Author + information
- Received April 4, 2017
- Revision received July 25, 2017
- Accepted July 27, 2017
- Published online November 30, 2017.
- Faris Khan, MD, MSa,∗ (, )
- Gustaf Sverin, MDa,
- Ulrika Birgersdotter-Green, MDa,
- Jennifer P. Miller, BS, BA, MBAb,
- Gautam Lalani, MDa,
- Travis Pollema, MDc and
- Victor Pretorius, MBChBc
- aSection of Electrophysiology, Division of Cardiology, Department of Medicine, School of Medicine, University of California, San Diego, La Jolla, California
- bCRHF, Medtronic, Mounds View, Minnesota
- cDivision of Cardiothoracic Surgery, Department of Surgery, University of California, San Diego, School of Medicine, La Jolla, California
- ↵∗Address for correspondence:
Dr. Faris Khan, Section of Electrophysiology, Division of Cardiology, Department of Medicine, University of California San Diego School of Medicine, UCSD Medical Center, 9454 Medical Center Drive, 3rd Floor, Room 3E-313; La Jolla, California 92037.
Objectives The study sought to assess the risk of collateral lead damage during cardiac implantable electronic device extraction.
Background With the increasing numbers of cardiovascular implantable electronic devices, there has been an increase in the number of percutaneous device and lead extractions. It is unknown how often collateral damage (defined as the need for unintended lead extraction, or loss of lead’s integrity or dislodgement) occurs in the planned retained leads.
Methods In this retrospective study, 108 patients who underwent incomplete cardiovascular implantable electronic device removal at the University of California, San Diego from September 2010 to September 2015 were included. The authors established the integrity of previously functioning leads at the end of each procedure as well as on follow-up visits using parameters including lead impedance change, threshold change, drop in P- or R-wave signal amplitude, or presence of lead noise.
Results Only 4 of 143 leads (2.7%) were found to have collateral damage. One right atrial (RA) lead had a clear insulation break, the second RA lead was found dislodged, and the third RA had a constant noise. The right ventricular lead was found to have a new high pacing threshold. Collateral lead age, extracted lead implantation site, collateral lead implantation site, and mode of lead extraction (laser, traction, or rotational dilator) did not have a significant correlation with the outcome of collateral lead damage.
Conclusions Lead extraction can be performed safely; however, there is a small risk of damaging adjacent leads. Close follow-up is needed, especially for the first few months, to assess for the reconnected leads’ integrity.
The use of cardiac implantable electronic devices (CIEDs) has been steadily increasing over the last few years. In the United States alone, there are more than 3 million patients with CIEDs and roughly 400,000 implantations each year (1). With the increasing numbers of CIEDs, there has been an increase in the number of percutaneous device and lead extractions. In many cases, the indication for CIED removal is infection, which totals two-thirds of all extractions and necessitates removal of the entire system. With other indications such as lead malfunction or abandoned leads, 1 or several leads may be left in place. The majority of extractions are performed by electrophysiologists and cardiac surgeons (2). There are several techniques and tools available for lead extraction including simple traction, traction with devices, mechanical sheaths, laser sheaths, electrosurgical sheaths, rotating threaded-tip sheaths, extraction snares, and telescoping sheaths (3). Binding sites are commonly encountered during extraction, with the most common binding sites being the venous entry, the subclavian vein, the superior vena cava, the right atrium (RA), and the tricuspid valve (2). In addition, lead-to-lead binding also occurs; however, there is a paucity of data regarding collateral damage (defined as the need for unintended lead extraction, or loss of lead integrity or dislodgement) that occurs in the retained leads. We conducted this study to assess for collateral lead damage.
A retrospective study of all patients who underwent incomplete CIED removal at the University of California, San Diego (UCSD) Health System from September 2010 to September 2015 was performed for a total of 108 patients. Electronic medical records were analyzed for baseline demographics, clinical characteristics including ischemic or nonischemic cardiomyopathy, left ventricular ejection fraction when available, comorbid conditions, oral anticoagulants, types of devices, extracted leads’ position, method of extraction, and procedural complications, if any. The integrity of previously functioning leads was established before the extraction and after the extraction prior to skin closure. The CIED system was then interrogated in standard follow-up and data collected up to 12 months of follow-up. Clinically relevant changes in lead parameters including impedance, threshold, and P- and R-wave amplitudes were assessed and documented (Table 1). Presence of lead noise was also assessed before and after the lead extraction as well as on follow-up visits up to 12 months. The study was approved by the UCSD institutional review board. The data were extracted by 2 coauthors and checked for interobserver variation by random assignment of 5% of cases for cross checking of extracted data and accuracy. The patients were excluded if follow-up data were not available (either in UCSD health records or with patients’ referring physicians) or if the complete CIED system was extracted. Leads from all main manufactures were represented among the patients included in the study.
All elective lead extraction patients underwent gated computed tomography scanning of the chest and a CXR within 2 weeks of their procedure date. Patients who were referred from other facilities underwent the studies on admission. Their CIEDs were interrogated before the procedure to determine pacemaker dependence and information was obtained regarding lead type, lead duration, and fixation mechanism (active vs. passive). This information was also obtained for any abandoned leads when possible. The procedure was conducted in a hybrid operating room, and the extraction team consisted of an electrophysiologist, cardiothoracic surgeon, anesthesiologist, cardiac electrophysiology fellow, device company representative, and operation room nursing team.
All patients underwent 5-F arterial sheath insertion in the femoral artery and 6-F venous sheath in the femoral vein, in addition to an internal jugular vein central line placement. If the patient was pacer dependent, temporary venous pacing was established via the femoral venous sheath. Tachytherapies were disabled for implantable cardioverter-defibrillators (ICDs) and rate response was turned off with the device programmed to VVI 40. A transesophageal echocardiogram probe was placed for continuous monitoring and high quality fluoroscopy was used throughout the procedure. PEAK PlasmaBlade (Medtronic, Minneapolis, Minnesota) was used for tissue dissection to reduce the risk of thermal injury to the surrounding structures including CIED leads (4). Regular stylets, locking stylets, clearing stylets, laser sheaths (Spectranetics, Colorado Springs, Colorado), and controlled mechanical rotational sheaths (Cook Medical, Bloomington, Indiana) were used where applicable. After opening the CIED pocket, the pulse generator was removed and regular stylets were placed in the leads to disengage the active fixation mechanism (when applicable) while applying traction (a stiff stylet was also placed in the collateral leads to reduce lead buckling during the extraction). The next step was to exchange the nonlocking stylet with a locking stylet and advance a laser sheath (along with an outer sheath) over the lead. The size of the sheath was determined by the size and type of lead (pacer lead, single- or dual-coil ICD lead) and was upsized if excessive resistance was encountered per the primary extractor’s discretion. If there was significant binding or evidence of significant calcifications, a mechanical rotational sheath was used as needed. After the lead was removed, transesophageal echocardiogram was used to assess for any effusion and all leads were sent to pathology for analysis.
We expressed continuous variables as mean ± SD and categorical variables as percentages. A Student t test with unequal variances and Fisher exact test were used to assess the association between baseline variables shown in Table 2. To assess the impact of various covariates on lead malfunction, we used logistic regression analysis and Fisher’s exact test as shown in Table 3. The following models were assessed using generalized estimating equations.
Model 1. Lead malfunction (binary outcome) with extracted lead site as the predictor variable (robust clustered for the unique identifier for each individual patient).
Model 2. Lead malfunction (binary outcome) with the mode of removal (extraction) as the predictor variable (robust clustered for the unique identifier for each individual patient).
Model 3. Lead malfunction (binary outcome) with the retained lead age as the predictor variable (robust clustered for the unique identifier for each individual patient).
Model 4. Lead malfunction (binary outcome) with the retained lead site as the predictor variable (robust clustered for the unique identifier for each individual patient).
The results were analyzed using Stata software version 13.1 (StataCorp LP, College Station, Texas).
Of the 350 patients who underwent CIED extraction from September 2010 to September 2015, 242 patients were excluded (133 were excluded due to infectious etiology and 109 either due to loss of follow-up [n = 22] or complete CIED system extraction [n = 87] for any reason other than infectious etiology). The remaining 108 patients who underwent incomplete CIED system extraction were included in the study. A total of 131 leads were extracted and a total of 143 leads were analyzed for any collateral damage. Median follow-up period was 5 months. We used the laser sheath in the majority of our extractions (117 of 131 extracted leads). There was no significant difference in baseline characteristics of patients who did or did not have collateral lead damage except for permanent pacemaker and their dependence on permanent pacemaker (Table 2). Immediate intra- or post-operative complications (other than collateral lead damage) were seen in 4 patients (3.7%). One patient developed a pocket hematoma requiring blood transfusion, 1 had a pneumothorax, 1 had a traumatic airway hemorrhage in the setting of thrombocytopenia, and 1 had a retained pacing lead fragment. Of the retained leads, 91 were in the RA, 22 were in the right ventricle (RV), and 30 were in the coronary sinus (Table 3). Only 4 leads, 3 RA leads and 1 RV lead, were found to have collateral damage (2.7%) (Tables 3 and 4). One RA lead had a clear insulation break as a result of an RV ICD lead extraction procedure, and was therefore extracted. The second RA lead was found to be dislodged at 1 month, and the third RA lead had a constant noise (Figure 1) at 4.5-month follow-up visit after undergoing RV pacemaker lead extraction. The lead noise was easily reproducible with isometrics, despite adjusting atrial lead sensitivity, suggesting insulation breach possibly in pectoral region. The RV pacemaker lead was found to have a new high pacing threshold after completion of an RA pacemaker lead extraction procedure and was replaced. On univariate analysis and Fisher’s exact tests, collateral lead age, extraction site (RA, RV, or coronary sinus), collateral lead implantation site, mode of lead extraction (laser vs. traction vs rotational dilator), and post-extraction complication did not have a significant correlation with the outcome of collateral lead damage (p > 0.05) (Table 3).
Collateral lead damage is an important complication that may not be recognized immediately but carries a significant clinical impact. This is especially true in pacer dependent patients or those who are at increased risk of sudden cardiac death from potentially fatal ventricular arrhythmias where the damaged lead may fail to deliver the appropriate therapy. In our study, we found the risk of collateral lead damage to be 2.7%. The study by Nichols et al. (5) using claims data reported 0.9% overall incidence of collateral lead damage within 1 year of a CIED replacement procedure. However, their study had several limitations including inability to differentiate patients who underwent pulse generator change only versus those who required lead replacement. The Danish registry showed 1.2% to 2.4% incidence of lead related complication; however, the reason for reimplantation was not defined (6). Other registry based retrospective studies also had similar limitations (7). In contrast, the study by McCanta et al. (8) showed that the collateral lead damage risk can be as high as 32%. However, their study comprised a pediatric population with median age of 15 years, whereas our study comprised an adult population with a median age of 65 years (mean 63 years of age). In addition, 9 of their 20 patients (42%) who underwent incomplete lead extraction had a history of congenital heart disease, which can increase the complexity of the procedure. In addition, they primarily utilized a mechanical rotational device (Evolution, Cook Medical) for 56% of the extracted leads. In our study we used a laser sheath (Spectranetics) for 89% of our extracted leads. Although their study concluded that duration of lead in situ before the extraction was significantly shorter in the lead that endured collateral damage, our study did not show a similar pattern (mean duration of 46.8 months in leads without collateral damage vs. 43.4 months in leads with collateral damage). Given the different population groups, we cannot compare the effectiveness of one extraction tool over another or the incidence of complication post-procedure. One of the limitations of their study was the lack of follow-up data beyond the initial hospital course after the sentinel lead extraction procedure, which we have addressed in our study. Two of 4 patients who had a collateral lead damage were identified on their follow-up visits. It is possible that the leads could have developed intrinsic malfunction (2); however, given the timeline, it appears unlikely and were most likely in the setting of their recent extraction procedure leading to a concomitant collateral lead damage.
As far as immediate collateral lead damage during an extraction procedure is concerned, it can occur either due to mechanical or thermal injury. In general, the CIED lead is constructed using up to 4 insulations; however, only the outer insulation provides the first line of defense for the retained lead. The 3 most common outer insulations are silicone elastomers, polyether urethane, and copolymer Optim (9). If the insulation is damaged due to mechanical cutting during an extraction procedure, silicone elastomers insulations will react differently than polyurethane because of their physical properties. Silicone, being a mechanically weaker material with a lower tensile strength, may continue to tear if cut, potentially leading to a wider breach in the insulation. The rate of progression of silicone elastomer–based lead damage depends on the lead’s initial condition, mechanical loading, and cycle frequency that the lead is undergoing relative to the initial damage. Also, the damage in low-stress areas will not propagate as quickly as the high-stress or high cycle bending areas. Polyurethane is much less susceptible to propagation of a tear or to experience lead–can or lead–lead wear (Table 4). Silicone–polyurethane copolymer Optim has demonstrated less abrasion resistance after implant or aging than polyurethanes (10–12). In our study, 3 of 4 collaterally damaged leads suffered an insulation breach. Two leads had a silicone–polyurethane copolymer Optim and 1 lead had a silicone elastomer insulation (Table 5).
Lead binding may also contribute to collateral damage either from mechanical damage due to proximity of the leads during extraction or due to abrasion from lead to lead contact. Abrasion may reduce the wall thickness of the outer insulation without breaching, and therefore, it may not be electrically detectable, but it can make the lead mechanically weaker. The rate of abrasion progression (lead on lead) depends on the amount of pressure and movement, age of the leads, type of outer insulation, and the type of materials in contact (polymer on polymer or polymer on metal).
Although polyurethanes are mechanically stronger than silicones, applied heat by electrocautery may lead to the insulation’s breach. Damage to the polyurethane and copolymer may occur as low as 10 W; however, silicone elastomers are less susceptible to applied heat by electrocautery (13). The most common site for the thermal injury is the pocket area during the CIED change-out. In contrast, radiofrequency energy produced by the plasma blade causes less damage to the polyurethanes compared with standard electrocautery (14). It is the practice of our group to use plasma blade for all such cases.
Another possible mechanism of late insulation breach can be due to the placement of a tightly cinched, nonabsorbable pursestring suture around a retained lead or leads to achieve hemostasis at the time of procedure.
Collateral lead damage can also occur as a result of passive manipulation of the lead during an extraction procedure by pre-mechanical loading. If the collateral lead is passively manipulated in such a way as to cause kinking or deformation of the coil or cables in the lead body, it may decrease the fatigue resistance of the cables resulting in early fractures. The rate of fracture would be dependent on the initial damage, amplitude of mechanical loading, and bending cycle count. A lead inside the heart will experience many more cycles than a lead section in the pectoral region. A lead in the shoulder region will experience greater mechanical loading due to skeletal movement of the shoulder region compared to mechanical loading in the heart (15). In addition, smaller leads have higher mechanical loads.
As the indications for lead extractions continue to evolve and expand, it is likely that more leads will be extracted for noninfectious causes. The most common causes of noninfectious incomplete lead extraction are lead fracture, malfunction, ICD lead recall, or upgrade of an existing CIED system (16,17).
Although operator experience can help to minimize the perioperative complications including collateral lead damage, it is extremely important to thoroughly check the retained leads’ integrity both immediately after an extraction procedure as well as on follow-up visits. It is the practice of our group to closely follow patients with incomplete CIED extraction for the first 6 to 12 months to avoid any potential lead malfunction as we believe that this is the vulnerable period where a newly diagnosed lead problem in the retained leads is most likely related to collateral lead damage from the previous extraction procedure.
First, this is a retrospective study with its potential inherent biases. Second, our center receives a number of externally referred complex cases with multiple abandoned leads, and as such our patient population may not be completely reflective of the cases encountered in lead extraction programs with smaller patient volume and the incidence of collateral lead damage in our setup would be expected to apply to lead extraction centers with similar experience and lead extraction techniques. Another possible limitation is that we do not have enough data to correlate collateral lead damage to type of lead. In addition, the period of follow-up in our study might not have allowed for all damaged leads to present clinically. Last, given the small number of cases who underwent the procedure using rotational mechanical extraction sheaths, we cannot compare the effectiveness of one extraction tool over another.
Lead extraction can be performed safely with the appropriately selected tools and techniques; however, there is a small risk of damaging adjacent leads. Close follow-up is needed post-procedure as well as for the first few months to assess for the reconnected leads’ integrity.
COMPETENCY IN MEDICAL KNOWLEDGE: Securing clinical training as well as proper hospital setup can minimize the risk of procedural complications. In addition, future research may be needed to identify new tools as well as to improve currently available tools and techniques to minimize the risk of complications associated with lead extraction procedure.
Ms. Miller is the senior engineering director at Medtronic. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
All authors attest they are in compliance with human studies committees and animal welfare regulations of the authors' institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information, visit the JACC: Clinical Electrophysiology author instructions page.
- Abbreviations and Acronyms
- cardiovascular implantable electronic device
- implantable cardioverter-defibrillator
- right atrium
- right ventricle
- University of California, San Diego
- Received April 4, 2017.
- Revision received July 25, 2017.
- Accepted July 27, 2017.
- 2017 American College of Cardiology Foundation
- Buch E.,
- Boyle N.G.,
- Belott P.H.
- Nichols C.I.,
- Vose J.G.,
- Mittal S.
- Krahn A.D.,
- Lee D.S.,
- Birnie D.,
- et al.
- McCanta A.C.,
- Tanel R.E.,
- Gralla J.,
- Runciman D.M.,
- Collins K.K.
- Cuvillier E.
- Chaffin K.A.,
- Wilson C.L.,
- Himes A.K.,
- et al.
- Kypta A.,
- Blessberger H.,
- Saleh K.,
- et al.
- Lau E.W.
- Zabek A.,
- Malecka B.,
- Haberka K.,
- et al.
- Bogossian H.,
- Miijic D.,
- Frommeyer G.,
- Winter J.