Author + information
- Received January 4, 2011
- Revision received February 16, 2011
- Accepted March 15, 2011
- Published online July 12, 2011.
- Thomas B. Morrison, MD⁎,
- Paul A. Friedman, MD‡,⁎ (, )
- Linda M. Kallinen, BS§,
- David O. Hodge, MS†,
- Daniel Crusan, BS†,
- Kapil Kumar, MD∥,
- David L. Hayes, MD‡,
- Robert F. Rea, MD‡ and
- Robert G. Hauser, MD§
- ↵⁎Reprint requests and correspondence:
Dr. Paul A. Friedman, Division of Cardiovascular Disease, Mayo Clinic, 200 1st Street SW, Rochester, Minnesota 55905
Objectives This study sought to compare all-cause mortality in patients with Fidelis leads (Medtronic, Minneapolis, Minnesota) to those with a nonadvisory lead.
Background Although Fidelis leads are prone to fracture, and rare deaths due to lead failure have been reported, it is unclear whether the presence of a Fidelis lead is associated with increased mortality. This study compares all-cause mortality in a large cohort of patients with Fidelis and Quattro implantable cardioverter-defibrillator (ICD) leads.
Methods All patients with Fidelis (Medtronic models 6931, 6948, and 6949) and Quattro (Medtronic model 6947) leads followed at 3 tertiary care centers were identified from the medical records (implant dates: November 19, 2001, to December 23, 2008). Clinical and device-specific data were collected into a common database. Deaths were identified from medical records and the Social Security Death Index. Survival was estimated using the Kaplan-Meier method.
Results A total of 2,671 patients (1,030 Fidelis and 1,641 Quattro) were identified. There were 398 deaths: 147 in the Fidelis group (mean follow-up: 34.4 months) and 251 in the Quattro group (mean follow-up: 39.9 months). No deaths were associated with 85 Fidelis and 23 Quattro failures. At 4 years, survival was diminished in patients with Fidelis compared with Quattro leads (80.7% vs. 83.9%, p = 0.025). After adjustment for factors associated with mortality, survival was similar between groups. One hundred percent pacing was not associated with mortality. Elective removal of nonfailed leads was performed in 5.1% of Fidelis and 0.9% of Quattro patients.
Conclusions In a conservatively managed cohort, in whom observation was predominantly utilized, adjusted survival is similar between patients with Fidelis and Quattro ICD leads.
Implantable cardioverter-defibrillators (ICDs) have consistently been shown to reduce mortality from sudden death in high-risk patients (1–5). Appropriate system function requires an intact lead to sense cardiac signals and deliver therapy. The Sprint Fidelis (models 6930, 6931, 6948, and 6949) high-voltage ICD lead (Medtronic, Minneapolis, Minnesota) is predisposed to premature failure (6–10), and death due to lead failure has been reported (11–13). In addition, recent studies have demonstrated that the risk of Fidelis failure increases over time (6,14,15), with greater failure rates reported by independent centers than by the manufacturer. Currently, an estimated 143,000 patients in the United States have a Fidelis lead implanted (16). Whether patents with implanted Fidelis leads have an increased mortality is not known. We therefore pooled data from 3 large centers to determine whether patients with a Fidelis lead in service are at increased risk of death compared with similar patients with nonadvisory leads (Quattro, Medtronic), and sought to determine optimal follow-up and management.
All patients with Medtronic Sprint Fidelis (model numbers 6931, 6948, and 6949) and Medtronic Quattro Secure (model number 6947) ICD leads followed at Minneapolis Heart Institute (Minneapolis, Minnesota), Beth Israel Deaconess Medical Center (Boston, Massachusetts), and Mayo Clinic (Rochester, Minnesota) that were implanted between November 19, 2001, and December 23, 2008, were included in the study. This study was approved by the institutional review board at each of the 3 participating institutions. Data regarding ICD lead implantation and follow-up were collected prospectively at each center as part of a local ICD database. Investigators from each site formed a committee that identified clinical variables of interest, which were given standardized definitions and then merged to create a single database for the purpose of this study. Study design, data collection and analysis, and manuscript preparation were performed and funded by the investigators.
All study leads were implanted by experienced cardiologists specializing in electrophysiology at 1 of the participating institutions. Lead placement was done via a left- or right-sided cephalic cutdown, axillary, or subclavian vein using standard introducer techniques. Leads were positioned in the right ventricular apex or ventricular septum. Defibrillation safety margins and pacing threshold and sensing measurements were obtained according to each participating center's protocol. Patients were followed every 3 to 4 months at clinic visits or remotely if appropriate.
A lead was considered implanted after it was tested, connected to the ICD pulse generator, and the incision was closed. A lead failure was defined as a lead removed from service due to an inability to meet its performance specifications or otherwise perform as intended (17). A lead failed if: 1) it exhibited abnormal impedance; 2) it exhibited electrical noise as manifested by nonphysiological signals on the electrogram or by pulse generator diagnostic data suggesting rapid oversensing, for example, nonphysiological short intervals and/or recurrent nonsustained ventricular tachycardia with intervals usually <220 ms; or 3) it could not sense R waves and/or provide effective electrical therapy due to an apparent structural defect such as a conductor fracture or insulation breach. Functional abnormalities, including exit block and physiological oversensing in the presence of an electrically intact lead, were not considered failures. Lead displacement was not a lead failure unless a fixation mechanism defect was identified. Leads removed from service were classified in accordance with the recommendations of the Heart Rhythm Society (17).
Patient deaths were identified from local medical records and from the Social Security Death Index (18). Appropriate and inappropriate device therapies were determined at the time of device interrogation by review of stored electrograms. A group of 100%-paced patients was defined as the presence of complete heart block in the absence of a stable escape rhythm or 100% pacing in patients not receiving cardiac resynchronization therapy (CRT).
Continuous variables are presented as a mean ± SD, and discrete variables are summarized using group percentages. Comparisons between lead types of patient characteristics were tested using the Pearson chi-square test for continuous variables and the Student 2-sample t test for discrete variables. The relationship between patient demographics and device characteristics with mortality was analyzed with a univariate analysis using Cox proportional hazards models. Utilizing the same variables, a multivariate Cox proportional hazard model was selected using the stepwise selection technique. Unadjusted patient survival was estimated utilizing the Kaplan-Meier method. An adjusted comparison of mortality between Fidelis and Quattro leads was performed using the variables identified from the multivariate Cox proportional hazard model. With our study population and an alpha of 0.05, we had 80% power to detect a hazard ratio of 1.38 in the multivariable model.
A total of 2,671 patients (1,030 Fidelis and 1,641 Quattro) were included. Baseline patient characteristics are shown in Table 1. Patients with Fidelis leads were more likely to be female and have resynchronization systems, a primary prevention indication, and dilated cardiomyopathy. Patients with Quattro leads were more likely to be male and have a secondary prevention indication, previous bypass surgery, ischemic cardiomyopathy, atrial fibrillation, and diabetes mellitus. Fidelis leads in service were model 6949 (n = 1,007, 97.8%), model 6948 (n = 2, 0.2%), and model 6931 (n = 21, 2%). Elective removal from service of nonfailed leads was performed in 53 (5.1%) of Fidelis and 15 (0.9%) of Quattro patients. Average follow-up was 34.4 months (range 3.1 to 65.6 months) and 39.9 months (range 3.0 to 95.2 months) for the Fidelis and Quattro patients, respectively.
During follow-up, 398 patients died: 147 in the Fidelis group and 251 in the Quattro group. No deaths were associated with 85 Fidelis and 23 Quattro lead failures. Unadjusted patient survival (Fig. 1) was diminished in the Fidelis group as compared with the Quattro group (p = 0.025). Patient survival at 48 months was 80.7% in patients with Fidelis leads compared with 83.9% in those with Quattro leads.
Inappropriate shocks were seen in 38 of 85 (45%) Fidelis failures. In addition to inappropriate shocks, 1 pacemaker-dependent patient presented with presyncope and ventricular asystole requiring emergent placement of a temporary pacemaker. Another patient receiving CRT showed evidence of failure to pace on interrogation but did not present with a clinical event. Extraction of a failed lead was complicated by superior vena cava perforation requiring emergent cardiac surgery in 1 patient.
Fidelis leads were significantly more likely to fail than Quattro leads. Lead survival at 48 months was 87.0% and 98.7% (p < 0.0001) in Fidelis and Quattro leads, respectively.
Univariate predictors of mortality
In univariate analysis, Fidelis leads were associated with increased mortality (hazard ratio [HR]: 1.28, 95% confidence interval [CI]: 1.03 to 1.60). Other device and patient factors associated with increased mortality in univariate analysis were the presence of an atrial lead, heart failure, CRT, increasing age, male sex, ischemic cardiomyopathy, previous coronary artery bypass grafting, history of stroke, history of peripheral arterial disease, severe valvular disease, atrial fibrillation, diabetes, increasing creatinine, dialysis, chronic obstructive pulmonary disease, and pulmonary hypertension (Table 2). Conversely, pulse generator changes, hypertrophic cardiomyopathy, and increasing ejection fraction were associated with decreased mortality.
Multivariate predictors of mortality
Secondary prevention indication, the presence of an atrial lead, heart failure, increasing age, history of peripheral artery disease, atrial fibrillation, chronic dialysis, and chronic obstructive pulmonary disease were independent predictors of mortality in a multivariate model (Table 3). Increased numbers of pulse generator replacements and increasing ejection fraction were associated with decreased mortality in the multivariate model. After adjustment for factors associated with mortality in the multivariate model, the presence of a Fidelis ICD lead was not significantly associated with increased mortality (HR: 1.06, 95% CI: 0.84 to 1.33).
The presence of 100% pacing was not associated with mortality in either the univariate (HR: 1.118, 95% CI: 0.86 to 1.46) or in the multivariate analyses.
Implantable defibrillators perform critical lifesaving functions (1–5). Whether a defibrillator lead at increased risk of failure adversely impacts mortality depends on the nature of the malfunction, the risk of death, and the ability of surveillance techniques to identify malfunction prior to a lethal event. We pooled raw data from 3 large tertiary centers and found no increased risk of mortality due to the presence of an active Fidelis advisory lead, as compared with a similar cohort treated with a nonadvisory (Quattro) lead, after adjusting for other factors. No patient deaths were associated with lead failure, and 100% pacing was not associated with increased mortality. Importantly, a conservative follow-up strategy was adopted, with elective removal from service of nonfailed leads in only 53 (5.1%) of Fidelis and 15 (0.9%) of Quattro patients, with 36% (n = 19) of elective Fidelis removals performed in 100%-paced patients.
Failure of a Fidelis ICD lead carries the potential risk of a fatal complication, and over 13 lead-related deaths have been reported (11–13). Three mechanisms of Fidelis lead failure may lead to a fatal outcome: 1) fracture of the pace-sense portion of the lead in a pacemaker-dependent patient may lead to failure to pace (12); 2) fracture of the pace-sense portion of the lead can result in oversensing and potentially fatal inappropriate shocks (11); and 3) fracture of the high-voltage portion of the lead may impair appropriate shock delivery.
Several factors likely account for the lack of increased mortality seen in our study. All participating centers utilize regular remote monitoring and thereby may identify lead failures early, minimizing the risk for an adverse event. Fracture of a pace-sense lead commonly presents with intermittent, transient oversensing of rapid nonphysiological make-break potentials, leading to inappropriate shock (19). A downloadable algorithm identifies these potentials, generates immediate audible and Internet-based alerts, and increases the number of intervals needed to detect ventricular arrhythmias, minimizing the risk of inappropriate detection of ventricular tachycardia (20). Since the sense amplifier in ICDs is continuously on to screen for ventricular arrhythmias, this surveillance in conjunction with aggressive use of Internet-based monitoring at participating centers may have prevented fatal events. Consequently, bradycardic sudden death—which would require output failure in the absence of significant nonphysiological noise—appears to be an uncommon manifestation of this mode of lead failure. Nonetheless, given the continuous lifesaving nature of appropriate lead function in patients who are otherwise asystolic (as opposed to an only intermittent critical lead function during the time of arrhythmia in other patients), a more aggressive approach towards lead replacement appears warranted in pacemaker-dependent patients. Indeed, of the nonfailed Fidelis leads removed from service, 36% (n = 19) were in 100%-paced patients. Although in our study 100% pacing was not associated with increased mortality, 1 pacemaker-dependent patient presented with pre-syncope due to inability to pace and required an emergent intervention for ventricular asystole. Another patient receiving CRT had evidence of failure to pace on interrogation but did not have any clinical events secondary to lead failure. These highlight the potential risk in this subpopulation.
Fracture of the high-voltage element accounts for 10% of Fidelis lead failures (16). Daily impedance measurements screen for impedance conductor failure, with reported sensitivity of impedance changes ranging from 83% to 92.9% (21,22). In order to result in an unwarned fatal event, a high-voltage failure and a life-threatening arrhythmic event must occur within 24 h, or the failure must be underdetected by daily impedance monitoring.
The most common adverse events associated with Fidelis lead failure are inappropriate device therapies. Inappropriate device therapies may result in death (11), diminished overall survival (23), pain, depression, and diminished quality of life (24). Lead failure presenting as inappropriate shock ranges from 21% (8) to 83% (7). In our study, 38 of 85 (45%) of Fidelis failures presented with inappropriate shocks. Downloadable algorithms that increase the number of intervals to detect are associated with a 79% relative reduction in the occurrence of inappropriate shocks and therefore represent an important component of preventing these events (20).
Our 48-month lead survival of 87.0% is lower than that reported by the manufacturer: 93.8% to 95.4% at 48 months (25). The discrepancy highlights the complexity of and resources necessary to identify failure events that exceed the background component failure present in all devices, and may also reflect differences in populations and in follow-up methodologies.
Managing patients with Fidelis leads
Management of patients with implanted Fidelis leads involves balancing the risk of malfunction, the ability of surveillance to detect malfunction before catastrophic clinical events, and the risk of intervention. Current manufacturer recommendations for managing patients with Fidelis leads entail continued clinical follow-up with the addition of the Lead Integrity Alert downloadable algorithm (16). Our data support this strategy for most patients, in that mortality was not increased utilizing a fairly conservative approach. This argues against prophylactic removal of a normally functioning Fidelis lead from service in a nonpacemaker-dependent patient.
However, management must be tailored to patient risk. The Lead Integrity Alert algorithm is limited to patients with Medtronic pulse generators, does not provide advanced warning in approximately 25% of patients, and is limited in part by the inability of older patients to hear alert tones (26). Clinical predictors of lead failure include higher ejection fraction (27), age <50 years, and, possibly, sports/physical activity (10). Additionally, the risk of lead failure appears to continue to increase over time (6,14,15). Thus, although an increased mortality was not seen in 100%-paced patients, the sample size may have limited detection of risk in this subpopulation, and prophylactic lead replacement in this population appears reasonable. Other groups in whom prophylactic lead replacement (typically at the time of pulse generator battery depletion) may be reasonable include young patients (under 50 years), those with preserved ejection fraction, those who are quite active physically, and those with spontaneous ventricular tachycardia/fibrillation events (who are thus at increased risk for recurrence). Whether advisory leads are abandoned or extracted depends on patient characteristics and physician and medical center skills and resources, and has been reviewed elsewhere (28–30).
ICD versus arrhythmia management system
Historically ICDs, were autonomously functioning implanted rhythm management devices, with periodic modifications (program changes) applied by caregivers. Our study highlights the fact that ICDs are part of a complex ecosystem that begins at implantation, and must include sophisticated follow-up entailing automated, direct patient and Internet-enabled alerts, practitioner visits, system modification with downloadable software, and revision when needed. Ongoing surveillance by heart rhythm experts not only may detect impending system malfunction, but may identify impending clinical changes (e.g., incipient pulmonary edema or asymptomatic atrial fibrillation) that warrant pharmacological or other therapy independent of device function (31,32). This is supported by our finding that the predictors of death were not device function but medical comorbidities, a finding consistent with other studies (33).
Although data were collected prospectively for ICD databases at each center and merged for the purposes of this study, the analysis performed was retrospective and is therefore subject to all of the limitations of a retrospective analysis. All participating centers are referral centers, and therefore, our study population may not be representative of an average practice population. Although we did compare the Fidelis to a nonadvisory lead, we did not compare it to leads from other manufacturers. Lastly, we were unable to adjudicate the mechanism of death to estimate the probability of death due to lead failure.
In a cohort managed with a relatively conservative approach that predominantly entailed observation, adjusted survival is similar between patients with Fidelis and Quattro ICD leads.
Dr. Friedman has received research support from Medtronic, Boston Scientific, Bard, St. Jude Medical, and Pfizer; has intellectual property rights for work with Bard EP, Hewlett Packard, Medical Positioning, Inc., Aegis Medical, and NeoChord; and has been a speaker/consultant (all <$10,000) for Medtronic, Boston Scientific, and St. Jude. Dr. Hayes has been a speaker at educational meetings for Medtronic, Boston Scientific, Sorin Medical, Biotronik, and St. Jude Medical; has served in as advisory capacity for St. Jude Medical, Boston Scientific, and Pixel Velocity; and has been a steering committee member for St. Jude Medical. All other authors have reported that they have no relationships to disclose.
- Abbreviations and Acronyms
- confidence interval
- cardiac resynchronization therapy
- hazard ratio
- implantable cardioverter-defibrillator
- Received January 4, 2011.
- Revision received February 16, 2011.
- Accepted March 15, 2011.
- American College of Cardiology Foundation
- Connolly S.J.,
- Gent M.,
- Roberts R.S.,
- et al.
- Groves R.
- Samsel T.
- ↵(2010) Medtronic. Sprint Fidelis® model 6930, 6931, 6948, and 6949 product performance report.
- Social Security Administration
- Swerdlow C.D.,
- Gunderson B.D.,
- Ousdigian K.T.,
- et al.
- Vollmann D.,
- Erdogan A.,
- Himmrich E.,
- et al.
- Gunderson B.D.,
- Patel A.S.,
- Bounds C.A.,
- Shepard R.K.,
- Wood M.A.,
- Ellenbogen K.A.
- Daubert J.P.,
- Zareba W.,
- Cannom D.S.,
- et al.
- Schron E.B.,
- Exner D.V.,
- Yao Q.,
- et al.
- ↵(2009) Medtronic. Sprint Fidelis model 6949 lead performance, November 2009 update.