonary angiography was considered the gold standard for diagnosing PE. However, the CT angiogram has been shown to be equivalent in clinical accuracy and is less invasive. The associated risks for death, major complications, and minor complications with angiography are approximately 0.5%, 1%, and 5%, respectively. Major complications include contrast-induced nephropathy, bleeding, and pulmonary arterial rupture.
k. Pretest probability score. Because PE can be deadly, high sensitivity is required to avoid missing the diagnosis. Adding a pretest clinical probability assessment can improve sensitivity (Table 23.1).
Table 23.1 Pretest Probability for Pulmonary Embolism: Wells Criteria
Pulmonary embolism as likely or more likely than alternate diagnosis
Clinical signs or symptoms of deep venous thrombosis
Prior history of deep venous thrombosis or pulmonary embolism
Heart rate >100 beats/min
Immobilization ≥3 days
Malignancy or hemoptysis (1 point each)
Clinical Probability: High >6.0, Moderate 2.0 to 6.0, Low <2.0
l. Diagnostic approach to PE:
1. Determine whether the patient has low, intermediate, or high pretest probability of PE (see Table 23.1)
a. If low (Wells Criteria <2), apply the pulmonary embolism rule out criteria (PERC) (see Table 23.2).
Table 23.2 Pulmonary Embolism Rule Out Criteria (PERC)
Age <50 years
Heart rate <100 beats/minute
Oxyhemoglobin saturation ≥95%
No estrogen use
No prior DVT or PE
No unilateral leg swelling
No surgery/trauma requiring hospitalization within the prior four weeks
i. If patient fulfills all 8 criteria: stop further workup (PE unlikely).
ii. If not all criteria fulfilled, perform D-dimer.
c. If high (Wells Criteria >6.0), perform CT angiography.
a. If negative: stop further workup (PE unlikely).
b. If positive: perform CT angiography.
3. CT angiography
a. If negative: stop further workup (likely no clinically significant PE).
b. If positive: treat for PE. Duration of anticoagulation depends on risk factors.
c. If uninterpretable: treat and repeat or perform other studies.
4. If CT angiography is contraindicated, perform other diagnostic studies (e.g., V/Q scan) or empirically anticoagulate until study can be performed.
E. Risk Stratification. The presence of the following risk factors portends an ominous prognosis. Close monitoring and potential escalation of care (thrombolysis, vasopressors, or embolectomy) might be required:
a. Hypotension (about 30% mortality; consider thrombolysis).
b. Elevated BNP, troponin (both elevated suggests mortality as high as 30%).
c. Evidence of right ventricular dysfunction on echocardiogram and/or right ventricle to left ventricle ratio >0.9 on CT angiogram.
a. The initial treatment of PE can be one of the following:
i. Unfractionated heparin. This is intravenous therapy so can only be given to inpatients.
ii. Low-molecular-weight heparin (LMWH) such as enoxaparin. Contraindicated in severe renal insufficiency but can be given to outpatients or those that may be discharged from the emergency department.
iii. Fondaparinux. Acceptable alternative to LMWH. More convenient than enoxaparin because it can be given once daily.
iv. Oral factor Xa inhibitors or direct thrombin inhibitors. Rivaroxaban and apixaban are currently the only direct oral anticoagulants approved for treatment of PE as monotherapy.
b. Subsequent long-term treatment:
i. The oral anticoagulants are excellent choices in nonpregnant patients with normal renal function and absence of active malignancy.
ii. Warfarin is preferred when newer direct oral anticoagulants are not available or if a patient has renal dysfunction and cannot take LMWHs.
iii. Inferior vena cava (IVC) filters should be reserved as an alternative, short-term management option for patients with absolute contraindications to anticoagulation or temporary heightened risk for thrombosis (e.g., surgical procedures).
c. Anticoagulation is the standard treatment for both DVT and PE.
i. Unfractionated heparin should be administered immediately and is preferred over other options in the following situations: renal failure, hemodynamic instability, extensive clot burden, anticipated need for reversal (e.g., patient at high risk for bleeding), obesity, or poor subcutaneous absorption (where the pharmacokinetic profile of other agents might be variable). The recurrence of venous thromboembolism increases with delayed or inadequate anticoagulation, so rapid achievement of a partial thromboplastin time (PTT) of 1.5–2.5 times control is desirable. Heparin should be continued until adequate oral anticoagulation with warfarin is achieved.
ii. LMWHs are preferred when oral anticoagulation is not tolerated (e.g., poor absorption) or in the presence of pregnancy and malignancy because it is superior and safer than warfarin in these settings. Dosing is weight based and can be given as a daily or twice-daily regimen. LMWH is contraindicated in patients with severe renal insufficiency.
iii. Fondaparinux. Use when LMW heparin is contraindicated or if a history of heparin-induced thrombocytopenia exists.
iv. Oral anticoagulants. Apixaban and rivaroxaban can be administered as monotherapy and do not require monitoring. Other factor Xa inhibitors and direct thrombin inhibitors should be preceded by 5 days of heparin.
v. Oral warfarin, when used, should be started when the diagnosis is made and should overlap with initial heparin therapy to avoid recurrent thrombosis. The efficacy and safety of early warfarin initiation have been well documented. The duration of therapy depends on risk factors and patient’s personal history of PE as follows:
1. First PE with a removable risk factor (i.e., trauma, surgery, immobilization): treat with warfarin, international normalized ratio (INR) of 2–3 for 3–6 months or until risk factor has been removed.
2. First PE, without risk factors (idiopathic PE): treatment is controversial because the stimulus for PE may not have been removed and epidemiologic studies show a high incidence of recurrent PE. Recommendations suggest 12 months to lifelong anticoagulation based on an assessment of bleeding risks.
3. First PE and irreversible risk factor (e.g., hypercoagulable disorder) or recurrent PE: patients should receive lifelong anticoagulation with warfarin, targeting an INR of 2–3.
d. IVC filter. For patients with absolute contraindications to anticoagulation and lower extremity clot, an IVC filter may be placed percutaneously through the femoral vein. For patients with short-term contraindication to anticoagulation (e.g., trauma, recent stroke), temporary IVC filters may be placed and should be removed as soon as anticoagulation is safe. Filters are also indicated in patients with massive or submassive PE who have poor cardiopulmonary reserve and in whom another embolic event might prove life-threatening. If it appears that the patient would likely die from another PE, a temporary filter should be placed to protect the patient during this vulnerable period. IVC filters prevent lower extremity thrombus from traveling to the lungs and lower the risk for recurrent PE. However, IVC filters increase lower extremity venous stasis and can increase risk for subsequent DVT. Hence, removal of IVC filters should always be planned when placing these devices.
e. Thrombolytic agents (e.g., urokinase, tissue plasminogen activator) dissolve clot, resulting in more rapid resolution of perfusion abnormalities than standard heparin therapy. Thrombolytics improve initial hemodynamics in massive PE but do not improve mortality rates. Furthermore, there is a substantial risk for bleeding associated with these agents. For these reasons, thrombolytic agents are not used routinely in the treatment of PE. In patients with massive PE characterized by hemodynamic instability, thrombolytics may be beneficial and should be considered in an ICU setting.
Suggested Further Readings
Aleva FE, Voets L, Simons SO, de Mast Q, van der Ven A, Heijdra YF. Prevalence and localization of pulmonary embolism in unexplained acute exacerbations of COPD: a systematic review and meta-analysis. Chest 2017;151:544–54.Find this resource:
Chopra V, Anand S, Hickner A, et al. Risk of venous thromboembolism associated with peripherally inserted central catheters: a systematic review and meta-analysis. Lancet 2013;382:311–25.Find this resource:
Di Nisio M, van Es N, Buller HR. Deep vein thrombosis and pulmonary embolism. Lancet 2016;388:3060–73.Find this resource:
Freund Y, Cachanado M, Aubry, A, et al. Effect of the pulmonary embolism rule-out criteria on subsequent thromboembolic events among low-risk emergency department patients: The PROPER randomized clinical trial. JAMA 2018;319:559–566.Find this resource:
Mismetti P, Laporte S, Pellerin O, et al. Effect of a retrievable inferior vena cava filter plus anticoagulation vs anticoagulation alone on risk of recurrent pulmonary embolism: a randomized clinical trial. JAMA 2015;313:1627–35.Find this resource:
Prandoni P, Lensing AW, Prins MH, et al. Prevalence of pulmonary embolism among patients hospitalized for syncope. N Engl J Med 2016;375:1524–31.Find this resource:
Righini M, Van Es J, Den Exter PL, et al. Age-adjusted D-dimer cutoff levels to rule out pulmonary embolism: the ADJUST-PE study. JAMA 2014;311:1117–24.Find this resource: