Pulmonary embolism

Overview 
Theory 
Diagnosis 
Management 
Follow up 
Resources 

Hemodynamically unstable patients require urgent primary reperfusion, anticoagulation, and supportive care.[19] Approaches include systemic thrombolysis, catheter-directed thrombolysis, pharmaco-mechanical intervention, mechanical thrombectomy, and surgical thrombectomy.[200][201][202] Extracorporeal membrane oxygenation (ECMO) may be considered to stabilize patients as the right heart recovers from strain due to severe pulmonary vascular resistance.[203]

For patients at intermediate risk of a poor outcome, anticoagulation and ongoing monitoring is required. Reperfusion is generally employed as a rescue therapy if decompensation occurs.

Classification of severity and prognostic stratification

Several classification systems have been employed to describe the severity of pulmonary embolism (PE), and short-term mortality may be assessed by clinical prediction tools such as the Pulmonary Embolism Severity Index (PESI) or simplified PESI scores.[204]

The term "submassive" has been applied to PE with significant anatomic extent, but with normotension, and massive describes anatomically extensive PE complicated by shock or hypotension.

The European Society of Cardiology categorizes severity and risk of early (in-hospital or 30 day) death from PE as:[3]

High risk when presenting with hemodynamic instability (shock or hypotension).
Intermediate-high risk when presenting without hemodynamic instability but with evidence of right ventricular (RV) dysfunction on imaging, abnormal cardiac biomarkers and clinical parameters of severity (such as a high PESI score).
Intermediate-low risk when presenting without hemodynamic instability, with either evidence of RV dysfunction on imaging or elevated cardiac biomarkers (but not both) and clinical parameters of severity (such as a high PESI score).
Low risk when none of these factors are present.

Patients with confirmed PE without shock or hypotension require further risk stratification: for example with the Pulmonary Embolism Severity Index (PESI) or the simplified Pulmonary Embolism Severity Index (sPESI).[3][124]

PESI classifies patients with confirmed PE without shock or hypotension into 1 of 5 risk categories associated with increasing 30-day mortality. PESI risk category is derived from the sum of points for 11 clinical criteria; sPESI has only 6 criteria, and reports risk stratification dichotomously (low [0 points] or high [≥1 point(s)] risk of 30-day mortality).

Studies indicate that PESI and sPESI predict short-term mortality with comparable accuracy, but the latter is easier to use.[205][206] One meta-analysis that evaluated the prognostic utility of PESI/sPESI for all-cause mortality reported pooled sensitivity and pooled specificity of 91% and 41%, respectively.[206]

PESI has been used to identify patients eligible for outpatient management in prospective studies.[207][208][209] Based upon social background and likely compliance with treatment, European Society of Cardiology guidelines suggest that low-risk patients (PESI class I or class II), and potentially those with a sPESI score of 0, can be considered for early discharge and outpatient management.[3][208][209]

Overall, a PESI of class I-II or an sPESI of 0 is a reliable predictor of low-risk PE. If the calculated PESI or sPESI is low but there are signs of cardiac dysfunction present on transthoracic echo (TTE) or cardiac biomarkers are elevated, the patients should be considered intermediate-risk.[3][Figure caption and citation for the preceding image starts]: PESI criteria and risk stratificationCreated by BMJ Knowledge Centre [Citation ends]. com.bmj.content.model.Caption@e35861 [Figure caption and citation for the preceding image starts]: sPESI criteria and risk stratificationCreated by BMJ Knowledge Centre [Citation ends].

RV function assessed by echocardiography and cardiac troponin testing should be considered in patients with PESI risk stratification ≥III or sPESI ≥1. In patients with confirmed PE without shock or hypotension: RV dysfunction is predictive of adverse outcome and enables further risk stratification; elevated troponin levels are associated with increased risk for short-term mortality, PE-related mortality, and serious adverse events.[176][177][178][210][211]

Intermediate high-risk patients

Patients with PESI risk stratification ≥III, or sPESI ≥1, with RV dysfunction and a positive cardiac troponin test belong to an intermediate high-risk group.[3] Rescue thrombolysis may be indicated in intermediate high-risk patients, and in patients with other clinical features of cardiopulmonary impairment (e.g., elevated heart rate, respiratory rate, jugular venous pressure) who have started anticoagulant therapy, and:[3][19][212]

are deteriorating (as seen by a decrease in systolic BP, increase in heart rate, worsening gas exchange, signs of inadequate perfusion, worsening RV function, or increasing cardiac biomarkers), but have not yet developed hypotension
are exhibiting signs of hemodynamic decompensation (e.g., systolic blood pressure [BP] <90 mmHg for at least 15 minutes, or drop of systolic BP by at least 40 mmHg for at least 15 minutes with signs of end organ hypoperfusion).

Consideration of bleeding risk will inform choice of thrombolytic therapy.

Intermediate low-risk patients[3]

Normotensive patients with PESI risk stratification ≥III, or sPESI ≥1, with normal echocardiogram and/or negative cardiac biomarkers troponin test are considered to be intermediate low-risk.
Intermediate low-risk patients should be hospitalized (even in the absence of RV dysfunction).

Suspected PE with shock or hypotension

High-risk patients (presenting with shock or hypotension [i.e., systolic BP <90 mmHg]) require aggressive treatment with supportive therapy, primary reperfusion, and anticoagulation.

Supportive therapies and empiric anticoagulation (unless contraindicated) should be instituted without delay.[19] Unfractionated heparin (UFH) may be preferred in this population; most clinical studies of interventional therapies have used heparin as the anticoagulant component of the regimen.[3]

Supportive therapies

Local resuscitation protocols should be followed.

Respiratory support

Supplemental high-concentration oxygen should be administered, targeting oxygen saturations of 94% to 98% (or 88% to 92% in patients at risk of hypercapnic respiratory failure).[213]
Intubation and mechanical ventilation may be necessary for patients with severe hypoxemia/respiratory failure. Mechanical ventilation can lead to hypotension, so BP should be monitored closely.
Tracheal intubation: animated demonstration
How to insert a tracheal tube in an adult using a laryngoscope.
Bag-valve-mask ventilation: animated demonstration
How to use bag-valve-mask apparatus to deliver ventilatory support to adults. Video demonstrates the two-person technique.
ECMO therapy may be employed in patients with high-risk PE, usually in conjunction with reperfusion therapies.[203]

Intravenous fluids

If systolic BP is <90 mmHg, intravenous fluids should be given. Acute RV failure with resulting low systemic output is the leading cause of death in patients with PE.[3]
Studies indicate that aggressive volume expansion is of no benefit, and may even worsen RV function by causing mechanical overstretch, or by reflex mechanisms that depress contractility. However, modest fluid challenge (i.e., 500 mL crystalloid) may be of benefit in patients with PE, a low cardiac index, and normal BP.[217]

Vasoactive agents

If systolic BP is <90 mmHg, vasopressors may be given in parallel with (or while waiting for) pharmacologic, surgical, or interventional reperfusion treatment.[3]
Norepinephrine (noradrenaline) may improve RV function and RV coronary perfusion.[3] However, its use should probably be limited to hypotensive patients.[3]
Dobutamine enhances contractility with an increase in stroke volume and cardiac output. However, its systemic vasodilatory effect can lead to hypotension.[3]
Epinephrine (adrenaline) combines the beneficial properties of norepinephrine (vasoconstriction with increased RV perfusion, positive inotropy) and dobutamine (positive inotropy), but without the vasodilatory effects associated with the latter.[218]

Primary reperfusion in patients with shock or hypotension

Systemic thrombolytic therapy

Systemic thrombolytic therapy is recommended in patients with hemodynamic compromise (shock, systolic BP <90 mmHg, or vasopressor requirement to maintain systolic BP >90 mmHg), as this patient group has a high mortality rate.[3][19][22][124][219][220][221][222][223] The American College of Chest Physicians (ACCP) recommends systemic thrombolytic therapy (unless contraindicated) using a peripheral vein for patients with acute PE associated with hypotension who do not have a high bleeding risk. The ACCP does not make specific recommendations on preferred agents due to lack of comparative data.[19]

The preferred thrombolytic agents are alteplase or reteplase; tenecteplase is an alternative option.[221][222][223]

Systemic thrombolytic therapy is associated with lower all-cause mortality than anticoagulation alone in patients with high-risk (massive) PE (acute PE with sustained hypotension [i.e., systolic BP <90 mmHg for at least 15 minutes]).[3][219][220][224]

Ideally, PE should be confirmed by imaging before thrombolytic therapy is administered.[124] However, if the patient is at risk of imminent cardiac arrest, treatment may be commenced on clinical grounds alone.[126]

Systemic thrombolytic therapy induces prompt clot dissolution and improves RV function, pulmonary blood flow, and lung perfusion.[3][224] Thrombolysis plus heparin was associated with significantly reduced 30-day mortality compared with heparin alone (2.3% [24/1033] vs. 3.9% [40/1024], respectively; pooled odds ratio [OR] 0.59, 95% CI 0.36 to 0.96, P=0.03) in a meta-analysis of patients with acute PE.[224]

Thrombolysis is, however, associated with a significantly increased risk of major and minor bleeding, including hemorrhagic stroke.[219][220][224] More patients receiving thrombolytic therapy plus heparin experienced a major bleeding episode compared with those taking an anticoagulant alone (9.9% [96/974] vs. 3.6% [35/961], respectively; OR 2.91, 95% CI 1.95 to 4.36).[224] The reported incidence of intracranial or fatal hemorrhage was 1.7% in the thrombolysis group and 0.3% in the anticoagulant group.[224]

Absolute contraindications to thrombolysis include: hemorrhagic stroke or stroke of unknown origin at any time; ischemic stroke in the preceding 6 months; central nervous system damage or neoplasms; recent major trauma/surgery/head injury (in the preceding 3 weeks); gastrointestinal (GI) bleeding within the last month; known bleeding risk.[3][225]

Relative contraindications to thrombolysis include: transient ischemic attack in the preceding 6 months; oral anticoagulant therapy; pregnancy, or within 1 week postpartum; traumatic resuscitation (in relation to this episode of PE); refractory hypertension (systolic BP >180 mmHg); advanced liver disease; infective endocarditis; active peptic ulcer.[3][225]

Thrombolytic therapy is not typically recommended for hemodynamically stable patients with acute PE.[3][19] In a randomized double-blind trial, primary reperfusion thrombolytic therapy plus heparin in normotensive patients with intermediate-risk PE (acute RV dysfunction and myocardial injury without overt hemodynamic compromise) prevented hemodynamic decompensation compared with heparin alone, but increased the risk of major hemorrhage and stroke.[221]

Following thrombolysis, anticoagulation therapy should be continued. Depending on the degree of concern for post-procedure bleeding, treatment with UFH can be resumed, followed by conversion to treatment-phase therapy when bleeding risk remits, or treatment-phase therapy can be started immediately.

Surgical embolectomy or catheter-directed therapies

Systemic thrombolytic therapy increases bleeding risk, including that of intracranial bleeding.[19][224] Surgical pulmonary embolectomy and catheter-directed therapy (which typically involves either or a combination of mechanical and pharmacotherapeutic thrombus fragmentation) likely have lower attendant bleeding risk than systemic therapy.[19][212][226][227][228]

Surgical pulmonary embolectomy is recommended for patients in whom systemic thrombolytic therapy has failed or is absolutely contraindicated.[3][227] Mortality rates following pulmonary embolectomy range from 4% to 27%.[229] In a small cohort of patients who underwent surgical pulmonary embolectomy for acute massive pulmonary thromboembolism, the 10-year survival rate was 84%.[230]

Catheter-directed therapy, which typically involves a combination of mechanical and pharmacotherapeutic thrombus fragmentation, may be considered for patients with acute PE associated with hypotension who also have a high bleeding risk, failed systemic thrombolysis, or shock that is likely to cause death before systemic thrombolysis can take effect (e.g., within hours), if appropriate expertise and resources are available.[19] Catheter-directed therapy uses a lower dose of thrombolytic drug (approximately one third of full-dose systemic thrombolytic therapy) and is believed to reduce the risks of bleeding at remote sites (e.g., intracranially or gastrointestinally).[19] One meta-analysis of nonrandomized trials of catheter-directed therapies reported a clinical success rate of 87% with an associated risk of major and minor complications of 2% and 8%, respectively.[231] Evidence is limited by small studies, study design (i.e., nonrandomized), and use of intermediate efficacy end points.[228][232]

Following catheter-directed therapy, anticoagulation therapy should be continued. Depending on the degree of concern for post-procedure bleeding, treatment with UFH can be resumed, followed by conversion to treatment-phase therapy when bleeding risk remits, or treatment-phase therapy can be started immediately.

Anticoagulation: general principles

Anticoagulation is the mainstay of therapy for the treatment of most patients with PE (including ongoing therapy for those with severe disease who undergo interventional treatments). Patients are treated with anticoagulants to:

Prevent propagation/progression of the thrombus in the pulmonary arteries
Reduce the risk of further PE from coexisting deep vein thrombosis (DVT)

Anticoagulant therapy for venous thromboembolism (VTE) has been described in three phases: initiation, treatment (also referred to as "long-term"), and extended.[3][19][45][233]

Initiation (from suspected diagnosis to up to 5-21 days following diagnosis): goals of care are to arrest the active prothrombotic state and to inhibit thrombus propagation and embolization
Treatment (initiation to 3 months): goals are to prevent new thrombus while the original clot is stabilized and intrinsic thrombolysis is under way
Extended (3 months to indefinite): goal is secondary prevention of new VTE

The recommended treatment regimens for patients with PE have changed rapidly as newer anticoagulants have become available. Care should be taken to minimize the risk of major hemorrhage throughout the treatment period and monitor for the development of heparin-induced thrombocytopenia (HIT) if UFH or low molecular weight heparin (LMWH) is used.[83][234]

Initiation phase of anticoagulation (from suspected diagnosis to up to 21 days)

Patients diagnosed with PE (or who are suspected of PE and have a high probability of disease) should receive an anticoagulant based on clinical stability and dosed according to the initiation phase of therapy, unless contraindicated.[3][19][235]

The choice of agent depends on patient factors such as clinical stability, bleeding risk, hepatic function, renal function, pregnancy, presence of cancer, obesity, concomitant drugs the patient is on and the ability to monitor drug-drug interactions, cost, and the risk of bleeding. Choice may also depend on individual physician or patient preference or recommendations in local guidelines.[19]

If treatment was initiated before diagnostic confirmation and PE is subsequently excluded, anticoagulation can be discontinued. In patients with confirmed PE, anticoagulation should be adjusted to the treatment-phase dose after completion of the initiation phase and should continue for at least 3 months.[3][19][124]

Recommendations for initial choice of anticoagulant

UFH is recommended in severe disease/clinical instability when thrombolysis may be utilized, or if the patient is at high-risk of bleeding.
In stable patients, the choice of initial anticoagulant is guided by the choice for the most appropriate longer-term therapy. Generally, this will be a direct oral anticoagulant (DOAC), but there are exceptions for specific patient populations.
DOACs (e.g., apixaban, edoxaban, rivaroxaban, dabigatran) are generally recommended over vitamin K antagonists (usually warfarin). If a DOAC is chosen, there is either an initiation phase at a higher oral dose (apixaban and rivaroxaban), or lead-in treatment with a parenteral anticoagulant for 5-10 days while treatment is established (edoxaban and dabigatran). This is then followed by oral monotherapy at treatment-phase dosing of the chosen agent.
For patients where warfarin is more appropriate, treatment with LMWH, fondaparinux or UFH alongside the starting dose of warfarin is needed in the initiation phase, while therapeutic anticoagulation is established.
In patients who present with massive PE/clinical instability, lead-in treatment with UFH or LMWH is continued for 5-10 days while treatment is established with a DOAC or warfarin. This is then followed by oral monotherapy at treatment-phase dosing of the chosen agent.
Fondaparinux, argatroban and bivalirudin are generally reserved for patients with HIT or those with a history of this condition.[236][237]

Considerations for specific anticoagulants

DOACs

DOACs are as effective as UFH, LMWH, and warfarin for the treatment of VTE, and are generally recommended over these drugs outside of special populations.[238] No monitoring of coagulation profile is necessary, and bleeding complications are similar to or less than those of warfarin, but there is a lower or similar incidence of VTE.[239][240] All have a longer half-life than UFH or LMWH and a shorter half-life than warfarin, and all have a rapid onset of action.
Apixaban and rivaroxaban are initiated at a higher initial oral dose with no need for lead-in therapy with a parenteral anticoagulant. Edoxaban and dabigatran require lead-in therapy with a parenteral anticoagulant for 5-10 days before oral monotherapy.
DOACs do not interact with food; however, they do undergo some drug-drug interactions. Notable drug interactions include: strong inhibitors or inducers of P-glycoprotein (with edoxaban and dabigatran); and strong inhibitors or inducers of P-glycoprotein and CYP3A4 (with apixaban and rivaroxaban).
Specific reversal agents for dabigatran (idarucizumab) and apixaban and rivaroxaban (recombinant coagulation factor Xa [andexanet alfa]) have been approved. Reversal of warfarin, in the setting of major or life-threatening bleeding, is recommended with vitamin K and prothrombin complex concentrates.[241]

Warfarin

In patients who will transition from UFH, LMWH, or fondaparinux to warfarin, warfarin should be started the same day as these drugs are started unless there is a very high-risk for bleeding. If bleeding risk is high, observing the patient for 1-2 days on UFH alone is advisable.
Three strategies can be used to select the initial dose of warfarin:[242][243]
- a clinical algorithm calculates the estimated stable and starting dose based on several patient characteristics
- a genetic algorithm calculates the estimated stable and starting dose based on the results of genetic tests such as CYP450-2C9 genotype and VKOR-C1 haplotype, as well as clinical variables
- a fixed-dose approach using initiation nomograms.
Use of an individualized nomogram for selecting the initial warfarin dose, and for subsequent titrations, is likely to result in better outcomes than a fixed-dose initiation, and is preferred.[243][244] Tests are available that determine the genotype of the patient for CYP2C9 variants and vitamin K epoxide reductase variants. However, overall, this information has not led to more rapid or safe anticoagulation compared with routine dosing. Genotyping is expensive and it takes several days to receive results.[245][246][247][248] When available, employing an individualized approach to warfarin initiation may be preferred. An online tool is available to assist with warfarin initiation dosing, which utilizes clinical variables with or without the addition of genetic information. WarfarinDosing.org: warfarin dosing Opens in new window
Once warfarin is started, it is continued concomitantly with the parenteral anticoagulant while the dose of warfarin is titrated. Subsequent dosing of warfarin is based on the international normalized ratio (INR). The therapeutic INR range is 2-3 (target 2.5, unless concomitantly being used for anticoagulation of mechanical heart valves). UFH, LMWH, or fondaparinux should be continued for a minimum of 5 days, and until INR is 2 or greater for at least 24 hours, at which point the parenteral anticoagulant can be discontinued.[19][249]

Heparin

UFH is preferred when a short-acting agent is needed due to concerns about bleeding, and in severe disease/clinical instability when thrombolysis may be utilized. UFH treatment is usually initiated with an intravenous weight-based loading bolus followed immediately by initiation of a weight-based continuous infusion. It requires monitoring of activated partial thromboplastin time (aPTT) or heparin-calibrated anti-Xa activity, which is used to titrate dose to the target range.
LMWH may be used in the initiation phase pending subsequent transition to a DOAC (edoxaban or dabigatran) or warfarin in the treatment-phase. LMWH is dosed subcutaneously according to patient weight.
Platelet count is regularly measured during treatment with UFH or LMWH because of the possibility of HIT as a complication.

Specific patient populations

Severe disease

For patients with high-risk (massive) PE or or high clinical probability of PE with shock or hypotension (i.e., systolic BP <90 mmHg), in whom interventional therapy is being planned or considered, intravenous UFH is preferred as most studies of interventional therapies were performed with this anticoagulant. It can also be adjusted if needed during intervention and has a relatively short half-life if bleeding occurs.[3][19][250]
Once stabilized the patient can be transitioned onto an anticoagulant guided by the choice for the most appropriate longer-term therapy.

Increased risk of bleeding

It may be preferable to treat patients who are at increased risk of bleeding (e.g., recent surgery, peptic ulceration) with intravenous UFH initially because it has a short half-life and its effect can be reversed quickly with protamine.[45] Once it is clear anticoagulation is tolerated, selection of an appropriate anticoagulation regimen can take place.

Active cancer

In patients with VTE and active cancer (cancer-associated thrombosis), guidelines from the ACCP, and the UK National Institute for Health and Care Excellence (NICE) recommend a DOAC (apixaban, edoxaban, rivaroxaban) over LMWH.[19][20]
Guidance from the American Society of Clinical Oncology suggest using LMWH, UFH, fondaparinux, rivaroxaban, or apixaban for initial anticoagulation.[46]
DOACs (particularly edoxaban and rivaroxaban) are associated with a higher risk of GI bleeding than LMWH. In patients with luminal GI cancer, the ACCP recommends apixaban or LMWH as the preferred agents.[19][251]

Renal impairment

For patients with renal impairment (i.e., creatinine clearance <30 mL/minute), intravenous or subcutaneous UFH, followed by warfarin, is the preferred anticoagulant regimen.
Apixaban is approved for use in severe renal disease, and has outcomes similar to UFH followed by warfarin, and represents an alternative option.[252]
LMWH has unpredictable renal clearance among patients with renal failure. For patients on LMWH, laboratory monitoring of the anticoagulant effect (i.e., by anti-factor Xa assay) is generally not necessary, but should be considered in patients with severe renal impairment and those with moderate renal impairment if use is prolonged (i.e., >10 days).[253]
Fondaparinux, edoxaban, rivaroxaban, and dabigatran are generally not recommended in people with severe renal impairment, and patients with creatinine clearance <25 to 30 mL/minute were excluded from large randomized controlled trials. Apixaban, edoxaban, and rivaroxaban may be used in some patients with renal impairment; however, consult your local guidance as recommendations vary between countries.

Hepatic impairment

UFH or LMWH are recommended in these patients, and should be overlapped with warfarin unless cancer is present.[19]
Warfarin should be used cautiously if the baseline INR is elevated; extended-duration LMWH may be preferred.[21][254]
DOACs are generally not recommended in patients with hepatic impairment, especially those with moderate-to-severe impairment (Child-Pugh class B or C).[19][45]

Obesity

UFH or LMWH are options for the initiation phase of treatment in patients living with obesity. The use of actual body weight is appropriate when calculating the therapeutic dose in obese patients. Laboratory monitoring of the anticoagulant effect of LMWH (i.e., by anti-factor Xa assay) is generally not necessary, but should be considered in patients with class III obesity (body mass index [BMI] 40 or above).[21][84][253]
There is no known weight limit for the use of DOACs; however, they have not been extensively studied in patients with extreme weights. The Scientific and Standardization Committee of the International Society on Thrombosis and Haemostasis recommends dabigatran and edoxaban are avoided in patients with BMI >40 kg/m² or weight >120 kg given the lack of clinical outcomes data. Rivaroxaban and apixaban can be considered in these patients.[255] Two large, retrospective, matched cohort studies showed similar outcomes in patients receiving rivaroxaban, apixaban, or dabigatran versus warfarin, though no prospective comparative evidence exists.[256][257] If DOACs are used in these patients, appropriate drug-specific monitoring may be considered, though there is limited evidence that drug-specific levels predict important clinical outcomes.[84]

Pregnancy

Women who develop VTE and who are pregnant or may become pregnant can be treated with subcutaneous UFH or LMWH monotherapy.[258] Because of changes in the pharmacodynamics of subcutaneous UFH during pregnancy, LMWH is preferred.[19][259] Routine measurement of peak anti-Xa activity for pregnant or postpartum patients on LMWH is not recommended except in women at extremes of body weight (i.e., <50 kg or >90 kg) or with other complicating factors (e.g., renal impairment or recurrent VTE) that put them at high-risk.
Warfarin is known to cause teratogenic effects when used in pregnancy and should be avoided.
If breast-feeding is planned, LMWH is the agent of choice. Warfarin is an alternative; it is minimally secreted in breast milk, but there is extensive clinical experience suggesting no ill effect in the breast-feeding infant.[197][260]
The safety of DOACs in pregnancy and lactation is unknown and they should be avoided in both situations (but can be used in the postpartum period if the patient is not breast-feeding).

HIT

In patients with HIT, the recommended anticoagulant is argatroban. Fondaparinux, apixaban, rivaroxaban, and dabigatran have also been suggested, although they are not approved for patients with active HIT.[236][237] Argatroban is preferred for patients with HIT with high bleeding risk or renal impairment. See Heparin-associated thrombocytopenia.

For more information on initiating anticoagulation, please see Anticoagulation management principles.

Treatment-phase of anticoagulation (initiation to 3 months)

The ACCP guidelines recommend that patients who do not have a contraindication are given a 3-month treatment-phase of anticoagulation. DOACs are recommended over warfarin.[19]

During the treatment-phase, follow-up and reevaluation are based on the patient’s level of risk for bleeding, comorbidities, and the anticoagulant selected.[19]

Patients taking edoxaban or dabigatran should remain on the same dose started during the initiation phase unless renal function substantially declines, warranting discontinuation.[45]
Patients taking apixaban and rivaroxaban should have their dose adjusted to the treatment-phase dose.[45]
Patients treated with warfarin should continue to have INR monitoring. The frequency of measurements depends on the stability of INR values at each visit. Commonly, INR is measured once or twice weekly after initial dose titration, with the time between measurements progressively extending if values remain in range. The target range of 2-3 (target INR 2.5) is maintained, unless concomitantly being used for anticoagulation of mechanical heart valves.[45]
If extended LMWH is used (e.g., in patients who cannot take an oral drug, patients with cancer with concomitant drugs that have significant drug-drug interaction that precludes use of DOACs, patients with an intraluminal GI malignancy and high-risk of GI bleeding, and patients with severe liver disease where neither warfarin nor DOACs can be used), the dose depends upon the agent:
- if dalteparin is chosen, the dose is reduced after 1 month
- if enoxaparin is chosen, some experts suggest reducing the initial dose after 1 month though this is based on opinion only, and the initial dose can be continued.

The treatment-phase of anticoagulation differs in pregnant patients. Patients with pregnancy-associated VTE undergo treatment-phase anticoagulation for at least 3 months, or until 6 weeks postpartum, whichever is longer.[258][259] At the conclusion of this phase in the postpartum, decisions are made according to whether the patient is planning to breast-feed. Guidelines differ on offering extended anticoagulation for VTE associated with pregnancy, as there is an intermediate risk of future unprovoked VTE.[22][250]

Once the treatment-phase has been completed, all patients should be evaluated for extended-phase therapy.[19]

Extended phase of anticoagulation (3 months to indefinite)

The goal for continuation of anticoagulant therapy into the extended phase (i.e., beyond the first 3 months and with no scheduled stop date) is secondary prevention The ACCP guidelines recommend that patients who are diagnosed with PE in the absence of transient provocation (unprovoked PE or provoked by a persistent risk factor) are given extended-phase anticoagulation.[19] These patients should be given a DOAC, unless contraindicated, in which case they should be given warfarin.

Extended-phase anticoagulation is not recommended in patients with PE diagnosed in the context of a major or a minor transient risk factor.[19]

The ACCP guidelines recommend using reduced-dose apixaban or rivaroxaban for patients receiving these drugs; the choice of a particular drug and dose should consider the patient’s BMI, renal function, and adherence to the dosing regimen.[19] The decision to start or continue extended therapy should be based on patient preference and the predicted risk of recurrent VTE or bleeding.[19]

The continued use of extended-phase anticoagulation should be reassessed at least annually, as well as at any time there is a significant change in the patient’s health status.[19]

The evidence to continue extended therapy beyond 4 years is uncertain. The ACCP recommends shared decision-making, taking into account the patient’s values and preferences. Patients should be periodically reassessed for bleeding risk, burdens of therapy, and any change in values and preferences.[19]

Provoked PE (minor or major transient risk factors)

Anticoagulation is discontinued after a course of at least 3 months. There is consensus that patients who have an index PE that occurs in the setting of a major transient provocation have a relatively low risk of developing recurrent VTE in the next 5 years, with estimates in the range of 15%.[19] In these patients, a time-limited course of anticoagulation of at least 3 months is suggested.[19]
The presence of a hereditary thrombophilia does not alter this recommendation, and guidelines recommend against testing for thrombophilias in patients with a PE occurring following a major transient provocation.[60]
The risk of recurrent VTE is modestly higher in patients who sustain PE in the setting of a minor transient provocation. Guidelines differ on offering extended anticoagulation for VTE associated with minor transient provoking risk factors.[22][250]

Unprovoked PE (no identifiable risk factor)

Patients with an unprovoked PE who have been started on anticoagulation therapy should be assessed after 3 months for continued treatment.[19]
For patients with a first PE that is unprovoked who have a low or moderate bleeding risk, extended anticoagulant therapy is recommended (with no scheduled stop date and reassessment of ongoing therapy at regular intervals, such as annually). For those patients with a high bleeding risk, 3 months’ treatment only is recommended.
For patients with a second unprovoked PE who have a low or moderate bleeding risk, extended anticoagulant therapy is recommended (with no scheduled stop date) over 3 months' treatment. In patients with a high bleeding risk, 3 months’ treatment only is recommended.
Many studies have attempted to identify subgroups of patients with unprovoked VTE who do not need to be treated indefinitely with oral anticoagulation. There is strong evidence that the risk of recurrent VTE is higher in the following patients: male sex; those with a diagnosis of a proximal DVT (versus isolated calf DVT); those with ultrasound evidence of residual clot; those who have an elevated D-dimer 1 month after stopping a 3- to 6-month course of oral anticoagulation; and those who had an unprovoked DVT.[19][45] Several risk assessment models have been developed for this purpose, including the DASH score, the Vienna Prediction Model, and the “Men Continue and HER-DOO2” model.[261] The latter model identifies a subset of women with low risk for recurrent VTE after an initial unprovoked event, and one prospective validation study of this model was published.[262]

Cancer-associated VTE

Cancer represents a persistent provocation for VTE until cured. Among patients who are diagnosed with PE and have an active cancer (e.g., cancer under any form of active therapy or palliation) there is a very high-risk for recurrent VTE and indefinite anticoagulation is recommended. Guidelines recommend using a DOAC (apixaban, edoxaban, rivaroxaban) or LMWH for at least the initial 6 months of therapy.[19][20]
A DOAC (apixaban, edoxaban, rivaroxaban) or LMWH is the preferred agent for patients with a higher risk of bleeding, especially those with GI cancers. LMWH is preferred for those with potential drug-drug interactions with DOACs.[19][251][263][264][265]

Bleeding risk

When assessing bleeding risk, the following factors should be considered:[19]
- Age >65 years
- Previous bleeding
- Cancer (especially GI cancer with DOACs)
- Renal failure
- Liver failure
- Thrombocytopenia
- Previous stroke
- Diabetes mellitus
- Anemia
- Antiplatelet therapy
- Poor anticoagulant control
- Comorbidity with reduced functional capacity
- Recent surgery
- Frequent falls
- Alcohol misuse
- Use of nonsteroidal anti-inflammatory drugs (NSAIDs)
- Uncontrolled hypertension
Patients with none of these risk factors are considered low risk; one risk factor renders a patient moderate risk; and two or more risk factors renders a patient high-risk.
Risk assessment models to assess bleeding risk derived from atrial fibrillation populations are not known to be accurate in patients with PE. VTE-specific bleeding risk assessment models have been developed.[266][267][268]
Drug-drug interactions may increase the risk of bleeding in patients receiving anticoagulants, and both the pharmacodynamic and pharmacokinetic interactions should be thoroughly evaluated prior to initiation.

Inferior vena cava filters (IVC)

An IVC filter can be placed in patients.[3][19][212]

With acute PE and an absolute contraindication to anticoagulant therapy, such as active major bleeding
With confirmed recurrent PE despite adequate anticoagulation.

The ACCP guidelines recommend using an IVC filter only for patients with acute PE (e.g., diagnosed in the preceding 1 month) and an absolute contraindication to anticoagulant therapy (e.g., active major bleeding, severe thrombocytopenia, high bleeding risk, central nervous system lesion). The ACCP recommends against the use of IVC filters in addition to anticoagulation in patients with acute PE.[19] Other guidelines consider relative indications for IVC filter use to include massive PE with residual deep venous thrombus in a patient at risk for further PE, free-floating iliofemoral or IVC thrombus, and severe cardiopulmonary disease and DVT (e.g., cor pulmonale with pulmonary hypertension).[269]

Some centers insert IVC filters intraoperatively or immediately postoperatively in patients who undergo surgical pulmonary embolectomy.[270][271][272]

IVC filter placement should take place as early as possible if it is the only treatment that can be initiated. There is little evidence available to suggest the ideal time for placement. Observational studies suggest that insertion of a venous filter might reduce PE-related mortality rates in the acute phase but with an associated increase in the risk of filter-related VTE.[273][274]

Complications associated with permanent IVC filters are common, although they are rarely fatal.[274] Early complications (including insertion-site thrombosis) occur in approximately 10% of patients. Late complications are more frequent and include recurrent DVT (approximately 20% of patients) and post-thrombotic syndrome (up to 40% of patients).[275][276][277] Occlusion of the IVC affects approximately 22% of patients at 5 years and 33% at 9 years, regardless of the use and duration of anticoagulation.[276]

Post-filter anticoagulation should be considered on a case-by-case basis according to relative and absolute contraindications.[278] Anticoagulation should be initiated if the contraindication resolves or if a risk/benefit analysis suggests this to be a reasonable course.[19] When retrievable filters are used, they should be removed if anticoagulation has been instituted and once it is clearly being tolerated.[3]

Hospitalization versus outpatient therapy

While hospitalization has historically been offered to most patients with PE, there is compelling data that patients at low risk of a poor outcome can be managed as outpatients, taking into account the patient’s personal circumstances, and as long as all the following criteria are met:[19][209]

Clinically stable with good cardiopulmonary reserve.
No contraindications to anticoagulation such as recent bleeding, severe renal or liver disease, or severe thrombocytopenia (i.e., <50,000/mm³).
No risk factors for bleeding that would require close observation in the hospital (e.g., chronic liver disease with or without varices, recent or prior GI bleeding, bleeding disorder, malignancy, recent stroke, or prior intracranial hemorrhage).
Expected adherence to treatment.
The patient feels well enough to be treated at home.
No concomitant illnesses requiring hospitalization.

Patients who are incidentally diagnosed with asymptomatic PE should receive the same initial and long-term anticoagulation as those with comparable symptomatic PE.[19]

The Registro Informatizado de la Enfermedad TromboEmbolica venosa (RIETE) and HESTIA criteria may be helpful in selecting patients (with VTE at low risk of adverse clinical outcome) who could be managed as outpatients.[209][279][280][281]

Antiplatelet therapy

If the decision is to stop extended-phase anticoagulation in patients with an unprovoked proximal PE, the ACCP guidelines recommend low-dose aspirin (unless contraindicated) to prevent recurrent VTE.[19][282][283] The benefits of using aspirin should be balanced against the risk of bleeding and inconvenience of use.[282][283] Aspirin should not, however, be considered a reasonable alternative for patients who are willing to undergo extended anticoagulation therapy, as aspirin is less effective. The use of aspirin should in any case be reassessed when patients stop anticoagulant therapy because it might have been stopped when anticoagulant therapy was started.[19]

Patients with recurrent VTE on anticoagulant therapy

Recurrent VTE is unusual among patients receiving therapeutic-dose anticoagulant therapy, with the exception of cancer (7% to 9% on-therapy recurrence with LMWH).[3][19][284] In addition to definitively establishing the presence of recurrent PE, consideration should be given to compliance with anticoagulant therapy or the presence of underlying malignancy.[19]

ACCP guidelines recommend a temporary switch to LMWH (for at least 1 month) for patients with recurrent PE who are thought to be compliant with a non-LMWH anticoagulant (or within the therapeutic range if receiving warfarin).[19] An increased dose of LMWH (one quarter to one third) is appropriate for patients with recurrent PE who have been receiving LMWH.[19]

An IVC filter can be placed in patients with confirmed recurrent PE despite adequate anticoagulation; however, direct evidence supporting this intervention is very limited.[3][212] Complications associated with permanent IVC filters are common, although they are rarely fatal.[274] Overall, early complications (including insertion-site thrombosis) occur in approximately 10% of patients. Late complications are more frequent and include recurrent deep vein thrombosis (approximately 20% of patients) and post-thrombotic syndrome (up to 40% of patients).[275][276] Occlusion of the IVC affects approximately 22% of patients at 5 years and 33% at 9 years, regardless of the use and duration of anticoagulation.[276]

Recurrent VTE following discontinuation of anticoagulant therapy

For patients who are no longer receiving anticoagulant therapy and experience a second VTE with no identifiable risk factor (i.e., unprovoked), guidelines recommend the following anticoagulant treatment durations:[3][19]

Low or moderate bleeding risk: extended anticoagulant therapy with periodic reassessment to review risk-benefit ratio
High bleeding risk: stop anticoagulation after 3 months.

Use of this content is subject to our disclaimer