Author + information
- Received August 24, 2018
- Revision received January 10, 2019
- Accepted January 11, 2019
- Published online June 17, 2019.
- Gordon Ho, MDa,∗∗ (, )
- Prerana Bhatia, MDa,∗,
- Ishan Mehta, BSa,
- Timothy Maus, MDb,
- Swapnil Khoche, MDb,
- Travis Pollema, DOc,
- Victor Gert Pretorius, MBChBc and
- Ulrika Birgersdotter-Green, MDa
- aDepartment of Medicine, Division of Cardiac Electrophysiology, University of California-San Diego, La Jolla, California
- bDepartment of Anesthesiology, University of California-San Diego, La Jolla, California
- cDepartment of Cardiothoracic Surgery, University of California-San Diego, La Jolla, California
- ↵∗Address for correspondence:
Dr. Gordon Ho, Department of Medicine, Division of Cardiac Electrophysiology, University of California-San Diego, 3350 La Jolla Village Drive, 111A, San Diego, California 92161.
Objectives This study sought to prospectively evaluate the prevalence, risk factors, and short-term major clinical outcomes of mobile thrombus detected on transvenous leads in patients undergoing lead extraction.
Background The prevalence and clinical significance of thrombus on transvenous leads in patients undergoing lead extraction is not well characterized.
Methods Consecutive patients undergoing transvenous lead extraction for noninfectious indications were enrolled. Preoperative transesophageal echocardiograms were performed prospectively for all patients to examine for mobile thrombus. Anticoagulation was not started for thrombus unless other indications were present. Clinical endpoints of mortality and cardiovascular morbidity (symptomatic pulmonary embolism, myocardial infarction, or cerebrovascular accident) were assessed at a minimum of 2-month follow-up.
Results A total of 108 patients underwent lead extraction for noninfectious indications. Lead thrombi were detected in 20 (18.5%) patients and all were <2 cm. Clinical and lead characteristics were not associated with formation of lead thrombi, except for younger patient age. In patients with detected thrombi, there were no short-term deaths, symptomatic pulmonary embolisms, or myocardial infarctions, except 1 patient with a stroke 3 months after lead extraction (7% vs. 5%; p = 1.00). Median follow-up was 9 months.
Conclusions Mobile thrombi on transvenous leads are commonly found in patients referred for transvenous lead extraction and are rarely associated with acute major adverse outcomes. Careful extraction of leads with small incidentally detected thrombi can likely be performed without major acute clinical sequelae. Larger studies with longer follow-up are needed to further assess the long-term clinical significance of lead thrombi.
- cardiac implantable devices
- transesophageal echocardiography
- transvenous lead extraction
The prevalence of mobile intracardiac thrombi detected on transvenous leads from cardiac implantable electronic devices varies widely in published data, from 1.4% to 30% using a variety of imaging modalities (1–4) in different patient populations. The clinical significance of these incidentally detected thrombi is unclear, with some studies reporting a low incidence of pulmonary embolism (PE), but there are cases of patients with recurrent PE requiring surgical lead and thrombus extraction. Furthermore, it is unclear whether anticoagulation is indicated in these patients who had incidental thrombi without embolic sequelae are found.
In patients undergoing transvenous lead extraction, the prevalence and clinical outcome of incidental mobile lead thrombi has not been reported, and it is unknown whether these patients have a higher incidence of embolic events. We performed a prospective cohort study of consecutive patients referred for transvenous lead extraction to evaluate the prevalence and short-term major clinical outcomes of thrombus detected on transvenous leads at time of lead extraction.
Consecutive patients with prior cardiac implantable devices with transvenous leads who were referred for transvenous lead extraction at the University of California, San Diego, between March 2015 and December 2016 for noninfectious indications such as lead malfunction were enrolled into this prospective study. Inclusion criteria included any patient >18 years of age who was referred for transvenous lead extraction for any indication except infection, bacteremia, or pocket infection. Exclusion criteria included any patient who could not tolerate a transesophageal echocardiogram (TEE).
A routine preoperative TEE was performed prospectively for all patients on the day of the lead extraction procedure, focusing on all transvenous leads from the superior vena cava to endocardial insertion points to examine for mobile thrombus located in the intracardiac portion of the leads. A mobile thrombus was defined as any echodensity seen on the intracardiac portion of a lead that appeared distinct (sharp, irregular edges) from the lead material, could move freely (not be affixed to a vessel wall), and either move along with or be independent of the direction of lead movement. Care was taken to exclude artifact or venous occlusion in the superior vena cava. Figure 1 shows 2 representative examples of differing mobile thrombi; both move freely in the right atrium with the lead. Figure 1A and Online Video 1 show a large thrombus with significant paradoxical movement with the RV lead. Figure 1B and Online Video 2 show a thrombus affixed to the RA lead that moves with the lead with a small component moving paradoxically to the lead.
Then, lead extraction was performed according to standard clinical protocol using traction and laser or mechanical sheaths. Relevant baseline clinical and lead characteristics were recorded such as patient age, sex, body mass index, left ventricular ejection fraction, comorbid conditions, basic labs, anticoagulant use at time of extraction, number of leads, lead age, lead recall status, presence of defibrillation coil, abandoned leads, type of malfunction, lead chamber, insulation material, and manufacturer.
Clinical endpoints of mortality and cardiovascular morbidity (symptomatic PE, myocardial infarction [MI], or cerebrovascular accident [CVA]) were assessed at a minimum 2 month follow-up clinic visit or phone call. Clinically indicated pulmonary imaging was only performed if patients expressed symptoms of new dyspnea or chest pain after the lead extraction. Of note, given unclear optimal anticoagulation approach in context of limited retrospective studies (4,5), anticoagulation was not routinely started for an incidental finding of a small thrombus <2 cm, unless other clinical indications such as atrial fibrillation (AF) with high CHA2DS2-VASc (congestive heart failure, hypertension, age ≥75 years, diabetes mellitus, prior stroke or transient ischemic attack or thromboembolism, vascular disease, age 65 to 74 years, sex category) score were present.
Baseline characteristics and outcomes were analyzed using R Commander version 3.5.0 (R Foundation for Statistical Computing, Vienna, Austria). Analysis of variance, Welch’s 2-sample Student's t test, Pearson’s chi-square, and Fisher’s exact test were used for statistical analysis depending on the frequency and variance of the sample. Results are expressed as mean ± SD, median (interquartile range), or frequency and proportion, when applicable. A statistically significant difference had a p value of <0.05.
Incidence of incidental mobile lead thrombi
Of the 166 patients referred for transvenous lead extraction, 108 patients with 237 transvenous leads had a noninfectious indication and were included in the analysis. A thrombus was found on 29 (12.2%) of the leads among 20 (18.5%) patients. The mean thrombus size was 1.4 ± 0.4 cm, with the smallest thrombus 0.6 cm and the largest thrombus 1.9 cm. They were predominantly located in the right atrium compared with the right ventricle (86% vs. 14%; p < 0.001). Clinical characteristics associated with thrombus formation are outlined in Table 1. Each patient had between 1 and 5 transvenous leads implanted at the time of evaluation, with a mean of 2.2 ± 0.7 leads/patient. To support the absence of infection, patients with thrombi did not have higher white blood cell counts nor higher maximum temperature compared with the group without thrombi.
Clinical and procedural outcomes
The median length of follow-up was 9 (interquartile range: 2 to 14) months. There were 4 deaths in the patients without thrombus compared with no deaths in the patients with detected lead thrombus (4.5% vs. 0%; p = 1.00). Two of the deaths in patients without thrombus occurred due to postoperative superior vena cava tear, 1 was due to pneumonia, and the cause of the fourth death is unknown, as the patient died at home 7 months postoperatively. Major short-term adverse cardiovascular events (pulmonary embolus, MI, CVA, or transient ischemic attack) occurred in 1 patient with a lead thrombus (ischemic stroke) compared with 2 patients without lead thrombi (transient ischemic attack and MI; 5% vs. 2%; p = 0.50). The patient with lead thrombi who had an ischemic stroke was on warfarin before and after lead extraction with a subtherapeutic preoperative international normalized ratio of 1.4 at the time of stroke. Similarly, the composite endpoint (mortality, pulmonary embolus, MI, CVA, and transient ischemic attack) was not significantly different between patients with and without lead thrombi (7% vs. 5%; p = 1.00). Short-term adverse major clinical outcomes are summarized in Table 2.
Of the 20 patients with a detected lead thrombus, 18 (90%) patients underwent extraction of a lead with a thrombus on it. In the leads with thrombi that were extracted, 85% were extracted with a laser sheath and 15% were extracted with a mechanical sheath; none were extracted using traction only. For all patients, regardless of thrombus status, extraction sheaths were used in the majority of patients, 81% using laser sheaths, 6% using mechanical sheaths, and 13% using traction only. Operative outcomes, such as fluoroscopy time per transvenous lead extracted were not significantly different among patients with and without lead thrombus (11 ± 8 min vs. 14 ± 9 min; p = 0.26). There was also no significant difference in postoperative hospital length of stay.
Clinical characteristics associated with lead thrombi formation
Out of all the baseline clinical characteristics, only patient age was significant, with younger age as more likely to be associated with a lead thrombus (56 ± 19 years vs. 64 ± 14 years; p = 0.04). Table 3 displays all the clinical risk factors analyzed. There was no difference in baseline anticoagulant use in patients with thrombi compared with patients without lead thrombi (30% vs. 38%; p = 0.47). There was no difference in the incidence of thrombus formation between the patients treated with warfarin versus direct oral anticoagulants (14% vs. 17%; p = 1.00).
Lead characteristics associated with lead thrombi
No specific lead characteristics were associated with the formation of incidental mobile thrombi (Table 3). The functioning status of the leads and the reasons for extraction for each lead are illustrated in Figure 2. There were 66 (28%) leads that were not extracted. The other 171 (72%) leads were extracted for various indications, which are all displayed in Figure 2. The most common indications were device upgrade (19%), followed by lead fracture (16%) and malposition (9%). There was no statistically significant association of thrombus with functioning versus malfunctioning lead (9% with thrombus vs. 15%; p = 0.19) or by extraction indication (p = 0.06). No specific insulating material was associated with presence of thrombus. Lead body insulation materials were comprised of 3 main groups—silicone, polyurethane, and combination. The combination materials included silicone and polyurethane copolymers (n = 44), proximal polyurethane and distal silicone (n = 3), and silicone with polyurethane sleeve (n = 6). Leads recalled for various reasons were issued on 20 (9%) of the leads at the time of this study, and overall, these recalled leads were not associated with a higher prevalence of lead thrombus (10% vs. 8%; p = 0.61). However, Riata leads (St. Jude Medical, St. Paul, Minnesota) appeared to be more thrombogenic, with 3 of 6 total Riata leads (50%) having thrombus. Lead thrombus was also associated with neither abandoned status of a lead (7% vs. 3%; p = 0.35) nor passive fixation (17% vs. 15%; p = 0.56).
There are 3 key findings in this study. First, there was a high prevalence (18.5%) of incidental mobile lead thrombi detected prospectively in a consecutive cohort of patients undergoing transvenous lead extraction. Second, we found that patients who undergo lead extraction of leads with small mobile lead thrombi experienced minimal short-term major adverse outcomes despite not routinely anticoagulating patients with lead thrombi. Last, we found that lead characteristics were not associated with the formation of incidental lead thrombus.
Insights into the prevalence of incidental lead thrombi
This study is the largest prospective series to date of 108 patients and is the only study to evaluate the prevalence of mobile thrombi and outcomes in patients undergoing lead extraction. Incidental thrombi were discovered in 20 (18.5%) patients. To our knowledge, only 2 prior studies have prospectively evaluated the prevalence of mobile lead thrombi, the first study in 66 patients 6 months after cardiac implantable electronic device implantation using TEE and venography (1) and the second study in 86 patients undergoing AF or ventricular tachycardia ablation using intracardiac electrocardiography (4). These studies identified a high incidence of thrombi in 20% and 30% of patients, respectively, similar to our findings. Conversely, 3 prior retrospective studies reported a prevalence of lead thrombi ranging from 1.4% to 12% (2,3,6), which is lower than that reported in the prospective studies, likely due to undersampling from retrospective study methodologies.
Short-term clinical outcomes in patients with incidental mobile lead thrombi
During the follow-up period (median 9 months), the presence of a lead thrombus was not associated with increased short-term combined adverse clinical outcomes (all-cause mortality, clinical PE, CVA, or MI). No deaths or symptomatic PE occurred in patients with a lead thrombus. Only 1 patient with a lead thrombus experienced an ischemic stroke 3 months after transvenous lead extraction of the lead with a thrombus. This patient also had AF and subtherapeutic international normalized ratio, which were a more likely etiology of his CVA than embolization of the lead thrombus. No patients with detected lead thrombus had a history of PE. Similar to our findings, the largest retrospective study to date of 1,833 patients undergoing ablation found no difference in clinical outcomes in patients with and without lead thrombi (3).
Short-term clinical outcomes in patients undergoing extraction of leads with a thrombus
The majority of patients (90%) with any detected lead thrombi underwent extraction of at least 1 lead to which a thrombus was attached. In all leads with thrombi that were extracted, extraction was performed using either a laser or mechanical sheath. Theoretically, it is possible that mobile thrombi are sheared off by the extraction sheath during extraction, and may be embolized, although this aspect was not specifically studied. However, no patients with detected lead thrombi acutely developed symptomatic PE after extraction. Our findings may suggest that despite this embolization risk, use of extraction sheaths is associated with minimal acute adverse outcomes. However, our study is underpowered to address this specific question due to low event rates, and our findings must be interpreted in a patient-specific approach.
It is important to note that although there were minimal acute major adverse events in patients with lead thrombi, the long-term effects of thrombi are unknown, particularly with subacute pulmonary thromboembolic disease. In a prospective study by Supple et al. (4) of 86 patients undergoing AF or ventricular tachycardia ablation, they found a high prevalence of patients (30%) with mobile lead thrombi, and these patients were found to have a higher pulmonary arterial systolic pressure by echocardiogram. Similar to our findings, none of their patients had a history of symptomatic clinical PE, and supports the hypothesis that small embolized thrombi may not be acutely hemodynamically significant. However, the long-term effects of thrombi are unknown, and the signal of elevated pulmonary arterial pressures may suggest the development of subacute pulmonary thromboembolic disease. Larger studies with longer follow-up in patients with lead thrombi are needed to assess the risk of developing chronic pulmonary thromboembolic disease.
Insights into the use of anticoagulation in patients with incidentally discovered lead thrombus
The effect of anticoagulation on lead-associated thrombus formation and clinical outcomes remains unclear. Numerous studies indicate that mobile and fixed thrombi can occur on transvenous leads despite patients being anticoagulated (2–4). The risk of pulmonary and systematic thromboemboli associated with implanted cardiac devices has not shown to be different among patients on anticoagulants (7–9). Although the majority of lead-associated thrombi can resolve with intensification or initiation of anticoagulation (2,6), it is unclear whether anticoagulation improves clinical outcomes in patients with incidentally discovered mobile lead thrombi.
In our study, no patient with an incidentally discovered lead thrombus was started on anticoagulation if he or she did not have any other indications for anticoagulation such as AF with high CHA2DS2VASc score. Although larger, randomized studies are needed to confirm this, these findings suggest that patients with lead thrombi may not need to be anticoagulated. Furthermore, all thrombi detected in our study were 2 cm or less, and thus these findings can only apply to patients with small lead thrombi.
Insights into clinical and lead characteristics associated with lead thrombi
There were no significant clinical characteristics associated with thrombus formation, except patient age; more lead thrombi were found in younger patients for an unknown reason. Otherwise, clinical comorbidities such as AF were not associated with lead thrombi, which is consistent with prior prospective studies (1,4,10,11), but inconsistent with 1 larger retrospective older study that found an association of AF with thrombus (2). This finding may be due to the fact that contemporary patients with AF are usually appropriately anticoagulated; in our study, 78% of patients with AF were on anticoagulation. Surprisingly, lead thrombi formation was not associated with anticoagulation status, which is consistent with prior studies (4).
Interestingly, lead characteristics such as number of leads, lead insulation material, cardiac chamber of lead fixation, abandoned lead status, and lead age were not associated with mobile thrombus formation on intracardiac transvenous leads. Although there have been small conflicting studies associating certain risk factors associated with venous occlusion in the subclavian veins such as number of leads (10–12), our findings support the assertion that differences in lead design and configuration may not influence formation of mobile thrombi in the intracardiac portion of transvenous leads. Finally, we found that abandoned leads were not associated with intracardiac lead thrombi formation. These aspects have not been evaluated comprehensively in prior work.
In our unique study population of patients referred for lead extraction, a high proportion of patients (51%) had functional problems with their leads. Despite this, the prevalence of lead thrombi was still generally lower than that in prior prospective studies, and this supports our finding that lead malfunction is not associated with formation of thrombi. Although overall Food and Drug Administration recall status was not associated with thrombus formation, there was a low number of Riata leads or externalized cables in our study population, in which 1 would expect higher risk of thrombus formation (8). Only 6 of the 20 recalled leads were Riata leads or externalized cables, and as expected, a thrombus was detected on high proportion of these leads (n = 3, 50%).
Several limitations of our study are noted, including the lack of statistical power to detect differences in clinical outcomes, lack of definitive pulmonary imaging, and lack of lead thrombi histology. First, although this was the largest prospective study to date, it still lacked statistical power to confidently report no difference in clinical outcomes between patients with and without lead thrombus. Given the small event rate, it would take a study population that is around 15 times our study population to achieve statistical power, which is not feasible for a prospective study. To improve the statistical power of our study, we reported a combined clinical adverse event rate. Of all retrospective and prospective studies reported in published reports, our study was the third largest, after 2 retrospective studies (2,3) with the largest study including 1,833 patients, which also found no difference in clinical outcomes between patients with and without lead thrombi, consistent with our findings.
Second, our study follow-up did not include routine performance of imaging studies to detect PE for all patients; these studies were only performed in patients who were symptomatic with clinical suspicion for PE. Given the uncertain clinical significance of asymptomatic PE suggested in prior studies (3,4), it was felt that these imaging studies were not necessarily indicated and did carry risk, in particular the adverse effects of potentially unnecessary radiation and risks of unnecessary anticoagulation in patients who are asymptomatic. Consistent with our methodology, routine pulmonary imaging was not performed in the other 5 studies of mobile thrombi in live patients (1–4,6).
Third, our study protocol was not designed to retrieve thrombi seen on TEE to verify the presence of thrombus detected on TEE. Despite performing extraction of the leads with thrombus, it was difficult to identify and collect thrombi from the leads during extraction, as they usually were not present on the lead after removal from the extraction sheath. Thus, histologic examination was not possible to elucidate the structure of the thrombus. Although the phenomenon of “ghosts” is well known after a lead extraction procedure, we did not include findings of mobile echogenic material noted after lead extraction. However, this study did not exclude patients that may have undergone prior lead extraction procedures, so visualization of ghosts remains a possibility.
This prospective study identified a high incidence of mobile intracardiac thrombi on transvenous leads in patients. No particular lead characteristic increased risk of thrombi formation. Extraction of these leads with small thrombi was rarely associated with short-term major adverse clinical outcomes despite not starting or intensifying anticoagulation. Larger studies with longer follow-up are needed to assess the long-term clinical effects of incidental lead thrombi.
COMPETENCY IN MEDICAL KNOWLEDGE: The findings from this original research report support lifelong learning skills and enhances several clinical competencies for professional caregivers. This study informs physicians that the incidental finding of thrombus on a transvenous lead can be expected in about 19% of patients presenting for lead extraction, and the careful extraction of leads with small thrombi under 2 cm diameter may be performed without acute major sequelae. Furthermore, clinical characteristics or lead characteristics rarely predict formation of lead thrombi. These findings add to the body of medical knowledge a series of patients who had successful extraction of leads with thrombi has been performed with extraction sheaths without major adverse events. When a clinician encounters clinical scenarios in which thrombi is detected before a lead extraction procedure, one can expect a low incidence of acute adverse events. Also depending on the patient, a strategy of not starting anticoagulation for a finding of incidental lead thrombus is not associated with acute adverse events.
TRANSLATIONAL OUTLOOK: This study produced several important clinical findings focusing on a specific patient population of lead extraction patients, but the results raise new interesting questions regarding thrombogenicity of transvenous leads. First, although extraction of small thrombi did not seem to be associated with major acute adverse events, the long-term effects of incidental lead thrombi are still unknown, such as the development of long-term subacute disease such as chronic thromboembolic pulmonary hypertension. Clinical studies are currently underway to address this particular question. Second, this study found that lead thrombi is rather common, consistent with prior studies. There is a paucity of published reports in bioengineering and materials science assessing factors in lead design that lead to thrombogenicity. Thus, future basic science studies are needed to evaluate improved structural and material lead design to reduce thrombogenicity of transvenous leads, such as lead structure to reduce turbulent flow particularly on uneven lead surfaces and improved biocompatibility of lead insulation material. Third, advancements in technology may provide tools to predict the personalized risk of lead thrombus. Computational modeling of blood flow in the heart and advanced dynamic cardiac imaging could potentially be used to predict where lead thrombi may form; in a similar fashion, these same tools could potentially predict where binding sites may potentially form and cause difficulties for lead extraction.
↵∗ Drs. Ho and Bhatia contributed equally to this paper and are joint first authors.
Dr. Ho has received fellowship support from Abbott, Boston Scientific, Biotronik, and Medtronic; and owns equity in Vektor Inc. for work unrelated to this manuscript. 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
- atrial fibrillation
- cerebrovascular accident
- myocardial infarction
- pulmonary embolism
- transesophageal echocardiogram
- Received August 24, 2018.
- Revision received January 10, 2019.
- Accepted January 11, 2019.
- Sugrue A.,
- DeSimone C.V.,
- Lenz C.J.,
- Packer D.L.,
- Asirvatham S.J.
- Supple G.E.,
- Ren J.-F.,
- Zado E.S.,
- Marchlinski F.E.
- Pedersen S.B.,
- Hjortshøj S.P.,
- Bøtker H.E.,
- et al.
- Khairy P.,
- Landzberg M.J.,
- Gatzoulis M.A.,
- et al.
- Noheria A.,
- Ponamgi S.P.,
- Desimone C.V.,
- et al.
- Bar-Cohen Y.,
- Berul C.I.,
- Alexander M.E.,
- et al.