Pulmonary Embolism (Opus)

Acute pulmonary embolism (PE) is the lodging of a thrombus, most often embolised from a Deep Vein Thrombosis of the proximal lower-limb or pelvic veins, within the Pulmonary Artery|pulmonary arterial tree, producing an acute rise in Pulmonary Vascular Resistance and pressure-overload of the right ventricle ^[1]A1a[2]A1a. It is the most dangerous manifestation of Venous Thromboembolism (VTE), and the central defining physiologic concept is acute Right Ventricular Afterload|right ventricular afterload mismatch ^[3]A1a[5]B2a.

Classification (severity, by haemodynamics and risk)

The European Society of Cardiology (ESC) classification, which governs management, stratifies confirmed PE not by anatomic clot burden but by early-mortality risk ^{[11]A1b[22]B2a}.

Nomenclature

• Synonyms and abbreviations: PE; acute pulmonary thromboembolism; "massive"/"submassive" PE (deprecated in favour of the ESC risk tiers); part of the Venous Thromboembolism spectrum alongside Deep Vein Thrombosis ^[1]A1a[3]A1a.

The defining sibling distinction is between acute PE and its chronic sequela, Chronic Thromboembolic Pulmonary Hypertension , a member of the post-PE syndrome family ^[20]B2a.

Acute PE sits on a continuous pathophysiologic spectrum with Deep Vein Thrombosis : in most cases the embolus originates from a proximal lower-limb or pelvic deep vein, travels through the right heart, and impacts the pulmonary arterial bed, which is why the two conditions are managed under the single umbrella of Venous Thromboembolism ^{[1]A1a[2]A1a[3]A1a}. Anatomically, emboli may be described as saddle (straddling the main pulmonary artery bifurcation), lobar, segmental, or subsegmental, but a key conceptual point reinforced by contemporary guidance is that anatomic location is a poor surrogate for danger: a proximal saddle embolus in a patient with preserved right ventricular function and normal biomarkers behaves as a lower-risk lesion than smaller bilateral emboli in a patient whose right ventricle is already failing ^{[11]A1b[13]A1a[14]A1a}.

The ESC four-tier framework is deliberately built on physiology rather than anatomy because the right ventricle, not the clot, determines survival in the first hours to days ^{[6]A1a[11]A1b}. High-risk PE is defined by overt haemodynamic compromise -- sustained hypotension (systolic blood pressure below 90 mmHg or a drop of at least 40 mmHg), shock, or the need for vasopressors -- and demands immediate reperfusion ^{[13]A1a[16]B2b}. The intermediate strata are subdivided by whether imaging shows right ventricular dysfunction and whether cardiac Troponin is elevated, while the low-risk tier is defined principally by a Simplified PESI|sPESI of zero ^{[11]A1b[12]B2b}. This graded scheme directly maps to graded therapy, from outpatient anticoagulation through inpatient monitoring to emergency reperfusion, which is why correct classification is the single most consequential cognitive step after confirming the diagnosis ^{[11]A1b[13]A1a[22]B2a}.

Pearl: The error clinicians make most is equating large anatomic clot burden with "high-risk" PE. Risk is defined by haemodynamics and RV function (shock, RV dysfunction, troponin), not by the size of the filling defect on CT; a saddle embolus in a normotensive patient with a normal RV is intermediate-low, not high, risk ^{[11]A1b[14]A1a}.

Venous thromboembolism, of which PE is the most lethal form, affects nearly 10 million people worldwide each year, and in Western populations roughly 1 in 12 individuals will be diagnosed with VTE in their lifetime ^[1]A1a[4]A1a. Incidence rises steeply with age and is reported to be increasing over time; in Germany the annual PE incidence rose from 85 to 109 per 100,000 between 2005 and 2015 ^[22]B2a. Rates are lower in Asian populations, and data from many regions remain sparse ^[4]A1a. Most events are unprovoked, but the strongest provoking factors are major surgery and active cancer ^[1]A1a.

Risk factors

The clinical importance of these epidemiologic figures is that PE is both common and rising, so the pretest probability of PE in a breathless or chest-pain presentation is rarely negligible, and the disease is heavily concentrated in older and hospitalised populations ^{[4]A1a[22]B2a}. Around 20% of individuals with a VTE event die within one year, though death is frequently attributable to the provoking illness (such as advanced Cancer ) rather than the embolism itself, and complications are common among survivors ^[4]A1a. Because rates are substantially lower in Asian than in Western populations and high-quality incidence data are lacking for many regions, absolute risk should be interpreted in the local context rather than extrapolated globally ^[4]A1a.

Risk factors are usefully divided into acute or subacute triggers (provoking factors that transiently raise risk, such as surgery, trauma, acute medical illness, or immobilisation) and basal or acquired factors that may be modifiable or static (age, obesity, prior VTE, inherited Thrombophilia , malignancy, estrogen exposure) ^[4]A1a. Most events are in fact unprovoked, which has direct implications for anticoagulation duration because unprovoked events carry a higher recurrence risk than those provoked by a now-resolved transient factor ^{[1]A1a[18]A1a}. Cancer deserves particular emphasis: it is both a strong provoking factor and an independent driver of worse outcomes, with active malignancy associated with higher in-hospital mortality in high-risk PE (OR 1.70, 95% CI 1.13-2.56) ^[16]B2b. All three arms of Virchow's Triad -- venous stasis, endothelial injury, and hypercoagulability -- can be mapped onto these clinical risk factors, providing a mechanistic framework that explains why surgery, immobility, malignancy, pregnancy and inherited thrombophilias converge on the same final pathway ^[5]B2a.

Pearl: The highest-yield modifiable lever is peri-operative and hospital thromboprophylaxis: surgery and hospitalisation are among the strongest provoking factors, and primordial/primary prevention opportunities are concentrated here ^[4]A1a.

Acute PE is fundamentally a disease of the right ventricle. The mechanistic cascade lets one derive both the presentation and the treatment from first principles ^{[3]A1a[5]B2a[6]A1a}.

Mechanistic cascade

1. A venous thrombus (driven by Virchow's Triad -- stasis, endothelial injury, hypercoagulability) detaches and embolises to the pulmonary arteries, mechanically obstructing flow ^[5]B2a.

2. Obstruction plus hypoxic and humoral Pulmonary Vasoconstriction (serotonin, thromboxane) acutely raises Pulmonary Vascular Resistance and RV afterload ^[5]B2a[6]A1a.

3. The thin-walled RV dilates to maintain stroke volume via the Frank-Starling Law -- initially adaptive, but RV wall tension and oxygen demand rise sharply ^[6]A1a.

4. RV dilatation bows the interventricular septum leftward ( Ventricular Interdependence ), impairing LV filling and dropping cardiac output ^[6]A1a.

5. Falling output and rising RV wall tension reduce right coronary perfusion, causing RV Myocardial Ischemia|ischaemia -- a vicious cycle of the failing, dilating RV ^[5]B2a[6]A1a.

6. Simultaneously, ventilation-perfusion mismatch and increased dead space produce Hypoxemia|hypoxaemia , the degree of which depends on clot location and baseline cardiopulmonary reserve ^[5]B2a.

7. The cycle terminates in obstructive shock, Cardiac Arrest|haemodynamic collapse and death, or -- if survived -- may leave persistent RV impairment and post-PE syndrome ^[3]A1a[6]A1a.

Mechanism -> therapeutic target

Because the RV, not the lung parenchyma, is the proximate cause of death, every effective therapy ultimately serves to unload or support the right ventricle ^{[6]A1a[13]A1a}.

Understanding this cascade explains the entire clinical picture. The acute, fixed obstruction means the right ventricle cannot recruit the time needed for the adaptive hypertrophy that protects the chronically pressure-loaded ventricle in pulmonary arterial hypertension; the thin-walled RV is suddenly asked to generate pressures it has never trained for, which is why even moderate clot burdens can be catastrophic in a previously normal heart, and why patients with prior cardiopulmonary disease tolerate PE poorly ^[5]B2a[6]A1a. The neurohormonal and inflammatory response -- catecholamine surge, release of vasoactive mediators, and local pulmonary vasoconstriction -- is initially compensatory, supporting systemic pressure, but it simultaneously raises right ventricular afterload further, tightening the vicious cycle of dilatation, wall stress, ischaemia and falling output ^[5]B2a[6]A1a.

The reason hypoxaemia is mechanistically secondary is instructive: ventilation-perfusion mismatch, intrapulmonary shunting and reduced mixed venous oxygen content from low cardiac output all contribute, but supplemental oxygen alone does not address the obstructive lesion or the failing right ventricle ^[5]B2a[6]A1a. This is the conceptual bridge to therapy: anticoagulation halts thrombus propagation and allows endogenous fibrinolysis to remodel the existing clot over days to weeks, whereas thrombolysis and mechanical therapies act on the existing obstruction to acutely reduce afterload, and circulatory support (cautious fluids, vasopressors, and ECMO ) buys time for the right ventricle to recover ^{[13]A1a[14]A1a[16]B2b}. Persistent right ventricular impairment after the acute event, or failure of the thrombus to resolve, seeds the chronic syndromes -- post-PE functional impairment and Chronic Thromboembolic Pulmonary Hypertension -- that are addressed in later sections ^{[3]A1a[20]B2a}.

Pearl: The single reframing insight is that acute PE kills through RV afterload failure, not through hypoxaemia -- so the decisive interventions reduce clot burden and support the RV, and aggressive fluid loading can worsen the bowing septum and precipitate collapse ^{[6]A1a[13]A1a}.

The cardinal symptom of acute PE is dyspnoea, typically of sudden onset, because abrupt vascular obstruction increases dead space and triggers reflex tachypnoea while a failing RV cannot augment cardiac output on exertion ^[3]A1a[5]B2a. Pleuritic chest pain arises when distal emboli cause pulmonary infarction and adjacent pleural inflammation; central emboli more often produce a dull, anginal pain from RV ischaemia ^[5]B2a. Haemoptysis, when present, reflects alveolar haemorrhage at an infarcted segment ^[3]A1a.

Patients may report presyncope or frank syncope; this is an ominous symptom because it signals a transient fall in cardiac output from significant RV afterload, and its presence raises the probability of haemodynamically significant PE ^{[3]A1a[13]A1a}. Many presentations are non-specific, which is precisely why PE is frequently missed -- anchoring on an alternative diagnosis such as heart failure increases the chance the diagnosis is overlooked ^[5]B2a.

On examination, tachycardia and tachypnoea are the most common signs; sinus tachycardia is the rule. Signs of RV pressure overload -- elevated jugular venous pressure, a left parasternal heave, a loud pulmonary component of the second heart sound (P2), and in severe cases hypotension and cool peripheries -- point to a high-risk presentation ^[3]A1a[6]A1a. Unilateral leg swelling suggests the source Deep Vein Thrombosis ^[1]A1a.

Examination signs and discriminating value

The temporal pattern of symptoms is itself diagnostically useful: the classic history is abrupt rather than gradual, because embolisation is a discrete event, in contrast to the more insidious progression of heart failure or the indolent course of malignancy-related effusions ^[3]A1a[5]B2a. Pulmonary infarction, which tends to occur with smaller distal emboli reaching the periphery where collateral bronchial supply is insufficient, produces the triad of pleuritic pain, haemoptysis and a low-grade fever, and may be accompanied by a pleural rub on auscultation ^[3]A1a[5]B2a. Central or large emboli, by contrast, tend to present with dyspnoea, dull retrosternal discomfort from right ventricular ischaemia, and haemodynamic features rather than pleuritic pain ^[5]B2a[6]A1a.

The link between symptom and mechanism is what makes the history clinically actionable. Exertional and then resting dyspnoea reflects the inability of the obstructed pulmonary circulation and the failing right ventricle to augment output and oxygenation on demand ^[5]B2a[6]A1a. Presyncope and syncope reflect transient global hypoperfusion when right ventricular output falls precipitously, and their presence should immediately raise the suspicion of a haemodynamically significant embolus even if blood pressure has recovered by the time of assessment ^{[3]A1a[13]A1a}. On examination, the most reliable findings are the least specific -- tachycardia and tachypnoea -- while the more specific signs of right ventricular pressure overload (raised jugular venous pressure, a left parasternal heave, a loud P2, and a right-sided third heart sound) carry greater weight when present but are insensitive ^[3]A1a[6]A1a. Because individual symptoms and signs perform poorly in isolation, they are formally combined into validated clinical probability rules rather than used as standalone discriminators, and a documented source Deep Vein Thrombosis meaningfully raises pretest probability ^[1]A1a[8]B2b.

Pearl: Syncope and sustained hypotension are the bedside findings that most change pretest probability and immediately reclassify the patient toward high-risk PE, mandating expedited imaging or empirical reperfusion if collapse is imminent ^[13]A1a.

Diagnosis follows a sequential strategy: assess clinical pretest probability with a validated rule, then apply D-dimer testing to exclude PE in non-high-probability patients, reserving imaging for the rest ^{[1]A1a[2]A1a[18]A1a}. This integrated approach safely withholds imaging and anticoagulation in patients with a non-high probability and a normal D-dimer ^{[1]A1a[21]A1a}.

The Wells Score and revised Geneva Score are the standard prediction rules; the YEARS Algorithm and age-adjusted/clinical-probability-adjusted D-dimer thresholds reduce unnecessary imaging ^[18]A1a. D-dimer is highly sensitive but poorly specific -- a rule-out, not a rule-in, test; a high-sensitivity ELISA D-dimer can exclude VTE with a negative predictive value (NPV) of 99-100% ^[9]B2a. The TEST OF CHOICE and GOLD STANDARD for confirmation is CT Pulmonary Angiography (CTPA) ^[7]B2b.

Diagnostic tests -- findings and performance

Diagnostic algorithm

This pathway mirrors the ESC/contemporary guideline approach ^{[1]A1a[18]A1a[21]A1a}.

Differential diagnosis

The logic of the diagnostic pathway is to use cheap, sensitive tests to safely exclude PE in the majority and to reserve definitive imaging for those who genuinely need it, thereby reducing radiation, contrast exposure and incidental overdiagnosis ^{[1]A1a[18]A1a[21]A1a}. D-dimer is the linchpin of this exclusion strategy: it is released by endogenous fibrinolysis of any thrombus and is therefore highly sensitive but non-specific, rising also with age, infection, malignancy, pregnancy and surgery, which is why it must always be interpreted against a clinical probability rule and never used as a rule-in test ^{[9]B2a[18]A1a}. Age-adjusted and clinical-probability-adjusted thresholds (as embodied in the YEARS Algorithm ) raise the usable cut-off in the populations where a fixed threshold would otherwise be unhelpful, increasing specificity and reducing imaging without sacrificing safety ^{[7]B2b[18]A1a}.

Comparative diagnostic data illustrate the trade-offs. In a study using CTPA as the reference standard, the YEARS algorithm was more sensitive than the Wells score (97.4% vs 74.4%) but less specific (14.0% vs 33.9%), with a higher negative predictive value (98% vs 92.4%), while both rules shared a poor positive predictive value of around 10.9% -- confirming that these tools exclude rather than confirm PE ^[7]B2b. An ECG-based score has been reported to slightly outperform Wells and clearly outperform the Geneva score for pretest probability estimation, and a sufficiently sensitive negative D-dimer alone performed well as a stand-alone exclusion test in one cohort, though this remains contrary to current guideline practice ^[8]B2b. The electrocardiogram in PE may show sinus tachycardia, the classic S1Q3T3 pattern, or signs of right ventricular strain; these are specific but insensitive and chiefly support risk stratification rather than diagnosis ^[8]B2b. Echocardiography is not used to confirm PE in stable patients but is invaluable at the bedside in the unstable patient, where right ventricular dilatation and dysfunction support empirical treatment while definitive imaging is arranged ^{[6]A1a[13]A1a}. The Ventilation-Perfusion Scan remains a useful alternative to CTPA, particularly in pregnancy or contrast allergy ^[23]B2a.

Pearl: At the bedside, the fastest rule-out is a non-high pretest probability plus a negative high-sensitivity D-dimer, which excludes PE without imaging; the fastest rule-in in an unstable patient is RV dysfunction on bedside echo supporting empirical treatment while CTPA is arranged ^{[9]B2a[18]A1a}.

Once PE is confirmed, risk stratification -- not clot size -- determines therapy. The ESC model combines haemodynamic status, a clinical score ( PESI or Simplified PESI|sPESI ), imaging RV dysfunction, and cardiac Troponin ^{[11]A1b[13]A1a}.

Severity / risk tools

The ESC scheme is anchored on sPESI: a score of 0 identifies a low-risk group with a 30-day death rate of only ~0.5%, while the combination of RV dysfunction AND elevated troponin defines the intermediate-high group whose death rate (~7.7%) warrants monitoring for deterioration ^{[11]A1b[12]B2b}.

The clinical power of this layered approach is that it separates two distinct questions: who is safe enough for outpatient or early-discharge management, and who is at sufficient risk of deterioration to warrant inpatient monitoring or pre-emptive escalation planning ^{[11]A1b[12]B2b[13]A1a}. The PESI and its simplified version were derived and validated primarily to identify the low-risk group, and the negative predictive value of a Simplified PESI|sPESI of zero for 30-day death is around 99.5%, which is what justifies considering these patients for ambulatory care ^[11]A1b. Crucially, in patients with sPESI of zero the ESC model allows right ventricular imaging and troponin testing to be omitted in roughly a fifth of patients while preserving this high negative predictive value ^[11]A1b.

The harder problem is the intermediate group, where the ESC model acknowledges that risk stratification "requires further improvement" because event rates in the intermediate-low and intermediate-high tiers (around 6.0% and 7.7% respectively in validation cohorts) overlap considerably ^[11]A1b. This is the rationale for refining stratification with additional biomarkers and organ-function measures: optimised high-sensitivity Troponin thresholds can identify both a very-low-risk group (cut-off around 12 pg/mL) and a higher-risk subgroup among those already flagged by sPESI or imaging (cut-off around 87 pg/mL), and renal dysfunction defined by eGFR adds incremental prognostic value on top of the ESC model, improving reclassification of intermediate- and high-risk patients ^[12]B2b. In practice, the PESI /sPESI score, imaging RV assessment and biomarkers are integrated rather than used sequentially, and many centres convene a PE response team for the most complex intermediate-high and high-risk cases ^[22]B2a.

Pearl: Management escalates at the intermediate-high threshold -- RV dysfunction on imaging PLUS elevated troponin in a normotensive patient -- because this is the group that may deteriorate and require rescue reperfusion despite initially "stable" vitals ^{[11]A1b[13]A1a}.

Acute PE has a distinct, time-critical acute presentation. The first decision is haemodynamic: is the patient in shock/hypotension (high-risk) or normotensive ^{[13]A1a[14]A1a}?

Acute management algorithm

Evidence-graded explanation

Prompt therapeutic anticoagulation is the cornerstone for ALL risk groups and is the MANAGEMENT OF CHOICE at first contact (1b; ESC/ACC Class I) ^{[14]A1a[21]A1a}. For most patients the DRUG OF CHOICE is a direct oral anticoagulant such as Apixaban or Rivaroxaban , which can be started without parenteral lead-in; alternatives are Low Molecular Weight Heparin bridging to Warfarin , or unfractionated Heparin when rapid reversibility or renal failure dictates ^{[1]A1a[18]A1a[21]A1a}.

For high-risk (massive) PE with shock, the reperfusion treatment of choice is systemic Thrombolysis with Alteplase unless contraindicated (1b; ESC Class I) ^{[18]A1a[19]A1a}. In contemporary PERT registry data, in-hospital mortality was 20.6% in high-risk and rose to 42.1% in catastrophic PE with collapse, underscoring the urgency ^[16]B2b. Routine systemic thrombolysis is NOT recommended for intermediate-risk PE because major bleeding outweighs benefit; it is reserved for those who deteriorate ^{[13]A1a[18]A1a}. Where thrombolysis fails or is contraindicated, Pulmonary Embolectomy|surgical embolectomy , Catheter-Directed Thrombolysis , or ECMO support are options ^{[14]A1a[16]B2b}. NNT not derivable from reported data for the registry mortality figures.

The defining feature of acute PE management is that the haemodynamic question must be answered before, or in parallel with, definitive imaging, because the unstable patient cannot safely wait ^{[13]A1a[16]B2b}. In the patient with shock, bedside Echocardiography showing a dilated, hypokinetic right ventricle is often sufficient to justify empirical reperfusion when transfer to CT is unsafe ^{[6]A1a[13]A1a}. For the normotensive patient, the priority is prompt therapeutic anticoagulation, which both prevents clot propagation and allows endogenous fibrinolysis to begin remodelling the existing thrombus ^{[13]A1a[18]A1a[21]A1a}.

The evidence base for reperfusion is nuanced. Systemic Thrombolysis reduces the combined endpoint of death and haemodynamic decompensation but at the cost of major and intracranial bleeding, so its net benefit is favourable only where the baseline risk of death is high -- that is, in high-risk PE -- and unfavourable in stable intermediate-risk PE, where guideline bodies recommend against routine use ^{[13]A1a[18]A1a}. This is why the management of choice in intermediate-high PE is anticoagulation with close monitoring and a pre-agreed plan for rescue reperfusion if the patient deteriorates, rather than upfront lysis ^{[13]A1a[16]B2b[18]A1a}. When systemic lysis is contraindicated or has failed, Pulmonary Embolectomy|surgical embolectomy , Catheter-Directed Thrombolysis and mechanical Catheter Embolectomy|thrombectomy are the alternatives, and refractory shock or arrest may be bridged with ECMO ^{[14]A1a[16]B2b}. Supportive measures matter: oxygen, cautious volume (because over-filling worsens septal bowing), and Norepinephrine as the preferred vasopressor to maintain right coronary perfusion are all part of initial stabilisation ^{[6]A1a[13]A1a}. Low-risk patients, by contrast, are increasingly managed with a DOAC as outpatients or after a short observation period, provided social and follow-up criteria are met ^{[18]A1a[22]B2a}.

Pearl: In high-risk PE, the time-sensitive action that most changes outcome is recognising shock early and giving systemic thrombolysis before haemodynamic collapse -- catastrophic PE with arrest carries ~42% in-hospital mortality versus ~17% for non-catastrophic high-risk PE ^[16]B2b.

After initial stabilisation, anticoagulation is the long-term backbone, every agent tied back to the mechanism -> target table (thrombus propagation) ^{[1]A1a[21]A1a}.

Guideline-directed therapy

Narrative

DOACs are first-line for almost all patients because of comparable efficacy and less major bleeding than vitamin K antagonists, and they simplify both initial and long-term therapy ^{[1]A1a[18]A1a[22]B2a}. Low Molecular Weight Heparin retains a defined role in active Cancer and pregnancy, where DOACs are avoided or used with caution, though factor Xa inhibitors are increasingly used in non-GI cancers ^{[22]B2a[25]B3a}. Minimum treatment duration is 3 months; PE provoked by a major transient factor can stop at 3 months, whereas unprovoked PE or PE with a persistent risk factor (e.g. cancer, antiphospholipid syndrome) is considered for extended, often reduced-dose, anticoagulation when recurrence risk outweighs bleeding risk ^{[18]A1a[19]A1a[21]A1a}. Decisions on extended therapy should use a recurrence-risk assessment and shared decision-making ^{[18]A1a[21]A1a}.

The choice of agent and the duration of therapy are two separate decisions. On the agent, Apixaban and Rivaroxaban can be initiated as monotherapy with an intensified loading phase (apixaban 10 mg twice daily for seven days, rivaroxaban 15 mg twice daily for three weeks) before stepping down to maintenance dosing, whereas Dabigatran and Edoxaban require an initial parenteral Low Molecular Weight Heparin lead-in ^{[13]A1a[18]A1a[21]A1a}. DOACs are preferred over Warfarin for almost all patients because they achieve comparable efficacy with a lower risk of major and intracranial bleeding and without the need for routine INR monitoring ^{[1]A1a[18]A1a[22]B2a}. The principal exceptions remain active Cancer -- where LMWH or selected factor Xa inhibitors are used, with DOAC caution in gastrointestinal malignancy -- and pregnancy, where LMWH is mandatory because DOACs cross the placenta and are contraindicated ^{[22]B2a[24]B2a[25]B3a}.

On duration, all patients receive a minimum of three months of therapeutic anticoagulation, which prevents early recurrence ^{[18]A1a[19]A1a[21]A1a}. Thereafter the decision turns on whether the event was provoked by a major transient factor (stop at three months), provoked by a persistent factor or unprovoked (consider extended therapy), with recurrent or cancer- or antiphospholipid-associated events warranting indefinite anticoagulation ^{[18]A1a[19]A1a}. For extended therapy, reduced-dose factor Xa inhibitors retain efficacy with a favourable bleeding profile, and the duration decision should be revisited periodically using a structured recurrence-versus-bleeding assessment and shared decision-making rather than set once and forgotten ^{[18]A1a[21]A1a}. Routine thrombophilia testing is not recommended and rarely changes management ^[19]A1a.

Pearl: The therapy most often under-applied is extended anticoagulation in unprovoked PE: stopping at 3 months in a truly unprovoked event leaves a high recurrence rate, yet patients are frequently taken off treatment prematurely without a structured recurrence-vs-bleeding assessment ^{[18]A1a[21]A1a}.

Interventional and device therapy is reserved for high-risk PE and selected deteriorating intermediate-high patients; anticoagulation alone suffices for most ^{[14]A1a[15]A1a}.

Procedures and devices

Catheter-based therapies are increasingly adopted by PE response teams, but high-quality randomised evidence is lacking: the only RCT of ultrasound-augmented catheter-directed thrombolysis (n=59) showed no clear benefit over heparin for mortality or bleeding, so current evidence does not support widespread routine use over established therapy ^[15]A1a. IVC filters should remain restricted to genuine contraindications to anticoagulation or recurrence despite adequate therapy ^[17]B2b.

The central tension in this field is the gap between rapidly expanding catheter technology and the relative scarcity of randomised evidence. The Cochrane systematic review of catheter-directed therapies identified only a single small randomised trial (n=59) of ultrasound-augmented catheter-directed thrombolysis in intermediate-risk PE and found no clear difference versus heparin alone in all-cause mortality, major or minor bleeding, recurrent PE, or length of stay, with the certainty of evidence rated very low ^[15]A1a. Much of the enthusiasm for catheter therapy rests on registries and single-arm series that report improvements in surrogate endpoints such as the right ventricular to left ventricular ratio or thrombus burden, but these have limited proven clinical utility and the review explicitly cautioned against widespread adoption on this basis ^[15]A1a.

Observational data nonetheless suggest a role in carefully selected patients. In a single-centre cohort, mechanical Catheter Embolectomy|thrombectomy was associated with lower in-hospital mortality than routine care in elevated-risk patients, and contemporary PERT registry data show that advanced therapies are used more often as risk rises, with systemic Thrombolysis and ECMO concentrated in catastrophic high-risk PE ^{[16]B2b[26]B2b}. Inferior Vena Cava Filter|IVC filters occupy a narrow, well-defined niche: they are indicated when anticoagulation is absolutely contraindicated or when PE recurs despite adequate anticoagulation, and registry data show they are actually used in only a small minority (around 2.7%) of haemodynamically unstable patients, with a suggestion of reduced short-term mortality when placed early in high-risk patients ^[17]B2b. Decisions of this complexity are increasingly made by multidisciplinary PE response teams ^{[14]A1a[22]B2a}.

Pearl: The selection criterion most often misapplied is using catheter-directed or systemic reperfusion for stable intermediate-risk PE on the basis of clot burden or RV:LV ratio alone -- surrogate endpoints that have not translated into proven mortality benefit; reperfusion should be driven by haemodynamic deterioration ^{[13]A1a[15]A1a}.

Complications of acute PE divide into disease-related sequelae and therapy-related (iatrogenic) harms ^{[3]A1a[20]B2a}.

Complications

Up to ~50% of VTE survivors have long-term morbidity, and roughly 20% of people with a VTE event die within a year (often from the provoking condition) ^{[4]A1a[15]A1a}.

The complications of PE can be grouped by timescale as well as by whether they stem from the disease or its treatment. In the acute phase, the dominant disease complication is obstructive shock progressing to Cardiac Arrest , which is the proximate cause of early death and the target of emergency reperfusion ^[16]B2b. Over the medium to long term, the most important sequelae are the post-PE syndrome family: persistent right ventricular impairment, post-PE functional limitation with chronic dyspnoea and reduced quality of life despite adequate anticoagulation, and -- most seriously -- Chronic Thromboembolic Pulmonary Hypertension , which arises when organised thrombus fails to resolve and the pulmonary vasculature remodels ^{[3]A1a[20]B2a}. CTEPH is the one chronic complication that is potentially curable, by pulmonary endarterectomy or balloon pulmonary angioplasty, so the key clinical discipline is to screen survivors with persistent dyspnoea for it before attributing symptoms to deconditioning ^[20]B2a.

On the iatrogenic side, the principal harm is bleeding from anticoagulant and thrombolytic therapy: in contemporary high-risk PE registries, major bleeding occurred in around 10.5% of high-risk versus 3.5% of intermediate-risk patients, reflecting both the severity of illness and the intensity of treatment ^[16]B2b. Heparin-induced thrombocytopenia is a less common but serious immune complication of Heparin exposure that paradoxically causes thrombosis and mandates a switch to a non-heparin anticoagulant. Recurrent VTE, driven by inadequate dose or duration of anticoagulation or by a persisting risk factor, links back to the long-term therapy decisions and is itself a precursor to the chronic syndromes ^{[18]A1a[21]A1a}.

A practical corollary is that complication surveillance should be built into structured follow-up rather than left to chance. Survivors warrant clinical review at three to six months to reassess symptoms, screen for the post-PE syndrome, and make the duration-of-anticoagulation decision; persistent breathlessness at this point should trigger objective evaluation for Chronic Thromboembolic Pulmonary Hypertension rather than reassurance alone, given that CTEPH is uncommon but potentially curable and is easily missed when symptoms are attributed to deconditioning or anxiety ^[20]B2a. Balancing this against the iatrogenic bleeding risk of continued anticoagulation is the core of long-term management, and it is why the recurrence-versus-bleeding assessment recurs throughout PE care ^{[18]A1a[21]A1a}.

Pearl: The most preventable complication is recurrent VTE and its downstream CTEPH/post-PE syndrome -- correct, adequately dosed and adequately durationed anticoagulation from the outset prevents most recurrences and the chronic sequelae they seed ^{[18]A1a[21]A1a}.

Untreated, PE carries a high mortality from RV failure and recurrence; with prompt anticoagulation and risk-adapted care, outcomes improve markedly, though long-term morbidity is common ^{[3]A1a[13]A1a}. By ESC risk tier, 30-day death was reported at 22% in high-risk, 7.7% in intermediate-high, 6.0% in intermediate-low, and 0.5% in low-risk patients, with sPESI 0 giving a 99.5% NPV for survival ^[11]A1b. In contemporary high-risk registry cohorts, in-hospital mortality was 20.6%, rising to 42.1% in catastrophic PE with haemodynamic collapse ^[16]B2b. Across all VTE, roughly 20% of patients die within one year, frequently from the underlying provoking illness rather than PE itself ^[4]A1a.

The strongest predictors of poor outcome are haemodynamic instability, RV dysfunction, elevated Troponin , higher PESI / Simplified PESI|sPESI class, renal dysfunction, vasopressor and ECMO requirement, clot-in-transit, and active Cancer (malignancy OR 1.70 for in-hospital death) ^{[11]A1b[12]B2b[16]B2b}. Up to half of survivors report persistent functional limitation ^[20]B2a.

The natural history of untreated PE is dominated by early death from right ventricular failure and by recurrent embolisation, which is precisely why prompt anticoagulation -- which prevents propagation and recurrence -- so dramatically alters the trajectory ^{[3]A1a[13]A1a[18]A1a}. With risk-adapted, guideline-directed care, the prognosis of low-risk PE approaches that of the background population, whereas high-risk PE remains lethal: the ESC validation cohort reported 30-day death rates of 22% in high-risk, 7.7% in intermediate-high, 6.0% in intermediate-low and 0.5% in low-risk patients, and contemporary registry data report in-hospital mortality of 20.6% in high-risk PE rising to 42.1% in catastrophic PE with haemodynamic collapse ^{[11]A1b[16]B2b}. Across the whole VTE spectrum, approximately one in five patients dies within a year, though often from the underlying provoking condition rather than the embolism itself ^[4]A1a.

The strongest predictors of poor outcome map directly onto the risk-stratification tools: haemodynamic instability, right ventricular dysfunction, elevated Troponin , higher PESI / Simplified PESI|sPESI class, renal dysfunction, and the need for vasopressors or ECMO , together with clot-in-transit and active Cancer (associated with an OR of 1.70 for in-hospital death) ^{[11]A1b[12]B2b[16]B2b}. Beyond survival, up to half of survivors report persistent functional limitation, so prognosis must be framed in terms of long-term morbidity as well as mortality ^{[4]A1a[15]A1a[20]B2a}.

Two further nuances shape prognostic counselling. First, much of the early mortality in PE is front-loaded into the first hours and days, when right ventricular failure is most likely to decompensate, so survival curves separate early and then flatten; this is the physiologic basis for the urgency of risk stratification and reperfusion in the high-risk group ^{[13]A1a[16]B2b}. Second, the cause of death differs by population: in unprovoked or cardiopulmonary-limited patients, death is more often directly attributable to the embolism and its haemodynamic consequences, whereas in cancer-associated VTE the underlying malignancy frequently drives one-year mortality, which is why the headline figure that roughly one in five VTE patients dies within a year must always be interpreted in light of the provoking context ^{[4]A1a[16]B2b}.

Pearl: The single modifiable factor that most improves prognosis is timely, correctly dosed reperfusion-and-anticoagulation guided by early risk stratification -- recognising and treating the high-risk patient before collapse converts a ~42% catastrophic mortality toward the far lower stable-PE figures ^{[11]A1b[16]B2b}.

Several populations change diagnostic and therapeutic decisions in PE ^{[23]B2a[24]B2a[25]B3a}.

Special populations

Prevention

• Primary prevention: pharmacologic and mechanical thromboprophylaxis during hospitalisation, surgery, and other high-risk states, where most preventable VTE occurs (reported) ^[4]A1a.

• Secondary prevention: appropriate-duration anticoagulation after the index event, with extended reduced-dose Apixaban / Rivaroxaban for unprovoked PE when recurrence risk outweighs bleeding ^{[18]A1a[21]A1a}.

• Postpartum thromboprophylaxis for at-risk women, with duration and dose individualised ^[24]B2a.

These special populations matter because they change one or more of the diagnostic threshold, the imaging modality, the anticoagulant, or the duration of therapy. In Pregnancy , PE is among the leading causes of maternal death in high-income countries, and the diagnostic challenge is compounded because physiological dyspnoea, tachycardia and leg swelling overlap with PE symptoms; the pregnancy-adapted YEARS Algorithm and the revised Geneva Score are reported to be safe and effective for risk stratification, and CT pulmonary angiography and ventilation-perfusion scanning have comparable safety, with the choice often guided by chest radiograph findings ^{[23]B2a[24]B2a}. Iodinated contrast for CTPA is not associated with neonatal adverse events, and counselling to address maternal anxiety about fetal exposure is an important part of care ^[23]B2a. Low Molecular Weight Heparin is the anticoagulant of choice throughout pregnancy and the postpartum period because DOACs and Warfarin are avoided ^{[24]B2a[25]B3a}.

In active Cancer , higher recurrence and bleeding risks and the independent mortality impact of malignancy favour LMWH or carefully selected factor Xa inhibitors and often longer or indefinite therapy ^{[16]B2b[22]B2a}. In the elderly, obese, and those with renal impairment, anticoagulant pharmacokinetics shift and bleeding risk rises, so dose adjustment is required; agents such as Tinzaparin have been used with favourable safety and efficacy in renal impairment, the elderly, obesity and pregnancy with once-daily dosing ^[25]B3a. Prevention is where the greatest population gains lie: most preventable VTE is hospital- and surgery-associated, making systematic thromboprophylaxis the highest-yield primary-prevention intervention, while appropriate-duration anticoagulation after the index event, including extended reduced-dose Apixaban or Rivaroxaban for unprovoked PE, constitutes secondary prevention ^{[4]A1a[18]A1a[21]A1a}.

Pearl: The caveat most often missed is in pregnancy -- DOACs are contraindicated, Low Molecular Weight Heparin is the agent of choice, and PE must be actively and promptly investigated because physiological pregnancy symptoms mask it and delayed recognition drives maternal mortality ^{[23]B2a[24]B2a}.