Endovascular tissue remnants in transoesophageal echocardiography after transvenous lead extraction: prevalence and impact on clinical course

Background Endovascular fibrous, floating remnants (EFR), referred to as "ghosts", have been described in several studies following the transvenous lead extraction (TLE) of pacemaker and ICD leads. Their clinical significance and potential associations with patient characteristics and device features remain inconclusive. Purpose This study investigates the prevalence of EFR in transoesophageal echocardiography (TOE) after lead extraction and their relationship with clinical parameters, device characteristics, and subsequent clinical outcomes. Methods In a single-centre, retrospective registry, patients with active cardiac implantable electronic devices (CIEDs) who underwent TLE for infectious indications (local infection or cardiac device-related infective endocarditis) were identified. TLE cases that were performed more than 12 months after initial device implant in a study period from January 2023 to December 2024 were included. Patients without follow-up TOE after TLE were excluded. Postoperative TOEs were used to differentiate between absent EFR, non-tricuspid valve-associated EFR, and tricuspid valve-associated EFR. Results A total of 126 patients met the selection criteria (median age 77.4 years, 19.8% female, 69.8% with vegetations on preoperative TOE). A median of 2 leads was to be extracted (interquartile range [IQR] 1; median age of oldest lead 8.5 years [IQR 9.2 years]), with complete procedural success being achieved in 93.7% (clinically successful in 97.7%). CIED reimplantation was performed in 64.3% of patients. Non-tricuspid valve-associated and tricuspid valve-associated EFR were detected in 38 (30.2%) and 23 (18.3%) patients, respectively. The number of extracted leads was significantly higher in patients with postoperative EFR (p < 0.05 compared to patients without EFR). A higher number of prior lead-associated interventions was associated with an increased prevalence of non-tricuspid valve-associated EFR (p < 0.001 vs. patients without EFR; p = 0.080 vs. tricuspid valve-associated EFR). Other baseline characteristics and mean or maximum lead dwell time did not differ between groups. The presence of EFR was associated with a prolonged duration of antibiotic treatment, a delayed reimplantation, and a longer length of hospital stay (all p < 0.05). In-hospital mortality was 10.32% with no significant differences between groups. Conclusion Both tricuspid and non-tricuspid valve-associated EFR significantly influence the in-hospital course after TLE in patients with CIED-associated infections. The impact on long-term prognosis requires further scientific investigation.