Transcatheter valve endocarditis

Article written and reviewed by Dr Luca Moretti

Transcatheter valve endocarditis is an infection of a prosthesis implanted percutaneously or of contiguous cardiac tissues. Most evidence concerns transcatheter aortic valve replacement, referred to as TAVI or TAVR; other sites do not necessarily share the same epidemiology. It is a form of prosthetic valve endocarditis, but the metal frame, displaced native leaflets, prosthetic covering, and relationships with the aortic root introduce specific difficulties. Infection can be clinically important even when echocardiography shows no vegetation on the leaflets and the transvalvular gradient has remained close to baseline.

The event is relatively uncommon, but carries substantial morbidity. The 2016 Regueiro registry reported approximately 1.1 cases per 100 person-years: this is an incidence rate, not a cumulative risk applicable to any duration of observation. Estimates vary according to diagnostic criteria, prosthesis generation, population, and follow-up methods. Comparison with surgical valve replacement is influenced by patient selection and the competing risk of death; data from randomized studies and registries, such as those analyzed by Lanz, must be interpreted in their proper context. No historical percentage automatically describes the risk of an individual patient treated today.

Etiology, pathogenesis, and pathophysiology

Unlike conventional surgical valve replacement, in TAVI the calcified native leaflets are displaced and remain in contact with the device. The prosthetic frame expands within the annulus and, depending on the platform, extends into the root and outflow tract. Irregular apposition, calcifications, and textile covering create a complex interface among material, blood, and tissue. Bacteremia can colonize biological leaflets, the frame, or adjacent structures; the presence of multiple surfaces explains why infection cannot be reduced to the image of a mobile valvular vegetation.

Contamination may be periprocedural or result from subsequent bacteremia. Vascular access, catheters, urinary procedures, hospitalizations, and soft-tissue infections contribute to varying degrees; the interval from implantation alone does not identify the portal of entry. The early-versus-late distinction guides empirical therapy, but it must be integrated with recent healthcare exposure and local microbiology. Diabetes, renal failure, previous endocarditis, and conditions predisposing to bacteremia affect risk. However, an observational association with a procedural characteristic does not prove that the characteristic caused infection in the individual case.

Enterococcus faecalis is particularly prominent in TAVI series, alongside S. aureus, coagulase-negative staphylococci, and streptococci. The frequency of enterococci makes assessment of the urinary tract, gastrointestinal tract, and healthcare exposures important, without automatically assuming a urinary origin. Staphylococci can colonize skin, access sites, and intracardiac materials; streptococci require species-specific interpretation. Candida and other less common pathogens become relevant when candidemia, immunosuppression, prolonged treatment, or lack of response is present. The definitive regimen must be based on microbiologic identification, not solely on membership in the TAVI category.

Biofilm on the prosthesis may maintain microorganisms that are poorly susceptible to therapy even in the absence of conventional resistance. Adhesion to biomaterials and reduced metabolic activity of some populations contribute to persistence and relapse. Clearance of bacteremia is an important sign of response, but does not prove sterilization of every surface. Infection confined to the leaflets may cause destruction and regurgitation; periprosthetic extension may instead involve the annulus, root, and mitral-aortic continuity while initially preserving relatively normal valve function.

The geometry of the prosthesis influences lesion distribution and how it should be sought. A frame extending higher into the root may provide surfaces above the valvular plane, while the ventricular portion and sealing skirt may conceal a site near the outflow tract. Vegetations or infection may also involve the mitral valve or a concomitant cardiac implantable electronic device. An examination focused only on leaflet opening therefore risks being incomplete. Different models should not be ranked in a universal hierarchy of infectious risk on the basis of this anatomic plausibility alone: clinical differences require comparative data.

Intraprosthetic regurgitation results from leaflet alteration, whereas a paravalvular leak may reflect destruction or modification of the landing-zone interface. Many patients already have residual leakage after implantation: attribution to infection requires temporal comparison and analysis of the mechanism. TAVI does not have the same sutured sewing ring as a surgical prosthesis, so the description of dysfunction must respect its anatomy. New severe regurgitation increases left ventricular volume load and may cause pulmonary edema, especially when a hypertrophied, poorly compliant ventricle tolerates an acute change poorly.

Invasion of the root can form abscesses, pseudoaneurysms, and fistulas and may involve the conduction system. Block occurring immediately after TAVI may result from mechanical interaction with the device; a new conduction disturbance during bacteremia or after a period of stability instead requires investigation for infectious extension. Left-sided vegetations send emboli into the systemic circulation, potentially affecting the brain, spleen, kidneys, and coronary arteries. Mass size is related to embolic risk, but the causative organism, mobility, and previous events also contribute: failure to visualize a large mass does not eliminate risk.

Clinical manifestations

Presentation may be subacute, with intermittent fever, fatigue, anorexia, and functional decline, or may be dominated by sepsis and heart failure. In older and frail patients, fever may be mild or absent; confusion, falls, worsening renal function, and loss of independence may precede cardiac suspicion. Enterococcal or staphylococcal bacteremia should not be attributed to an extracardiac source without asking whether the prosthesis is involved. This is particularly important when the initial source is treated but cultures remain positive or become positive again after treatment is stopped.

Dyspnea may result from new aortic regurgitation, ventricular dysfunction, prosthetic obstruction, sepsis, or fluid overload in renal failure. Distinction requires clinical and echocardiographic assessment, because response to diuretics does not exclude a mechanical lesion requiring correction. A murmur may be difficult to interpret when a known residual leak is present. An increase in gradients during fever and anemia may reflect increased flow rather than obstruction: a change in a Doppler value must be related to the hemodynamic picture and leaflet morphology.

An embolic event may precede diagnosis: stroke, abdominal pain, limb ischemia, or low-back pain requires consideration of the prosthesis as a potential source. Vertebral involvement can maintain symptoms even after bloodstream clearance. Conduction disturbances, a new paravalvular leak, or hemolysis instead suggest a local complication. In patients with a pacemaker implanted after TAVI, the electronic system must also be assessed: two intracardiac foreign materials can be involved in the same episode and require a coordinated plan.

Investigations and diagnosis

Blood cultures should precede antibiotics whenever possible: at least three adequate sets are collected, without waiting for fever peaks and without delaying treatment of instability. Species and susceptibility are identified and clearance is documented with subsequent cultures. Microorganisms compatible with prosthetic infection should not be dismissed as contaminants without examining reproducibility and context. When cultures are negative, previous antibiotics, epidemiologic features, and tests for pathogens not routinely recovered guide the pathway described for blood culture-negative endocarditis.

It is essential to retrieve the prosthesis model, size, and implantation date, procedural report, any overlapping implants, and the baseline echocardiogram. Comparison makes it possible to distinguish pre-existing leaks, prosthesis-patient mismatch, and new abnormalities. Transthoracic echocardiography assesses ventricular function, regurgitation, and gradients; transesophageal echocardiography searches for vegetations and involvement of adjacent structures. The frame and calcifications may create shadowing and reverberations, while the lesion may lie in a poorly accessible area. Absence of vegetations on the first echocardiogram is therefore not sufficient to exclude an infection supported by microbiologic evidence.

Cardiac CT is useful for defining abscesses, pseudoaneurysms, fistulas, and the relationships of the device with the annulus, root, and coronary arteries. ECG-gated acquisitions can describe leaflet motion and thickening, but distinguishing thrombosis from infection requires clinical integration. Hypoattenuated leaflet thickening may indicate subclinical thrombosis and is not equivalent to a vegetation; conversely, identifying a plausible thrombotic mechanism does not justify ignoring persistent bacteremia. Renal function, image quality, and expected benefit guide the use of contrast medium, particularly in unstable patients.

The diagnosis of prosthetic dysfunction integrates velocities and gradients with Doppler velocity index, effective orifice area, ejection duration and contour, and comparison with baseline flow. A high gradient already present immediately after implantation suggests a problem different from a sudden increase during fever; mismatch, thrombosis, degeneration, and infection may nevertheless coexist. In paravalvular regurgitation, multiple eccentric jets make color extent in a single view insufficient. Assessment requires multiple planes, ventricular consequences, and, when necessary, other modalities. A hemodynamically stable prosthesis may remain infected, because function and microbiologic activity are related but nonidentical dimensions.

18F-FDG PET/CT can support the diagnosis when echocardiography and cultures do not resolve the case and can identify distant foci. Metabolic preparation, interpretation of artifacts, and uptake distribution are critical. Post-implant inflammation may produce a noninfectious signal; antibiotics and small lesions may instead reduce sensitivity. The 2023 ESC criteria accept characteristic prosthetic uptake as a major criterion regardless of the interval from implantation, with expert interpretation; Duke-ISCVID gives weight to a three-month interval. The system being applied must be stated, without combining only the elements most favorable to diagnosis.

Suspected leaflet thrombosis requires particular caution when inflammation and bacteremia coexist. CT evidence of reduced motion or thickening can support a thrombotic component, but anticoagulation does not sterilize a vegetation and may complicate a cerebral lesion. Before intensifying antithrombotic therapy, the two hypotheses should be clarified as far as possible and contraindications sought. The hypothesis of biological degeneration must also respect timing and morphology: a perforation that develops rapidly during sepsis has a different meaning from progressive calcification over years. Diagnosis derives from the trajectory of the case, not solely from the type of device.

Echocardiographic, tomographic, and metabolic evidence complement one another, but do not automatically constitute independent major criteria every time they describe the same lesion. Diagnostic criteria are classification tools, not reasons to delay treatment of sepsis with strong suspicion of prosthetic infection. In uncertain cases, anatomy, cultures, and alternative sources are reassessed serially. Pre-existing regurgitation or structural degeneration must not be converted into infectious findings merely to reach a threshold, just as failure to meet a threshold initially does not prove absence of disease.

Assessment of complications includes ECG, renal and hepatic function, complete blood count, urinalysis, and extracardiac imaging guided by symptoms and surgical planning. In the presence of neurologic deficits, brain imaging distinguishes ischemia, hemorrhage, and infectious vascular lesions. If explantation is performed, allocation of prosthesis and tissue between culture, histology, and molecular diagnostics should be agreed before surgery; all material must not be fixed in formalin. A positive PCR can identify the causative organism after antibiotics, but by itself does not measure residual viability.

Treatment and prognosis

Management requires an Endocarditis Team with structural and cardiac surgical expertise in transcatheter valve explantation. Assessment cannot stop at the statement that the patient was “already considered high risk”: the initial choice of TAVI concerned a different clinical situation and must be reconsidered in light of infection. The benefit of surgical source control, explant risk, possibility of recovery, and the patient's goals must be estimated. In frail patients, previous functional status, cognition, nutrition, and reversible organ injury help distinguish high risk from a truly low probability of benefit.

Empirical therapy takes into account time from implantation, healthcare acquisition, local epidemiology, and previous antibiotics. It must cover plausible pathogens, including enterococci and staphylococci, and should be broadened when the context suggests resistance or Gram-negative organisms. A regimen designed for community-acquired native valve infection is not automatically adequate after a recent hospitalization. Renal function affects dosing and monitoring and may change rapidly during sepsis and diuresis. After identification, therapy is targeted; the presence of the device does not justify maintaining broad combinations without a microbiologic rationale.

For susceptible Enterococcus faecalis, ampicillin plus ceftriaxone is an important strategy, including in the presence of high-level aminoglycoside resistance. A usual adult regimen, to be adapted to the patient's condition, combines ampicillin 12 g/day in divided doses and ceftriaxone 2 g every twelve hours for six weeks. Comparative experience is predominantly observational and suggests a favorable renal profile compared with aminoglycoside-containing combinations. The combination should not be extrapolated to E. faecium or isolates with different susceptibility: species and susceptibility testing remain decisive.

In staphylococcal infections, the principal agent is selected according to methicillin susceptibility, favoring a beta-lactam when appropriate and using vancomycin or daptomycin-based strategies when indicated. For staphylococcal prosthetic valve endocarditis, regimens including rifampin and gentamicin should be discussed in light of the limited certainty of clinical benefit and the risk of toxicity. If rifampin is chosen, it is introduced after several days of effective therapy and clearance of bacteremia, not as monotherapy. Interactions with anticoagulants and other medications are particularly relevant in older patients receiving multiple drugs.

Duration is generally at least six weeks, with extensions for particular organisms or foci. Streptococcal, HACEK, or fungal infections require their own regimens; an infected TAVI is not treated with the short regimen used for uncomplicated native valve disease. Bloodstream sterilization and the clinical course guide counting and response, while surgery does not automatically require restarting a complete antibacterial course if valve culture is negative. Candida poses problems of prolonged antifungal therapy and suppression, discussed on the page on fungal endocarditis.

Surgery is indicated according to the principles of prosthetic valve endocarditis: heart failure due to valvular dysfunction, uncontrolled infection, and prevention of embolism in appropriate circumstances. Shock or refractory pulmonary edema may require emergency surgery; abscesses, fistulas, persistent bacteremia, or local progression require an urgent decision. Stabilization with antibiotics and diuretics may be necessary, but does not repair irreversible anatomic damage. The absence of an easily visible vegetation does not eliminate an indication based on periprosthetic destruction and instability.

TAVI explantation may require more than simple valve replacement. The frame may be incorporated into the aortic wall; removal can damage the annulus, root, or adjacent structures and necessitate extensive reconstruction. Prosthesis type, frame height, time since implantation, any valve-in-valve procedure, and infectious involvement determine complexity. Preoperative CT, when feasible, helps plan the relationships with the coronary arteries and aorta. Debridement must remove infected tissue and material, while the choice of replacement depends on reconstructable anatomy and individual conditions.

The EXPLANT-TAVR registry describes patients who actually underwent explantation and documents the complexity of infectious cases, but does not include a randomized comparison with all nonoperated patients. Its results therefore cannot be used either to promise a uniform benefit or to conclude that surgery is futile when operative risk appears high. Patients who reach surgery are selected by initial survival, anatomy, and center decisions. The prognosis without source control must be weighed in the same balance as procedural mortality.

A new valve-in-valve procedure during active infection may correct part of the hemodynamic dysfunction but leaves the infectious focus in place and introduces additional foreign material. It is not a routine eradication strategy and does not replace explantation and debridement; any rescue use belongs to exceptional, individually discussed cases. A previously documented and cured infection presents a different problem, to be assessed according to residual dysfunction. Percutaneous closure of an infected paravalvular leak can likewise mask the defect without resolving the underlying destruction.

Coexistence of a pacemaker requires integrated management of two potential sources. A normal pocket and absence of lead vegetations do not exclude colonization during persistent bacteremia; conversely, a sterile mass does not necessarily make the system the primary source. Species, duration of positivity, and imaging guide discussion of extraction, especially if valve surgery is planned. A new conduction disturbance may require temporary pacing while avoiding implantation of new material during uncontrolled infection. Cardiac surgery and electrophysiology must therefore agree on sequence and electrical protection.

A neurologic complication changes the timing balance. Ischemic stroke without hemorrhage does not always require postponement of urgent surgery, whereas intracranial hemorrhage or infectious aneurysm requires neurovascular assessment and updated imaging. The risk of cardiopulmonary bypass must be compared with sepsis, heart failure, and embolism during the waiting period. Antithrombotic therapy is not intensified to prevent septic emboli: atrial fibrillation, a recent coronary procedure, and other pre-existing indications must be reassessed against bleeding, renal function, and planned interventions.

When conservative management is chosen, it requires an explicit rationale and a verifiable plan: a treatable pathogen, microbiologic response, stable prosthetic function, and absence of a complication that makes drug therapy insufficient. If indicated surgery is not feasible, goals of care are defined and, in selected situations, suppressive therapy is considered, with its limitations made clear. Follow-up does not end with completion of infusions: it monitors relapse, toxicity, nutrition, functional recovery, and the need for new decisions. Stability achieved in hospital must be sustainable in the actual care setting.

Geriatric assessment distinguishes pre-existing conditions from the potentially reversible effects of acute illness. Sarcopenia, malnutrition, cognitive impairment, and dependence in activities of daily living influence recovery and the ability to sustain prolonged therapy; however, delirium and immobility developing during sepsis do not automatically describe baseline status. Comparison with relatives and prior documentation helps estimate realistic goals. If treatment remains conservative, home care, medication management, and access to urgent assessment must be organized before discharge. A formally appropriate decision that cannot be implemented outside the hospital exposes the patient to interruptions, toxicity, and delayed recognition of relapse.

Out-of-hospital continuation, parenteral or oral in selected cases, requires source control, known susceptibility, a validated regimen, and access to close monitoring. The presence of a prosthesis in the POET trial does not amount to evidence dedicated to every infected TAVI, especially when complicated or in very frail patients. At the end of the active phase, a new echocardiographic baseline is documented and clinical surveillance is planned. New fever requires cultures before antibiotics whenever possible, because repeated empirical treatment can make relapse difficult to identify.

Prevention includes oral and skin care, vascular access management, prompt identification of bacteremia, and dental prophylaxis for indicated procedures, because a transcatheter valve is among the high-risk substrates. Before elective implantation, active infections should be resolved and procedural prophylaxis should follow the relevant protocols. No prophylaxis eliminates the possibility of subsequent bacteremia. Patients should know that persistent fever, chills, new dyspnea, or unexplained deterioration warrants early assessment, even years after the procedure.

Complications

Prosthetic and periprosthetic destruction can cause acute regurgitation, abscess, fistula, and shock. The frame may make lesions less immediately demonstrable and delay recognition of progression; new conduction abnormalities or hemodynamic worsening therefore require repeat multimodality assessment. An unchanged gradient does not exclude infection extending beyond the leaflet plane. Residual damage may sustain heart failure and hemolysis even after microbiologic control, making a subsequent structural or surgical strategy necessary.

Systemic emboli can cause permanent neurologic deficits, visceral or peripheral ischemia, and metastatic foci. Their presence affects surgical risk and recovery, but the extent of injury must be defined without equating every silent lesion with a devastating stroke. A splenic or vertebral abscess can maintain bacteremia and reinfect a reconstruction; the sequence of source control depends on cardiac urgency. Renal failure may simultaneously result from sepsis, congestion, emboli, and drugs and requires a multifactorial interpretation.

Renal injury affects several steps at once: antimicrobial clearance, contrast tolerance, fluid balance, and operative risk. Creatinine may rise because of hypoperfusion, venous congestion, glomerulonephritis, emboli, or toxicity and must be interpreted in relation to urinary sediment and hemodynamic trends. Reducing dosage without verifying exposure may make treatment inadequate, whereas maintaining unadjusted doses may worsen injury. Monitoring must accompany every relevant change in renal function and support. Even improvement after diuresis or correction of sepsis does not eliminate the need to monitor the cumulative effects of combination therapy.

Relapse is favored by persistent infected material, a pathogen that is difficult to eradicate, and uncontrolled foci; new bacteremia may instead cause reinfection after successful treatment. Microbiologic comparison and anatomic evolution help distinguish the two scenarios. In older adults, deconditioning, loss of independence, and treatment toxicity affect outcome as much as some cardiac parameters. Survival to hospital discharge therefore does not describe the overall result: assessment should include function, quality of life, and sustainability of the subsequent care pathway.