According to European guidelines,8 patients are classified based on their hemodynamic stability, clinical score (PE severity index [PESI] or simplified PESI [sPESI]), right ventricular (RV) dysfunction or dilatation, and biomarkers (table 2) into the following categories: a) HR-PE: defined as patients with RV involvement on imaging (echocardiography or computed tomography), elevated biomarkers (troponin, B-type natriuretic peptide [proBNP]), and hemodynamic instability (divided into 3 subcategories: persistent hypotension, obstructive shock, and cardiac arrest)8; about 4% of patients have HR-PE, with short-term mortality of up to 27% within the first 24hours and up to 49% within 72hours; b) intermediate-risk PE: in contrast to HR-PE, these patients are hemodynamically stable; this category is itself subdivided into 2 categories: IHR-PE, characterized by the presence of RV involvement on imaging and elevation in any biomarker; and low–intermediate-risk PE, characterized by the presence of 1 or none of the previous criteria (but PESI III-V or sPESI ≥ 1); this group represents 55% of patients; although most improve with anticoagulation, the IHR-PE subgroup should be closely monitored due to the risk of clinical, respiratory, or hemodynamic deterioration; and c) low-risk PE: these patients are hemodynamically stable, with a PESI I-II or sPESI score of 0 and without RV involvement on imaging or biomarker elevation; they comprise about 40% of all patients with PE.

In recent decades, efforts have been made to identify a subgroup of IHR-PE patients with short-term risk of hemodynamic deterioration. Although the guideline-recommended therapy is anticoagulation, for patients with a risk of death ≥ 10% according to the various scales and markers, this group of experts recommends considering elective CDI because this technique is associated with ≤3% mortality in various studies, mainly observational.15–20 To identify this subgroup, we propose a series of markers and prognostic scales based on parameters detectable at admission (table 2).

Patients with low-intermediate risk can be admitted to a standard hospital floor. Those with IHR-PE require close monitoring to detect early signs of deterioration and should therefore be admitted to an intermediate care unit or CICU. Patients at high risk should be admitted to a CICU.

Anticoagulation is the cornerstone of PE treatment. Treatment with low-molecular-weight heparin or fondaparinux is recommended over unfractionated heparin (UFH) because they more rapidly reach therapeutic concentrations and are associated with a lower risk of heparin-induced thrombocytopenia. A meta-analysis showed that patients with PE initially treated with low-molecular-weight heparin exhibited fewer thrombotic complications, a better safety profile, and lower mortality vs UFH.21 UFH is the drug of choice for patients with chronic kidney disease (clearance <30mL/min), high bleeding risk, morbid obesity, high risk of hemodynamic deterioration, or those who are candidates for CDI. In patients with an absolute contraindication, a vena cava filter can be considered (table 1 of the supplementary data).

Reperfusion is defined as targeted therapy to rapidly restore pulmonary flow by reducing thrombus burden and relieving RV pressure overload, typically in situations of obstructive shock (table 1). Although we report the main international recommendations in table 2 of the supplementary data, the present document diverges from the indications in the European guidelines and summarizes the most common treatments, as follows.

Catheter-directed intervention

The use of percutaneous techniques for the management of PE has grown significantly in the last 10 years.14,26,27 The techniques are divided into 2 groups—local thrombolysis (LT) and MT—and are summarized in figure 1.

Figure 1.

The most common reperfusion treatments for pulmonary embolism. CDI, catheter-directed intervention; HR-PE, high-risk pulmonary embolism; IHR-PE, intermediate–high-risk pulmonary embolism; rtPA, recombinant tissue-type plasminogen activator.

(0.56MB).

In both groups, ultrasound-guided puncture is essential for optimizing safety and ruling out venous thrombosis. Navigating to the pulmonary artery is recommended using a Swan-Ganz or pigtail catheter for a central pass through the tricuspid valve. Pressures should be recorded, including those of the right atrium (preload score) and pulmonary artery, in addition to the initial and final saturation (or pre- and postprocedure values in LT). Pulmonary artery angiography can be performed to confirm the diagnosis and assess the anatomy. An arterial line can be considered for gasometry and invasive blood pressure monitoring, particularly in unstable patients.

Catheter-directed mechanical thrombectomy

This percutaneous procedure permits thrombus removal from the pulmonary circulation through aspiration and aims to restore flow in the pulmonary arteries and decrease the RV overload less invasively than surgical embolectomy (SE).

Dedicated devices. Two main MT systems are currently available. The FlowTriever system (Inari Medical, United States) comprises various catheters up to 24 Fr, a system for blood recovery, filtration, and return, and 2 self-expanding nitinol systems facilitating the organized extraction of thrombi. The system is advanced coaxially on an extra-support guidewire that should be positioned as distally as possible in the inferior lobes. In contrast, the 8-and 12-Fr Lightning Indigo and 16-Fr Lightning Flash systems (Penumbra, United States) are connected to a suction pump with a flow sensor that allows optimization of the aspiration efficacy, reducing blood loss. The system moves freely within the pulmonary artery anatomy. These 2 systems are described in figure 1. Table 5 summarizes the ideal candidate characteristics, precautions, and contraindications for this technique while table 6 reviews the evidence on MT.

Table 5.

Characteristics of patients with pulmonary embolism who are candidates for mechanical or catheter thrombectomy

Good candidates
HR-PE	IHR-PE (rescue)	IHR-PE (elective)
• Contraindication for ST• Without contraindication for ST but with high bleeding risk with ST (BACS scale37)• ST or LT failure	• Clinical or hemodynamic deterioration according to ESC guidelines with anticoagulation as an alternative to ST due to its better safety profile	• In selected patients with IHR-PE, consider elective CDI if several poor prognostic factors (table 2) are present to prevent clinical or hemodynamic deterioration

Precautions

• Inferior vena cava and superior vena cava abnormalities• Presence of a bioprosthesis in the tricuspid or pulmonary valve• Previous thoracic radiation (higher perforation risk)• Presence of RV pacing leads• Right-sided congenital heart diseases• Chronic thromboembolic pulmonary hypertension (acute thrombus over chronic)• Presence of inferior vena cava filter• Thrombus-in-transit and patent foramen ovale• Avoid accessing distal branches with large-caliber devices• Monitor the amount of blood during extraction. Request cross-matching tests

Contraindications
• Absence of venous access• Presence of a mechanical prosthesis in the tricuspid or pulmonary valve• Disseminated pulmonary metastases• Active bleeding that contraindicates intraprocedural anticoagulation with sodium heparin• Peripheral thrombus

ESC, European Society of Cardiology; HR-PE high-risk pulmonary embolism; IHR-PE, intermediate–high-risk pulmonary embolism; LT, local thrombolysis; RV, right ventricular; ST, systemic thrombolysis.

Table 6.

Evidence on percutaneous mechanical thrombectomy devices

Study	Population	tPO	Description	Efficacy	Safety	NCT
FlowTriever
FLARE19	IR-PE106 patients	48 h	Single-arm studyFibrinolysis 2%	RV/LV reduction 0.38	3.8% MAEs1% MB0% ICB	NCT02692586
FLAME38	HR-PE115 patients	At admission	ObservationalFlowTriever vs medical therapy (ST 68.9%) vs performance goal	Composite MAEs17% vs 32% (P<.01)FlowTriever vs performance goal	Mortality1.9% with FlowTriever29.5% with medical therapy	NCT04795167
FLASH16	HR-PE (8%)IHR-PE (77%)800 patients	48 h	ObservationalFibrinolysis 2.3%	mPAP reduction 7.6 mmHgRV/LV reduction 0.25	1.8% MAEs at 48 h0.3% mortality at 48 h	NCT03761173
PEERLESS	IR-PE550 patients	7 d	RandomizedFlowTriever vs LT	Composite MAEs at 7 d	Enrollment completed	NCT05111613
PEERLESS 2	IHR-PE1200 patients	30 d	RandomizedFlowTriever vs heparin	Composite MAEs at 30 d	Currently enrolling patients	NCT06055920

Penumbra Indigo
EXTRACT-PE20	IR-PE119 patients	48 h	Single-arm studyIndigo 8 Fr	RV/LV reduction 0.43	1.7% MAEs at 48 h0.8% mortality at 48 h	NCT03218566
STORM-PE	IHR-PE100 patients	48 h	RandomizedIndigo 12 Fr vs heparin	RV/LV reduction	Currently enrolling patients	NCT05684796
STRIKE-PE	IHR-PE600 patients	48 h	Single-arm studyLightning 12 and 16 Fr vs heparin	RV/LV reduction	Currently enrolling patients	NCT04798261

HR-PE, high-risk pulmonary embolism; ICH, intracranial hemorrhage; IHR-PE, intermediate–high-risk pulmonary embolism; IR-PE, intermediate-risk pulmonary embolism; LT, local thrombolysis; MAEs, major adverse events; MB, major bleeding; mPAP, mean pulmonary artery pressure; NCT, identifier at www.clinicaltrials.gov; RV/LV, ratio of the diameters of the right ventricle and left ventricle; tPO, time to assessment of the primary objective (PO).

Nondedicated devices and other peripheral thrombectomy systems used to treat PE are described in table 4 of the supplementary data.

Technical details. The vascular access for all of these procedures is the femoral vein, unless it is thrombosed or the patient has a vena cava filter. Once access is obtained, anticoagulation should be maintained with UFH to achieve an activated clotting time of 250 to 300seconds.

There are no standardized criteria for MT completion. The objective of MT comprises clinical and hemodynamic parameters, meaning that a complete disappearance of the thrombus is not necessary to achieve a stable situation. Table 1 proposes objectives that can often be achieved with partial anatomical recanalization alone. Vascular closure can be performed with Perclose Prostyle or FemoStop (Abbott, United States), with Z- or figure of 8- suturing, or with the FlowStasis system included with the FlowTriever system.

Combination of techniques. Mechanical (fragmentation or MT) and pharmacological (LT) techniques can be used sequentially. There is little systematic evidence beyond registry data39 on the combined treatment and no randomized trials or comparisons between techniques. Some MT registries show anecdotal use of thrombolytics (2.3% in the FLASH trial).16 However, thrombolytics were used in 23.3% of cases in the corresponding Spanish registry.26 Theoretically, the combination of the techniques would be complementary, with MT immediately improving hemodynamic and respiratory parameters by eliminating large proximal thrombi and LT aiming to completely resolve the thrombus by acting on smaller branches. However, this strategy could increase the bleeding risk by combining large-caliber access with thrombolytic therapy. When this technique is used (particularly in MT as rescue from LT or ST), safe vascular access is vital (ultrasound-guided puncture, micropuncture).

Currently, combination treatment is not the strategy of choice in CDI but is reserved as rescue therapy. It is essential to select the appropriate technique for each individual patient from the outset to maximize benefits and minimize complications, taking into account the local experience and resources (figure 1 and figure 2).

Figure 2.

Modality selection for catheter-directed intervention. ECMO, extracorporeal membrane oxygenation; LT, local thrombolysis; MT, mechanical thrombectomy; rtPA, recombinant tissue-type plasminogen activator.

(0.25MB).

Pulse oximetry should be used to continuously monitor oxygen saturation. An appropriate blood oxygen level should be maintained (arterial oxygen pressure >80mmHg and oxygen saturation >90%) using a nasal cannula, oxygen reservoir, high-flow nasal cannula, and noninvasive or invasive mechanical ventilation, in accordance with the recommendations in table 7 and figure 3.27,42,43

Figure 3.

Management of hypoxemia in patients with pulmonary embolism admitted to the intensive care unit and effects of mechanical ventilation on cardiac function. Modified from Arrigo et al.42 LV, left ventricle; PE, pulmonary embolism; PFO, patent foramen ovale; RV, right ventricle; SatO2, oxygen saturation.

(0.3MB).

Vital signs (blood pressure, heart rate, and electrocardiogram) should be continuously monitored, as well as right heart catheterization parameters. After the procedure, monitoring is recommended for at least 24hours for IHR-PE and 48hours for HR-PE. An echocardiogram is recommended immediately after the procedure and at 24hours. If the patient is hypotensive, the measures included in table 7, table 5 of the supplementary data, and figure 4 are recommended.8,42,44,45

Figure 4.

Integrated management of right ventricular failure in patients with PE. BP, blood pressure; PE, pulmonary embolism; RRT, renal replacement therapy; VA-ECMO, venoarterial extracorporeal membrane oxygenation.

(0.3MB).

Although CDI is a safe procedure compared with ST, it is not free of complications.46 The most frequent are generally mild bleeding complications related to the vascular access and the anticoagulation. Although infrequent, iatrogenic perforation of the pulmonary arterial tree may range between mild and self-limiting hemoptysis and massive hemoptysis with hemodynamic and respiratory deterioration. The management of these patients necessitates reversal of the anticoagulation, balloon occlusion of the affected branch, and pulmonary angiography-mediated localization of the bleeding site to perform intravascular mechanical hemostasis (eg, prolonged balloon inflation, embolization). In life-threatening cases, orotracheal intubation is necessary with a dual-lumen tube to isolate the bleeding lung and permit complete ventilation of the contralateral lung.

Other rare but serious complications are RV perforation, tricuspid valve injury, and paradoxical embolism in patients with patent foramen ovale.

Early mortality in HR-PE exceeds 15% and is concentrated in the first hours of admission.47 Accordingly, emergency reperfusion therapy is recommended, with ST as the treatment of choice,8 despite limited evidence. MT is an increasingly popular technique due to its safety and efficacy (in the FLAME trial,38 MT was superior to the control medical therapy arm including ST, with in-hospital mortality of 1.9%). MT is currently indicated in patients with contraindications for fibrinolysis or after its failure,8 although the opinion of this expert panel is that it may also be useful in patients with high bleeding risk with ST, as assessed using the BACS scale.37 If MT is indicated in this context, it should be performed within a maximum of 60minutes from the decision if the patient is in a center with MT availability or within 90minutes if the patient requires transfer to a referral center (table 1).1 In addition, appropriate respiratory and hemodynamic support should be ensured at all times. LT can be recommended in selected patients if MT is unavailable (table 3).

In IHR-PE, anticoagulation without reperfusion therapy is sufficient for most patients, although normotensive patients with an elevated risk of hemodynamic deterioration require close monitoring8 because some may experience hemodynamic deterioration and require reperfusion (∼10% mortality with medical therapy). Current data indicate that CDI-related mortality in this setting is <3%,15–20,26 which is why this expert panel proposes the consideration of elective CDI in selected patients with IHR-PE with several poor prognostic factors to prevent this clinical deterioration. Patients with IHR-PE who are candidates for elective reperfusion are those meeting the criteria listed in table 2. Particularly relevant are the presence of syncope, elevated lactate, borderline blood pressure, and elevated heart rate.

In these patients, the first treatment should be anticoagulation, with consideration of elective reperfusion therapy during the first 24 to 48hours (table 1). The PERT should develop an individualized plan that includes the initial treatment, the alternatives, and the logistics necessary for reperfusion, including CDI.47

The distinction between transient and permanent comorbidities, and major and minor risk factors, influences the risk of recurrence and the need for chronic oral anticoagulation after a PE (table 7 of the supplementary data).8,48

After a PE, anticoagulation should be maintained for at least 3 months. Subsequent treatment depends on the risk of recurrence, the risk of bleeding, and patient preferences. The established recommendations (independently of an eventual diagnosis of thrombophilia)48 are as follows:

• •

  Withdraw the anticoagulation 3 months after a first PE event triggered by a major and resolved transient risk factor.

• •

  Indefinite anticoagulation for a major permanent risk factor.

• •

  Indefinite anticoagulation for men with idiopathic PE due to intermediate risk between the 2 previous groups.

• •

  Additional studies (table 8 of the supplementary data) to determine whether anticoagulation should be maintained in women with idiopathic PE and for PE secondary to a minor and resolved transient risk factor, patients who wish to discontinue anticoagulation, and patients with an unknown risk/benefit relationship for indefinite anticoagulation.

Many patients have persistent dyspnea 3 months after anticoagulation for PE.49 Screening for CTEPD and chronic thromboembolic pulmonary hypertension (CTEPH) has prognostic and therapeutic value.

CTEPD is characterized by dyspnea on exertion, persistent perfusion deficits on ventilation/perfusion scintigraphy, normal resting pulmonary pressures, and exercise-induced pulmonary hypertension. Its true incidence is unknown. Screening should include cardiopulmonary exercise testing, in addition to echocardiography and lung scintigraphy. Confirmation requires resting and exercise catheterization.49,50

In addition, the diagnosis of CTEPH requires pulmonary arterial hypertension confirmed with resting right heart catheterization. Lifelong anticoagulation is necessary. Management comprises pulmonary thromboendarterectomy or pulmonary artery angioplasty, as well as specific pulmonary vasodilator therapy.50

The incidence of CTEPH after acute PE ranges between 2% and 5% in observational studies. However, the percentage of patients who develop CTEPD after acute PE is unknown. Although small studies indicated that ST can reduce the risk of CTEPH, a 3-year follow-up of 706 patients from the PEITHO trial revealed similar rates of CTEPH with ST and anticoagulation (2.1% vs 3.2%, P=.79).51 Studies with LT have found a greater short-term reduction in pulmonary artery systolic pressure,35,39 equaling that of the anticoagulation group during follow-up. No randomized studies have compared pulmonary pressure during follow-up in patients who remain symptomatic after acute PE treated with ST or CDI. Accordingly, there is insufficient evidence to state that reperfusion reduces the risk of CTEPH in patients with acute PE compared with anticoagulation.

CDI in patients with PE is not free of risks; a crucial factor is interventional cardiologists’ ability to obtain safe vascular access, navigate to the right heart, and initiate ECMO. Other essential factors are operators’ understanding of acute patients and of the therapeutic options and specific devices. The initial concentration of experience in a small number of operators and the standardization of the procedure may improve outcomes but the experience should be disseminated within the team for occasional emergency situations.