RESULTS
Between July 2009 and June 2018, a total of 566 endovascular leads were extracted from 302 patients [Evolution® group (133 patients with 233 leads) and TightRailTM group (169 patients with 333 leads)]. The baseline demographic, clinical and laboratory data of the study groups were represented in Table 1. Diabetes mellitus, hypertension, coronary artery disease, and HFrEF were more prevalent in the TightRailTM group (p<0.05).
The details of CIEDs and TLE procedure were shown in Table 2. The mean number of extracted leads per patient was 1.8±0.73, and the median lead dwell time was 5.0 (0.6-33) years. All ICD leads were of dual coils, and all coronary sinus electrodes have a passive fixation mechanism. Re-implantation was performed in 216 (71.5%) patients during the index TLE procedure. Leads with active fixation mechanism and coronary sinus leads were more common in the TightRailTM group (p<0.05). The most common TLE indications in the Evolution® and TightRailTM groups were lead malfunction (57.9%) and CIED-related infection (49.1%), respectively (p<0.05). Among different sheath sizes, 13F in the TightRailTM group and 9F in the Evolution® group were more frequently used (p<0.05).
Complete procedural and clinical success were achieved in 259 (85.8%) and 270 (89.4%) patients using mechanical dilator sheaths alone, respectively. There was no statistically significant difference between Evolution® and TightRailTM groups in complete procedural (86.9% vs. 84.2%, p>0.05) and clinical success (90.6% vs. 80%, p>0.05) (Figure 1). Rescue snaring system was used because of lead fracture in 29 patients (9.6%) (11.2% in Evolution® vs. 8.2% in TightRailTM group, p=0.286). Totally, clinical success was achieved in 298 patients (98.6%) using both mechanical dilator sheath and snaring system (99.2% for Evolution® group vs. 98% for TightRailTM group). Rescue snaring system was only failed in one patient in the TightRailTM group due to trapping of 3 electrodes in the right femoral vein that were extracted surgically. Three patients failed with mechanical dilator sheaths without using rescue snaring system (n=2 in TightRailTM group vs. n=1 in Evolution® group). Among 2 patients in the TightRailTM group, the procedure was terminated in one patient due to the loss of consciousness during TLE in whom intracranial hemorrhage and subsequent death was observed, and in another patient the electrode was left in the right ventricle due to failure with rescue tools. In one patient in the Evolution® group, a major vascular injury that required surgical intervention developed, and the one electrode was extracted completely by surgically. Major and minor complications were observed in 2.3% (n=7) and 9.3% (28) of patients which were similar among study groups (p>0.05) (Figure 2 & 3). Efficacy and safety outcome data of the study groups were shown in Table 3.
At median 36.6 (0.2-117.5) months follow-up duration [74.6 (0.6-117.5) months for Evolution® group vs. 25.8 (0.2-63.5) months for TightRailTM group, p=0.001], all-cause mortality was observed in 73 patients (25.6% in Evolution®group vs. 23.1 in TightRailTM group, p=0.616). When the survival curves of both groups are evaluated with log-rank test, there is a statistically significant difference between Evolution® and TightRailTM groups; because of the survival duration of Evolution® group was more prolonged than TightRailTM group due to that Evolution® group had longer follow-up (p<0.001) (Figure 4)
Baseline demographic, clinical, laboratory and procedural characteristics of patients according to the procedural failure and all-cause mortality were shown in Supplementary File 2 and 3, respectively . In multivariable logistic regression analysis, the presence of HFrEF (OR: 5.73, 95% C.I.: 1.49-22.0, p=0.011), increased baseline aPTT level (OR: 1.09, 95% C.I.: 1.03-1.17, p=0.006) and baseline GFR level <60 ml/min/1.73m2 (OR: 2.81, 95% C.I.: 1.12-7.08, p=0.028) were associated with a higher risk of all-cause mortality. Furthermore, baseline leukocyte count (OR: 1.20, 95% C.I.: 1.02-1.42, p=0.028), number of extracted leads (OR: 2.91, 95% C.I.: 1.45-5.86, p=0.003) and lead dwell time (OR: 1.08, 95% C.I.: 1.003-1.162, p=0.041) were found as independent predictors of TLE procedural failure.