Non-infective endocarditis

Article written and reviewed by Dr Luca Moretti

Non-infective endocarditis comprises conditions in which vegetations and other endocardial lesions form without a microorganism being the cause. The term encompasses heterogeneous processes: in thrombotic forms, deposition of platelets and fibrin on a predisposed endothelium predominates, whereas in immune-mediated forms valve injury also includes inflammation, necrosis, and subsequent fibrous repair. Not all sterile valvular masses belong to this group: tumors, Lambl's excrescences, and degenerative changes are alternative diagnoses, not subtypes of endocarditis. Likewise, a negative culture does not make a vegetation non-infective.

The two principal entities are nonbacterial thrombotic endocarditis, frequently associated with malignancy and hypercoagulability, and Libman-Sacks endocarditis, characteristic of the lupus and antiphospholipid setting. Terminology may overlap: NBTE is sometimes used broadly for sterile thrombotic vegetations, including autoimmune forms, whereas “marantic” refers mainly to the neoplastic and wasting-disease context. The phenotype and associated condition should therefore be specified rather than allowing the label alone to dictate treatment. Eosinophilic or fibrosing endomyocardial diseases instead have their own anatomic distribution and pathophysiology and should not automatically be assimilated to the valvular vegetations considered here.

There is no single reliable estimate of the incidence of the entire group. Autopsy series, oncology cohorts, and echocardiographic studies in lupus examine different populations and use tools with different sensitivities. An increase in diagnoses during life may reflect greater availability of transesophageal echocardiography and more careful investigation of embolic sources, without necessarily demonstrating an increase in biological incidence. Clinical relevance derives mainly from systemic emboli and valvular dysfunction: inconspicuous lesions can cause severe neurologic injury, whereas a valve that has become fibrotic can remain hemodynamically impaired even when inflammatory activity is controlled.

Etiology, pathogenesis, and pathophysiology

Formation of a sterile vegetation requires interaction between the endocardial surface and the blood environment. Healthy endothelium limits platelet adhesion and coagulation activation; mechanical injury, inflammatory mediators, or immune-mediated injury can alter its properties and expose thrombogenic surfaces. Platelet deposition and fibrin generation produce the core of the vegetation. In conditions of systemic hypercoagulability, a relatively limited local stimulus may be sufficient, and the valve may have no important pre-existing functional abnormality. This explains why absence of known heart disease does not exclude a valvular embolic source.

In malignancy, particularly certain adenocarcinomas, tumor cells and extracellular vesicles can activate coagulation through tissue factor and other signals; inflammation, platelets, and leukocytes amplify the response. Mucins and selectin-mediated interactions contribute to the prothrombotic setting of some tumors, while advanced disease may be associated with consumption of clotting factors and platelets. Not all patients with cancer develop vegetations and not every cancer-associated thrombosis is cardioembolic: venous or microvascular thrombosis can form independently of the valve. The neoplastic state should therefore be regarded as a plausible explanation for thrombogenicity, not as automatic proof of the embolus's site of origin.

In lupus, immune complexes, complement, and inflammation can damage the endothelium and valvular matrix, favoring secondary thrombotic deposition. In antiphospholipid syndrome, antibodies directed against phospholipid-binding proteins contribute to activation of endothelium, monocytes, and platelets and to loss of the antithrombotic balance. The two conditions can coexist but are not synonymous: lupus may be present without antiphospholipid syndrome, and the latter may occur without lupus. The valvular lesion may reflect a combination of immune-mediated injury and thrombosis, followed by fibrous organization. The resulting spectrum ranges from a recent, potentially modifiable vegetation to a retracted and irreversibly deformed valve apparatus.

The composition and attachment of the deposit determine clinical behavior. Marantic vegetations, often poor in inflammatory cells, may be friable and detach without causing obvious leaflet destruction. In Libman-Sacks endocarditis, the lesion may extend to both valvular surfaces and the subvalvular apparatus, with thickening and scarring. The distinction is useful but not absolute: no single gross or echocardiographic feature establishes the etiology with certainty. In particular, relative preservation of the valve favors a thrombotic form, but does not exclude early infective endocarditis.

Vegetations on left-sided valves drive arterial embolization. Fragmentation may affect different vascular territories at different times, causing cerebral, renal, or splenic infarctions and limb or intestinal ischemia. A small residual mass after an event does not demonstrate that the initial risk was low, because part of the deposit may already have detached. Embolic potential also does not coincide with the severity of regurgitation: a vegetation may leave coaptation nearly intact while remaining clinically dangerous. When deformity and fibrosis instead impair closure, regurgitation causes volume overload, chamber remodeling, and increased upstream pressures.

It is essential to distinguish damage caused by the valve from that caused by the systemic disease. In antiphospholipid syndrome, an occlusion may be a local thrombosis; in cancer, stroke and ischemia may have multiple mechanisms; in lupus, renal disease, atherosclerosis, and myocardial involvement may explain part of the cardiovascular dysfunction. The presence of a vegetation does not eliminate these alternatives. The most useful pathophysiologic framework therefore separates embolic source, systemic activity, and mechanical damage, because each may require a different intervention.

Clinical manifestations

The history should begin with the concrete reason that led to discovery of the lesion. Stroke, transient ischemic attack, amaurosis, or cognitive decline may precede any cardiac symptom. Onset, duration, recovery, and recurrences should be reconstructed, distinguishing episodes that preceded therapy from those that occurred during genuinely adequate drug exposure. Sudden flank pain, hematuria, abdominal pain, or cooling of a limb suggest possible extracerebral sites. The chronology of events in different territories may make a disseminated embolic process plausible without yet defining whether it is sterile or infective.

The search for the predisposing condition includes oncologic history, disease course and treatments, weight loss, and organ-specific symptoms; on the autoimmune side it includes lupus, documented thromboses, obstetric complications, nephropathy, cytopenias, and cutaneous or articular manifestations. It is important to distinguish clinically defined antiphospholipid syndrome from isolated antibody positivity. Medication history should document anticoagulants, antiplatelet agents, immunosuppressants, and antibiotics, including interruptions for procedures and adherence problems. This information modifies both the probability of the different diagnoses and the meaning of subsequent tests.

Fever, chills, vascular access, prostheses, intracardiac devices, recent procedures, and possible portals of entry must be sought even when an autoimmune or neoplastic explanation already exists. A patient may simultaneously have cancer and infective endocarditis, or a sterile vegetation may subsequently become infected. Absence of fever does not exclude infection in older or immunosuppressed patients; fever itself may have neoplastic or inflammatory causes. It is therefore an error to build the diagnosis on a rigid contrast between febrile and afebrile patients.

Cardiopulmonary symptoms include dyspnea, orthopnea, reduced functional capacity, and palpitations when regurgitation, heart failure, or arrhythmias are present. Attribution must also consider anemia, thromboembolism, effusions, treatment toxicity, and myocardial involvement. Physical examination assesses vital signs, perfusion, and neurologic status, followed by pulses and signs of peripheral ischemia. Auscultation looks for murmurs, but their absence does not exclude an emboligenic lesion; crackles, edema, and venous pressure help identify hemodynamic consequences. General examination searches for infectious foci, signs of autoimmune disease, and oncologic findings while maintaining the distinction between a clue and etiologic proof.

Investigations and diagnosis

The diagnostic pathway must answer four questions: whether a true valvular lesion exists, whether it is infected, what condition sustains it, and what damage it has produced. There is no single test for the entire group and no universally validated criteria for all non-infective endocarditis. Diagnosis is generally clinical and integrative; histologic confirmation is possible when surgical tissue is available, without this justifying an operation performed solely for diagnosis. Explicitly stating residual uncertainty is particularly important when previous antibiotics have reduced microbiologic yield.

Transthoracic echocardiography is the first examination for assessing masses, valvular function, chambers, and ventricles. Transesophageal echocardiography further evaluates an equivocal finding or high suspicion despite a negative examination, particularly in the presence of emboli and small lesions. The report should describe number, location, mobility, attachment base, and consequences of the finding; merely stating that a mass is present is insufficient for a treatment decision or subsequent comparison. Three-dimensional imaging improves anatomic description in selected cases, but does not identify microorganisms or prove autoimmunity. CT and other modalities contribute to specific questions while retaining limitations in characterization of small vegetations.

Morphologic comparison includes infected vegetation, thrombotic deposit, fibroelastoma, Lambl's excrescence, and degeneration. A thin, filamentous structure on the closure line has a different meaning from a pedunculated mass or diffuse thickening, but none of these appearances eliminates the need for clinical context. Abscess, fistula, perforation, and rapidly progressive destruction strengthen the hypothesis of infection; their absence does not exclude it. A decrease in mass during anticoagulation can support a thrombotic component, but does not retrospectively correct an incomplete microbiologic evaluation.

Microbiologic investigation should precede antibiotics whenever possible, with multiple blood cultures and adequate volumes. When endocarditis is suspected, three peripheral sets are usually obtained; in an unstable patient, organizing collection must not delay necessary treatment. Negativity must be interpreted in relation to previous antibiotics and difficult-to-culture pathogens. If suspicion persists, targeted serology and molecular methods are guided by exposure and clinical context, including testing for Coxiella and Bartonella when indicated. Duke-ISCVID criteria classify infective endocarditis: failure to reach the “definite” category is not equivalent to proving sterile endocarditis.

Complete blood count, platelet count, renal and hepatic function, urinalysis, inflammatory markers, and coagulation testing characterize the context and treatment safety. Fibrinogen and D-dimer help assess activation and consumption but do not diagnose a vegetation. In autoimmune assessment, complement and anti-DNA antibodies can contribute to evaluation of lupus, whereas lupus anticoagulant, anticardiolipin, and anti-beta2-glycoprotein I define the antiphospholipid profile. Persistence at least twelve weeks apart, titer, and interference from anticoagulants must be considered. A prolonged aPTT due to lupus anticoagulant does not automatically imply low thrombotic risk.

The 2023 ACR/EULAR criteria include valvular manifestations in classification of antiphospholipid syndrome, but they are high-specificity research tools and do not replace individual diagnostic judgment. On the oncology side, history, examination, baseline testing, and imaging guided by findings direct the search for malignancy and any histologic confirmation. Reliance on an indiscriminate panel of tumor markers is inappropriate, as is interpreting a vegetation as proof of metastatic disease. If no cause is identified, follow-up should keep open the possibility that it may become recognizable later.

Assessment of organ damage proceeds according to urgency: CT and vascular imaging in the stroke pathway, MRI to define small infarcts, distribution, and hemorrhagic component, and targeted examinations for visceral or peripheral ischemia. A multiterritorial cerebral pattern suggests embolization but does not establish its etiology. Rhythm monitoring and evaluation of alternative sources help avoid excessive attribution to the valve. When tissue is available, appropriate samples for histology and microbiology should be prepared without fixing all material in formalin: final interpretation integrates thrombosis, inflammation, organization, and the search for microorganisms.

Treatment and prognosis

Therapy cannot be uniform because different mechanisms produce similar images. A thrombotic vegetation in active adenocarcinoma, an autoimmune lesion with systemic activity, and a fibrotic leaflet causing regurgitation require different priorities. The plan should include causal treatment, embolic prevention, and hemodynamic control. Evidence directly addressing sterile vegetations is often observational; recommendations for thrombosis, lupus, and valvular disease provide complementary references without automatically becoming specific protocols for every non-infective endocarditis.

In neoplastic NBTE with embolization, unfractionated or low-molecular-weight heparin is frequently used on the basis of available experience. Renal function, platelet count, bleeding, body weight, and procedures guide choice and monitoring; the need for rapid titration may favor unfractionated heparin. Evidence for direct oral anticoagulants in cancer-associated venous thromboembolism does not automatically demonstrate equivalent efficacy against arterial embolism from vegetations. Tumor control is essential to reduce the procoagulant stimulus, and treatment duration should not be decided solely on echocardiographic disappearance of the mass when an active cause persists.

In thrombotic antiphospholipid syndrome, vitamin K antagonists remain the usual reference treatment. Intensity depends on the type of event and the balance between recurrence and bleeding; direct oral anticoagulants are not equivalent in arterial or high-risk profiles. Isolated antibody positivity does not carry the same indication as documented thrombosis. Aspirin may have a role in specific profiles, but does not universally replace anticoagulation. Before defining an event as recurrence during therapy, the nature of the event, adherence, drug exposure, and possible alternative sources should be verified.

In lupus, hydroxychloroquine and other immunomodulators follow the systemic clinical picture. Glucocorticoids may be necessary for organ activity, with the aim of limiting long-term burden; a stable valvular scar is not by itself an indication to intensify immunosuppression. Small series in Libman-Sacks endocarditis suggest that a recent component may improve with combined medical treatment, but do not identify a single universally effective regimen. It is crucial to distinguish active lesion from organized damage, avoiding the expectation of pharmacologic regression of a mechanical deformity that has already become irreversible.

After stroke, the timing of antithrombotic therapy is determined by infarct size, hemorrhage, thrombocytopenia, and residual probability of infection. Applying a fixed interval to the entire group is inappropriate. Surgery is discussed for severe valvular dysfunction and recurrent emboli despite adequate therapy, taking into account the source of the events and systemic prognosis. There is no universal size threshold for sterile vegetations that mandates intervention. Repair feasibility, prosthesis choice, and postoperative management depend on anatomy, the causative disease, and thrombotic and bleeding risks.

Antibiotics do not treat a truly sterile lesion, but the need for empirical therapy during the work-up depends on the probability and severity of an infection that remains possible. Follow-up monitors embolic events, valvular and ventricular function, activity of the underlying cause, and drug toxicity. Echocardiographic comparison is not limited to vegetation size, and the schedule is adapted to evolution. Prognosis may be dominated by advanced malignancy, neurologic injury, or a correctable valvular lesion: the same outcome should not be attributed to all non-infective endocarditis. The ability to control the predisposing condition is one of the principal elements distinguishing the different pathways.

Complications

Cerebral emboli can cause major stroke or repeated injury with progressive loss of independence. The risk of hemorrhagic transformation interferes with prevention of new emboli and requires reassessment rather than a definitive decision based only on the initial diagnosis. Renal, splenic, mesenteric, or limb emboli cause consequences that depend on site, collaterals, and duration of occlusion; mesenteric ischemia and acute limb ischemia are emergencies. In antiphospholipid and neoplastic disease, thrombosis formed directly within the vessel may coexist and should not automatically be described as a fragment originating from the vegetation.

Valvular dysfunction can progress to heart failure through persistent regurgitation, remodeling, and increased pulmonary pressures. In autoimmune forms, fibrosis may maintain the defect after control of systemic disease. Rapidly destructive worsening should instead prompt a search for superimposed infection: initial sterility does not prevent subsequent colonization. Oral hygiene, treatment of foci, and vascular access care are relevant measures, whereas procedural antibiotic prophylaxis follows the categories defined by infective endocarditis guidelines and not membership in this group alone.

Bleeding may result from antithrombotic therapy, thrombocytopenia, renal failure, or consumptive coagulopathy and may coexist with marked thrombogenicity. Immunosuppression adds infectious and toxic risks that must be monitored separately from valvular activity. After surgery, prosthetic thrombosis, emboli, repair deterioration, and recurrence remain possible if the systemic cause persists. Removal of the vegetation does not automatically eliminate the process that generated it: continuity of care, planning of treatment interruptions, and rehabilitation for neurologic injury complete cardiologic management.