SYMPTOMS

Dyspnoea, pleuritic chest pain, and haemoptysis are the classic symptoms of PE. Most patients will have at least one of these symptoms, with dyspnoea being the most common. The combination of pleuritic chest pain and haemoptysis reflects a late presentation where pulmonary infarction has occurred. If syncope occurs there is a high likelihood that there has been a massive PE. A family history of venous thrombosis raises the possibility of inherited thrombophilia.

PHYSICAL SIGNS

Physical signs can be absent, but the most frequent sign is tachypnoea. Others include tachycardia, fever and signs of RV dysfunction (raised jugular venous pressure, parasternal heave and loud pulmonary component of the second heart sound). If massive PE is present, signs may include hypotension, pale mottled skin and peripheral or even central cyanosis. It is important also to examine for signs of DVT in the legs and arms.

D-DIMER

The serum D-dimer level is useful for exclusion of VTE, particularly when it is normal and combined with a low-risk clinical assessment.⁶ A variety of different assays are available. Negative tests, particularly from enzyme-linked immunosorbent assays (ELISA), are highly predictive of the absence of both DVT and PE.⁷ A high D-dimer concentration is also an independent predictive factor associated with mortality.⁸

Unfortunately, positive D-dimer tests are common in intensive care unit (ICU) patients who often suffer from one of the variety of conditions (such as trauma, surgery, diffuse intravascular coagulation, malignancy, acute myocardial infarction, pneumonia and heart failure) which also lead to a raised D-dimer, meaning that other tests are required to confirm or refute the diagnosis of VTE.

TROPONIN, BRAIN NATRIURETIC PEPTIDE (BNP) AND NT-TERMINAL PRO-BNP

Although of little use for diagnosis, measurement of troponin, BNP or NT-terminal Pro-BNP can add to the risk stratification of patients with known PE. The presence of a raised troponin is associated with haemodynamic instability in patients with non-massive PE independently of clinical, echocardiographic and laboratory findings.⁹ Raised troponin also predicts a higher mortality.¹⁰ Low levels of BNP and NT-terminal Pro-BNP have been shown to correlate with an uneventful course in patients with known PE.¹¹ NT-terminal Pro-BNP appears to be a better predictor of outcome than troponin.¹²

ARTERIAL BLOOD GASES

A normal arterial blood gas profile does not rule out the diagnosis; however hypoxaemia (with a widened alveolar–arterial oxygen gradient), hypocapnia and an increased end-tidal CO₂ gradient¹³ should raise the suspicion of PE, even though there are many other causes of these findings in critically ill patients.¹⁴ Metabolic acidosis may be present if shock from a large PE occurs.

ELECTROCARDIOGRAPH

A normal electrocardiograph (ECG) is found in about one-third of patients. Apart from sinus tachycardia (which is non-specific), the most frequent ECG abnormalities are non-specific S-T depression and T-wave inversion in the anterior leads, reflecting right heart strain. The pattern of a deep S-wave in lead I, and a Q-wave and inverted T-wave in lead III (S1Q3T3) is classical but is infrequently present. Other possible abnormalities include left- or right-axis deviation, P pulmonale, right bundle-branch block and atrial arrhythmias. The ECG is also useful in excluding acute myocardial infarction and pericarditis.

CHEST X-RAY

The chest X-ray is often normal or only slightly abnormal, with non-specific signs such as cardiac enlargement, pleural effusion, elevated hemidiaphragm, atelectasis and localised infiltrates. More specific findings, including focal oligaemia, a peripheral wedge-shaped density above the diaphragm or an enlarged right descending pulmonary artery,¹⁵ are uncommon and difficult for non-radiologists to identify. The chest X-ray is also useful in identifying an alternative diagnosis such as pneumothorax, pneumonia, acute pulmonary oedema, rib fracture and pleural effusion.

COMPUTED TOMOGRAPHY

As CT technology has improved, CT angiography (CTA) has emerged as a cost-effective and clinically reliable alternative to the V/Q scan.¹⁶ The single-detector row CT has been superseded by multidetector row CT, which allows imaging of the entire chest with high-resolution images in ‘in-plane’ and ‘through-plane’ resolution. High-resolution images to the level of segmental and in some cases subsegmental pulmonary arteries can be obtained in a short time period (often a single breath-hold).

When CTA is compared to conventional angiography it appears reliable, with sensitivity, specificity and accuracy of 100%, 89% and 91% respectively.⁵ Adding venography of the leg veins to the CTA (CTA-CTV) further increases the diagnostic certainty.¹⁷ It is therefore recommended that CTA or CTA-CTV should be the principal radiological test for patients with high and moderate probability of PE.¹⁸

Although the ability of many CT scanners to detect PE at the subsegmental level is limited, it is debatable whether this is of clinical significance. Patients who have negative or indeterminate CT scans and in whom anticoagulation is withheld have low subsequent rates of thromboembolic events.¹⁹

The advantage of CT is that it can not only diagnose PE but can also be used to assess severity of the condition. Increased RV/LV ratio (> 0.9)²⁰ and clot in the proximal branches of the pulmonary artery²¹ correlate with the clinical severity of PE. Severity stratification is further increased by combining CT with other tests such as troponin⁹ and BNP or NT-terminal ProBNP.¹¹ CT scanning may also identify the causative DVT in the veins of the legs, pelvis and abdomen or detect alternative or additional diagnoses such as a pulmonary mass, pneumonia, emphysema, pneumothorax, pleural effusion or mediastinal adenopathy (Figure 30.2).

Figure 30.2 Computed tomography (CT) scanning and pulmonary embolism (PE). Three CT images from the same patient demonstrating the ability of CT to detect pulmonary emboli, assess severity and confirm additional diagnoses. Image 1 (mediastinal window): filling defects in the right pulmonary artery and inferior branch of the left pulmonary artery. A pulmonary artery catheter is also in situ. Image 2 (lung window): a left pneumothorax with pleural adhesions, and a small area of left-sided posterior consolidation. Image 3 (mediastinal window): dilatation of the right ventricle relative to the left ventricle.

ECHOCARDIOGRAPHY

Because of its portability, echocardiography has become increasingly useful in critically ill patients with possible PE. Many patients with PE have an echocardiographic abnormality, the most common being RV dilatation, RV hypokinesis, paradoxical interventricular septal motion toward the LV, tricuspid regurgitation and pulmonary hypertension.²² The pattern of RV hypokinesis with apical sparing is considered pathognomonic for PE.²³ The presence of RV dysfunction correlates with mortality.²⁴ Echocardiography is poor at excluding PE, as a negative echocardiogram can miss up to 50% of PEs.²⁵

Transthoracic echocardiography will also allow estimation of pulmonary arterial pressure, identification of intracardiac thrombi (which usually requires surgical embolectomy) and aids in differential diagnosis by excluding aortic dissection and pericardial tamponade. Transoesophageal echocardiography has the additional benefit of directly identifying embolus in the proximal pulmonary arteries, which is common in patients with haemodynamically significant PE.²⁶

Echocardiography has its best application in haemodynamically unstable patients, where it can be rapidly brought to the patient. If the patient has RV dilatation and hypokinesis in the right clinical setting, PE is extremely likely.

VENTILATION/PERFUSION SCANNING

The lung ventilation/perfusion (V/Q) scan is becoming less commonly used with the advent of multidetector row CT. The perfusion scan identifies defects in perfusion which may be classified as single or multiple, and as subsegmental, segmental or lobar. By combining this with ventilation scanning, these perfusion defects can then be labelled as mismatched (normal ventilation in a zone of perfusion defect) or matched defects (ventilation defect corresponds to perfusion defect). The probability of PE is then classified as high, intermediate or low probability for PE, or normal.²⁷

A normal lung scan effectively excludes PE and treatment can be safely withheld. A high-probability scan is considered diagnostic for PE. The majority of patients do not have such clear-cut scan results, and even a low-probability scan cannot satisfactorily exclude PE, as a patient with a low-probability scan but with a high level of clinical suspicion has a 40% chance of having a PE.

Despite the increased use of CT, the V/Q scan retains a role when CT is either unavailable or contraindicated (e.g. significant renal impairment, anaphylaxis to intravenous contrast or pregnancy).¹⁵V/Q scanning also allows quantification of regional blood flow within the lungs, which may be required in the assessment of chronic pulmonary venous embolism.

MAGNETIC RESONANCE IMAGING

PE may also be diagnosed by gadolinium-enhanced magnetic resonance pulmonary angiography. So far, this has proven more useful in the assessment of chronic VTE.

PULMONARY ANGIOGRAPHY

The pulmonary angiogram is regarded as the only test which can exclude PE with relative certainty. It will detect most PEs even at the subsegmental level. However, it is unavailable in many centres.

SEARCH FOR DEEP VENOUS THROMBOSIS

Doppler ultrasound has been recommended to search for DVT in the leg veins, from where over 90% of emboli originate. If a leg DVT is confirmed, anticoagulation is required unless the DVT is entirely below the knee, where the associated morbidity is low. Ultrasound is highly accurate in symptomatic or proximal DVT, although in asymptomatic patients, ultrasound is much less likely to find DVT, meaning the absence of a DVT does not exclude PE.

Leg venography is a more sensitive test; however, it is invasive and now uncommonly performed.