A 72-year-old Caucasian man presents to the hospital in Columbus, Ohio. He reports having two recent evaluations at an emergency department for generalized weakness and cough, the cause of which has not been determined. He arrives from another facility after a fall, presumably as a result of tripping or slipping.
His medical history includes implantation of a remote cardiac resynchronization therapy defibrillator. The cardiology report notes that while the patient’s left ventricular ejection fraction temporarily improved to 55-60%, it was followed by a decline to 25% at 1-year prior to presentation.
The transferring facility notes that at the time of the presentation, the patient was hemodynamically stable (blood pressure 105/70). The initial physical examination revealed perianal and genital excoriations with associated suprapubic swelling, and generalized weakness.
Further evaluation found that the patient had methicillin-resistant Staphylococcus aureus bacteremia, and subsequent transesophageal echocardiography revealed multiple vegetations on the aortic and mitral valves, and the right atrial device lead.
The patient was diagnosed with definite infective endocarditis per the modified Duke criteria (two major criteria), and transferred to the Columbus hospital to undergo transvenous lead extraction (TLE).
His medical history up until the time of transfer includes:
- Chronic non-ischemic heart failure with reduced ejection fraction (NYHA class III, AHA class C)
- Ventricular tachycardia
- Atrioventricular node ablation due to atrial fibrillation refractory to medical treatment
- Chronic kidney disease (stage IIIa)
- Remote embolic stroke
- Non-Hodgkin’s lymphoma in remission following treatment via a right sided tunneled catheter placed 4 years previously
Clinicians initially remove the patient’s indwelling right subclavian port, which they suspect is where the infection originated, and then complete full extractions of the cardiac device system.
Because the patient’s atrioventricular node ablation left him dependent upon pacing, he is bridged with a temporary transvenous wire until blood cultures show no evidence of residual bacteremia. Clinicians then implant a Micra leadless pacemaker into the apical septal region of the right ventricle (QRS duration 200 ms) and remove the temporary wire.
To allow long-term intravenous treatment with antibiotics, the team places a peripherally inserted central catheter. The patient receives intravenous vancomycin (1.5 g daily) for 6 weeks, and is subsequently discharged.
Seven weeks later, as part of the long-term treatment plan, the patient is slated to receive a subcutaneous implantable cardioverter defibrillator (SICD). Before surgery, clinicians re-evaluate the vegetations via transesophageal echocardiogram, which identifies a 1.3 cm tubular and mobile echo-dense ghost in the right atrium.
The patient is subsequently discharged without any change in management or further intervention.
Three months after SICD placement, the patient presents with inappropriate shock due to lead displacement, and is readmitted for lead revision. Assessment with transthoracic echocardiogram during that admission shows no remnant of the ghost or valvular vegetations. The patient remains infection-free at follow-up, 1 year after the ghost was identified.
Clinicians reporting this case of a “ghost” observed in a patient following TLE note that this novel complication has important clinical significance, as it portends poor outcomes and affects 8-14% of patients who undergo the procedure.
Because infective endocarditis is associated with significant morbidity and mortality, it is a class 1 indication for removal of all intra-cardiac hardware. The ghost — i.e., a persistent remnant or sheath that encapsulated the previous lead – is defined as a mobile mass that remains after TLE and often follows the lead’s intracardiac route into the right cardiac cavities.
When the phenomenon was first reported in 2008, ghosts were thought to represent fibrous sheaths, but they may also be infectious vegetations, and distinguishing between the two can be challenging.
Ghosts remain a rare entity and as such, there are no specific guidelines for their detection or management. Given that infection in lead-dependent infective endocarditis can be reliably confirmed or excluded using positron emission tomography/computed tomography (PET-CT) and single photon emission computed tomography (SPECT)-CT with radioisotope-labeled leukocytes, the case authors suggest that these modalities can also be used to evaluate ghosts.
Two sequential mechanisms are thought to give rise to ghosts: thrombosis with subsequent collagenous fibrosis, followed by endothelialization of the fibrous capsule surrounding the lead. This encapsulation prevents lead migration and subsequent thrombus formation.
When indicated for lead malfunction, wire fractures, insulation damage, or lead migration, TLE rarely results in residual ghosts. However, they are easily identified after TLE is performed due to infection. Infected leads tend to be easier to extract than those that are noninfected, because the infectious process theoretically breaks the seal between collagenous adhesions, endocardial surfaces, and the lead itself. The case authors note that despite the important role of infection, the presence of ghosts does not constitute a criterion for the diagnosis of infective endocarditis.
Although significant strides have been made in the understanding of ghost pathogenesis, many questions remain regarding the therapeutic and prognostic implications, and further studies are warranted to evaluate the need for interval monitoring with blood cultures and/or echocardiography after encountering a ghost.
In this patient’s case, the ghost was identified after a 6-week course of antibiotic therapy for infective endocarditis, leaving the potential role of active infectivity requiring prolonged antibiotic treatment unclear.
Emerging evidence does suggest, however, that ghosts are harbingers of poor outcomes. The presence of a ghost has been linked with a 3.4-fold higher all-cause mortality compared with patients without ghosts, and given their independent association with death, the presence of ghosts may represent a marker of a constellation of high-risk features such as older age, endocarditis, and co-morbidities. Other prospective data suggests that ghosts may also be associated with cardiac device-related infective endocarditis (CDRIE) relapse/recurrence (hazard ratio 4.594; P=0.046).
The presence of ghosts may thus identify a sicker patient who may be at higher risk of death, and would thus require careful monitoring, closer follow-up, and prompt individualized treatment to avoid worsening or complications.
While transesophageal echocardiography has been the cornerstone in the diagnosis of infective endocarditis, intracardiac echocardiography increases diagnostic accuracy by specifically identifying vegetations not otherwise evident in an estimated 15% of patients with possible infective endocarditis. In addition, the case authors state, given the superior ability of intracardiac echocardiography to evaluate right-sided cardiac structures, the procedure can not only help detect ghosts intra-procedurally during TLE, but may also play an important role in ghost management.
In the absence of specific diagnostic guidelines, advanced imaging such as PET-CT and SPECT-CT shows promise in patients with CDRIE, having shown high sensitivities (65% and 73.7%, respectively) and specificities (88% and 81%, respectively) for the diagnosis of CDRIE.
For now, diagnosis of cardiac device pocket infection is primarily based on the judgment of experienced physicians. C-reactive protein (CRP), high sensitivity-CRP, and procalcitonin tests with specific cut-offs for cardiac device infections may provide objective evidence to assist with diagnosis.
In the case of procalcitonin, the optimized cut-off reference value of 0.05 ng/mL (one tenth the standard cut-off) provided a sensitivity of 60% and specificity of 82%, according to the authors of a paper on biomarker diagnosis of cardiac device pocket infection. In contrast, white blood count, lipopolysaccharide binding protein, presepsin, polymorphonuclear-elastase, and interleukins-1ß, -6, -8, -10, -23 were found to have less sensitivity and were less helpful in diagnosis.
The case authors conclude that key questions yet to be answered include predisposing risk factors for the development of ghosts, the role of serial imaging, and the appropriate duration of antibiotic therapy. Further description of the role of ghosts toward clinical outcomes may avert their residual risk, as a large body of evidence confirms that ghosts are to be feared rather than ignored.
The case report authors noted no conflicts of interest.