Stroke in patients with cardiovascular implantable electronic device infection undergoing transvenous lead removal

      Background

      Stroke can be a devastating complication in patients with cardiovascular implantable electronic device (CIED) infection. Paradoxical septic embolism can occur in the presence of device leads and patent foramen ovale (PFO) via embolic dislodgment during transvenous lead removal (TLR).

      Objective

      The purpose of this study was to examine stroke and its associated factors in patients undergoing TLR for CIED infection.

      Methods

      We performed a retrospective analysis of all patients undergoing TLR for CIED infection from January 1, 2000, to July 30, 2017, from all 3 tertiary referral centers at the Mayo Clinic (Rochester, Phoenix, and Jacksonville). The primary outcome was stroke and was further categorized into preprocedural and postprocedural stroke. Associated risk factors were analyzed.

      Results

      A total of 774 patients (mean age 67.6 ± 14.9 years) underwent TLR for CIED infection. The stroke rate in this cohort was 1.9% (95% confidence interval [CI] 1.1%–3.2%). The preprocedural and postprocedural stroke rate was 0.9% (95% CI 0.4%–1.9%) and 1.0% (95% CI 0.4%–2.0%), respectively. PFOs were identified in 46.7% of patients with stroke and in 12.9% of patients without stroke, and were independently associated with stroke (P = .0002). This was especially in patients with right-sided vegetations with right-to-left shunting (odds ratio 6.4; 95% CI 1.3–31.0; P = .022).

      Conclusion

      In patients with CIED infection undergoing TLR, the presence of PFO, especially with right-sided vegetation with right-to-left shunting, was associated with an increased risk of stroke. This finding suggests that PFO screening before TLR warrants meticulous attention.

      Keywords

      Attention HRS Members and Journal Subscribers

      Visit the HRS Learning Center at www.hrsonline.org/HRJ-CME to earn CME credit through an online activity related to this article. Certificates are available for immediate access upon successful completion of the activity.

      Introduction

      The incidence of cardiovascular implantable electronic device (CIED) infection is increasing.
      • Greenspon A.J.
      • Patel J.D.
      • Lau E.
      • Ochoa J.A.
      • Frisch D.R.
      • Ho R.T.
      • Pavri B.B.
      • Kurtz S.M.
      16-year trends in the infection burden for pacemakers and implantable cardioverter-defibrillators in the United States 1993 to 2008.
      • Baddour L.M.
      • Epstein A.E.
      • Erickson C.C.
      • et al.
      Update on cardiovascular implantable electronic device infections and their management: a scientific statement from the American Heart Association.
      The mainstay of treatment of CIED infection is targeted antimicrobial therapy along with removal of device hardware, in most cases by transvenous lead removal (TLR). Unfortunately, among all procedures performed in the electrophysiology laboratory, TLR carries the highest risk of stroke.
      • Lee J.Z.
      • Ling J.
      • Diehl N.N.
      • et al.
      Mortality and cerebrovascular events after heart rhythm disorder management procedures.
      Patients with CIED infection often have more medical comorbidities,
      • Klug D.
      • Balde M.
      • Pavin D.
      • et al.
      Risk factors related to infections of implanted pacemakers and cardioverter-defibrillators: results of a large prospective study.
      and stroke can be a devastating complication in this group of patients.
      Previous studies have shown that pulmonary embolism may occur in patients with CIED infection before or after the TLR procedure.
      • Grammes J.A.
      • Schulze C.M.
      • Al-Bataineh M.
      • Yesenosky G.A.
      • Saari C.S.
      • Vrabel M.J.
      • Horrow J.
      • Chowdhury M.
      • Fontaine J.M.
      • Kutalek S.P.
      Percutaneous pacemaker and implantable cardioverter-defibrillator lead extraction in 100 patients with intracardiac vegetations defined by transesophageal echocardiogram.
      • Meier-Ewert H.K.
      • Gray M.E.
      • John R.M.
      Endocardial pacemaker or defibrillator leads with infected vegetations: a single-center experience and consequences of transvenous extraction.
      Stroke in patients with CIED infection undergoing TLR has not been evaluated in a systematic fashion. Therefore, we sought to assess all strokes in patients with CIED infection undergoing TLR and to further examine factors associated with stroke in this population.

      Methods

      We performed a retrospective cohort study of all patients undergoing TLR for CIED infection at the Mayo Clinic Enterprise Heart Rhythm Practice. The study included patients at the 3 academic campuses in Rochester, Minnesota; Phoenix, Arizona; and Jacksonville, Florida. The study was approved by the Mayo Clinic Institutional Review Board.

      Study population

      Adult patients (age ≥18 years) who underwent TLR for CIED infection between January 1, 2000, and July 30, 2017, were included in the study (Figure 1). CIED infection was defined using International Classification of Diseases, Ninth Revision (ICD-9) and International Classification of Diseases, Tenth Revision (ICD-10) codes for device infection, bacteremia, and endocarditis (Supplemental Table S1). These diagnostic codes were selected based on definitions of CIED infection in the Heart Rhythm Society expert consensus statement on CIED lead management and extraction.
      • Kusumoto F.M.
      • Schoenfeld M.H.
      • Wilkoff B.L.
      • et al.
      2017 HRS expert consensus statement on cardiovascular implantable electronic device lead management and extraction.
      TLR procedures were identified using Current Procedural Terminology (CPT) codes 33234, 33235, and 33244, which identifies transvenous lead explant and extraction procedures.
      • Kusumoto F.M.
      • Schoenfeld M.H.
      • Wilkoff B.L.
      • et al.
      2017 HRS expert consensus statement on cardiovascular implantable electronic device lead management and extraction.
      We excluded patients with concurrent procedural codes for orthotropic heart transplantation, ventricular assist device placement, or total artificial heart implantation, because these patients frequently have concomitant device removal procedural codes.
      Figure thumbnail gr1
      Figure 1Patient inclusion and categorization CONSORT (Consolidated Standards of Reporting Trials) flow diagram. CIED = cardiovascular implantable electronic device; VAD = ventricular assist device.

      Data collection

      Patient demographic parameters and comorbidities were collected. Comorbidities were identified using ICD-9 and ICD-10 codes (Supplemental Table S2).
      All transthoracic echocardiogram (TTE) and transesophageal echocardiogram (TEE) reports were reviewed to collect predefined echocardiographic parameters (vegetation, patent foramen ovale [PFO], shunt direction). Two study personnel independently reviewed the echocardiographic reports. Data on vegetation and shunt direction were collected based on echocardiogram performed before or during the TLR procedure during the same hospitalization. Presence of a PFO was determined by review of all echocardiogram reports within a time period of 6 months before to 6 months after the TLR procedure. Diagnosis of PFO used standard published criteria as used at our institution.
      • Silvestry F.E.
      • Cohen M.S.
      • Armsby L.B.
      • et al.
      Guidelines for the echocardiographic assessment of atrial septal defect and patent foramen ovale: from the American Society of Echocardiography and Society for Cardiac Angiography and Interventions.
      Vegetations were identified based on visualization of mobile echodense masses and were further categorized into right-sided and left-sided vegetations.

      Lead removal protocol

      After CIED infection was confirmed, patients underwent transvenous lead explant or extraction.
      • Kusumoto F.M.
      • Schoenfeld M.H.
      • Wilkoff B.L.
      • et al.
      2017 HRS expert consensus statement on cardiovascular implantable electronic device lead management and extraction.
      The procedures were performed under monitored anesthesia care or under general anesthesia, in the electrophysiology laboratory or the hybrid operating room, with cardiovascular surgery backup. In device leads that were extracted >1 year after implantation, specialized tools, such as locking stylets, mechanical sheaths, and laser sheaths, were used to extract the device. Use of intraprocedural TEE was based on the discretion of the operator. In patients who were on anticoagulation, international normalized ratio was permitted to return to normal and direct-acting oral anticoagulants were withheld per guidelines before the procedure. In patients who were bridged with therapeutic heparin, heparin was stopped 6 hours before the procedure. Anticoagulation was restarted after the procedure based on the patient’s underlying stroke risk, balanced against the bleeding risk from a débrided pocket.

      Assessment of outcomes

      The primary outcome was stroke during the period of illness (defined as 4 weeks before TLR procedure until hospital discharge). Stroke events were identified based on individual chart review to ensure that the diagnosis was made according to the standardized definition from the Valve Academic Research Consortium, which defined stroke based on occurrence of acute onset of a focal or global neurologic deficit with at least 1 neurologic sign or symptom consistent with stroke, lasting >24 hours (or less if available neuroimaging documented a new infarct, or the neurologic deficit resulted in death), no readily identifiable nonstroke cause for clinical presentation, and confirmation by either a neurology specialist or neuroimaging procedure.
      • Kappetein A.P.
      • Head S.J.
      • Genereux P.
      • et al.
      Updated standardized endpoint definitions for transcatheter aortic valve implantation: the Valve Academic Research Consortium-2 consensus document.
      Transient ischemic attack was defined as focal or global neurologic deficit lasting <24 hours, with neuroimaging not demonstrating a new infarct. We did not include hemorrhagic stroke.
      Strokes were adjudicated independently by 2 reviewers, with a third reviewer casting the deciding vote in cases of disagreements. Patients with possible stroke were initially identified using ICD-9 and ICD-10 codes (Supplemental Table S3) before individual chart review. Stroke was further categorized into preprocedural, if the event occurred within 4 weeks before the TLR procedure, and postprocedural, if the event occurred during or after the procedure until hospital discharge.
      Secondary outcomes were duration of hospitalization and 30-day mortality, defined as death within 30 days of the TLR procedure.

      Statistical analysis

      Categorical variables are expressed as percentages, whereas continuous variables were expressed as mean ± SD. Baseline patient demographics were compared between stroke and nonstroke groups by using the χ2 test for categorical variables and the Student t test for continuous variables. Wilcoxon rank sum test was used to compare median length of stay. Logistic regression was used to look at the association of various factors to stroke. Variables that showed a univariate relationship with stroke were also put into a multivariate model to assess their relative importance to stroke. All P values were 2-sided, and P <.05 was considered significant. In cases of multiple hypothesis testing, Bonferroni correction was used.
      • Streiner D.L.
      • Norman G.R.
      Correction for multiple testing: is there a resolution?.
      Kappa was calculated to determine the degree of agreement between independent reviewers on stroke outcome. All statistical analysis was performed using SAS version 9.4 (SAS Institute, Cary, NC).

      Results

      The derivation of the patient cohort is summarized in Figure 1. A total of 774 patients (mean age 67.6 ± 14.9; 26% female) underwent TLR for CIED infection during the 17-year study period.
      There were no significant differences in demographics and comorbidities between patients with and those without stroke (Table 1). However, patients who had a stroke were more likely to have a history of stroke (33.3% vs 7.4%; P = .0002).
      Table 1Characteristics of patients with CIED infections who underwent transvenous lead removal
      All stroke (n = 15)Preprocedural stroke (n = 7)Postprocedural stroke (n = 8)No stroke (n = 759)P value
      Comparing all stroke and no stroke.
      Age (yr)67.4 ± 15.862.7 ± 17.671.5 ± 13.867.6 ± 14.9.9
      Female5 (33.3)3 (42.9)2 (25.0)198 (26.1).56
      LVEF (%)53.1 ± 13.358 ± 6.248.3 ± 17.146.4 ± 16.5.14
      BMI (kg/m2)28.7 ± 7.129.6 ± 9.827.8 ± 3.529.5 ± 6.8.52
      Comorbidity
       Atrial fibrillation9 (60.0)4 (57.1)5 (62.5)376 (49.5).45
       Congestive heart failure8 (53.3)4 (57.1)4 (50.0)506 (66.7).28
       Hypertension10 (66.7)5 (71.4)5 (62.5)581 (76.5).36
       Coronary artery disease13 (86.7)6 (85.7)7 (87.5)523 (68.9).17
       Diabetes mellitus3 (20.0)2 (28.6)1 (12.5)294 (38.7).18
       Chronic kidney disease4 (26.7)1 (14.3)3 (37.5)292 (38.5).43
       Chronic obstructive pulmonary disease3 (20.0)2 (28.6)1 (12.5)64 (8.4).13
       History of stroke5 (33.3)3 (42.9)2 (25.0)56 (7.4).0002
      Echocardiographic characteristics
       Patent foramen ovale7 (46.7)2 (28.6)5 (62.5)98 (12.9).002
       Right-sided vegetations7 (46.7)5 (71.4)2 (25.0)232 (30.6).29
       Left-sided vegetations6 (40.0)5 (71.4)1 (12.5)80 (10.5).001
      Outcomes
       Duration of hospitalization [median (IQR)]19 days (10)14 days (16)20 days (11)13 days (10).022
       30-day mortality2 (13.3)1 (14.3)1 (12.5)45 (5.9).23
      Values are given as mean ± SD or n (%) unless otherwise indicated.
      BMI = body mass index; CIED = cardiovascular implantable electronic device; IQR = interquartile range; LVEF = left ventricular ejection fraction.
      Comparing all stroke and no stroke.
      Overall, the stroke rate associated with TLR in CIED infection was 1.9%. Among the 15 patients who experienced stroke events, 7 strokes (0.9%) occurred before the TLR procedure, and 8 strokes (1.0%) occurred after the TLR procedure (Figure 2). Among the postprocedural stroke group, 62.5% of strokes occurred within the first 2 days after TLR. Kappa for agreement on stroke outcomes was 0.96 (95% confidence interval [CI] 0.89–1.00).
      Figure thumbnail gr2
      Figure 2Timing of stroke in relation to transvenous lead removal procedure.
      Detailed demographic and clinical characteristics of each patient who experienced a stroke event are given in Supplemental Table S4. There were no occurrences of transient ischemic attack. All strokes seemed to be cardioembolic in nature, with neuroimaging findings ranging from acute infarct in a single focus to multiple bihemispheric foci. There were no strokes with a clear nonembolic cause, such as findings of severe ipsilateral carotid stenosis, small subcortical stroke with lacunar presentation, or severe intracranial stenosis in the relevant vessel.

      Echocardiographic characteristics

      Echocardiography with the use of agitated saline was performed in 326 patients (42.0%), and TEE was performed in 585 patients (75.6%). PFOs were found in 105 patients (13.6%). Most of the PFOs were only detected after TTE with agitated saline or TEE with or without agitated saline (Supplemental Figures S1 and S2), with no difference in types of echocardiography performed in stroke patients with and without PFO (Supplemental Table S5). Patients with stroke had higher rates of PFO than did patients without stroke (46.7% vs 12.9%; P = .002). Multivariable analysis revealed that PFO was independently associated with stroke (odds ratio [OR] 7.75l 95% CI 2.60–23.15; P = .0002) (Table 2).
      Table 2Multivariable analysis of risk factors associated with stroke
      Variable
      Selected based on variables that were significant on univariate analysis.
      Odds ratioP value
      Patent foramen ovale7.75 (2.60–23.15).0002
      Left-sided vegetation7.68 (2.50 – 23.60).0004
      Selected based on variables that were significant on univariate analysis.
      Among patients with PFO, a total of 7 strokes occurred. Two (28.6%) occurred before the TLR procedure, and 5 (71.4%) occurred after the TLR procedure. Patients with right-sided vegetation and right-to-left shunting had an overall stroke rate of 19% (OR 6.4; 95% CI 1.3–31.0; P = .022) (Table 3). All postprocedural strokes in patients with PFOs occurred within 2 days of the TLR procedure.
      Table 3Subgroup analysis of stroke rate in patients with patent foramen ovale
      Overall strokeOdds ratio (95% CI)P value
      P <.025 was determined to be significant, based on Bonferroni correction for multiple hypothesis testing.
      n (%)95% CI
      Right-sided vegetation (n = 32)4 (12.5%)3.5–28.93.3 (0.7–15.9).130
      Right-to-left shunt (n = 66)6 (9.1%)3.4–18.73.8 (0.44–32.8).225
      Right-sided vegetation and right-to-left shunt (n = 21)4 (19.0%)5.5–41.96.4 (1.3–31.0).022
      CI = confidence interval.
      P <.025 was determined to be significant, based on Bonferroni correction for multiple hypothesis testing.
      Patients with stroke also had a higher occurrence of left-sided vegetations than did patients without stroke (40% vs 10.5%; OR 5.7; 95% CI 2.0–16.3]; P = .0016). Strokes in patients with left-sided vegetation were predominantly preprocedural in nature, with 5 strokes (83.3%) occurring before the TLR procedure, and only 1 (16.7%) occurring after the TLR procedure (P = .04). Multivariable analysis revealed that left-sided vegetation was independently associated with stroke (OR 7.68; 95% CI 2.50–23.60; P = .0004) (Table 2).
      Among all postprocedural strokes, there were 3 patients who had a stroke but did not have a PFO or left-sided vegetation. All 3 strokes occurred late in the hospital course (3–11 days after lead extraction). Two patients had atrial fibrillation with dense spontaneous echocontrast in the left atrial appendage and were not undergoing anticoagulation before the stroke. One patient had atrial fibrillation and mechanical mitral valve, and was without anticoagulation for around 48 hours, with anticoagulation restarted 1 day before the stroke.

      Length of stay and 30-day mortality

      Median length of stay was 13 days (interquartile range 3–23). Patients who suffered from a stroke had a significantly increased length of hospital stay than did those who did not have a stroke (19 vs 13 days; P = .022). Overall 30-day mortality was 6.2%. There was no statistical difference in 30-day mortality between patients with and those without a stroke (13.3% vs 5.9%; P = .23).

      Discussion

      Our analysis of 774 patients with CIED infection who underwent TLR procedure showed that (1) there was an overall stroke rate of 1.9%; (2) PFO was identified in 105 patients (13.6%); (3) PFOs were independently associated with stroke, especially with the presence of right-sided vegetations with right-to-left shunting; and (4) left-sided vegetations were also independently associated with stroke.
      Our data showed that the postprocedural stroke rate after TLR of infected CIED was 1.0%. Data on stroke rate specifically after TLR of patients with CIED infections are limited. Existing registry data of stroke rate after transvenous lead extraction procedures ranges from 0.07%–0.11%.
      • Bongiorni M.G.
      • Kennergren C.
      • Butter C.
      • et al.
      The European Lead Extraction ConTRolled (ELECTRa) study: a European Heart Rhythm Association (EHRA) Registry of Transvenous Lead Extraction Outcomes.
      • Wazni O.
      • Epstein L.M.
      • Carrillo R.G.
      • et al.
      Lead Extraction in the Contemporary Setting: the LExICon study. An observational retrospective study of consecutive laser lead extractions.
      Patients with CIED infection may have a higher stroke risk because of the risk of mobilization of right-sided vegetations during TLR leading to paradoxical embolism in the presence of a PFO (Figure 3), or because of the presence of concomitant left-sided vegetations, which itself carries the risk of embolic stroke.
      Figure thumbnail gr3
      Figure 3A: Transvenous lead removal may lead to dislodgment of endocardial lead vegetation. B: Paradoxical embolism of vegetation across a patent foramen ovale into the systemic circulation. C: Even after removal of the pacemaker lead, a mobile cast may still remain, leading to persistent risk of embolization.
      Our data suggest that PFOs were associated with a higher occurrence of stroke in patients with CIED infection undergoing TLR. Subgroup analysis of patients with PFO revealed that the risk of stroke was highest in patients with both right-sided vegetations with right-to-left shunting. Embolization of vegetative material on endocardial leads is typically confined to the pulmonary circulation.
      • Grammes J.A.
      • Schulze C.M.
      • Al-Bataineh M.
      • Yesenosky G.A.
      • Saari C.S.
      • Vrabel M.J.
      • Horrow J.
      • Chowdhury M.
      • Fontaine J.M.
      • Kutalek S.P.
      Percutaneous pacemaker and implantable cardioverter-defibrillator lead extraction in 100 patients with intracardiac vegetations defined by transesophageal echocardiogram.
      • Meier-Ewert H.K.
      • Gray M.E.
      • John R.M.
      Endocardial pacemaker or defibrillator leads with infected vegetations: a single-center experience and consequences of transvenous extraction.
      However, in the presence of a PFO, there is a risk for paradoxical embolism. In a study of patients after endocardial lead implantation, PFOs have been shown to be associated with an increased risk of stroke.
      • DeSimone C.V.
      • Friedman P.A.
      • Noheria A.
      • et al.
      Stroke or transient ischemic attack in patients with transvenous pacemaker or defibrillator and echocardiographically detected patent foramen ovale.
      There have also been multiple case reports on paradoxical embolism in patients with CIED infection, as well as paradoxical embolism during TLR.
      • Allie D.E.
      • Lirtzman M.D.
      • Wyatt C.H.
      • Vitrella D.A.
      • Walker C.M.
      Septic paradoxical embolus through a patent foramen ovale after pacemaker implantation.
      • Thompson J.J.
      • McDonnell K.M.
      • Reavey-Cantwell J.F.
      • Ellenbogen K.A.
      • Koneru J.N.
      Paradoxical septic emboli secondary to pacemaker endocarditis: transvenous lead extraction with distal embolization protection.
      • Larsen J.M.
      • Theuns D.A.
      • Thogersen A.M.
      Paradoxical thromboembolic stroke during extraction of a recalled St Jude Medical Riata defibrillator lead with conductor externalization.
      During TLR procedure, there is a possibility of dislodgment of debris or vegetation that travels across the PFO because of right-to-left shunting, leading to paradoxical embolism (Figure 3). Independently, right-sided vegetation or right-to-left shunting was not found to be significantly associated with an increased stroke rate. This is likely because of the small number of patients within each specific subgroup. More data will be needed to delineate their independent risk of stroke.
      Within our cohort, there was no transient ischemic attack. This may suggest that cerebrovascular events occur due to a solid particulate embolus rather than a transient thrombus in a hypercoagulable state.
      Strokes that occur after TLR could also be related to residual fibrous tissue called ghosts (Figure 3C).
      • Narducci M.L.
      • Di Monaco A.
      • Pelargonio G.
      • et al.
      Presence of ‘ghosts’ and mortality after transvenous lead extraction.
      This is a remnant of a fibrous sheath that encapsulates a lead and may remain even after TLR. It has been shown to be an independent predictor of mid-term mortality after TLR.
      • Narducci M.L.
      • Di Monaco A.
      • Pelargonio G.
      • et al.
      Presence of ‘ghosts’ and mortality after transvenous lead extraction.
      These ghosts may be a potential cause of persistent embolic risk even after TLR if vegetations are attached to it. Persistence of intracardiac vegetations despite TLR and the residual possibility for embolic phenomenon have previously been described in a case report but have not been systematically studied.
      • Novaro G.M.
      • Saliba W.
      • Jaber W.A.
      Images in cardiovascular medicine. Fate of intracardiac lead vegetations after percutaneous lead extraction.
      The risk of stroke in patients with CIED infection may also be due to concomitant presence of aortic or mitral valve vegetations and systemic embolism. Our study found that preprocedural stroke is mainly associated with left-sided vegetation, whereas postprocedural stroke is mainly associated with PFO. Stroke risk due to left-sided vegetation may decrease after TLR procedure because of the reduction of burden of infection after TLR as well as ongoing antibiotic therapy, which may have resulted in resolution of left-sided vegetations and reduction of risk of embolization.
      • Dickerman S.A.
      • Abrutyn E.
      • Barsic B.
      • et al.
      The relationship between the initiation of antimicrobial therapy and the incidence of stroke in infective endocarditis: an analysis from the ICE Prospective Cohort Study (ICE-PCS).
      The duration of hospital stay was longer in patients with stroke than in those without stroke, which may be an indirect indicator of the clinical impact of stroke. In all cases of stroke, there were neuroimaging findings of infarcts, more commonly in the distribution of the middle cerebral artery. Unfortunately, we were not able to obtain accurate objective data on the severity and persistence of neurologic deficits. However, comparison of 30-day mortality between those with and those without stroke did not reveal any statistical difference. Our average 30-day mortality of 6.2% was similar to published 30-day mortality of 3.8%–10% for patients with lead-related infective endocarditis undergoing TLR.
      • Grammes J.A.
      • Schulze C.M.
      • Al-Bataineh M.
      • Yesenosky G.A.
      • Saari C.S.
      • Vrabel M.J.
      • Horrow J.
      • Chowdhury M.
      • Fontaine J.M.
      • Kutalek S.P.
      Percutaneous pacemaker and implantable cardioverter-defibrillator lead extraction in 100 patients with intracardiac vegetations defined by transesophageal echocardiogram.
      • Polewczyk A.
      • Jachec W.
      • Tomaszewski A.
      • Brzozowski W.
      • Czajkowski M.
      • Opolski G.
      • Grabowski M.
      • Janion M.
      • Kutarski A.
      Lead-related infective endocarditis: Factors influencing early and long-term survival in patients undergoing transvenous lead extraction.
      With the increased risk of stroke in patients with PFO, especially in those with right-sided vegetation and right-to-left shunting, preprocedural PFO screening may be beneficial. PFO may be screened with either TTE with agitated saline or TEE with or without agitated saline (Figure 4A).
      Figure thumbnail gr4
      Figure 4A: Transesophageal echocardiogram with agitated saline showing patent foramen ovale with right-to-left shunt (arrow). B: Patent foramen ovale closure with 25-mm Gore Helex device (WL Gore & Associates, Flagstaff, AZ). C: SpiderFX filter (Covidien, Mansfield, MA) in bilateral common carotid arteries (arrows). D: Balloon occlusion of the patent foramen ovale. Arrow points to waist of the balloon. E: Transesophageal echocardiogram showing the balloon crossing the patent foramen ovale into the left atrium (arrowhead). New clot was visualized in the right atrial portion of the balloon after balloon deployment (arrow). F: Transesophageal echocardiogram after balloon removal showing new clot across the mitral valve (arrow). LA = left atrium; RA = right atrium.
      Further studies will be needed to evaluate the importance of management strategies in patients with PFO. The first option is surgical evacuation, either from conventional median sternotomy or via a limited atriotomy approach. However, because of the invasive nature of the surgical procedure, it is associated with postoperative mortality rates ranging from 9%–40%.
      • del Rio A.
      • Anguera I.
      • Miro J.M.
      • Mont L.
      • Fowler Jr., V.G.
      • Azqueta M.
      • Mestres C.A.
      Hospital Clinic Endocarditis Study Group. Surgical treatment of pacemaker and defibrillator lead endocarditis: the impact of electrode lead extraction on outcome.
      • Camboni D.
      • Wollmann C.G.
      • Loher A.
      • Gradaus R.
      • Scheld H.H.
      • Schmid C.
      Explantation of implantable defibrillator leads using open heart surgery or percutaneous techniques.
      Furthermore, the high doses of heparin needed during cardiopulmonary bypass surgery may be contraindicated if a patient already has an acute embolic stroke due to risk of hemorrhagic conversion. The second option is deployment of a PFO closure device before TLR (Figure 4B). However, because of the potential risk of infection of the PFO closure device, the procedure may have to be delayed until blood cultures are negative. The risk of infection of the PFO closure device will have to be balanced against the risk of delaying TLR. The third option is deployment of bilateral carotid artery filters during TLR (Figure 4C).
      • Thompson J.J.
      • McDonnell K.M.
      • Reavey-Cantwell J.F.
      • Ellenbogen K.A.
      • Koneru J.N.
      Paradoxical septic emboli secondary to pacemaker endocarditis: transvenous lead extraction with distal embolization protection.
      However, heparinization would still be required during deployment of the cerebral embolic protection device. The fourth option is temporary occlusion of the PFO with a balloon during lead extraction (Figure 4D). However, if the procedure is performed without anticoagulation to reduce the risk of the lead extraction procedure, there is a risk of development of clot on the balloon, both on the right atrial side (Figure 4E) and the left atrial side (Figure 4F) of the balloon.
      Our study focused on the first 3 aspects of the public health approach to problem-solving: (1) define the problem; (2) measure the magnitude of the problem; and (3) develop a conceptual framework for the key determinants of the problem.
      • Guyer B.
      Problem-solving in public health.
      We anticipate that as more evidence accumulates, the remaining aspects of the problem-solving methodology, which includes identification and development of intervention strategies, will be touched on.

      Study limitations

      Our results are best interpreted in the context of several limitations. Some patients with PFO likely went undetected because of the limitations of echocardiography. Within our patient cohort, 13.6% of patients were found to have PFO. This percentage was lower than data from autopsy series that found PFO to be present in 25% of the population but was similar to the 15% reported in a population-based study with TTE and agitated saline.
      • Hagen P.T.
      • Scholz D.G.
      • Edwards W.D.
      Incidence and size of patent foramen ovale during the first 10 decades of life: an autopsy study of 965 normal hearts.
      • Di Tullio M.R.
      • Jin Z.
      • Russo C.
      • Elkind M.S.
      • Rundek T.
      • Yoshita M.
      • DeCarli C.
      • Wright C.B.
      • Homma S.
      • Sacco R.L.
      Patent foramen ovale, subclinical cerebrovascular disease, and ischemic stroke in a population-based cohort.
      There is also variability in the extent for which maneuvers for detection of shunting was performed, such as agitated saline or Valsalva, which were performed at the discretion of each operator. We also could not analyze the relationship of stroke to the amount of shunting across the PFO or the concomitant presence of atrial septal aneurysm given the inconsistent reporting of these occurrences during this retrospective study. Accuracy of detection of vegetations was also limited based on the imaging technique used. Histologic and microbiological analyses of the extracted leads were not performed routinely. As a result, although only patients with bacteremia, infective endocarditis, and device infections were included in our study, it is possible that some of the vegetations identified within our study were simply thrombi or fibrin casts. The overall small absolute number of stroke outcome events also limited further statistical analysis on other procedural factors that may possibly be associated with stroke.
      Furthermore, this study was based on data from 3 tertiary referral centers, and there may be limited extrapolation to hospitals with a different patient population, procedural technique mix, and clinical expertise. We also excluded patients with CIED infection who did not undergo TLR; therefore, our results may only be applicable to patients with device infection who meet TLR criteria and did not have any contraindications to the procedure. However, we suspect that the number of patients who did not undergo TLR is small, with a previous study reporting that 98% of patients with CIED infection underwent complete device removal.
      • Sohail M.R.
      • Uslan D.Z.
      • Khan A.H.
      • Friedman P.A.
      • Hayes D.L.
      • Wilson W.R.
      • Steckelberg J.M.
      • Stoner S.
      • Baddour L.M.
      Management and outcome of permanent pacemaker and implantable cardioverter-defibrillator infections.
      Our study also did not include patients in whom the device was extracted via open heart surgery as the primary approach.
      Finally, because our cerebrovascular outcome definition was based on standardized definitions, we did not include subclinical stroke. Also, no systematic neurocognitive testing was performed to detect more subtle central nervous system injury in patients without clinically apparent neurologic symptoms.

      Conclusion

      In patients with CIED infection undergoing TLR, the overall stroke risk was 1.9%. Our study revealed that the presence of a PFO was associated with an increased risk of stroke, which mostly occurred after the TLR procedure. This was especially so if there was a concomitant finding of right-sided vegetation and right-to-left shunting, which suggests that paradoxical embolism of lead vegetation may occur in patients with CIED infection undergoing TLR. This finding highlights the importance of preprocedural PFO screening in patients with CIED infections undergoing TLR. Further studies will be needed to assess management strategies in patients with PFO undergoing TLR for CIED infection, such as PFO closure or use of cerebral embolic protection devices.

      Acknowledgment

      The authors would like to thank David J. Cheney for help with illustrations.

      Appendix. Supplementary data

      References

        • Greenspon A.J.
        • Patel J.D.
        • Lau E.
        • Ochoa J.A.
        • Frisch D.R.
        • Ho R.T.
        • Pavri B.B.
        • Kurtz S.M.
        16-year trends in the infection burden for pacemakers and implantable cardioverter-defibrillators in the United States 1993 to 2008.
        J Am Coll Cardiol. 2011; 58: 1001-1006
        • Baddour L.M.
        • Epstein A.E.
        • Erickson C.C.
        • et al.
        Update on cardiovascular implantable electronic device infections and their management: a scientific statement from the American Heart Association.
        Circulation. 2010; 121: 458-477
        • Lee J.Z.
        • Ling J.
        • Diehl N.N.
        • et al.
        Mortality and cerebrovascular events after heart rhythm disorder management procedures.
        Circulation. 2018; 137: 24-33
        • Klug D.
        • Balde M.
        • Pavin D.
        • et al.
        Risk factors related to infections of implanted pacemakers and cardioverter-defibrillators: results of a large prospective study.
        Circulation. 2007; 116: 1349-1355
        • Grammes J.A.
        • Schulze C.M.
        • Al-Bataineh M.
        • Yesenosky G.A.
        • Saari C.S.
        • Vrabel M.J.
        • Horrow J.
        • Chowdhury M.
        • Fontaine J.M.
        • Kutalek S.P.
        Percutaneous pacemaker and implantable cardioverter-defibrillator lead extraction in 100 patients with intracardiac vegetations defined by transesophageal echocardiogram.
        J Am Coll Cardiol. 2010; 55: 886-894
        • Meier-Ewert H.K.
        • Gray M.E.
        • John R.M.
        Endocardial pacemaker or defibrillator leads with infected vegetations: a single-center experience and consequences of transvenous extraction.
        Am Heart J. 2003; 146: 339-344
        • Kusumoto F.M.
        • Schoenfeld M.H.
        • Wilkoff B.L.
        • et al.
        2017 HRS expert consensus statement on cardiovascular implantable electronic device lead management and extraction.
        Heart Rhythm. 2017; 14: e503-e551
        • Silvestry F.E.
        • Cohen M.S.
        • Armsby L.B.
        • et al.
        Guidelines for the echocardiographic assessment of atrial septal defect and patent foramen ovale: from the American Society of Echocardiography and Society for Cardiac Angiography and Interventions.
        J Am Soc Echocardiogr. 2015; 28: 910-958
        • Kappetein A.P.
        • Head S.J.
        • Genereux P.
        • et al.
        Updated standardized endpoint definitions for transcatheter aortic valve implantation: the Valve Academic Research Consortium-2 consensus document.
        J Am Coll Cardiol. 2012; 60: 1438-1454
        • Streiner D.L.
        • Norman G.R.
        Correction for multiple testing: is there a resolution?.
        Chest. 2011; 140: 16-18
        • Bongiorni M.G.
        • Kennergren C.
        • Butter C.
        • et al.
        The European Lead Extraction ConTRolled (ELECTRa) study: a European Heart Rhythm Association (EHRA) Registry of Transvenous Lead Extraction Outcomes.
        Eur Heart J. 2017; 38: 2995-3005
        • Wazni O.
        • Epstein L.M.
        • Carrillo R.G.
        • et al.
        Lead Extraction in the Contemporary Setting: the LExICon study. An observational retrospective study of consecutive laser lead extractions.
        J Am Coll Cardiol. 2010; 55: 579-586
        • DeSimone C.V.
        • Friedman P.A.
        • Noheria A.
        • et al.
        Stroke or transient ischemic attack in patients with transvenous pacemaker or defibrillator and echocardiographically detected patent foramen ovale.
        Circulation. 2013; 128: 1433-1441
        • Allie D.E.
        • Lirtzman M.D.
        • Wyatt C.H.
        • Vitrella D.A.
        • Walker C.M.
        Septic paradoxical embolus through a patent foramen ovale after pacemaker implantation.
        Ann Thorac Surg. 2000; 69: 946-948
        • Thompson J.J.
        • McDonnell K.M.
        • Reavey-Cantwell J.F.
        • Ellenbogen K.A.
        • Koneru J.N.
        Paradoxical septic emboli secondary to pacemaker endocarditis: transvenous lead extraction with distal embolization protection.
        Circ Arrhythm Electrophysiol. 2014; 7: 1271-1272
        • Larsen J.M.
        • Theuns D.A.
        • Thogersen A.M.
        Paradoxical thromboembolic stroke during extraction of a recalled St Jude Medical Riata defibrillator lead with conductor externalization.
        Europace. 2014; 16: 240
        • Narducci M.L.
        • Di Monaco A.
        • Pelargonio G.
        • et al.
        Presence of ‘ghosts’ and mortality after transvenous lead extraction.
        Europace. 2017; 19: 432-440
        • Novaro G.M.
        • Saliba W.
        • Jaber W.A.
        Images in cardiovascular medicine. Fate of intracardiac lead vegetations after percutaneous lead extraction.
        Circulation. 2002; 106: e46
        • Dickerman S.A.
        • Abrutyn E.
        • Barsic B.
        • et al.
        The relationship between the initiation of antimicrobial therapy and the incidence of stroke in infective endocarditis: an analysis from the ICE Prospective Cohort Study (ICE-PCS).
        Am Heart J. 2007; 154: 1086-1094
        • Polewczyk A.
        • Jachec W.
        • Tomaszewski A.
        • Brzozowski W.
        • Czajkowski M.
        • Opolski G.
        • Grabowski M.
        • Janion M.
        • Kutarski A.
        Lead-related infective endocarditis: Factors influencing early and long-term survival in patients undergoing transvenous lead extraction.
        Heart Rhythm. 2017; 14: 43-49
        • del Rio A.
        • Anguera I.
        • Miro J.M.
        • Mont L.
        • Fowler Jr., V.G.
        • Azqueta M.
        • Mestres C.A.
        Hospital Clinic Endocarditis Study Group. Surgical treatment of pacemaker and defibrillator lead endocarditis: the impact of electrode lead extraction on outcome.
        Chest. 2003; 124: 1451-1459
        • Camboni D.
        • Wollmann C.G.
        • Loher A.
        • Gradaus R.
        • Scheld H.H.
        • Schmid C.
        Explantation of implantable defibrillator leads using open heart surgery or percutaneous techniques.
        Ann Thorac Surg. 2008; 85: 50-55
        • Guyer B.
        Problem-solving in public health.
        in: Armenian H. Shapiro S. Epidemiology and Health Services Research. Oxford University Press, New York1998: 15-26
        • Hagen P.T.
        • Scholz D.G.
        • Edwards W.D.
        Incidence and size of patent foramen ovale during the first 10 decades of life: an autopsy study of 965 normal hearts.
        Mayo Clin Proc. 1984; 59: 17-20
        • Di Tullio M.R.
        • Jin Z.
        • Russo C.
        • Elkind M.S.
        • Rundek T.
        • Yoshita M.
        • DeCarli C.
        • Wright C.B.
        • Homma S.
        • Sacco R.L.
        Patent foramen ovale, subclinical cerebrovascular disease, and ischemic stroke in a population-based cohort.
        J Am Coll Cardiol. 2013; 62: 35-41
        • Sohail M.R.
        • Uslan D.Z.
        • Khan A.H.
        • Friedman P.A.
        • Hayes D.L.
        • Wilson W.R.
        • Steckelberg J.M.
        • Stoner S.
        • Baddour L.M.
        Management and outcome of permanent pacemaker and implantable cardioverter-defibrillator infections.
        J Am Coll Cardiol. 2007; 49: 1851-1859