Prosthetic valve endocarditis

Article written and reviewed by Dr Luca Moretti

Prosthetic valve endocarditis is an infection involving a replacement valve and, frequently, the tissues in which it is implanted. The target therefore is not limited to the moving elements of the prosthesis: the sewing ring, fixation sutures, annulus and perivalvular structures may be the main site of disease. This distribution explains why a subtle vegetation may coexist with extensive destruction and why disappearance of bacteremia alone does not demonstrate eradication of the focus.

The disease affects both mechanical prostheses and bioprostheses. In the former, the interface between material and tissue and the consequences for anticoagulation are particularly relevant; in the latter, infection may also destroy biological leaflets. Transcatheter valves share some of these problems but have anatomical, diagnostic and operative peculiarities described on the page about endocarditis after TAVI. Defining the substrate helps predict patterns of extension and select investigations; it does not automatically identify the microorganism.

The distinction between early and late disease retains clinical usefulness provided it is not interpreted as an absolute biological separation. For some empirical treatment decisions, European guidelines use a twelve-month threshold after surgery, but definitions used in studies are not uniform. Time since implantation must be integrated with recent hospitalizations, vascular access, prior antibiotics, implantation method and local microbiology. A prosthesis present for years may become infected during hospital-acquired bacteremia, whereas an episode relatively close to surgery may have a demonstrable community origin.

Etiology, pathogenesis and pathophysiology

Implanted material is rapidly coated with host proteins, creating adhesion surfaces for circulating microorganisms. Inoculation may occur during surgery or through later bacteremia, even long afterward. Sutures and poorly vascularized areas favor bacterial persistence. Once infection is established, the focus may extend along the annulus without initially producing a large intracavitary mass: anatomical damage and vegetation size are not necessarily proportional.

Biofilm formation changes the relationship among pathogen, immune defenses and antibiotic therapy. Within the matrix adherent to the device, microorganisms with different metabolic activity coexist, some of which are less vulnerable to drugs active against growth. This tolerance is not the same as resistance demonstrable by susceptibility testing: an isolate classified as susceptible may persist within infected material. The microbiological result must therefore be interpreted together with the ability to achieve effective concentrations, treatment duration and the need to remove the focus.

Coagulase-negative staphylococci are particularly important because of their ability to colonize prosthetic materials; their identification cannot be dismissed as contamination without examining the number of positive samples, concordance of isolates and clinical context. Staphylococcus aureus may cause rapidly destructive infection with persistent bacteremia and dissemination. The probability of methicillin resistance depends on healthcare exposure and local epidemiology; the label prosthetic valve endocarditis alone does not reliably predict it.

Enterococci, particularly Enterococcus faecalis, and streptococci may cause community-acquired or healthcare-associated disease. Their isolation requires a reasoned search for the portal of entry, considering the urinary tract, gastrointestinal tract, oral cavity and recent procedures according to species and clinical history. Slow-growing organisms, including Cutibacterium acnes, may cause minimally febrile presentations and diagnostic delays. Here too, coherence among cultures, the prosthesis and anatomical lesions is more informative than a generic classification as skin flora.

Fungal infection, especially Candida, should be considered in the presence of major healthcare exposures, candidemia, immunosuppression or otherwise unexplained prosthetic infection. It may be associated with bulky vegetations, emboli and relapses after apparent response. Conversely, blood culture-negative disease is not necessarily fungal: previous antibiotic administration remains a common explanation, alongside pathogens requiring serology or molecular methods. The etiologic pathway should follow concrete epidemiological clues and avoid indiscriminate panels that are difficult to interpret.

Extension into the annulus causes necrosis and loss of suture integrity, leading to prosthetic dehiscence. Blood may pass through the defect between prosthesis and tissue, producing acute paravalvular regurgitation; a significantly detached prosthesis may show abnormal movement of the entire sewing ring. Hemodynamic overload depends on the site and magnitude of the leak. A small leak already present after surgery is not equivalent to infection, whereas a newly appearing or rapidly worsening leak requires investigation of its mechanism.

At the aortic position, perianular invasion may form an abscess, a pseudoaneurysm or a fistulous communication with an adjacent chamber. Involvement of conduction tissue may present with a new atrioventricular disturbance; extension to the aortomitral continuity makes surgical reconstruction particularly complex. In bioprostheses, leaflet destruction may add intraprosthetic regurgitation. More rarely, an infectious mass limits movement of prosthetic components and contributes to obstruction, which must be distinguished from thrombosis, pannus and structural degeneration.

Clinical manifestations

Fever remains an important sign, but the picture may be attenuated in older adults, immunosuppressed patients or after antibiotics. Asthenia, weight loss, anemia and functional decline may precede cardiac signs. In the postoperative period, the expected inflammatory response must be distinguished from deviation from the expected course: new fever after initial improvement, bacteremia, unexplained persistence of inflammatory markers or development of prosthetic dysfunction have a different meaning from a single isolated temperature rise.

Heart failure may develop rapidly when a paravalvular leak or destruction of a bioprosthesis causes major regurgitation. Dyspnea, pulmonary edema, hypotension and reduced peripheral perfusion reflect the hemodynamic consequence even when conventional systolic function appears preserved. In acute valvular insufficiency, the chambers have not had time to dilate and adapt: absence of marked remodeling is not reassuring. Assessment must determine how much instability is due to sepsis and how much to the mechanical lesion, which often coexist.

A new murmur, a change in auscultatory findings or a new conduction disturbance increases suspicion but their absence does not exclude disease. New atrioventricular block, especially with aortic infection, requires investigation for perianular extension. Hemolysis may accompany a paravalvular leak through an irregular high-velocity orifice: anemia, increased lactate dehydrogenase and reduced haptoglobin should be interpreted together with the blood smear and other possible mechanisms of anemia.

Systemic emboli may be the first manifestation: neurological deficit, abdominal pain, peripheral ischemia or splenic infarction require investigation for a cardiac focus in a patient with a prosthetic valve. Persistent low-back pain may indicate spondylodiscitis, while renal lesions may result from embolization, hemodynamic injury or immunological phenomena. A focal symptom is not merely a complication to record: it may identify an infected collection that sustains bacteremia and affects the order of necessary interventions.

Investigations and diagnosis

Before antibiotics, when conditions permit, at least three sets of blood cultures are obtained from separate peripheral venipunctures, with attention to volume and asepsis. There is no need to wait for a fever peak. In septic instability, collection should be rapid and must not delay urgent therapy. It is essential to inform the laboratory of the prosthesis and clinical suspicion because apparently contaminating or slow-growing species may have a different significance. Follow-up cultures document clearance and help identify uncontrolled infection.

Transthoracic echocardiography initially assesses ventricular function, gradients, regurgitation and hemodynamic consequences; comparison with the post-implant baseline study is often decisive. The type, position and size of the prosthesis, as well as any previously documented leaks, must be known. A high gradient during fever, anemia or a hyperdynamic state does not by itself prove obstruction. Likewise, small physiological jets of a mechanical prosthesis should not be confused with pathological regurgitation: location, direction and Doppler features require prosthesis-specific interpretation.

Transesophageal echocardiography has a central role because acoustic shadowing from the device limits the transthoracic study. Multiplanar and three-dimensional acquisitions help localize vegetations, dehiscence and paravalvular leaks and describe their relationships with adjacent structures. A negative examination does not exclude an early lesion or one hidden by artifacts; if suspicion remains high, echocardiography should be repeated, generally within five to seven days and earlier if the clinical picture changes. The report should provide an anatomical map useful for decision-making rather than merely stating whether vegetations are present or absent.

Cardiac CT complements echocardiography particularly for perivalvular assessment: it may define abscess cavities, pseudoaneurysms, fistulas and relationships with the aortic root. Quality depends on the protocol and artifacts; not every small mobile vegetation is well represented. CT may also help clarify abnormal movement of prosthetic components and plan reoperation. Distinguishing infected tissue, thrombus and pannus should not, however, be entrusted to a single densitometric value separated from microbiology and clinical evolution.

18F-FDG PET/CT is particularly useful when microbiology and echocardiography leave significant unresolved suspicion. Preparation to suppress physiological myocardial uptake and expert interpretation of signal distribution are required. Focal or heterogeneous uptake around the prosthesis may support infection; postoperative inflammation, surgical materials and adhesives may instead produce false positives. Already effective antibiotic treatment may reduce sensitivity. The value of the test comes from integrating the metabolic pattern with anatomy, not from a universal intensity threshold.

Classification systems do not assign exactly the same weight to time since surgery. In the 2023 Duke-ISCVID criteria, characteristic prosthetic uptake acquires major-criterion status from three months after implantation; at earlier stages it is a minor criterion. The 2023 ESC criteria accept characteristic prosthetic or periprosthetic uptake as a major criterion irrespective of the interval, with appropriate specialist interpretation. This difference should be stated when classifying the case. Radiolabeled leukocyte scintigraphy may provide a selective adjunct where available and indicated.

A clinical diagnosis of definite endocarditis according to Duke-ISCVID requires two major criteria, or one major and three minor criteria, or five minor criteria; combinations for possible disease are less stringent. In prosthetic-valve disease, predisposition, microbiological documentation and imaging findings carry particular weight, but multiple tests describing the same category of lesion do not automatically become independent major criteria. The diagnostic category organizes evidence; it does not replace judgment in the presence of infected dehiscence, an abscess or persistent bacteremia with high clinical probability.

If cultures are negative, previous antibiotics and sampling times are reconstructed precisely; serology and molecular tests are then selected for pathogens such as Coxiella burnetii, Bartonella and Tropheryma whipplei according to the clinical picture. If surgery is performed, the prosthesis and tissues must be correctly distributed among microbiology, histology and any molecular analyses: fixing everything in formalin prevents culture. Positive PCR on tissue may identify the organism after antibiotics, but persistence of DNA does not necessarily mean that viable microorganisms remain.

Extracardiac assessment includes renal function, urinalysis, complete blood count, inflammatory markers and targeted search for metastatic foci. Brain imaging is urgent in the presence of neurological symptoms and may be necessary during preoperative planning depending on context. Suspected infectious aneurysm requires a dedicated vascular pathway. No isolated C-reactive protein value certifies cure; concordance among clinical evolution, bloodstream sterilization, anatomical stability and control of secondary infected sites is more informative.

Treatment and prognosis

Management requires an Endocarditis Team with infectious-disease, cardiology, cardiac-surgery, microbiology and imaging expertise. Surgical assessment should begin early even when medical treatment is started: postponing it until antibiotic failure may allow anatomical progression that makes surgery more complex. The choice depends on hemodynamic stability, local extent, organism, blood-culture clearance, emboli and realistic reconstructive options. Age and comorbidities contribute to risk but do not alone describe either reversibility or potential benefit.

Empirical therapy should cover likely pathogens without losing connection with the context. In early or healthcare-associated disease, resistant staphylococci, enterococci and Gram-negative bacilli must be considered according to local epidemiology; in late community-acquired disease, the expected spectrum may differ. Allergies, renal function, known colonization and previous isolates modify the choice. As soon as identification and susceptibility are available, the regimen is narrowed and optimized; presence of a prosthesis does not justify maintaining broad empirical coverage indefinitely.

For methicillin-susceptible staphylococci, the backbone is an antistaphylococcal beta-lactam, or cefazolin when appropriate; for resistant isolates, vancomycin with exposure monitoring or specialist-selected high-dose daptomycin strategies are used. Treatment duration for prosthetic-valve disease is generally at least six weeks. The choice depends not only on the susceptibility report: tissue distribution, toxicity, persistent bacteremia and surgical control of infected material determine whether a microbiologically active regimen is also clinically adequate.

Traditional recommended regimens for staphylococcal prosthetic-valve endocarditis include rifampin for at least six weeks and gentamicin during the first two weeks, in addition to the primary drug. Clinical evidence supporting benefit from these combinations is limited, however, while nephrotoxicity, hepatotoxicity and interactions are important. Observational studies of rifampin have not shown a uniform outcome benefit. It is therefore necessary to distinguish a regimen recommendation from certainty of effect in an individual patient and discuss indication and risk within the team.

When used for staphylococcal infection involving foreign material, rifampin is introduced after several days of effective therapy, generally three to five, and after clearance of bacteremia. It must not be used alone: high inoculum and an inadequate combination favor selection of resistance. Monitoring includes liver function and systematic review of concomitant medications. The indication does not automatically extend to every prosthetic-valve endocarditis regardless of the isolated species and does not replace removal of a detached prosthesis or an uncontrolled perianular focus.

For susceptible Enterococcus faecalis, ampicillin combined with ceftriaxone is an important strategy, including in the presence of high-level aminoglycoside resistance. In adults with adequate organ function, ampicillin 12 g/day in divided doses and ceftriaxone 2 g every twelve hours are commonly used for six weeks in prosthetic-valve disease. The combination exploits complementary activity at bacterial targets and should not be automatically transferred to Enterococcus faecium. Regimens require individual adjustment and susceptibility confirmation; not every combination of two beta-lactams should be considered equivalent.

In infections caused by susceptible streptococci, penicillin, amoxicillin or ceftriaxone are selected according to species, susceptibility and tolerability; the usual adult ceftriaxone dose is 2 g every twenty-four hours. A six-week duration is used in prosthetic-valve disease: abbreviated regimens intended for selected native-valve cases are not transferable. For HACEK organisms, ceftriaxone is likewise a reference option with prolonged treatment. In fungal disease, liposomal amphotericin B with optional flucytosine or high-dose echinocandins are part of a specialist strategy that considers surgery and possible prolonged antifungal suppression.

Duration is counted from the start of therapy that is actually active, taking into account clearance of blood cultures when initially positive. Surgery performed during treatment does not automatically require restarting from zero; a positive valve culture may instead require a new complete course. Histology and molecular tests must be interpreted together with culture because they do not measure the same phenomenon. A formally completed calendar is insufficient if an undrained abscess, bacteremia or extracardiac source continues to sustain infection.

Anticoagulation requires an assessment separate from the antibiotic indication. Endocarditis itself is not a reason to start anticoagulants to prevent septic emboli. A mechanical prosthesis, however, carries a pre-existing indication that must be balanced against cerebral hemorrhage, procedures and imminent surgery; in some settings, temporary switching from a vitamin K antagonist to heparin is considered. Direct oral anticoagulants do not replace vitamin K antagonists for mechanical prosthetic valves. Rifampin may substantially reduce the effect of warfarin; discontinuation of rifampin may raise the INR and therefore requires close monitoring and dose adjustment.

The main surgical indications include heart failure due to prosthetic dysfunction, uncontrolled infection and substantial embolic risk. Refractory pulmonary edema or shock may require emergency surgery, generally within twenty-four hours. Abscess, fistula, dehiscence, progressive local extension or persistent bacteremia not explained by other foci point toward an urgent strategy, often within three to five days. Timing is not a mandatory waiting period: it should be shortened when hemodynamic or anatomical progression requires it.

Surgery aims to remove prosthesis, sutures and infected tissue, drain collections and reconstruct a stable foundation for valve function. Destruction of the annulus or root may require complex reconstruction, sometimes extending to the aortomitral continuity. Choice of substitute depends on anatomy, bleeding risk, reoperation prospects and center experience; no material compensates for inadequate debridement. Previous bypass grafts, adhesions and relationships with the sternum make planning of re-entry important and contribute to operative risk.

A neurological complication changes the decision but does not always require postponement. After ischemic stroke without hemorrhage, an urgent cardiac indication may remain dominant; intracranial hemorrhage instead requires a more detailed assessment of site, volume, stability and need for surgery. The risk associated with cardiopulmonary bypass and anticoagulation must be compared with the risk of waiting in the presence of destructive infection. The decision requires updated imaging and neurological and neurosurgical discussion when relevant, not automatic application of the same interval to every patient.

An exclusively medical strategy may be appropriate in selected stable cases without perivalvular invasion or significant dysfunction, with a controllable organism and documented response. It requires close surveillance and explicit criteria for reconsidering surgery. When an indicated operation is not feasible, suppressive therapy may be considered individually, but it does not carry the same meaning as demonstrated eradicative cure. In observational studies, comparisons between operated and non-operated patients are affected by patient selection and the time required to reach surgery; crude mortality differences alone do not measure the effect of surgery.

Transition to oral therapy or outpatient parenteral treatment concerns only stabilized patients with a defined organism and susceptibility, no uncontrolled foci, appropriate regimens and reliable follow-up. The POET trial included patients with prosthetic valves, but its findings do not justify early discharge of complicated or microbiologically uncertain disease. Follow-up assesses recurrence, function of the new or existing prosthesis, renal recovery, adverse effects and anticoagulation management. Oral hygiene, prevention of vascular-access infections and prophylaxis for at-risk dental procedures in patients with an indication form part of subsequent protection.

Complications

Prosthetic dehiscence may cause severe regurgitation, hemolysis and instability up to shock. Severity is not judged only by the visible extent of detachment: regurgitant volume, chamber pressures, ventricular reserve and speed of onset all matter. A paravalvular leak associated with active infection requires a different pathway from a chronic sterile leak; percutaneous closure is not an ordinary solution to infected material and may leave the destructive mechanism intact.

Perianular extension may progress to fistulas, pseudoaneurysms and lesions of the aortomitral continuity. New conduction block or hemodynamic deterioration during antibiotics requires anatomical reassessment even if fever has fallen. Bloodstream sterilization may precede control of the local lesion because drugs reduce the circulating component without restoring destroyed tissue. This dissociation explains why monitoring must integrate microbiology, ECG and imaging rather than treating one dimension as proof of cure.

Neurological complications include ischemic emboli, hemorrhagic transformation, hemorrhage from infectious aneurysm and, less commonly, abscesses. In a patient with a mechanical prosthesis, the conflict between anticoagulation requirement and hemorrhagic risk is added. Intravenous thrombolysis is not recommended for stroke associated with endocarditis; thrombectomy may be considered in appropriate cases by a neurovascular team. The decision must also account for infection of embolic material and possible concomitant vascular lesions.

Acute kidney injury may result from hypoperfusion, sepsis, embolization, glomerulonephritis or drug toxicity. Automatically attributing it to antibiotics risks overlooking persistent infection; automatically attributing it to sepsis may delay discontinuation of a nephrotoxic drug. Creatinine trend, urinary sediment, medication exposure and hemodynamic status guide interpretation. Renal impairment changes doses and operative risk, making selection of combinations with plausible benefit and proportionate monitoring even more important.

Recurrence may represent persistence of the same focus, whereas a new infection may result from another episode of bacteremia. Identity of the organism, time interval, microbiological comparison when available and residual anatomy help distinguish relapse from reinfection without relying on a single chronological cutoff. New fever after treatment requires blood cultures before antibiotics when possible and reassessment of the prosthesis. Follow-up should therefore provide the patient with practical instructions about symptoms to report and ensure rapid access to specialist evaluation.