+++
History and Physical Examination
++
Erythema, warmth, pain, swelling, and tenderness are not specific for DVT, but suggest the need for further evaluation. PE must always be considered when unexplained dyspnea is present. Pleuritic chest pain and hemoptysis are also common in patients with PE. Coughing may be present, and although it is sometimes caused by PE, it more commonly occurs with bronchitis or pneumonia. Anxiety and lightheadedness are symptoms that may be caused by PE, but may also be caused by a number of other entities that result in hypoxemia or hypotension. Severe dyspnea and syncope are the principal symptoms that may suggest massive, life-threatening PE.67,68,69 Tachypnea and tachycardia are the most common signs of PE, but they are also nonspecific. A pleural rub may suggest pulmonary infarction, and accentuated pulmonic component of the second heart sound may suggest PE, but can also be explained by other disorders. With embolism of sufficient magnitude to cause RV dysfunction, a murmur of tricuspid regurgitation, systemic hypotension, or jugular venous distension might be present.
++
A major advance in the diagnostic approach to both venous thrombosis and PE has been a transition from a technique-oriented approach to one that uses Bayesian analysis. In doing so, the pretest probability of the disease, calculated independently of a particular test result through either empiric means or through a standardized prediction rule, is calculated. This pretest probability aids in the selection and interpretation of further diagnostic tests to create a post-test probability of the disease. This post-test probability can then be used as a basis for clinical decision making. For PE, three such scores have been developed and validated. Wells and coworkers70 prospectively tested a rapid seven-item bedside assessment to estimate the clinical pretest probability for PE. An alternative scoring system, the Geneva score, involved seven variables and required gas exchange and radiographic information.71 Recently, a revised Geneva score requiring eight clinical variables without gas exchange or radiographic information was validated and published.72 Although such scoring systems have not proven to be more accurate than implicit assessment, they have been adequately validated73,74,75 and do provide a means of standardization that compensates for variability in physician experience and judgment. Both the Wells score and the revised Geneva rule were simplified in an attempt to increase their adoption into clinical practice,76,77 and these simplified versions have been adequately validated as well.74,78 Thus, in the absence of hemodynamic instability at presentation, the diagnostic workup of a patient with suspected PE should begin with the assessment of the clinical or pretest probability. Table 75–2 summarizes clinical prediction rules for PE.
++
+++
Differential Diagnosis
++
PE may mimic a large spectrum of diseases. The most common differential diagnoses are pneumonia, musculoskeletal pain, pneumothorax, costochondritis, congestive heart failure, chronic lung disease, asthma/bronchitis, acute myocardial infarction, aortic dissection, pericarditis, and anxiety states.
+++
Nonimaging Studies for Pulmonary Embolism
++
The plasma d-dimer is a specific derivative of cross-linked fibrin. Measurement of circulating plasma d-dimer has been comprehensively evaluated as a diagnostic test for acute VTE.79 A normal enzyme-linked immunosorbent assay (ELISA) or ELISA-derived assays result is highly sensitive in excluding PE and DVT. When an ELISA d-dimer level is below an established cutoff level, the sensitivity and negative predictive value for VTE are 95% or above.79 Therefore, it can be used to exclude PE in patients with either a low or a moderate pretest probability. Importantly, d-dimer measurement is not recommended in patients with high clinical probability, because a normal result does not safely exclude PE. Increased levels of cross-linked fibrin degradation products are an indirect, but suggestive, marker of intravascular thrombosis, indicating endogenous fibrinolysis. An increased d-dimer level is nonspecific for PE and may be seen with advancing age and in patients with various conditions, including infections and other inflammatory states, cancer, myocardial infarction, the postoperative state, and the second and third trimesters of pregnancy. Thus, the specificity of d-dimer in suspected PE decreases steadily with age, to almost 10% in patients older than 80 years.80 Recent evidence suggests using age-adjusted cutoffs (instead of the standard 500 μg cutoff) to improve the performance of d-dimer testing in the elderly.81,82 In a contemporary meta-analysis, age-adjusted cutoff values (age × 10 μg/L above 50 years) resulted in increased specificity while retaining sensitivity above 95%.83 Similar results were obtained in a recent prospective study that showed increased specificity and no additional false-negative results after using an age-adjusted cutoff.84 In addition, it has been suggested that the height of d-dimer level correlates with the likelihood of PE85 and with increased mortality, especially in noncancer patients.86
++
It is important to recognize that the usefulness of d-dimer testing in inpatients is still limited as a result of comorbidities in this population that elevate d-dimer levels. However, the negative predictive value of a (negative) d-dimer test remains high in these situations.
+++
Arterial Blood Gas Analysis
++
Although hypoxemia is common in acute PE, some patients, particularly young individuals without underlying cardiopulmonary disease, may have a normal arterial oxygen partial pressure (PaO2). In a retrospective study of hospitalized patients with PE, the PaO2 was greater than 80 mm Hg in 29% of patients who were younger than 40 years compared with 3% in the older group.68 However, the alveolar-arterial difference was elevated in all patients. An important tenet should be that unexplained hypoxemia, particularly in the setting of risk factors for DVT, suggests the possibility of PE.
++
Electrocardiography (ECG) findings in acute PE are generally nonspecific and include T-wave changes, ST-segment abnormalities, incomplete or complete right bundle branch block, right-axis deviation in the extremity leads, and clockwise rotation of the QRS vector in the precordial leads. The changes that do occur are likely caused by right-heart dilation. In milder cases, the only anomaly may be sinus tachycardia, present in approximately 40% of patients. Atrial arrhythmias, most frequently atrial fibrillation, may be associated with acute PE. Approximately 20% of patients with PE have no ECG changes. Therefore, ECG cannot be relied upon to rule in or rule out PE, although ECG proof of a clear alternative diagnosis, such as myocardial infarction, is useful when PE is among the possible diagnoses. The “classic” S1Q3T3 pattern described by McGinn and White87 in 1935 in seven patients with acute cor pulmonale secondary to PE was subsequently demonstrated to be present in approximately 10% of PE cases.88 In patients without underlying cardiac or pulmonary disease from the Urokinase Pulmonary Embolism Trial, ECG abnormalities were documented in 87% of patients with proven PE.89 These findings were not specific for PE, however. In this clinical trial, 26% of patients with massive or submassive PE and 32% of those with massive PE had manifestations of acute cor pulmonale, such as the S1Q3T3 pattern, right bundle branch block, a P-wave pulmonale, or right-axis deviation. The low frequency of specific ECG changes associated with PE was confirmed in the PIOPED study.68
++
Despite its lack of diagnostic accuracy, ECG may be helpful in predicting adverse clinical outcomes in patients with PE. It was recently suggested that a T-wave inversion in V2 or V3 is the most frequent ECG sign of massive PE.90 In another PE study, both the pseudoinfarction pattern (Qr in V1) (Fig. 75–1) and T-wave inversion in V2 were closely related to the presence of RV dysfunction and were independent predictors of adverse clinical outcome.91 In a recent trial, abnormal ECG at presentation proved to be an independent predictor of an adverse outcome, although no individual abnormality appeared capable of predicting such an outcome after being adjusted for the patients’ clinical symptoms and findings on admission and for the presence of pre-existing cardiac or pulmonary disease.92
++
+++
Imaging Studies for Pulmonary Embolism
++
The chest radiograph is abnormal in the majority of patients with PE, but the findings are nonspecific and often subtle. Atelectasis, cardiomegaly, pulmonary infiltrates, small pleural effusions, and mild elevation of a hemidiaphragm may be present.68,93 Classic radiographic evidence of pulmonary infarction (Hampton hump) or decreased vascularity (Westermark sign) is suggestive, but uncommon. A normal chest radiograph in the presence of significant dyspnea and hypoxemia without evidence of bronchospasm or anatomic cardiac shunt is strongly suggestive of PE. In most situations, however, the chest radiograph cannot be used to definitively diagnose or exclude PE. Although exclusion of other processes such as pneumonia, congestive heart failure, pneumothorax, or rib fracture (which may cause symptoms similar to acute PE) is important, it is crucial to recognize that PE often coexists with other underlying lung diseases.
+++
Computed Tomography of the Chest
++
Since the introduction of multidetector computed tomography (MDCT) angiography, contrast-enhanced CT of the chest has become the most useful imaging test in patients with clinically suspected acute PE. It allows adequate visualization of the pulmonary arteries down to at least the segmental level.
++
First-generation scanners had poor resolution in the segmental pulmonary arteries and a limited sensitivity for subsegmental clots. However, a negative scan appeared to predict a benign clinical course over the ensuing 3 months.94
++
Second-generation scanners involve continuous movement of the patient through the scanner with concurrent scanning by a constantly rotating gantry and detector system.95 A helix of projecting data is obtained. Continuous volume acquisitions can be obtained during a single breath. Rapid scans can be obtained, facilitating imaging in critically ill patients.
++
The latest generation of MDCT scanners (Figs. 75–1 and 75–2) permits image acquisition of the entire chest with 1-mm or submillimeter resolution with a breath hold of less than 10 seconds.96
++
++
Limitations of chest CT in early clinical studies included poor visualization of horizontally oriented vessels in the right middle lobe and lingula because of volume averaging.95 The peripheral areas of the upper and lower lobes were inadequately scanned, and the presence of intersegmental lymph nodes resulted in false-positive study results. Multiplanar reconstructions in coronal, sagittal, or oblique planes aid in distinguishing lymph nodes from emboli. Based on studies with earlier generation scanners, CT was capable of revealing emboli in the main, lobar, or segmental pulmonary arteries with more than 90% sensitivity and specificity.95,96 However, for subsegmental emboli, the sensitivity and specificity were, and still remain, lower. The incidence of isolated subsegmental emboli with first- and second-generation scanners appeared to be approximately 6% to 30%.97
++
When MDCT is used for PE diagnosis, it is ideal to incorporate the scan into an overall diagnostic strategy that includes pretest probability, d-dimer testing, and in certain specific settings, ultrasound examination of the deep leg veins.96 An additional advantage of MDCT is the ability to evaluate a patient for the entire spectrum of VTE in one imaging session by scanning the legs and pelvis as well as the lungs. In PIOPED II, combining CT venography with CT angiography increased sensitivity for PE from 83% to 90% and had a similar specificity (around 95%)98,99; however, the corresponding increase in negative predictive value was not clinically significant. CT venography adds a significant amount of irradiation, which may be a concern, especially in younger women.100 As CT venography and ultrasonography yielded similar results in patients with signs or symptoms of DVT in PIOPED II,98 ultrasonography should be used instead of CT venography if indicated.
++
Outcome studies have demonstrated that PE can be safely excluded using a clinical assessment tool, d-dimer testing, and MDCT except in patients who present with a high clinical likelihood of embolism.99,101,102,103 In the PIOPED II trial, the sensitivity of CT was 83%, a finding somewhat at odds with published outcome data.99 PIOPED II did confirm the usefulness of clinical assessment in that the negative predictive value of a normal CT scan was 96% in patients with a low probability of embolism and 89% for those with an intermediate probability, but only 60% in those with a high clinical probability. In 2016, MDCT alone, with “wide detectors” of 256 to 320 rows, appears to be highly sensitive and specific for acute PE, and additional studies to confirm or refute the diagnosis are rarely required unless the quality of the CT is suboptimal. As demonstrated in the initial PIOPED trial, a negative V/Q scan or contrast pulmonary angiogram would also achieve this end.104 Sequential, noninvasive, lower extremity examinations in patients with adequate cardiopulmonary reserve, although not confirming that embolism did not occur, would render the probability of recurrence unlikely. This approach is exceedingly impractical nowadays. Additional testing can be considered in patients with a low clinical probability of PE and a CT scan that is suggestive, but not clearly diagnostic, of PE. Another important advantage of chest CT is the concomitant ability to define nonvascular and vascular structures such as airway, parenchymal, and pleural abnormalities; lymphadenopathy; and cardiac and pericardial disease. This is very important for rapid detection of alternative diagnoses (eg, aortic dissection, pneumonia, pericardial tamponade) in patients with acute “chest syndromes” in the emergency setting. Disadvantages of MDCT include the potential for contrast-induced renal failure in patients with renal disease. Patients with a history of allergy to contrast agents should receive preprocedure treatment with steroids administered at least 6 hours before the procedure if possible. H1 and H2 histamine blockers may be used in addition to corticosteroids.
++
CT can expose patients to clinically significant doses of radiation. Younger women who have a known elevated incidence of PE are particularly at risk from radiation damage to the breast and lungs. In addition, the amount of contrast required for CT scanning (100-150 mL) poses a substantial risk of radiocontrast-induced nephropathy for patients with pre-existing renal disease (serum creatinine ~2 mg/dL or creatinine clearance < 60 mL/min), especially when it is associated with diabetes mellitus.105 In such patients, a strategy using duplex ultrasonography and V/Q scanning would appear prudent, followed by selective conventional pulmonary angiography if the noninvasive techniques do not yield a diagnosis. Although multiple pharmacologic agents have been used, only periprocedural hydration, the use of nonionic, iso-osmolar contrast agents, and perhaps N-acetylcysteine are of proven benefit in preventing dye nephropathy.105
+++
Ventilation-Perfusion Scanning
++
V/Q scanning was the pivotal diagnostic test for suspected PE for many years. Although clinical indications for the study remain (contrast allergy, renal insufficiency, pregnancy), chest CT has now virtually replaced lung scanning. When the chest radiograph is normal or near normal, a V/Q scan can be performed; any perfusion defect in this situation will be considered to be a mismatch. Otherwise (eg, in the common clinical scenario involving an abnormal baseline chest x-ray), CT should be obtained. An additional advantage of the V/Q scan is that a portable study can be done in very ill patients who are too unstable to undergo CT.
++
The test is based on the intravenous injection of technetium (Tc)-99m–labeled albumin particles, which block a small fraction of the pulmonary capillaries and thereby enable scintigraphy assessment of lung perfusion. Perfusion scans are combined with ventilation studies, for which multiple tracers can be used. The purpose of the ventilation scan is to increase specificity; in acute PE, ventilation is expected to be normal in hypoperfused segments (mismatch). The radiation exposure from regular lung V/Q scanning is significantly lower than that of CT angiography.106
++
V/Q scanning is nondiagnostic in up to 70% of patients with suspected PE, which has been a cause for criticism, since this necessitates further diagnostic testing. Normal and high-probability scans are considered diagnostic. A normal perfusion scan excludes the diagnosis of PE with enough certainty that further diagnostic testing is unnecessary.
++
In the PIOPED study, the usefulness of lung scanning combined with clinical assessment of patients with suspected PE was prospectively evaluated.104 Patients with PE had scans that were of high, intermediate, or low probability, but so did most individuals without PE. Although the specificity of high-probability scans was 97%, the sensitivity was only 41%. Of interest, 33% of patients with intermediate-probability scans and 12% of those with low-probability scans were diagnosed definitively with PE by pulmonary arteriography. Forty percent of low-probability scans in patients with high clinical suspicion were followed by documentation of PE at angiography. When the clinical suspicion of PE was considered very high, the positive predictive value of high-probability scans for PE was 96%. In patients with nondiagnostic lung scans, further diagnostic testing for PE should be undertaken. Recent studies suggest that data acquisition in single-photon emission CT imaging may reduce the frequency of nondiagnostic scans106; however, large-scale prospective studies are lacking and are needed to validate these new approaches.
+++
Pulmonary Angiography
++
Standard contrast pulmonary arteriography (Fig. 75–3) has been considered the established “gold standard” imaging test for the diagnosis of PE. However, it is rarely performed now because MDCT offers similar diagnostic accuracy. Contrast injections may be useful, however, as a guide to percutaneous catheter-directed treatment of acute PE. The risk of subsequent VTE in patients with a negative test result and without anticoagulation has been shown to be less than 2%. In the current era when pulmonary arteriography is so rarely done, it is unclear whether this technique still remains this accurate.
++
++
Pulmonary angiography is not free of risk, with estimated rates of procedure-related mortality, major nonfatal complications, and minor complications of 0.5%, 1%, and 5%, respectively,107 with the majority of deaths occurring in patients with hemodynamic compromise or respiratory failure.
+++
Magnetic Resonance Imaging
++
Gadolinium-enhanced magnetic resonance angiography (MRA) is also occasionally used to evaluate clinically suspected PE.108 Earlier studies demonstrated that when MRA was performed under optimal conditions, it appeared to be highly sensitive and specific even for segmental PE compared with pulmonary angiography.108,109 MRA has several attractive advantages over chest CT, including no requirement of ionizing radiation or iodinated contrast agents. Furthermore, magnetic resonance technology also allows detailed assessment of RV size and function, which is potentially important for the risk stratification of PE patients. Large-scale studies have been recently published.110,111 The results show that MRA, although promising, is not ready for current clinical practice in suspected PE as a result of its low sensitivity, the high proportion of nondiagnostic scans, required scanning time, cost, and low availability in most emergency settings. Notably, the substantially longer time required for image acquisition is an important limitation with current MRA sequences. This is especially relevant in the acute setting when acutely ill patients may not be able to tolerate the imaging protocol.
++
Transthoracic echocardiography (Fig. 75–4A-C) has emerged as a potentially important tool for risk assessment and treatment guidance in patients with acute PE. Acute PE may lead to RV pressure overload and dysfunction, which can be detected by echocardiography. Because of the reported negative predictive value of 40% to 50%, a normal result cannot exclude PE.112,113 On the other hand, signs of RV overload or dysfunction may also be found in the absence of acute PE and be a result of concomitant cardiac or respiratory disease.114 The presence of RV dysfunction on a baseline echocardiogram in normotensive patients appears to represent an independent predictor of an adverse outcome or early death.112,115,116 Patients with severe RV dysfunction may demonstrate McConnell’s sign, which is severe hypokinesis of the RV free wall combined with preserved systolic contraction of the RV apex.117 A disturbed RV ejection pattern (“60/60 sign”) consisting of an RV acceleration time ≤ 60 ms in the presence of a tricuspid insufficiency pressure gradient ≤ 60 mm Hg may suggest PE, but is not highly sensitive or specific. When combined, the two latter signs were 94% specific and 36% sensitive in diagnosing acute PE, even in the presence of pre-existing cardiorespiratory disease.118
++
++
Approximately 40% of normotensive patients with symptomatic PE have echocardiographic evidence of RV dysfunction. Echocardiography is also useful to diagnose patent foramen ovale in patients with suspected paradoxical embolism, directly visualizing thrombi in the main pulmonary artery (Fig. 75–4D), right heart chambers, and vena cava.115 In hemodynamically unstable (hypotensive) patients with suspected acute PE, the absence of echocardiographic signs of RV overload or dysfunction significantly reduces the likelihood of acute PE as the cause of the hemodynamic instability.113 In the latter case, echocardiography may be of further help in the differential diagnosis of the cause of shock, by detecting pericardial tamponade, acute valve dysfunction, severe global or regional LV dysfunction, aortic dissection, or hypovolemia. In a carefully evaluated hemodynamically compromised patient with suspected PE, unequivocal signs of RV pressure overload and dysfunction may justify the consideration of aggressive therapy for PE if immediate CT angiography is not feasible.119
+++
Imaging Studies for Deep Venous Thrombosis
++
In the majority of cases, PE originates from lower extremity DVT. A number of diagnostic techniques can be used to evaluate patients with suspected DVT. Compression ultrasonography (duplex ultrasonography) is the most commonly used technique. Impedance plethysmography is rarely used now even though number of important clinical trials have been performed using this now-outdated technique. Magnetic resonance imaging (MRI) appears to be very accurate, but has not generally been used as a first-line test because of cost and inconvenience. CT venography used alone or in conjunction with CT angiography appears to have a sensitivity and specificity equivalent to that of duplex ultrasonography, but exposes the patient to additional radiation. Contrast venography remains the gold standard, but is rarely used. Each diagnostic technique has advantages and limitations. Although diagnostic algorithms may be suggested for suspected DVT, these are institution specific, depending on resources and available expertise with certain techniques.
++
Compression ultrasonography with venous imaging is a portable, accurate, and widely available diagnostic technique for DVT.120 The primary criterion to diagnose DVT by ultrasonography is the noncompressibility of the vein. Combined with a Doppler reading, this technique is referred to as duplex ultrasonography. Ultrasound technology has been further improved by the development of color duplex instrumentation that displays Doppler frequency shifts as color superimposed on the grayscale image. The color duplex images display both mean blood flow velocity, expressed as a change in hue or saturation, and direction of blood flow, displayed as red or blue. Ultrasound imaging techniques can also identify or suggest the presence of pathology other than DVT, such as Baker cysts, hematomas, lymphadenopathy, arterial aneurysms, superficial thrombophlebitis, and abscesses.121 The sensitivity and specificity of ultrasonography for symptomatic proximal DVT has been well above 90% in most recent clinical trials.122,123,124,125 However, there are limitations, including lower sensitivity for asymptomatic DVT, operator dependence, the inability to accurately distinguish acute from chronic DVT in symptomatic patients, and the lower sensitivity for calf vein thrombosis. Compared with other technologies, ultrasonography is relatively inexpensive and is the preferred diagnostic modality for symptomatic suspected proximal DVT.
++
Proven DVT by ultrasonography is strongly suggestive of PE in patients with symptoms suggesting PE, but it is not diagnostic. However, approximately 50% of patients with CT-confirmed embolism have no imaging evidence of residual lower extremity venous thrombosis.122,123,126 Thus, PE cannot safely be ruled out in patients with suspected PE on the basis of a negative lower extremity duplex ultrasonography. In patients with suspected PE and a negative MDCT examination of the chest, the addition of duplex ultrasonography appears to only minimally increase diagnostic yield. Recent evidence suggests that multiorgan (lung, heart, and leg vein) ultrasonography might increase the accuracy of clinical pretest probability estimation in patients with suspected PE and may safely reduce the need for MDCT.127 This strategy seems to be especially helpful for patients with multiorgan ultrasonography negative for PE plus an alternative ultrasonographic diagnosis or a negative d-dimer result.
++
Contrast venography is a costly and invasive procedure that may result in superficial phlebitis or hypersensitivity reactions, but it is generally safe and accurate. Although contrast venography is the gold standard for DVT diagnosis, it is now rarely performed, except in clinical trials because of its higher sensitivity in detecting asymptomatic thrombi than duplex ultrasonography.128 Venography is far more often done during procedures such as insertion of an inferior vena cava (IVC) filter than to simply diagnose suspected DVT.
+++
Impedance Plethysmography
++
Impedance plethysmography has been used in patients presenting with suspected acute DVT, but is rarely obtained in the era of ultrasonography. It measures the change in blood volume in the calf while a thigh cuff is inflated and deflated. Its sensitivity for proximal DVT ranges around 65%. Impedance plethysmography may not detect nonocclusive proximal DVT or occlusive proximal DVT present in parallel venous systems, such as duplicated femoral or popliteal veins, and cannot detect DVT isolated to the calf veins.120
+++
Magnetic Resonance Imaging
++
Magnetic resonance venography (MRV) has clear advantages as a diagnostic test for suspected DVT, and appears to be an accurate, noninvasive alternative to venography.129 A major feature of this technique is excellent resolution of the IVC and pelvic veins.108 Preliminary experience with MRV suggests that it is at least as accurate as contrast venography or ultrasound imaging, and more sensitive than ultrasonography, for pelvic vein and calf-limited thrombosis.129,130 Simultaneous bilateral lower extremity imaging can be accomplished, and MRV appears to accurately distinguish acute from chronic DVT. This technique is also useful in differentiating other entities, such as cellulitis or a Baker cyst, from acute DVT. As with many other diagnostic techniques, its usefulness depends to a certain degree on the experience of the reader. As is the case with PE diagnosis, required scanning time and expense lessen its utility.