Pulmonary embolism

Pulmonary embolism (PE) is a blockage of the main artery of the lung or one of its branches by a substance that has travelled from elsewhere in the body through the bloodstream (embolism). Usually this is due to embolism of a thrombus (blood clot) from the deep veins in the legs, a process termed venous thromboembolism. A small proportion is due to the embolization of air, fat, talc in drugs of intravenous drug abusers or amniotic fluid. The obstruction of the blood flow through the lungs and the resultant pressure on the right ventricle of the heart leads to the symptoms and signs of PE. The risk of PE is increased in various situations, such as cancer or prolonged bed rest.

Symptoms of pulmonary embolism include difficulty breathing, chest pain on inspiration, and palpitation. Clinical signs include low blood oxygen saturation and cyanosis, rapid breathing, and a rapid heart rate. Severe cases of PE can lead to collapse, abnormally low blood pressure, and sudden death.

Diagnosis is based on these clinical findings in combination with laboratory tests (such as the D-dimer test) and imaging studies, usually CT pulmonary angiography. Treatment is typically with anticoagulant medication, including heparin and warfarin. Severe cases may require thrombolysis with drugs such as tissue plasminogen activator (tPA) or may require surgical intervention via pulmonary thrombectomy.

Signs and symptoms
Symptoms of PE are sudden-onset dyspnea (shortness of breath), tachypnea (rapid breathing), chest pain of a "pleuritic" nature (worsened by breathing), cough and hemoptysis (coughing up blood). More severe cases can include signs such as cyanosis (blue discoloration, usually of the lips and fingers), collapse, and circulatory instability due to decreased blood flow through the lungs and into the left side of the heart. About 15% of all cases of sudden death are attributable to PE.

On physical examination, the lungs are usually normal. Occasionally, a pleural friction rub may be audible over the affected area of the lung (mostly in PE with infarct). A pleural effusion is sometimes present that is transudative, detectable by decreased percussion note, audible breath sounds and vocal resonance. Strain on the right ventricle may be detected as a left parasternal heave, a loud pulmonary component of the second heart sound, and raised jugular venous pressure. A low-grade fever may be present, particularly if there is associated pulmonary hemorrhage or infarction.

More rarely, inability of the right ventricle to remove fluid from the tissues leads to fluid accumulation in the legs (peripheral edema), congestion of the liver with mild jaundice and tenderness, and ascites (fluid in the abdominal cavity).

Risk factors
The most common sources of embolism are proximal leg deep venous thrombosis (DVTs) or pelvic vein thromboses. Any risk factor for DVT also increases the risk that the venous clot will dislodge and migrate to the lung circulation, which happens in up to 15% of all DVTs. The conditions are generally regarded as a continuum termed venous thromboembolism (VTE).

The development of thrombosis is classically due to a group of causes named Virchow's triad (alterations in blood flow, factors in the vessel wall and factors affecting the properties of the blood). Often, more than one risk factor is present.
 * Alterations in blood flow: immobilization (after surgery, injury or long-distance air travel), pregnancy (also procoagulant), obesity (also procoagulant), cancer (also procoagulant)
 * Factors in the vessel wall: of limited direct relevance in VTE
 * Factors affecting the properties of the blood (procoagulant state):
 * Estrogen-containing hormonal contraception
 * Genetic thrombophilia (factor V Leiden, prothrombin mutation G20210A, protein C deficiency, protein S deficiency, antithrombin deficiency, hyperhomocysteinemia and plasminogen/fibrinolysis disorders)
 * Acquired thrombophilia (antiphospholipid syndrome, nephrotic syndrome, paroxysmal nocturnal hemoglobinuria)
 * Cancer (due to secretion of pro-coagulants)

Diagnosis
The diagnosis of PE is based primarily on validated clinical criteria combined with selective testing because the typical clinical presentation (shortness of breath, chest pain) cannot be definitively differentiated from other causes of chest pain and shortness of breath. The decision to do medical imaging is usually based on clinical grounds, i.e. the medical history, symptoms and findings on physical examination, followed by an assessment of clinical probability.

The most commonly used method to predict clinical probability, the Wells score, is a clinical prediction rule, whose use is complicated by multiple versions being available. In 1995, Wells et al. initially developed a prediction rule (based on a literature search) to predict the likelihood of PE, based on clinical criteria. The prediction rule was revised in 1998 This prediction rule was further revised when simplified during a validation by Wells et al. in 2000. In the 2000 publication, Wells proposed two different scoring systems using cutoffs of 2 or 4 with the same prediction rule. In 2001, Wells published results using the more conservative cutoff of 2 to create three categories. An additional version, the "modified extended version", using the more recent cutoff of 2 but including findings from Wells's initial studies were proposed. Most recently, a further study reverted to Wells's earlier use of a cutoff of 4 points to create only two categories.

There are additional prediction rules for PE, such as the Geneva rule. More importantly, the use of any rule is associated with reduction in recurrent thromboembolism.

The Wells score:
 * clinically suspected DVT - 3.0 points
 * alternative diagnosis is less likely than PE - 3.0 points
 * tachycardia - 1.5 points
 * immobilization/surgery in previous four weeks - 1.5 points
 * history of DVT or PE - 1.5 points
 * hemoptysis - 1.0 points
 * malignancy (treatment for within 6 months, palliative) - 1.0 points

Traditional interpretation
 * Score >6.0 - High (probability 59% based on pooled data )
 * Score 2.0 to 6.0 - Moderate (probability 29% based on pooled data )
 * Score <2.0 - Low (probability 15% based on pooled data )

Alternate interpretation
 * Score > 4 - PE likely. Consider diagnostic imaging.
 * Score 4 or less - PE unlikely. Consider D-dimer to rule out PE.

Blood tests
Early primary research has shown that in low/moderate suspicion of PE, a normal D-dimer level (shown in a blood test) is enough to exclude the possibility of thrombotic PE. This has been corroborated by a recent systematic review of studies of patients with low pre-test probability (PTP) of PE and negative D-dimer results that found the three month risk of thromboembolic events in patients excluded in this manner was 0.14%, with 95% confidence intervals from 0.05 to 0.41%, though this review was limited by its use of only one randomized-controlled clinical trial, the remainder of studies being prospective cohorts. D-dimer is highly sensitive but not very specific (specificity around 50%). In other words, a positive D-dimer is not synonymous with PE, but a negative D-dimer is, with a good degree of certainty, an indication of absence of a PE.

When a PE is being suspected, a number of blood tests are done, in order to exclude important secondary causes of PE. This includes a full blood count, clotting status (PT, aPTT, TT), and some screening tests (erythrocyte sedimentation rate, renal function, liver enzymes, electrolytes). If one of these is abnormal, further investigations might be warranted.

Imaging
The gold standard for diagnosing pulmonary embolism (PE) is pulmonary angiography. Pulmonary angiography is used less often due to wider acceptance of CT scans, which are non-invasive.

CT pulmonary angiography (CTPA) is a pulmonary angiogram obtained using computed tomography (CT) with radiocontrast rather than right heart catheterization. Its advantages are clinical equivalence, its non-invasive nature, its greater availability to patients, and the possibility of identifying other lung disorders from the differential diagnosis in case there is no pulmonary embolism. Assessing the accuracy of CT pulmonary angiography is hindered by the rapid changes in the number of rows of detectors available in multidetector CT (MDCT) machines. According to a cohort study, single-slice spiral CT may help diagnose detection among patients with suspected pulmonary embolism. In this study, the sensitivity was 69% and specificity was 84%. In this study which had a prevalence of detection was 32%, the positive predictive value of 67.0% and negative predictive value of 85.2% (click here to adjust these results for patients at higher or lower risk of detection). However, this study's results may be biased due to possible incorporation bias, since the CT scan was the final diagnostic tool in patients with pulmonary embolism. The authors noted that a negative single slice CT scan is insufficient to rule out pulmonary embolism on its own. A separate study with a mixture of 4 slice and 16 slice scanners reported a sensitivity of 83% and a specificity of 96%. This study noted that additional testing is necessary when the clinical probability is inconsistent with the imaging results. CTPA is non-inferior to VQ scanning, and identifies more emboli (without necessarily improving the outcome) compared to VQ scanning.
 * Non-invasive imaging

Ventilation/perfusion scan (or V/Q scan or lung scintigraphy), which shows that some areas of the lung are being ventilated but not perfused with blood (due to obstruction by a clot). This type of examination is used less often because of the more widespread availability of CT technology, however, it may be useful in patients who have an allergy to iodinated contrast or in pregnancy due to lower radiation exposure than CT.

Tests that are frequently done that are not sensitive for PE, but can be diagnostic.
 * Low probability diagnostic tests/non-diagnostic tests
 * Chest X-rays are often done on patients with shortness of breath to help rule-out other causes, such as congestive heart failure and rib fracture. Chest X-rays in PE are rarely normal, but usually lack signs that suggest the diagnosis of PE (e.g. Westermark sign, Hampton's hump).
 * Ultrasonography of the legs, also known as leg doppler, in search of deep venous thrombosis (DVT). The presence of DVT, as shown on ultrasonography of the legs, is in itself enough to warrant anticoagulation, without requiring the V/Q or spiral CT scans (because of the strong association between DVT and PE). This may be valid approach in pregnancy, in which the other modalities would increase the risk of birth defects in the unborn child. However, a negative scan does not rule out PE, and low-radiation dose scanning may be required if the mother is deemed at high risk of having pulmonary embolism.

Electrocardiogram
An electrocardiogram (ECG) is routinely done on patients with chest pain to quickly diagnose myocardial infarctions (heart attacks). An ECG may show signs of right heart strain or acute cor pulmonale in cases of large PEs - the classic signs are a large S wave in lead I, a large Q wave in lead III and an inverted T wave in lead III ("S1Q3T3"). This is occasionally (up to 20%) present, but may also occur in other acute lung conditions and has therefore limited diagnostic value. The most commonly seen signs in the ECG is sinus tachycardia, right axis deviation and right bundle branch block. Sinus tachycardia was however still only found in 8 - 69% of people with PE.

Echocardiography
In massive and submassive PE, dysfunction of the right side of the heart can be seen on echocardiography, an indication that the pulmonary artery is severely obstructed and the heart is unable to match the pressure. Some studies (see below) suggest that this finding may be an indication for thrombolysis. Not every patient with a (suspected) pulmonary embolism requires an echocardiogram, but elevations in cardiac troponins or brain natriuretic peptide may indicate heart strain and warrant an echocardiogram.

The specific appearance of the right ventricle on echocardiography is referred to as the McConnell's sign. This is the finding of akinesia of the mid-free wall but normal motion of the apex. This phenomenon has a 77% sensitivity and a 94% specificity for the diagnosis of acute pulmonary embolism in the setting of right ventricular dysfunction.

Algorithms
Recent recommendations for a diagnostic algorithm have been published by the PIOPED investigators; however, these recommendations do not reflect research using 64 slice MDCT. These investigators recommended:
 * Low clinical probability. If negative D-dimer, PE is excluded. If positive D-dimer, obtain MDCT and based treatment on results.
 * Moderate clinical probability. If negative D-dimer, PE is excluded. However, the authors were not concerned that a negative MDCT with negative D-dimer in this setting has an 5% probability of being false. Presumably, the 5% error rate will fall as 64 slice MDCT is more commonly used. If positive D-dimer, obtain MDCT and based treatment on results.
 * High clinical probability. Proceed to MDCT. If positive, treat, if negative, additional tests are needed to exclude PE.

The Pulmonary Embolism Rule-out Criteria, or PERC rule, helps assess people in whom pulmonary embolism is suspected, but unlikely. Unlike the Wells Score and Geneva score, which are clinical prediction rules intended to risk stratify patients with suspected PE, the PERC rule is designed to rule out risk of PE in patients when the physician has already stratified them into a low-risk category.
 * Pulmonary Embolism Rule-out Criteria

Patients in this low risk category without any of these criteria may undergo no further diagnostic testing for PE: Hypoxia - Sa02 <95%, unilateral leg swelling, hemoptysis, prior DVT or PE, recent surgery or trauma, age >50, hormone use, tachycardia. The rationale behind this decision is that further testing (specifically CT angiogram of the chest) may cause more harm (from radiation exposure and contrast dye) than the risk of PE. The PERC rule has a sensitivity of 97.4% and specificity of 21.9% with a false negative rate of 1.0% (16/1666).

Treatment
In most cases, anticoagulant therapy is the mainstay of treatment. Acutely, supportive treatments, such as oxygen or analgesia, are often required.

Anticoagulation
In most cases, anticoagulant therapy is the mainstay of treatment. Heparin, low molecular weight heparins (such as enoxaparin and dalteparin), or fondaparinux is administered initially, while warfarin, acenocoumarol, or phenprocoumon therapy is commenced (this may take several days, usually while the patient is in the hospital). Low molecular weight heparin may reduce bleeding among patients with pulmonary embolism as compared to heparin according to a systematic review of randomized controlled trials by the Cochrane Collaboration. The relative risk reduction was 40.0%. For patients at similar risk to those in this study (2.0% had bleeding when not treated with low molecular weight heparin), this leads to an absolute risk reduction of 0.8%. 125.0 patients must be treated for one to benefit.

It is possible to treat low risk patients (risk class I or class II) as outpatients. A randomised trial of 344 patients (171 outpatients and 168 inpatients) found that outcomes were equivalent whether patients were treated in hospital or at home (there was one death at 90 days in each group). This confirms the findings of an earlier systematic review of observational studies.

Warfarin therapy often requires frequent dose adjustment and monitoring of the INR. In PE, INRs between 2.0 and 3.0 are generally considered ideal. If another episode of PE occurs under warfarin treatment, the INR window may be increased to e.g. 2.5-3.5 (unless there are contraindications) or anticoagulation may be changed to a different anticoagulant e.g. low molecular weight heparin. In patients with an underlying malignancy, therapy with a course of low molecular weight heparin may be favored over warfarin based on the results of the CLOT trial.

Similarly, pregnant women are often maintained on low molecular weight heparin to avoid the known teratogenic effects of warfarin, especially in the early stages of pregnancy.

People are usually admitted to hospital in the early stages of treatment, and tend to remain under inpatient care until INR has reached therapeutic levels. Increasingly, low-risk cases are managed on an outpatient basis in a fashion already common in the treatment of DVT.

Warfarin therapy is usually continued for 3–6 months, or "lifelong" if there have been previous DVTs or PEs, or none of the usual risk factors is present. An abnormal D-dimer level at the end of treatment might signal the need for continued treatment among patients with a first unprovoked pulmonary embolus.

Thrombolysis
Massive PE causing hemodynamic instability (shock and/or hypotension, defined as a systolic blood pressure <90 mmHg or a pressure drop of 40 mmHg for>15 min if not caused by new-onset arrhythmia, hypovolemia or sepsis) is an indication for thrombolysis, the enzymatic destruction of the clot with medication. It is the best available medical treatment in this situation and is supported by clinical guidelines.

The use of thrombolysis in non-massive PEs is still debated. The aim of the therapy is to dissolve the clot, but there is an attendant risk of bleeding or stroke. The main indication for thrombolysis is in submassive PE where right ventricular dysfunction can be demonstrated on echocardiography, and the presence of visible thrombus in the atrium.

Surgery
Surgical management of acute pulmonary embolism (pulmonary thrombectomy) is uncommon and has largely been abandoned because of poor long-term outcomes. However, recently, it has gone through a resurgence with the revision of the surgical technique and is thought to benefit selected patients.

Chronic pulmonary embolism leading to pulmonary hypertension (known as chronic thromboembolic hypertension) is treated with a surgical procedure known as a pulmonary thromboendarterectomy.

Inferior vena cava filter
If anticoagulant therapy is contraindicated and/or ineffective, or to prevent new emboli from entering the pulmonary artery and combining with an existing blockage, an inferior vena cava filter may be implanted.

Prognosis
Mortality from untreated PE is said to be 26%. This figure comes from a trial published in 1960 by Barrit and Jordan, which compared anticoagulation against placebo for the management of PE. Barritt and Jordan performed their study in the Bristol Royal Infirmary in 1957. This study is the only placebo controlled trial ever to examine the place of anticoagulants in the treatment of PE, the results of which were so convincing that the trial has never been repeated as to do so would be considered unethical. That said, the reported mortality rate of 26% in the placebo group is probably an overstatement, given that the technology of the day may have detected only severe PEs.

Prognosis depends on the amount of lung that is affected and on the co-existence of other medical conditions; chronic embolisation to the lung can lead to pulmonary hypertension. After a massive PE, the embolus must be resolved somehow if the patient is to survive. In thrombotic PE, the blood clot may be broken down by fibrinolysis, or it may be organized and recanalized so that a new channel forms through the clot. Blood flow is restored most rapidly in the first day or two after a PE. Improvement slows thereafter and some deficits may be permanent. There is controversy over whether or not small subsegmental PEs need to be treated at all and some evidence exists that patients with subsegmental PEs may do well without treatment.

Once anticoagulation is stopped, the risk of a fatal pulmonary embolism is 0.5% per year.

Predicting mortality
The PESI and Geneva prediction rules can estimate mortality and so may guide selection of patients who can be considered for outpatient therapy.

Underlying causes
After a first PE, the search for secondary causes is usually brief. Only when a second PE occurs, and especially when this happens while still under anticoagulant therapy, a further search for underlying conditions is undertaken. This will include testing ("thrombophilia screen") for Factor V Leiden mutation, antiphospholipid antibodies, protein C and S and antithrombin levels, and later prothrombin mutation, MTHFR mutation, Factor VIII concentration and rarer inherited coagulation abnormalities.