Coronary embolism

Article written and reviewed by Dr Luca Moretti

Coronary embolism is the partial or complete occlusion of a coronary artery by material arriving from a distant site. The material is most often thrombotic but may contain infected vegetation, neoplastic tissue, valvular fragments, calcium, fat, air, or foreign material; its extracoronary origin distinguishes an embolus from locally formed thrombosis. The distinction is not merely semantic. It determines the etiological investigation, antithrombotic strategy, need for stenting, and prevention of recurrence.
The most recognizable presentation is myocardial infarction with angiographic occlusion and otherwise minimally diseased arteries, but the embolus may dissolve before coronary angiography and present as MINOCA. It may also lodge distally, fragment in the microcirculation, or involve multiple coronary territories; the angiographic phenotype depends on the size, consistency, and trajectory of the embolus. A normal coronary artery after aspiration is suggestive, not conclusive. Intracoronary imaging and investigation of the source complete the diagnosis.

In registries, embolism accounts for a minority of infarctions, with estimates varying according to selection and diagnostic criteria. In the original Shibata cohort, it accounted for 2.9 percent of first infarctions; contemporary series have produced different values and confirm underdiagnosis; atrial fibrillation is the most common cause. Endocarditis, prosthetic valves, and ventricular thrombi retain clinical importance. The data derive mainly from observational studies rather than dedicated therapeutic trials.
The patient should initially be treated as having an acute coronary syndrome because symptoms and ECG do not identify the origin of the occlusion. Etiological diagnosis proceeds in parallel with reperfusion and continues after flow has been restored; the dual objective is to salvage myocardium and interrupt the embolic source. Technically successful PCI does not complete the pathway. Without identification of the source, the risk of stroke, systemic emboli, and recurrent embolic infarction persists.

Etiology, risk factors, pathogenesis, and pathophysiology

In atrial fibrillation, stasis in the left atrial appendage, endothelial dysfunction, and hypercoagulability promote thrombus formation. A fragment passes through the mitral valve, left ventricle, and aortic root before entering a coronary ostium. Flow geometry makes the event less common than cerebral or peripheral embolism; atrial appendage thrombus nevertheless remains the dominant source in modern cohorts. Paroxysmal fibrillation may not appear on the initial ECG. Absent, interrupted, or underdosed anticoagulation increases plausibility without proving it.
A left ventricular thrombus may develop after a large anterior infarction, apical aneurysm, dilated cardiomyopathy, or severe systolic dysfunction. Akinetic segments produce stasis, while endocardial injury and inflammation complete the thrombogenic triad; a ventricular source may embolize to a coronary artery, although it more often affects other territories. A thrombus arising as a consequence of the index infarction must not be mistaken for its cause. Timing and previous imaging are therefore essential.

Infective endocarditis produces friable vegetations composed of fibrin, platelets, and microorganisms. A fragment may occlude a coronary artery, cause local arteritis, an abscess, or a mycotic aneurysm, and disseminate infection into the myocardium. A septic embolus radically changes treatment. Fever, positive blood cultures, and vegetations increase suspicion. Systemic thrombolysis creates a risk of intracranial hemorrhage, especially in the presence of cerebral emboli or infectious aneurysms, and is not an automatic solution.
Nonbacterial vegetations associated with cancer, lupus, or antiphospholipid antibody syndrome may produce sterile emboli. Thrombosed mechanical prostheses, degenerated bioprostheses, and perivalvular thrombotic material are other sources; valvular disease requires transesophageal echocardiography and urgent specialist assessment. A normal prosthetic gradient does not exclude small mobile thrombi. The choice among anticoagulation, selected fibrinolysis, and surgery depends on the valve, size, obstruction, infection, and stability.

Cardiac tumors may embolize directly or promote surface thrombi. A villous atrial myxoma is particularly friable; papillary fibroelastoma of a valve, sarcomas, and metastases are rarer sources. Histology of the aspirate may reveal tumor cells or nonthrombotic matrix; neoplastic tissue transforms a coronary finding into a systemic diagnosis. Echocardiography and MRI define location and mobility. Resection is considered in relation to embolic risk, operability, and oncological prognosis.
In paradoxical embolism, a venous thrombus crosses a patent foramen ovale, atrial septal defect, pulmonary arteriovenous fistula, or another right-to-left communication. Passage requires a favorable gradient, often transient during pulmonary embolism, Valsalva, or increased right-sided pressure; the paradoxical route is not demonstrated by the shunt alone. A plausible venous source, communication, and absence of a better alternative are required. PFO is common in the population and may be incidental.

Iatrogenic causes include thrombi formed in catheters, atheromatous or calcific fragments mobilized from the aortic root, air, valvular tissue, and materials introduced during PCI, cardiac surgery, or structural interventions. TAVI and mitral procedures may direct debris toward the coronary ostia. Procedural embolism occurs in an immediate or early temporal relationship. Inadequate procedural anticoagulation and insufficient catheter flushing are preventable factors. The embolized material guides retrieval and support.
An embolus may enter the right or left coronary artery according to ostial orientation and aortic dynamics. It lodges when vessel diameter becomes smaller than its caliber, causing an abrupt occlusion or mobile filling defect; a distal location is common but not obligatory. Spontaneous fragmentation or fragmentation caused during manipulation may transform a proximal occlusion into diffuse microembolization. Final injury depends on duration, collaterals, territory, and successful reperfusion.

Ischemia activates anaerobic metabolism, contractile dysfunction, and electrical instability; if prolonged, it produces necrosis according to a coronary territory. Septic emboli add inflammation and microbial invasion, while cholesterol crystals or calcium cause obstruction and a vascular reaction; the ischemic mechanism is shared, but the biological substrate varies; for this reason, angiographic appearance alone is insufficient to select definitive treatment. Analysis of the material and investigation of the source are causal rather than ancillary components.

Pathological anatomy, classification, and evolution

The most useful classification separates direct, paradoxical, and iatrogenic embolism. The direct form originates in the left-sided chambers, valves, pulmonary veins, or aorta and reaches the coronary artery without crossing an abnormal communication. Direct embolism includes thrombi, vegetations, and tumors. The paradoxical form originates in the venous compartment. The iatrogenic form results from material or thrombus mobilized during a procedure, although the boundaries may overlap.
Fresh thrombotic material is rich in platelets, fibrin, erythrocytes, and leukocytes in variable proportions. Over time, retraction, organization, fibroblastic proliferation, endothelialization, and canalization appear. A preexisting venous thrombus may therefore appear more organized than a thrombus newly formed in a left-sided chamber. Histological age provides clues, not exact dating. Antithrombotic therapy and manipulation alter the sample. A small aspirate may not represent the entire embolus.

Infective vegetations show fibrin, inflammation, and microbial aggregates, but previous antibiotics may reduce microbiological yield. Tumor material requires immunohistochemistry when morphology does not identify its origin. Calcific or valvular fragments may be sparse and rigid; pathological diagnosis depends on proper specimen preservation. One portion should be sent in formalin for histology. When infection is suspected, appropriate material should also be sent for culture and molecular methods.
Angiographically, an embolus may produce an abrupt cutoff in a segment without progressive narrowing, an intraluminal filling defect, a convex contour, or contrast flowing along the sides of the material. Simultaneous emboli in different branches or territories are particularly suggestive; the multiterritorial sign is a major Shibata criterion. No morphology is pathognomonic. Plaque rupture, dissection, spasm, and in situ thrombosis may mimic the same image.

Distal migration during contrast injection, guidewire passage, or aspiration may reopen the proximal segment while leaving slow flow, branch cutoff, and loss of myocardial blush. An apparently normal artery after the procedure may therefore conceal residual microvascular embolization; embolic no-reflow results from arteriolar obstruction, edema, vasoconstriction, and ischemic injury. Epicardial TIMI 3 flow does not guarantee normal tissue perfusion. ECG, blush, echocardiography, and MRI assess complementary consequences.
The embolus may dissolve spontaneously or under treatment before angiography. In this scenario, infarction with nonobstructive arteries remains, and an ischemic distribution on cardiac MRI may be the only tissue evidence; if coronary embolism is confirmed as the cause, the final diagnosis is primary myocardial infarction due to coronary embolism; MINOCA remains only the initial working diagnosis in the presence of nonobstructive coronary arteries. Myocarditis and Takotsubo syndrome must be excluded. Investigation of the source retains value even without visible thrombus, provided the reasoning does not become purely presumptive.

More rarely, the embolus remains nonocclusive, organizes in the wall, or becomes incorporated, creating late stenosis. Infected material may weaken the wall and form an aneurysm or pseudoaneurysm. A tumor embolus may grow locally or recur. Chronic evolution depends on the nature of the material. Anatomical follow-up is not necessary in every resolved case. It is indicated when stenosis persists, infection is suspected, an aneurysm develops, or symptoms remain.
The Shibata diagnostic classification distinguishes definite and probable embolism by combining major and minor criteria. It was developed retrospectively in patients with a first infarction and is not a prospectively validated standard in every population; the Shibata score improves uniformity and clinical suspicion. It does not replace OCT, IVUS, MRI, or multidisciplinary judgment. Concomitant atherosclerosis may reduce its applicability without truly excluding a superimposed embolus.

Clinical manifestations

The most common presentation is an acute coronary syndrome with sudden chest pain, ischemic ECG changes, and a dynamic rise in troponin. STEMI is frequent in embolic series because occlusion may be complete and not preceded by progressive collateral stenosis; the acute presentation is indistinguishable from atherothrombosis before angiography. Young age or few atherosclerotic risk factors increase suspicion. They must not, however, delay the reperfusion pathway dictated by risk.
A distal or transient occlusion may produce NSTEMI, brief pain, or silent ischemia. Troponin may be modest if the embolus dissolves rapidly or affects a small branch, whereas embolism of the left main or a proximal segment may cause immediate collapse; clinical variability reflects anatomy more than the nature of the embolus. A nondiagnostic initial ECG does not exclude the event. Serial tracings and biomarker kinetics remain necessary.

Dyspnea, pulmonary edema, hypotension, and shock indicate extensive loss of contractility, papillary ischemia, or multivessel involvement. Arrest may result from ventricular fibrillation, atrioventricular block, or electromechanical dissociation. In these cases, the etiological history is obtained without delaying stabilization and reperfusion. Hemodynamic instability requires emergent coronary angiography. Bedside ultrasonography identifies dysfunction and mechanical complications. The presence of sepsis or a prosthetic valve simultaneously modifies support and the differential diagnosis.
Signs of concomitant embolization have high diagnostic value. A neurological deficit, abdominal pain, limb ischemia, or renal or splenic infarctions suggest an active systemic source. The absence of clinical manifestations does not exclude small emboli detectable on imaging. Simultaneous systemic embolism is a major criterion when not explained by ventricular thrombus resulting from the index infarction. The neurological and vascular examinations should be documented. Extracardiac imaging is guided by symptoms rather than performed indiscriminately.

Known atrial fibrillation, previous palpitations, or discontinuation of anticoagulation points toward an atrial appendage origin. Sinus rhythm on admission does not exclude paroxysmal episodes. A history of recent infarction, cardiomyopathy, or apical aneurysm instead points toward ventricular thrombus. The rhythm history includes reports, devices, and episode duration. Adherence is reconstructed without judgment. Dose, renal function, interactions, and recent procedures may explain insufficient anticoagulant protection.
Fever, a new murmur, immunological phenomena, multiple emboli, or persistent infection should raise suspicion of endocarditis. In patients with prostheses, fever may be absent and acute valve dysfunction may dominate. The combination of infarction and hemorrhagic stroke makes treatment particularly complex. Embolic endocarditis requires blood cultures before antibiotics whenever possible. A life-threatening emergency justifies immediate treatment. Assessment involves cardiology, infectious diseases, cardiac surgery, and neurology.

In paradoxical embolism, dyspnea, hypoxemia, limb pain or edema, and signs of right-sided overload may coexist. An acute increase in right-sided pressures facilitates passage through a PFO. Venous thrombosis is not always still demonstrable after migration; the thromboembolic context supports a causal chain, not an automatic certainty. Contrast echocardiography defines the shunt. Pulmonary CT angiography and venous ultrasonography are indicated according to clinical probability and safety.
Iatrogenic embolism presents during or immediately after catheterization, TAVI, surgery, or another procedure. A sudden ST-segment deviation, hypotension, arrhythmia, or angiographic loss of a branch requires immediate recognition. The nature of the material may be inferred from the preceding procedural step; procedural timing is a strong causal clue. Air, thrombus, and a calcific fragment require different maneuvers. Angiographic recordings and retrieved materials should be preserved for review and prevention.

Investigations and diagnosis

Initial assessment follows the acute coronary syndrome pathway: ECG within minutes, repeated tracings, high-sensitivity troponin, complete blood count, coagulation tests, renal function, and electrolytes. Urgent echocardiography assesses function, territory, and complications without delaying reperfusion in STEMI; the diagnostic profile broadens when embolic clues emerge. Blood cultures precede antibiotics if endocarditis is suspected. INR, time since the last dose, and concomitant drugs guide anticoagulation management.
Coronary angiography may show an abrupt occlusion, filling defect, mobile thrombus, or emboli in multiple branches with little atherosclerotic burden elsewhere. Comparison with previous angiograms is valuable. An underlying lesion cannot be assessed correctly until the material is removed or reduced; causal angiography should describe location, TIMI flow, and thrombus burden. An intracoronary vasodilator helps identify reversible epicardial spasm; a negative response does not exclude a microvascular vasomotor component. Repeated manipulation may push the embolus distally.

Shibata criteria for the diagnosis of coronary embolism

• Major criteria: angiographic evidence of embolus or thrombus without an atherosclerotic component; concomitant emboli in multiple coronary territories; concomitant systemic embolization without ventricular thrombus attributable to the index infarction.
• Minor criteria: stenosis below 25 percent in nonculprit coronary arteries; embolic source demonstrated by echocardiography, CT, or MRI; embolic risk factors such as atrial fibrillation, cardiomyopathy, rheumatic valvular disease, a prosthetic valve, PFO, atrial septal defect, cardiac surgery, endocarditis, or a hypercoagulable state.
• Definite embolism: at least two major criteria, or one major and at least two minor criteria, or at least three minor criteria.
• Probable embolism: one major and one minor criterion, or two minor criteria.
• Proposed exclusions: pathological anatomy consistent with atherosclerotic thrombus, history of coronary revascularization, coronary ectasia, or plaque rupture or erosion demonstrated by IVUS or OCT.

The Shibata criteria are a classification tool derived from a retrospective cohort, not independent etiological proof. Some criteria are not independent, and the 25 percent threshold may exclude patients with incidental atherosclerosis but a true embolus; external validity remains incomplete. The original exclusions reflect the study design more than absolute biological impossibilities. The score should therefore organize evidence rather than replace clinical reconstruction.
After aspiration or recanalization, OCT may identify cap rupture, a plaque cavity, findings consistent with erosion over an intact cap, a calcified nodule, dissection, or stent malapposition. An intact wall without a thrombogenic substrate supports embolism, particularly if a source coexists. Coronary OCT has high resolution but requires contrast flushing. A large residual thrombus limits visualization. Instability, renal failure, or an uncrossable occlusion may prevent its use.

IVUS penetrates more deeply and defines plaque, remodeling, intramural hematoma, and stent geometry with less dependence on clearing the lumen. Surface resolution is lower than OCT, but it may be preferable in large vessels or renal failure when contrast is limited. Coronary IVUS does not directly identify the origin of the embolus. It demonstrates or excludes a plausible local substrate. The two techniques are complementary and should be used when they change the strategy.
Thrombectomy offers a unique diagnostic opportunity. The specimen should be photographed, described, and sent for histology; sterile microbiological handling is also required if infection is suspected. Fibrin and platelets confirm thrombus but do not automatically establish its site of origin. Histology of the aspirate may identify organization, vegetation, neoplastic tissue, or foreign material. A nonspecific result does not exclude embolism. Absence of a specimen may result from fragmentation or ineffective aspiration.

Transthoracic echocardiography assesses ventricular thrombus, aneurysm, global and regional function, valves, and right-sided pressures. Echocardiographic contrast improves visualization of the apex when images are suboptimal. Cardiac MRI is more sensitive for ventricular thrombus and characterizes the infarction; an intracardiac thrombus must be distinguished from trabeculae and tumors. A negative first echocardiogram does not end the investigation. The choice of method depends on the suspected source and stability.
Transesophageal echocardiography visualizes the left atrial appendage, prostheses, vegetations, thoracic aorta, and atrial septum in greater detail. Agitated saline contrast and the Valsalva maneuver document a right-to-left shunt, while Doppler and three-dimensional images define anatomy; an atrial source may be a thrombus or marked spontaneous contrast. Sedation and instability limit immediate examination. In prosthetic endocarditis, an early negative TEE may require repetition or multimodality imaging.

Telemetry during hospitalization and ambulatory monitoring look for paroxysmal fibrillation. Duration is calibrated to pretest probability and therapeutic consequences; brief monitoring has limited sensitivity for rare episodes. Implantable devices are reserved for selected cases, not every unexplained embolism; rhythm monitoring must distinguish clinical fibrillation from artifact. A documented episode does not necessarily prove causality. Duration, burden, atrial disease, and competing sources enter the interpretation.
If paradoxical embolism is suspected, compression ultrasonography looks for venous thrombosis and CT angiography assesses pulmonary embolism when indicated. PFO, venous thrombus, and consistent timing strengthen the diagnosis; without a venous source, the shunt remains a possible incidental route; the paradoxical chain requires all plausible steps. Elevated right-sided pressure supports the gradient. PFO closure after coronary embolism does not have the same randomized evidence available for some cryptogenic strokes.

The differential diagnosis includes atherosclerotic rupture or erosion, nonatherosclerotic in situ coronary thrombosis, SCAD, vasospasm, ectasia, trauma, and chronic occlusion. Cardiac MRI separates infarction from myocarditis and Takotsubo syndrome when coronary arteries are nonobstructive; the differential diagnosis should be positive, not based on summary exclusion. OCT or IVUS clarifies many ambiguous lesions. The concomitant presence of atrial fibrillation does not justify labeling every infarction as embolic.

Treatment and prognosis

In STEMI with suspected embolism, the priority is to restore flow rapidly through primary PCI. Aspirin, P2Y12 inhibition, and procedural anticoagulation are initially administered according to the acute coronary syndrome presentation, except for specific contraindications such as endocarditis with intracranial hemorrhage; the initial strategy often precedes etiological certainty. During the procedure, findings may reduce or increase the need for prolonged antiplatelet therapy. Every modification simultaneously considers thrombosis and bleeding.
Manual aspiration is not routinely recommended in STEMI because large trials have not demonstrated clinical benefit and have signaled a possible increase in stroke. In embolism with large accessible material, however, it may be used selectively to restore perfusion and obtain a specimen; selective thrombectomy does not contradict the prohibition against systematic use. The technique avoids deep aspiration and reinjection of material. Multiple passes increase trauma and distal dispersion.

After removal of the embolus, a normal coronary artery without dissection or causal plaque does not automatically require balloon angioplasty or stenting. Avoiding an unnecessary implant reduces future thrombosis and the need for dual antiplatelet therapy. Significant residual stenosis or a procedural complication may instead justify imaging-guided PCI. A no-stent strategy is a documented anatomical decision. It does not mean forgoing reperfusion. OCT or IVUS strengthens the choice when the vessel is stable and assessable.
Distal emboli or organized material may not be amenable to aspiration. Parenteral anticoagulation, vasodilators for a microvascular component, and hemodynamic support are individualized; intracoronary fibrinolytics have been used in exceptional cases, but evidence derives from small series; rescue therapy has no universal protocol. Glycoprotein IIb/IIIa inhibitors are reserved for selected thrombotic complications or no-reflow. Bleeding risk and the nature of the embolus limit intensification.

In endocarditis, targeted antibiotics and cardiac surgical assessment treat the source. Mechanical thrombectomy may be preferable for an accessible occlusion, while systemic fibrinolysis is generally avoided because of intracranial bleeding risk. A stent in a contaminated segment may become infected or develop an aneurysm; septic reperfusion requires immediate multidisciplinary discussion. Blood cultures and histology guide the antibiotic. Surgical indications also depend on valvular heart failure, uncontrolled infection, and embolic risk.
Atrial fibrillation with an embolic event is a strong indication for long-term anticoagulation unless contraindicated. DOACs are generally preferred to vitamin K antagonists in atrial fibrillation, but mechanical valves and moderate-to-severe rheumatic mitral stenosis require a vitamin K antagonist. Cardioembolic prevention must use an appropriate dose. Renal function, age, weight, and interactions are reassessed. An event during treatment requires verification of adherence and dosage before switching agents.

For ventricular thrombus, the duration and type of anticoagulation depend on etiology, morphology, ventricular function, and documented resolution. The AHA scientific statement considers a course of approximately three months after infarction reasonable, with follow-up imaging, whereas nonischemic cardiomyopathy may require at least three to six months. Anticoagulation duration is not identical for every source. Persistent thrombus or severe dysfunction may prolong it. Bleeding risk is reassessed periodically.
In paradoxical embolism, treatment includes anticoagulation of venous thromboembolism for the duration required by the provoked or persistent cause. PFO closure after isolated coronary embolism is supported mainly by pathophysiological reasoning and cases rather than dedicated trials; shunt closure is considered individually after causal confirmation. Shunt size, septal aneurysm, recurrences, and the need for anticoagulation influence the choice. Incidental PFO is insufficient.

If no stent was implanted and no causal plaque exists, the duration of antiplatelet therapy cannot automatically be extrapolated from atherothrombotic infarction. When anticoagulation is required, adding antiplatelet agents increases bleeding and requires a precise indication; antithrombotic simplification follows the diagnosis and anatomy. A recent stent requires combination treatment for the minimum effective duration. The plan should specify drugs, duration, and the time of reassessment.
The iatrogenic source is treated by removing or aspirating the material when possible, correcting anticoagulation, and preventing further introduction. Air embolism requires aspiration from the catheter, high-concentration oxygen, support, and maneuvers to disperse air according to location and stability; procedural prevention includes flushing, de-airing, and device checks. A structural fragment may require percutaneous or surgical retrieval. Review of the event identifies preventable process defects.

Immediate prognosis is dominated by territory, final flow, shock, and complications of infarction. In the long term, the disease that generated the embolus and the effectiveness of prevention carry weight. Recent cohorts confirm a substantial burden of heart failure, stroke, and mortality, especially with atrial fibrillation; residual risk persists even with anticoagulation; this does not demonstrate drug inefficacy but the complexity of the phenotype. Integrated follow-up monitors rhythm, source, ventricular function, and adherence.

Complications

Infarction may be extensive because embolic occlusion is sudden and the territory has not developed collaterals. Transmural necrosis creates a risk of remodeling, heart failure, and mechanical complications. The embolus may also prevent complete reperfusion despite an apparently open epicardial artery; myocardial injury is quantified with echocardiography and MRI. Heart failure prevention follows function and the clinical presentation. A nonatherosclerotic cause does not make the resulting infarction less severe.
Fragmentation and migration may occlude arterioles, causing no-reflow, reduced blush, and persistent ST-segment elevation. Microvascular injury increases infarct size and worsens functional recovery. Further catheter passes may clinically aggravate it; distal microembolization is difficult to remove mechanically. Vasodilators help only the vasomotor component. Prevention through gentle technique remains more effective than many rescue treatments.

Acute ischemia may cause ventricular tachycardia, ventricular fibrillation, bradycardia, or atrioventricular block. Location determines the type, while residual scar sustains late risk. Treatment follows resuscitation protocols and correction of the ischemic cause. Electrical instability does not prove an embolic mechanism. Once stabilized, ventricular function and residual arrhythmias are assessed. Indications for a defibrillator do not depend on the presence of the embolus alone.
Papillary muscle rupture, ventricular septal defect, and free-wall rupture are possible after transmural embolic infarction. In embolic cohorts, some mechanical complications appear more frequent than in nonembolic infarctions, but numbers remain limited; a mechanical complication causes a new murmur, shock, or pulmonary edema. Urgent echocardiography is decisive. Circulatory support and surgery are organized without awaiting complete definition of the embolic source.

The same source may send emboli to the brain, kidneys, spleen, mesentery, or limbs. Concomitant ischemic stroke complicates antithrombotic choices and may undergo hemorrhagic transformation, especially in endocarditis. Abdominal pain or a peripheral deficit must not be attributed to the stress of infarction; systemic dissemination requires targeted imaging and multidisciplinary coordination. The timing of anticoagulation depends on extent and bleeding. A second embolism may occur even after successful coronary reperfusion.
Infected emboli may cause a myocardial abscess, an infected coronary aneurysm, pericarditis, or persistent sepsis. A device implanted at a contaminated site may become another focus. Persistent fever or new coronary dilation requires timely reassessment; an infectious complication may emerge after the ischemic phase. CT, echocardiography, and angiography define extent. Antibiotic duration and surgery follow the microorganism, valve, and source control.

Recurrence may be coronary or extracoronary and reflects a persistent source, ineffective anticoagulation, uncontrolled infection, or active cancer. In the Shibata cohort, recurrent thromboembolic events mainly affected patients with atrial fibrillation; prevention of recurrence requires more than a prescription. Adherence, dose, renal function, and interactions are checked. A confirmed recurrence requires a renewed search for the source and collegial review of the strategy.
Anticoagulants and antiplatelet combinations increase gastrointestinal, intracranial, and procedural bleeding. Endocarditis with cerebral lesions, oncological thrombocytopenia, and recent surgery make the balance particularly difficult. Indiscriminately reducing treatment may nevertheless promote new emboli or stent thrombosis. Bleeding risk is reassessed dynamically. Modifiable factors are corrected and combination therapy is limited. Decisions, duration, and follow-up responsibility must be explicit at discharge.