Endocarditis and cardiac device infections

Article written and reviewed by Dr Luca Moretti

Cardiac implantable electronic device infections encompass different clinical entities: infection confined to the generator pocket, lead infection, associated valvular endocarditis, and systemic infection without demonstrable vegetations. Pacemakers, defibrillators and cardiac resynchronization systems all contain material capable of sustaining biofilm, but they differ in the number and course of their leads and in the consequences of removal. Distinguishing the compartments involved helps define investigations, duration of therapy and the reimplantation strategy; it does not mean that an apparently isolated pocket infection can be cured while leaving the remainder of an infected system in place.

Superficial wound infection, by contrast, is confined to the skin and subcutaneous tissue, without communication with the pocket. This distinction can prevent unnecessary extraction, but it requires careful examination and close reassessment: deep dehiscence, a sinus tract, progressive skin adherence or exposure of hardware indicates a different problem. An exposed generator or lead must be considered infected even without pus or positive cultures. Diagnosis cannot depend solely on the intensity of erythema, because indolent infections may present as late erosion with almost no systemic symptoms.

Etiology, pathogenesis and pathophysiology

Inoculation may occur during implantation or revision, through a later breach in tissue integrity, or during bacteremia from another site. Perioperative contamination may remain clinically silent and present later, whereas hematogenous spread may first involve the endovascular component. The pocket, connectors and leads form a continuum along which infection can propagate. Consequently, a local manifestation does not reliably define the biological boundary of the focus, and replacing only the generator may leave colonized surfaces that fuel recurrence.

Coagulase-negative staphylococci and Staphylococcus aureus are the main causative organisms. The former readily exploit adhesion to biomaterials and may cause indolent infections; the latter is more often associated with bacteremia, invasion and dissemination. Cutibacterium, enterococci, Gram-negative organisms and fungi are additional possibilities, whose relevance depends on exposures and population. A single positive bottle for skin flora requires verification, whereas repeated concordant isolates in a patient with intracardiac material should not be dismissed. The risk of system involvement during bacteremia varies by species: it is not identical for every Gram-negative organism.

Within biofilm, microorganisms adhere to the material and organize into a matrix that alters their metabolism and susceptibility to antimicrobials. Reduced metabolic activity and persistent subpopulations permit regrowth after treatment is stopped, even without resistance detected in vitro. Therapy may clear the circulating component and reduce pocket signs without sterilizing the entire system. This mechanism explains the central role of complete removal in definite infection and why local irrigation, drainage or repeated courses of antibiotics are not equivalent to source control.

Pocket hematoma increases infection risk through tissue tension, impaired healing and a greater likelihood of reintervention. Advanced renal failure, diabetes, immunosuppression, previous infections, implant complexity and repeated procedures contribute further. Risk does not depend only on the number of factors: some are modifiable, such as anticoagulation management, hemostasis and the need for early revision. The BRUISE CONTROL INFECTION study linked clinically significant hematoma to subsequent infection, supporting the importance of procedural prevention rather than indiscriminate addition of antibiotics.

The tissue response to a lead produces adhesions and fibrous sheaths over time, particularly at points of contact with the venous wall, valvular structures and endocardium. These adhesions are not themselves infection, but they explain why removal and source control become more complex with longer implant duration. In systems with multiple leads, the surfaces may adhere to each other and hinder independent mobilization. Colonization may involve segments not visible on echocardiography: the clinical significance of biofilm is not limited to an intracardiac vegetation. Abandoned leads must also be included in the anatomical inventory and extraction plan.

Vegetations may adhere to leads in the vena cava, right atrium or right ventricle and involve the tricuspid valve. Fragmentation causes septic pulmonary emboli; bacteremia may also disseminate to the spine, joints and left-sided valves. Leads can mechanically impair tricuspid coaptation even without infection, so regurgitation alone does not identify endocarditis. Assessment must distinguish pre-existing damage, infectious destruction and possible extraction-related injury, because prognosis and the need for repair depend on the actual mechanism.

Clinical manifestations

Pocket infection may present with pain, warmth, erythema, swelling, drainage, fluctuance or a sinus tract. In indolent forms, skin adherence, contour deformity or impending erosion may predominate. Exposure of hardware is a decisive finding even in an afebrile patient. Dermatitis from adhesives or antiseptics, suture reaction and hematoma may mimic some signs, but an alternative diagnosis must explain the depth and evolution of the lesion. A pocket should not be aspirated through the skin in an attempt to demonstrate sterility, because the procedure may inoculate microorganisms and does not exclude biofilm.

Endovascular infection may present with persistent or intermittent fever, chills, asthenia and bacteremia while the pocket appears normal. Cough, pleuritic pain, hemoptysis and recurrent pulmonary infiltrates suggest right-sided embolization; spondylodiscitis or arthritis may draw attention before the device does. In frail patients, confusion, worsening renal function or functional decline may replace fever. The presence of a plausible extracardiac source does not exclude secondary colonization of the system: the duration of bacteremia, the organism and the response after control of the source determine how extensively the suspicion should be investigated.

Sepsis with instability requires immediate management, but extraction planning should not wait for failure of multiple antimicrobial regimens. Signs of heart failure, new regurgitation or conduction disturbances prompt assessment for valvular involvement and perivalvular invasion. Deterioration after initial bloodstream sterilization may indicate an unremoved focus, an extracardiac complication or a new vascular-access problem. The temporal evolution is therefore essential: pocket findings, cultures, imaging and hemodynamic status must be interpreted as parts of the same process.

Investigations and diagnosis

At least three sets of blood cultures are obtained before antibiotics when conditions allow, without delaying treatment of sepsis. Persistent or recurrent positivity for S. aureus and other compatible pathogens raises suspicion even when initial echocardiography is negative. The number of positive samples and concordance of isolates help interpret coagulase-negative staphylococci; it is incorrect either to ignore all of them or to treat every contaminant as proof of system infection. Microbiology consultation can guide prolonged incubation or additional methods for slow-growing organisms.

At extraction, pocket tissue and removed components are submitted for appropriate microbiological testing, avoiding exclusive reliance on superficial swabs from a colonized sinus tract. Sonication, when available within a validated pathway, may increase recovery of adherent microorganisms, but a positive result must be interpreted in light of contamination risk during removal. A lead pulled through an infected pocket may yield a positive culture without by itself proving valvular endocarditis. Previous antibiotics may reduce sensitivity; a negative result does not negate unequivocal clinical findings such as exposure or deep drainage.

Transthoracic echocardiography defines cardiac function and valvular disease, whereas transesophageal echocardiography better assesses leads, the right atrium, tricuspid valve and left-sided valves. The examination searches for vegetations, valvular destruction and complications that alter treatment duration or the extraction approach. A lead-associated mass is not automatically infected: fibrin sheaths, thrombi and other adhesions are common and may be indistinguishable on morphology alone. A negative study does not exclude infection, because part of the device is extracardiac or poorly visualized; repetition and integration with other tests depend on clinical probability.

18F-FDG PET/CT may demonstrate activity along the pocket and lead course, clarify equivocal cases and identify metastatic foci. Sensitivity is not uniform across all compartments: small lead vegetations may be missed, especially after treatment. Postoperative inflammation and inadequate metabolic preparation limit specificity; uptake must be interpreted according to location, distribution and time since implantation. Radiolabeled leukocyte scintigraphy may provide complementary information in experienced centers. Neither test should be used to deny a clinically evident infection merely because metabolic signal is absent.

If suspicion of endovascular infection remains high despite an initial negative transesophageal study, repeating the examination within a few days, generally five to seven, is reasonable, sooner if the clinical picture changes. Small initial dimensions and lesion location may limit recognition. In selected cases, intracardiac echocardiography adds information about the leads and may accompany the procedure, but it requires invasive access and does not resolve every etiologic uncertainty. A positive finding must always be linked to cultures and local findings: increasing imaging sensitivity can also reveal more sterile masses, creating a risk of overdiagnosis if this integration is lacking.

CT helps identify pulmonary emboli and collections, perivalvular complications and anatomical relationships relevant to extraction. Lead age, number and type, calcifications, venous occlusions and previous procedures influence procedural complexity. Device interrogation clarifies pacing dependence, treated arrhythmias and whether defibrillation or resynchronization remains necessary. These are not merely technical details: they determine what temporary protection is required when the system is removed and what device, if any, should subsequently be reimplanted.

Criteria specific to device infections and the Duke-ISCVID criteria for endocarditis address partially different questions. A definitely infected pocket may require extraction without meeting the clinical definition of endocarditis; conversely, valvular endocarditis in a patient with a device requires assessment of the risk that the device is also colonized. The diagnosis should explicitly describe the pocket, bacteremia, leads and valves, indicating which sites are documented and which are suspected. This prevents the generic label “pacemaker infection” from obscuring differences that are decisive for treatment and follow-up.

Treatment and prognosis

In definite system infection, standard treatment combines antibiotics with complete removal of the generator and leads, including infected abandoned components when technically feasible. Pocket revision alone leaves potentially colonized material behind; repeated courses of antibiotics may suppress symptoms without eradicating the source. Extraction should be arranged promptly in an experienced center with surgical management of complications available. The need for planning does not justify unnecessary delay, particularly in sepsis or persistent bacteremia.

Empirical therapy, after cultures are obtained, generally covers methicillin-resistant staphylococci with vancomycin or an appropriate alternative such as daptomycin; Gram-negative coverage depends on the systemic picture and context. Targeted therapy may favor a beta-lactam for MSSA and requires specific regimens for enterococci or fungi. Rifampin is not an automatic addition when all material is removed; a different infected prosthesis that cannot be extracted represents a separate problem. Dosing, drug exposure monitoring and renal function must be reassessed during treatment, without using pharmacological therapy as a substitute for extraction.

In a superficial incisional infection genuinely confined to skin and subcutaneous tissue, a short course of antistaphylococcal therapy with close follow-up may be sufficient without removing the system. It must be documented that there is no communication with the pocket, exposed hardware or systemic signs, and response must be reassessed because an apparently superficial lesion may be the onset of deep infection. Failure or recurrence after treatment requires reconsidering the classification, not merely repeating the same prescription. If hardware is exposed, the superficial-incision framework no longer applies even when the patient remains afebrile.

Antibiotic duration is determined by the deepest site involved and the microbiology. After complete removal, a bacterial infection confined to the pocket with negative blood cultures is generally treated for ten to fourteen days; bacteremia, lead vegetations, valvular endocarditis or metastatic foci require longer courses, often in the range of four to six weeks. Selected situations with isolated lead vegetation, rapid clearance and absence of S. aureus may allow discussion of shorter post-extraction durations while maintaining an adequate total course. It is incorrect to apply a pocket-infection regimen to bacteremia merely because echocardiography does not show masses.

Transvenous extraction is the usual approach when feasible, but it requires expertise and a plan for hemorrhage, cardiac or vascular injury and tamponade. Risk increases with long lead dwell time, adhesions, calcification and anatomical complexity; the probability of injury cannot be inferred from chronological age alone. The ELECTRa registry documented outcomes and complications in European centers and underscores the importance of organization. The need for valve surgery, certain epicardial locations and specific anatomical relationships may favor surgical removal or combined strategies.

Pocket debridement accompanies removal of infected components and includes treatment of compromised tissue and involved suture material. Closure method and wound management depend on contamination, tissue viability and extent; simply moving the generator to another subcutaneous site while retaining colonized leads is insufficient. In the rare situation in which complete removal is not feasible, any salvage treatment must be explicitly distinguished from an eradicative extraction. The 2026 consensus on lead management allows selected conservative strategies when risk is prohibitive or the patient refuses removal, with clear information about limitations and surveillance.

Large vegetations require balancing the embolic risk of transvenous extraction, respiratory status, the feasibility of surgery and the presence of shunts. Echocardiographic size alone does not establish a universal threshold between techniques. In selected cases, percutaneous aspiration before or during extraction may be considered to reduce the mass; evidence remains limited and procedural risks persist. Incomplete removal must be explicitly recognized, with identification of residual material and whether retrieval remains possible: a retained infected fragment should not be described as complete source control.

Bacteremia without pocket signs or vegetations requires a decision based on species, persistence, recurrence and alternative sources. The likelihood of colonization is particularly relevant with S. aureus, coagulase-negative staphylococci and Candida; the balance may differ for other organisms. Disappearance of bacteremia after control of an extracardiac source is favorable but not an absolute guarantee. Valvular endocarditis, even without a lead vegetation, warrants discussion of extraction because retaining intracardiac material may compromise eradication. Conservative decisions should have a documented rationale and surveillance plan.

After removal, the indication for reimplantation is reassessed first: some patients no longer need the same device or may benefit from a different configuration. If needed, the new system is placed at a different site after resolution of local and systemic signs and favorable microbiological documentation. The AHA statement gives at least seventy-two hours of negative blood cultures as a general reference and a longer interval, around fourteen days after extraction, when valvular endocarditis is present. These conditions must be integrated with metastatic foci, organism and clinical course; they are not an automatic date that makes implantation safe despite ongoing active infection.

Pacing dependence requires planned temporary protection, for example an active-fixation lead connected to an external generator in appropriate settings. A patient at arrhythmic risk may require monitoring or a wearable defibrillator as indicated. Leadless pacemakers and subcutaneous defibrillators reduce some components exposed to infection, but they are not interchangeable: they do not necessarily provide all required pacing, resynchronization or antitachycardia functions. Device selection must be based on residual electrical needs as well as infectious risk.

When extraction risk is judged disproportionate or goals of care make removal inappropriate, suppressive antimicrobial therapy may be considered after adequate initial treatment. This is an individualized strategy with risks of recurrence and toxicity and is supported by limited evidence; it should not be presented as equivalent to removal. Discussion includes life expectancy and quality of life, technical feasibility in an experienced center and informed preferences. Frailty should not be conflated with procedural impossibility before concrete alternatives have been assessed.

Prevention begins before implantation: appropriate indication, postponement of elective procedures in the presence of active infection, pre-incision antibiotic prophylaxis, asepsis and meticulous hemostasis. Anticoagulant management should avoid strategies that unnecessarily increase hematoma; in appropriate patients, BRUISE CONTROL showed an advantage of continuing warfarin over heparin bridging, without establishing a single approach for every anticoagulant or bleeding-risk profile. A stable hematoma should not routinely be aspirated. Reducing early revisions and unnecessary manipulation protects the skin barrier.

The antibacterial envelope reduced major infections in the WRAP-IT trial in a selected population undergoing device procedures; the result supports risk-based use, not indiscriminate application or replacement of asepsis. The PADIT trial did not show a statistically significant benefit of the incremental antibiotic regimen studied compared with conventional prophylaxis. Prolonging antibiotics after every implant is therefore not a universal solution. Procedural prophylaxis, hematoma prevention and implant quality act at different steps and should be planned together.

Antibiotic prophylaxis for implantation primarily targets skin flora and must ensure adequate drug exposure at the time of incision. An antistaphylococcal cephalosporin is commonly used when appropriate; true allergies, colonization and MRSA risk may alter drug choice and infusion timing. Simply administering any antibiotic after the procedure is not sufficient. Infection at a distant source, including ongoing bacteremia, should be managed before elective implantation. The presence of an electronic device alone is not a routine indication for infective-endocarditis prophylaxis for dental procedures; such an indication depends on additional high-risk cardiac conditions.

Resynchronization may be essential for stability in some patients with heart failure, whereas others depend on pacing only because of a block that is no longer persistent. Before and after extraction, it is therefore necessary to distinguish the need for a minimum heart rate, the hemodynamic benefit of biventricular pacing and protection against arrhythmic death. A temporary system that maintains the heart rate does not necessarily reproduce all these effects. Serial assessment helps determine the timing and architecture of the new implant and recognize heart failure due to loss of electrical therapy, rather than attributing it solely to sepsis or fluid balance.

Prognosis is influenced by the presence of endocarditis, sepsis, renal failure, metastatic foci and the possibility of prompt removal. Improvement after extraction requires microbiological and clinical confirmation, with assessment of the wound and valvular function. A residual image along the course of the old lead may represent a fibrous sheath, but it must be interpreted against cultures and clinical evolution; it is neither automatically infected nor irrelevant in every circumstance. Follow-up should also verify adequacy of the reimplanted device and functional recovery after the septic phase.

Complications

Septic dissemination may involve the lungs, valves, spine and joints and may sustain infection after system removal. Control of the primary source does not eliminate the need to drain an empyema or treat spondylodiscitis. Persistent fever requires distinction among residual collections, infection of a new vascular access, drug toxicity and progression of valvular disease. A sequence of negative blood cultures is necessary in many pathways, but by itself does not demonstrate control of every infected compartment.

Extraction complications include injury to the vena cava or cardiac structures, hemopericardium, hemothorax, embolization and tricuspid damage. Their possibility makes organization of the center essential; it does not justify systematically leaving an infected device in place. Procedural risk must be weighed against sepsis and recurrence if hardware remains. After the procedure, hypotension, dyspnea or new anemia require immediate assessment for bleeding; a new murmur or congestion may instead indicate injury to the valvular apparatus.

Reinfection may result from retained material, reimplantation that is too early, new bacteremia or persistent risk factors. Comparison with the first isolate and reconstruction of procedures help distinguish relapse from a new episode. Loss of pacing or antitachycardia protection during the period without a device represents a parallel risk that must be anticipated, while prolonged antibiotic therapy may cause renal injury, cytopenias or access-related infections. Continuity between infectious-disease and electrophysiology teams is therefore necessary until the pathway is completed, not only on the day of extraction.