Prognostic Value of Preoperative Echocardiographic Findings in Patients Undergoing Transvenous Lead Extraction

(1) Background: In patients referred for transvenous lead extraction (TLE) transesophageal echocardiography (TEE) often reveals abnormalities related to chronically indwelling endocardial leads. The purpose of this study was to determine whether the results of pre-operative TEE might influence the long-term prognosis. (2) Methods: We analyzed data from 936 TEE examinations performed at a high volume center in patients referred for TLE from 2015 to 2019. The follow-up was 566.2 ± 224.5 days. (3) Results: Multivariate analysis of TEE parameters showed that vegetations (HR = 2.631 [1.738–3.983]; p < 0.001) and tricuspid valve (TV) dysfunction unrelated to the endocardial lead (HR = 1.481 [1.261–1.740]; p < 0.001) were associated with increased risk for long-term mortality. Presence of fibrous tissue binding sites between the lead and the superior vena cava (SVC) and/or right atrium (RA) wall (HR = 0.285; p = 0.035), presence of penetration or perforation of the lead through the cardiac wall up to the epicardium (HR = 0.496; p = 0.035) and presence of excessive lead loops (HR = 0.528; p = 0.026) showed a better prognosis. After adjustment the statistical model with recognized poor prognosis factors only vegetations were confirmed as a risk factor (HR = 2.613; p = 0.039). A better prognosis was observed in patients with fibrous tissue binding sites between the lead and the superior vena cava (SVC) and/or right atrium (RA) wall (HR = 0.270; p = 0.040). (4) Conclusions: Non-modifiable factors may have a negative influence on long-term survival after TLE. Various forms of connective tissue overgrowth and abnormal course of the leads modifiable by TLE can be a factor of better prognosis after TLE.


Introduction
Recently, due to the rising incidence of infectious and non-infectious complications related to cardiac implantable electronic devices (CIED), the number of transvenous lead extraction (TLE) procedures has also been increasing [1]. TLE is considered as a first-line strategy for the management of CIED-associated complications [2,3]. The rate of major complications associated with TLE has been estimated to range from 0.9 to 4.0%, and most often there is damage to the heart or venous vessels; the lead extraction procedure carries a 0 to 0.4% risk of death [2,3]. Due to the continuous improvement in the extraction strategy, most patients with major complications are discharged from hospital in a good general state [4]. Therefore, theoretically the fate of patients after TLE should not differ from those who did not undergo TLE. There is a large volume of published studies describing TLE outcomes, however the results are still unsatisfactory, because mortality is 5-25% at one year, 8-38% at three years, 8-44% at 5 years and 10-60% at 10 years, with the lowest values encountered in patients with non-infectious indications and highest in those with lead-related infective endocarditis (LRIE) [5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20][21][22][23]. Previous studies have not analyzed the effect of echocardiographic phenomena on long-term survival of patients undergoing TLE, and few reports have only assessed their relationship with the risk of procedure. The main echocardiographic parameter considered in order to estimate the risk of surgery was the value of the left ventricular ejection fraction (LVEF) [6,7,12,16] as well as the presence/size of vegetation [7,[9][10][11][13][14][15]18,19]. Only a few studies based on small sample sizes considered a possible impact of asymptomatic masses on endocardial leads (AMEL) on the length of survival following TLE [24][25][26][27][28][29]. This paper provides an in-depth analysis of preoperative TEE findings and their usefulness for predicting long-term outcomes of TLE.

Study Population
A prospective analysis was carried out on data from preoperative TEE performed at a high-volume center during 936 TLE procedures from June, 2015 to October 2019. All patients gave their written informed consent to TLE and analysis of anonymized medical records, approved by the Bioethics Committee of the Regional Chamber of Physicians and Dentists in Lublin no. 288/2018/KB/VII.

Factors Potentially Affecting Long-Term Survival after TLE
In order to identify the factors that may influence long-term survival the following variables have been analyzed: Patient-dependent factors: age (during TLE and at first CIED implantation), gender, NYHA class, LVEF, atrial fibrillation, chronic renal failure, diabetes, arterial hypertension, a history of coronary artery bypass graft (CABG), previous sternotomy, CHA2DS2-VAsc score, Charlson comorbidity index, chronic anticoagulation, and antiplatelet therapy.
CIED-related factors: the number of leads in the device before TLE, the number of leads the patient had before TLE, abandoned leads, excessive lead loops before TLE, high voltage (HV) leads, leads in the coronary sinus (CS), dwell time of the oldest lead in the patient, mean implant duration before TLE, cumulative dwell time of the extracted leads, and the number of CIED procedures before TEE.
Indication-related data: diagnosis of LRIE certain or probable with or without pocket infection or only local pocket infection.
TLE efficacy and complications: the rate of complete radiographic success, partial radiographic success, lack of radiographic success, clinical success, complete procedural success and presence of any major complication, hemopericardium, severe tricuspid valve damage during TLE, rescue cardiac surgery.
Most important preoperative TEE findings: tricuspid valve dysfunction, lead-dependent tricuspid valve dysfunction (LDTD), shadowing from the leads before TLE, fibrous tissue binding the lead to the heart structures, AMEL (fibrous tissue encasing the lead, lead thickening, clots and vegetation-like masses), vegetations, excessive lead loops and perforation or penetration of the lead through the cardiac wall up to the epicardium.

Lead Extraction Procedure
TLE was defined according to EHRA consensus document as intervention with removal of at least one lead that has been implanted for more than one year or a lead regardless of duration of implant requiring the assistance of specialized equipment that is not included as part of the of the typical implant package and/or removal of a lead from a route other than the implant vein [30].
Complete procedural success was defined as removal of all targeted leads and material, with the absence of any permanently disabling complication or procedure-related death. Clinical procedural success was defined as retention of a small portion of a lead (<4 cm) that does not negatively impact the outcome goals of the procedure and with absence of any permanently disabling complication or procedure-related death [30].
Extraction procedures were performed in a hybrid operating room or in an operating room, using mechanical systems such as polypropylene Byrd dilators (Cook ® Medical, Leechburg, PA, USA), making use of the oblique cutting edge of the tip to dissect leads from fibrous sheaths that immobilized the lead in the intravascular and/or intracardiac segment [11,28]. Procedures were performed in patients under general anesthesia and after preparation of the surgical field as for subjects coming in for cardiac surgery. Continuous invasive blood pressure monitoring from radial artery was used. The composition of the surgical team and the course of the extraction procedure have been described in detail elsewhere [31][32][33].

Preoperative TEE
TEE was performed using the Philips iE33 (Phillips Healthcare, Andover, MA, USA) or the GE Vivid S70 (General Electric Company, Boston, MA, USA) ultrasound machine equipped with X7-2t Live 3D or 6VT-D probes. Images and recordings were obtained before the procedure, after general anesthesia and tracheal intubation, during preparation of the surgical field, and dissection and stabilization of the leads in the region of the device pocket. Leads were evaluated in the mid-esophageal, inferior esophageal and modified transgastric views to visualize the right heart chambers and the tricuspid valve. In order to obtain complete visualization of the structures (and assessment of lead/heart interface) non-standard imaging planes were sometimes required. After the procedure the results were entered into a computer database. The TEE examination was described in detail in previous publications-we followed the methods of Nowosielecka et al. [31][32][33].

Echocardiographic Findings Associated with Endocardial Leads: Definition and Classification According to the Anatomy and Characteristic Features
Asymptomatic masses on endocardial leads (AMEL) [31]: Additional masses on the leads classified as clots (varying degrees of organization), components of connective tissue (so-called accretions), masses resembling vegetations (vegetation-like masses), probably the remnants after infections: Old fibrous vegetations or clots ( Figure 1).
Bacterial vegetations [31], i.e., multishaped, mobile masses of inhomogeneous echogenicity. Vegetations were diagnosed only if they were accompanied by signs of a general infection (positive inflammatory markers, positive blood cultures) or a regional infection (pocket infection) ( Figure 2).  Bacterial vegetations [31], i.e., multishaped, mobile masses of inhomogeneous echogenicity. Vegetations were diagnosed only if they were accompanied by signs of a general infection (positive inflammatory markers, positive blood cultures) or a regional infection (pocket infection) ( Figure 2). Hyperechoic segmental thickening of the leads defined as connective tissue overgrowth (undergoing fibrosis, mineralization, crystallization and even ossification) [31].
Buildup: Fibrous connective tissue sheath around the lead causing adherence to the endocardium and vessel walls producing images similar to segmental lead thickening but moving along with the cardiac wall. The term encompasses also segmental lead-to-lead adhesion (two or three leads) moving along together with the cardiac walls. Immobile masses binding the lead to the vein or heart wall most frequently represent a sign of preexisting asymptomatic inflammatory response triggered by the endocardial lead (foreign body reaction). Over time, fibrosis ensues with the presence of calcifications (mineralization, crystallization and ossification). This type of reaction may occur in patients with and without device infections [31] (Figure 3). Hyperechoic segmental thickening of the leads defined as connective tissue overgrowth (undergoing fibrosis, mineralization, crystallization and even ossification) [31].
Buildup: Fibrous connective tissue sheath around the lead causing adherence to the endocardium and vessel walls producing images similar to segmental lead thickening but moving along with the cardiac wall. The term encompasses also segmental lead-to-lead adhesion (two or three leads) moving along together with the cardiac walls. Immobile masses binding the lead to the vein or heart wall most frequently represent a sign of preexisting asymptomatic inflammatory response triggered by the endocardial lead (foreign body reaction). Over time, fibrosis ensues with the presence of calcifications (mineralization, crystallization and ossification). This type of reaction may occur in patients with and without device infections [31] (Figure 3). Other, separately classified lead-associated phenomena: Lead-dependent tricuspid dysfunction: valve regurgitation (very rarely TV stenosis) unquestionably caused by the lead (lead impingement, lead entanglement with tendinous chords, lead adhesion to the leaflet, leaflet perforation) [31] (Figure 4). Other, separately classified lead-associated phenomena: Lead-dependent tricuspid dysfunction: valve regurgitation (very rarely TV stenosis) unquestionably caused by the lead (lead impingement, lead entanglement with tendinous chords, lead adhesion to the leaflet, leaflet perforation) [31] (Figure 4). Cardiac wall perforation by the lead: visualization of the lead tip outside the heart contour, frequently with fluid in the pericardial sac; placement of the lead tip close to the border of the pericardium is referred to as penetration ( Figure 5). Excessive lead loops as a result of too weak fixation during implantation or lead fracture with insulation breach in the subclavian region. Excessive lead loops may be encountered in the right atrium or the right ventricle, and in the tricuspid valve orifice ( Figure 6). Excessive lead loops as a result of too weak fixation during implantation or lead fracture with insulation breach in the subclavian region. Excessive lead loops may be encountered in the right atrium or the right ventricle, and in the tricuspid valve orifice ( Figure 6).

Statistical Analysis
The Shapiro-Wilk test was used to test the distribution of continuous variables. A non-parametric distribution of all continuous variables was found. For a clearer presentation of the results, all continuous variables are presented as the mean ± standard deviation and some of them (patient age during TLE and during first CIED implantation, left ventricle ejection fraction, dwell time of the oldest lead in the patient) additionally as median with the first and the third quartile. The categorical variables are presented as number and percentage. The study population was divided into two groups depending on TLE outcomes (survival versus death) at two-year follow-up. The significance of differences between the groups was determined using the nonparametric "U" Mann-Whitney test and Chi square tests. The relationship between the echocardiographic parameters and mortality after TLE was analyzed using Cox regression analysis. All variables reached p < 0.1 in univariate analysis were included into a multivariate model. Two multivariate regression models were defined. Model 1 was built to assess the prognostic value of the echocardiographic variables only. Model 2 included echocardiographic variables from model 1 and adjusted by clinical parameters of known prognostic values (patient age at first implantation, patient age during TLE, gender, LVEF, NYHA functional class, presence of diabetes mellitus, renal failure, arterial hypertension, infectious indications for TLE, ICD and CRTD prior to TLE). Moreover, the impact of the binding sites between leads and VCS wall and/or right atrial wall and survival after TLE was presented as the

Statistical Analysis
The Shapiro-Wilk test was used to test the distribution of continuous variables. A nonparametric distribution of all continuous variables was found. For a clearer presentation of the results, all continuous variables are presented as the mean ± standard deviation and some of them (patient age during TLE and during first CIED implantation, left ventricle ejection fraction, dwell time of the oldest lead in the patient) additionally as median with the first and the third quartile. The categorical variables are presented as number and percentage. The study population was divided into two groups depending on TLE outcomes (survival versus death) at two-year follow-up. The significance of differences between the groups was determined using the nonparametric "U" Mann-Whitney test and Chi square tests. The relationship between the echocardiographic parameters and mortality after TLE was analyzed using Cox regression analysis. All variables reached p < 0.1 in univariate analysis were included into a multivariate model. Two multivariate regression models were defined. Model 1 was built to assess the prognostic value of the echocardiographic variables only. Model 2 included echocardiographic variables from model 1 and adjusted by clinical parameters of known prognostic values (patient age at first implantation, patient age during TLE, gender, LVEF, NYHA functional class, presence of diabetes mellitus, renal failure, arterial hypertension, infectious indications for TLE, ICD and CRTD prior to TLE). Moreover, the impact of the binding sites between leads and VCS wall and/or right atrial wall and survival after TLE was presented as the Kaplan-Meier survival curves. The log rank test was used to compare the survival distributions of the groups. A two-tailed p value < 0.05 was considered statistically significant.

Prognostic Factors Related to Implanted Devices
Of CIED-related factors, HV therapy (ICD lead presence) (p = 0.016), leads in the CS (LV pacing) (p < 0.001) and a higher number of leads the patient had before TLE (p = 0.029) were associated with lower survival rates. Similarly, infectious indications for TLE (LRIE) were associated with worse long-term survival (p < 0.001) ( Table 1).

Prognostic Factors Related to TLE Procedure
The factors related to procedure efficacy and major complications did not affect significantly long-term outcomes after TLE (Table 1).

Prognostic Factors Related to TEE Findings
Of preoperative TEE variables, tricuspid valve dysfunction (degree of regurgitation) (p < 0.001) and vegetations (p < 0.001) were significantly more common among those who died after TLE. In contrast, the signs of connective tissue overgrowth occurred significantly more often among those who survived: fibrous tissue binding the lead to the superior vena cava and heart structures (p = 0.024), fibrous tissue binding the lead to the heart structures (all) (p = 0.008), fibrous tissue binding the lead to the RA wall (p < 0.001), lead-to-lead adhesion (p < 0.009), asymptomatic masses on endocardial leads (all) (p = 0.022), and fibrous tissue encasing the lead (p = 0.021). Similarly, the presence of (any) lead loops in the heart before TLE (p = 0.037) and perforation or penetration of the lead through the cardiac wall up to the epicardium (p = 0.038) were associated with better chances of long-term survival ( Table 2).

Cox's Regression Analysis Results (Model-1; Echocardiographic Data)
Univariate Cox regression analysis showed a negative relationship between the chances of two-year survival and disease-related parameters only: lead-unrelated TV dysfunction (HR = 1.528; p < 0.001) and vegetations (HR = 3.078; p < 0.001). However, on the other hand, there was a positive relationship between the chances of two-year survival and the following variables (related to implant duration): fibrous tissue encasing the lead (HR = 0.442; p = 0.014), fibrous tissue binding the lead to the SVC and RA wall (HR = 0.208; p = 0.007), and lead-to-lead adhesion (HR = 0.484; p = 0.022). Additionally, perforation or penetration of the lead through the cardiac wall up to the epicardium (HR = 0.474; p = 0.024) and excessive lead loops (HR = 0.543; p = 0.033) were suggestive of better prognosis (Table 3). Abbreviations: LDTD-Lead dependent tricuspid valve dysfunction, RV-right ventricle, SVC-superior vena cava, TV-tricuspid valve.
Multivariable Cox regression analysis of TEE variables confirmed the negative relationship between the chances of two-year survival and lead-unrelated TV dysfunction (HR = 1.481; p < 0.001) and vegetations (HR = 2.631; p < 0.001), and the positive relationship between fibrous tissue binding the lead to the SVC and/or RA wall (HR = 0.285; p = 0.036), perforation or penetration of the lead through the cardiac wall up to the epicardium (HR = 0.469; p = 0.035) and excessive lead loops (HR = 0.528; p = 0.026) ( Table 3).

Discussion
Predicting long-term survival after various procedures, especially those related to the cardiovascular system, is an extremely important element of planning a therapeutic strat- Results of multivariate model-2 Cox regression analysis are also presented in the Figure S1.
Echocardiographic phenomena have been very rarely analyzed in terms of their impact on long-term survival after TLE A couple of papers showed only the role of so-called ghosts, i.e., post-removal, tubular, mobile masses following the lead's intracardiac route in the right-sided heart chambers [15,20]. The present study set out to investigate the usefulness of preoperative transesophageal echocardiography in the assessment of the long-term survival after TLE. The results are consistent with those observed in earlier studies [5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20][21][22] which showed that patient-dependent variables (demographic, related to the underlying disease, comorbidities, systemic infection) were the main risk factors for death at long-term follow-up. CIED-related factors, including the number of leads, left ventricular lead and ICD lead play a significant role, but secondary to the underlying condition [5,6,[11][12][13][14][15]19,21]. Of the previously identified echocardiographic factors [7,12,16], this study confirms the prognostic value of left ventricular ejection fraction. It should, however be emphasized that the present study was designed to investigate the role of the factors that have not been considered in previous analyses of long-term survival after TLE. Because of the complexity of relationships the preoperative TEE findings and abnormalities were divided into: (1) Non-modifiable factors related to the patient's general condition; (2) non-modifiable factors related to the underlying disease (indication-dependent); (3) factors that have no effect on the procedure course and chances of long-term survival; (4) factors that may increase the complexity of the extraction procedure and the development of complications but which per se do not decrease chances of long-term survival; and (5) abnormalities that can be corrected during the extraction procedure. 1. Non-modifiable factors related to the patient's general condition: Tricuspid valve regurgitation (TVR), excluding LDTD, was associated with worse long-term survival in multivariate analysis, but after adjustment for common risk factors for poor prognosis this variable lost its prognostic value. In patients referred for lead extraction TVR is a non-modifiable factor because of right ventricular status.

3.
Factors that have no effect on the procedure course and chances of long-term survival: AMEL (clots, vegetation-like masses) had no influence on chances of long-term survival in our analysis.

4.
Factors that may increase the complexity of the extraction procedure and the development of complications but which, per se, do not decrease chances of long-term survival: fibrous tissue binding the lead to the heart structures, lead-to-lead adhesion. The degree of connective tissue overgrowth in response of the endothelium to longterm irritation by the lead depends on implant duration, stiffness and the number of leads, but first of all on patient age (inverse relationship). This phenomenon has been better documented in papers describing lead removal in children and young patients and in adults with leads implanted in childhood [17]. Surprisingly, the current study found that various forms of connective tissue overgrowth (fibrous tissue binding the lead to the heart structures, lead-to-lead adhesion, lead thickening, scar tissue surrounding the lead) were associated with better long-term survival, although based on previous observations [31,32], connective tissue overgrowth was a predictor of TLE technical difficulty and major complications. This proves that the course of the procedure does not affect prognosis after TLE.

5.
Abnormalities that can be corrected during the extraction procedure: a. Lead-dependent tricuspid dysfunction was not significantly associated with the length of survival. This can be attributed to the fact that most patients with LDTD were referred for the intervention because of the lead propping one of the leaflets open, which was corrected to varying extent during TLE. b.
Excessive lead loops (loop in the right atrium, loop crossing the TV, loop in the right ventricle or pulmonary artery) in univariate and multivariate Cox analysis were significantly associated with better survival odds. The reason was that this abnormality was the indication (main or accompanying) for lead replacement and no patient left our facility with abandoned leads. c.
Perforation, penetration-as was the case of lead loops, perforation/penetration was the main or accompanying indication for lead extraction (most of them were "dry" and caused lead dysfunction or, less frequently, it was an incidental finding or the cause of fluid accumulation in the epicardial space). All perforating/penetrating leads were replaced, thus eliminating their influence on survival and future quality of life. Similar to lead loops, perforations in univariate and multivariate Cox analysis of Model 1 were significantly related to better survival odds.

Limitations
The current study is a single center observational prospective study. The lead extraction procedure was performed using mechanical tools and not laser sheaths.

Conclusions
The main factors predicting shorter survival among patients undergoing TLE were those related to the patient (patient age during TLE, male gender, higher NYHA class, low LVEF, atrial fibrillation, or chronic renal failure), and those related to the underlying disease, comorbidities and systemic infection.
Non-modifiable factors (patient-dependent and indication/infection-dependent) may have a negative influence on the postoperative course and long-term survival.
The exacerbation of the foreign body reaction resulting in fibrous tissue binding the lead to the vena cava superior or heart structures (especially right atrium wall) seemingly improves chances of longer survival.

Institutional Review Board Statement:
The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the institutional Bioethics Committee at Regional Physicians Chamber in Lublin no. 288/2018/KB/VII.

Informed Consent Statement:
Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
Readers can access the data supporting the conclusions of the study at www.usuwanieelektrod.pl.

Conflicts of Interest:
Authors declare no conflict of interest.