Time-Dependent Risk of Fidelis Lead Failure
Article Outline
The Medtronic Sprint Fidelis leads (models 6930, 6931, 6948, 6949) are 6.6-F bipolar high-voltage implantable cardioverter–defibrillator electrodes that were first introduced in September 2004. In October 2007, Fidelis leads were removed from the market. We sought to determine the time-dependent hazard of the Fidelis failure rate to date. A retrospective chart review was conducted in all patients who underwent implantation of a Sprint Fidelis lead (426 leads) at our center. We primarily implanted models 6931 and 6949. With 1,056 years of combined follow-up (average 2.3 ± 1), 38 of 426 (8.92%) Sprint Fidelis leads failed (3.6%/year). The hazard of fracture increased exponentially over time by a power of 2.13 (95% confidence interval [CI] 1.98 to 2.27, p <0.001) and the 3-year survival was 90.8% (95% CI 87.4 to 94.3). If a Fidelis lead was functioning normally at 1 year, the chance it would survive another year was 97.4% (95% CI 95.7 to 99.1); if functioning at 2 years, the chance of surviving another year was 94.7% (95% CI 91.8 to 97.7); and if functioning at 3 years, the chance of surviving 1 more year was 86.7% (95% CI 78.8 to 95.5). Other commonly used implantable cardioverter–defibrillator leads showed no evidence of increased failure rates. In conclusion, to date, the hazard of Fidelis lead fracture is increasing exponentially with time and, based on our data, occurring at a higher rate than the latest manufacturer's performance update. Further accumulative data are needed because it remains unknown if the fracture rate will level off or continue to increase.
During the previous year, 3 clinical studies have reported a higher Fidelis lead fracture rate than was indicated in the Medtronic performance update of 2009.1, 2, 3 The primary objectives of this study were to determine the performance of the Fidelis leads at a tertiary referral center and to determine if the hazard of Fidelis failure is changing with time. A secondary objective was to determine the failure rate of Fidelis leads compared to other implanted implantable cardioverter–defibrillator (ICD) leads used at our hospital during the same period.
Methods
A retrospective chart review was conducted on all patients who underwent implantation of a Sprint Fidelis (Minneapolis, Minnesota) lead at our center. We primarily implanted models 6931 and 6949. A lead was considered fractured based on the definition proposed by Farwell et al2: a sudden increase in long-term pacing or defibrillation impedance (>20% increase over a 24-hour period) and/or inappropriate shock(s) secondary to sensing of electrical noise artifacts from nonphysiologic make-break potentials. Clinical and device data were obtained from the patient clinic and hospital record. Age, gender, vein of access, previous ipsilateral device implantation, most recent left ventricular ejection fraction (LVEF), and lead fracture and survival data were recorded for fractured leads. For nonfractured leads, age, gender, most recent LVEF, and survival data were recorded.
Survival function and cumulative hazard for lead fracture were estimated using the Nelson-Aalen estimator with confidence intervals (CIs) based on log transformation.4 To assess whether the cumulative hazard was changing over time, the log of the cumulative hazard was plotted against the log of follow-up time in months. An ordinary least squares regression model was fit to these data to quantify the rate of change of the cumulative hazard of lead fracture over time. As another approach to depict the change in risk over time, conditional survival curves, calculated based on the Nelson-Aalen estimator, which began follow-up at 1 year, 2 years, and 3 years, were derived.4
Results
From September 2004 to July 2007, 426 Sprint Fidelis leads (292 model 6949, 132 model 6931, and 2 model 6948) were implanted at our center. Median follow-up time was 2.3 ± 1.0 years and 78 patients died. Total follow-up was 1,056 years. As of March 28, 2009, 38 lead fractures occurred 233 to 1,466 days after implantation (Table 1). Four of the 38 patients with lead fracture died; however, no known deaths related to lead fracture were realized. From January 2004 to January 2009, a spectrum of other ICD leads was implanted, with a very low lead-fracture rate. Table 2 lists fracture rates of individual lead models where >50 implantations were performed. Among ICD leads with <50 implantations, no fractures were noted.
Table 1. Details of Fidelis lead failures
| Patient | Age at Implantation (years)/Gender | Age of Lead at Fracture (days) | Presentation | Most Recent LVEF | Indication | Outcome |
|---|---|---|---|---|---|---|
| 1 | 56/M | 1,152 | IS–1 | 25 | CAD | SC, LR, Ext |
| 2 | 56/M | 344 | alert–1 | 25 | CAD | SC, LR |
| 3 | 59/M | 1,293 | IS–2 | 25 | CAD–2 | DC, LR |
| 4 | 63/M | 1,060 | IS–12 | 30 | CAD | SC, LR |
| 5 | 65/M | 356 | IS–1 | 25 | CAD | BIV, LR⁎ |
| 6 | 67/M | 959 | alert | 35 | CAD | SC, LR |
| 7 | 68/M | 885 | IS–2 | 25 | CAD | SC, LR |
| 8 | 68/M | 1,385 | IS–8 | 50 | CAD | DC, LR |
| 9 | 68/F | 1,271 | IS–7 | 35 | CAD–2 | DC, LR |
| 10 | 69/M | 829 | alert | 55 | CAD | SC, LR, Ext |
| 11 | 70/M | 372 | IS–21 | 25 | CAD | BIV, LR⁎ |
| 12 | 71/M | 233 | failure to pace | 32 | CAD | DC, LR, Ext |
| 13 | 72/M | 619 | failure to pace | 35 | CAD | SC, LR, PM Dep |
| 14 | 73/M | 326 | IS–9 | 45 | CAD | SC, LR, Ext |
| 15 | 73/M | 313 | IS–15 | 45 | CAD | SC, LR |
| 16 | 74/F | 797 | alert | 33 | CAD | SC, LR |
| 17 | 74/M | 1,292 | IS–54 | 30 | CAD–2 | DC, OFF⁎ |
| 18 | 75/M | 781 | alert | 20 | CAD | DC, LR, Ext |
| 19 | 77/M | 938 | IS–5 | 50 | CAD–2 | SC, LR |
| 20 | 82/M | 593 | failure to pace | 30 | CAD | DC, LR, PM Dep |
| 21 | 31/M | 926 | IS–2 | 55 | IDC | SC, LR, Ext |
| 22 | 52/F | 533 | — | 65 | IDC | BIV, LR |
| 23 | 58/M | 1,061 | IS–11 | 50 | IDC | BIV, OFF |
| 24 | 60/M | 727 | IS–8 | 28 | IDC | DC, LR |
| 25 | 64/M | 1,219 | alert | 20 | IDC | SC, LR, Ext |
| 26 | 65/M | 598 | — | 10 | IDC | BIV, LR |
| 27 | 67/F | 1,112 | alert | 28 | IDC | BIV, LR, Ext |
| 28 | 70/F | 961 | alert | 30 | IDC | SC, LR⁎ |
| 29 | 71/M | 260 | oversensing | 25 | IDC | DC, LR |
| 30 | 74/F | 818 | — | 35 | IDC | DC, LR |
| 31 | 76/M | 1,313 | IS–4 | 23 | IDC | BIV, LR |
| 32 | 82/F | 1,466 | IS–42 | 35 | IDC | BIV, LR |
| 33 | 21/M | 662 | IS–43 | 60 | HC | SC, LR |
| 34 | 22/F | 948 | IS–1 | 55 | HC | SC, LR, Ext |
| 35 | 54/F | 1,336 | IS–1 | 70 | HC–2 | SC, LR |
| 36 | 14/F | 728 | oversensing | 60 | LQTS–2 | SC, LR, Ext |
| 37 | 35/M | 692 | alert | 65 | VT–2 | SC, LR, Ext |
| 38 | 51/M | 1,237 | IS–6 | 45 | VT–2 | SC, LR |
⁎Deceased. |
Table 2. Fracture rates of commonly implanted implantable cardioverter–defibrillator leads: January 2004 to January 2009
| Implanted Lead Model | No. Implanted | Total Fractured (%) |
|---|---|---|
Model 6947 (Medtronic) | 221 | 1(0.45%) |
| Model 0185(Boston Scientific) | 290 | 0(0.0%) |
| Model 0184(Boston Scientific) | 67 | 0(0.0%) |
| Model 0158(Boston Scientific) | 72 | 0(0.0%) |
| Model 1581(St. Jude Medical) | 113 | 2(1.76%) |
| Model 7121(St. Jude Medical) | 71 | 0(0.0%) |
Table 3 lists clinical characteristics of the fractured and nonfractured Sprint Fidelis leads. Among the 38 Sprint Fidelis lead fractures, 8 patients (21.1%) had previous ipsilateral device implantation and 35 (92.1%) underwent implantation by axillary vein access. Twenty patients (52.6%) underwent implantation due to ischemic heart disease, 12 (31.6%) for idiopathic dilated cardiomyopathy, 1 (2.6%) for long QT syndrome, 3 (7.9%) for hypertrophic cardiomyopathy, and 2 (5.3%) for sustained idiopathic ventricular tachycardia with normal heart function. Of the 38 patients, there were 8 in whom a secondary prevention strategy was employed. There were 4 deaths (10.5%) in this population, none of which was felt to be related to the Fidelis fracture. There were 74 deaths (19.1%) in the nonfractured population during the same period.
Table 3. Clinical characteristics of fractured and nonfractured Sprint Fidelis leads
| Fractured Fidelis | Nonfractured Fidelis Leads | |
|---|---|---|
| No. | 38 | 388 |
| Age at implantation (years), mean ± SD | 61.7±16.7 | 67.6±14.8 |
| Model 6949 (%) | 25(66%) | 267(69%) |
| Model 6931 (%) | 13(33%) | 119(31%) |
| Model 6948 (%) | 0(0%) | 2(0.5%) |
| Men (%) | 28(74%) | 289(74%) |
| Most recent LVEF (%), mean ± SD | 37±14.9(average 169 days after implantation) | 30±14(average 294 days after implantation |
| Age of lead at most recent follow-up (days), mean ± SD | 852±355 | 897±379 |
| Deaths (%) | 4(10.5%) | 74(19.1%) |
The observed Sprint Fidelis fracture rate to date is 8.92% (after mean follow-up 2.3 ± 1 years) with a failure rate of 3.60%/year. The cumulative hazard of Fidelis lead failure is shown in Figure 1. The hazard of fracture increased exponentially over time in our sample by a power of 2.13 (95% CI 1.98 to 2.27, p <0.001). Figure 2 shows the results of a linear regression of log cumulative hazard by log time (months).
Figure 1.
Cumulative hazard of Sprint Fidelis implantable defibrillator lead failure (solid line) ± 95% CIs (dashed lines).
Figure 2.
Log time plotted against log cumulative hazard (slope 2.13, 95% CI 1.98 to 2.27, p <0.001).
The 3-year survival for Fidelis leads was 90.8% (95% CI 87.4%, 94.3%). Conditional survival probabilities for Fidelis leads functioning normally after 1 year, 2 years, and 3 years are shown in Figure 3. If a Fidelis was functioning normally at 1 year, the chance it would survive another year was 97.4% (95% CI 95.7 to 99.1); if functioning at 2 years, the chance of surviving another year was 94.7% (95% CI 91.8 to 97.7); and if functioning at 3 years, the chance of surviving 1 more year was 86.7% (95% CI 78.8 to 95.5).
Figure 3.
Conditional survival probabilities for Sprint Fidelis implantable defibrillator leads that are functioning normally 1 year, 2 years, and 3 years after implantation.
Fracture of the pace sense conductor caused 35 of the 38 (92%) Fidelis lead fractures. Fracture of the high-voltage conductor was not specifically detected in our population; however, detailed interrogation data were not available in 3 patients. Oversensing of nonphysiologic signals resulted in 22 patients (58%) receiving inappropriate shocks. The average number of shocks was 11.6 (range 1 to 54). Oversensing without inappropriate shocks was observed in 5 patients. Three of the 5 had oversensing that resulted in inhibition of pacing (Table 1). Two of the 3 patients reported presyncope or syncope. No patient was found to have died from causes specifically related to the Fidelis lead fracture.
Twenty (52.6%) of 38 fractured leads were single chamber, 10 (26.3%) were part of a dual-chamber system, and 8 (22.2%) were part of a biventricular system. Eleven patients (28.9%) underwent extraction of the fractured lead and no significant complications were reported with any extraction procedure. Thirty-six of 38 patients (94.7%) underwent placement of a new right ventricular ICD lead. In 2 patients, based on clinical circumstances, the decision was made to not revise the known fractured lead.
Discussion
ICD lead failure represents a major clinical event often resulting in multiple inappropriate shocks that affect quality of life and may have fatal consequences.5, 6, 7 Recently, the Fidelis lead has been found to have a higher than expected failure rate, although considerable controversy exists as to the frequency of fracture.1, 2, 3 Our large, single-center experience found a 3.60% fracture rate per year similar to findings by Hauser and Hayes1 and Farwell et al.2 The rate we observed is considerably higher than the approximately 3% fracture rate at 3 years seen in the Medtronic CareLink study.
Hauser and Hayes,1 in their recently reported analysis on 848 Sprint Fidelis leads found that, compared to other defibrillator leads in their population, the Sprint Fidelis failure rate was 3.75% versus 0.58%/year. The 3-year estimated survival was 87.9% compared to 95% for other leads. Eighty-seven percent of their patients had fractures in the pace sense conductor and 1/2 of the patients received inappropriate shocks. Overall, 72 of 848 (8.5%) Sprint Fidelis leads failed over 1.9 ± 1.3 years of follow-up. They noted that the hazard of Sprint Fidelis lead failure in their patients was increasing.3 Our data show a strikingly similar overall fracture rate (8.92%), with a 3-year survival of 90.8%. The work by Farwell et al2 described a similar scenario. Seventeen of 480 model 6949 leads failed and they concluded that the risk of lead fracture increased with time by a power of 2.74 over 19.8 months.
We observed an increasing rate of fracture with increased time since implantation. Our data show that the risk of Sprint Fidelis lead failure increased exponentially with time. In particular, the chance of failure in year 2 was 2.6%, but increased to 13.3% for a lead in the fourth year after implantation. These data confirm the findings of other recent studies also observing that the fracture rate increased with longer time after implantation.3, 4 Our data confirmed that the Fidelis leads' failure rate was higher than that for other ICD leads implanted concurrently. The Fidelis lead fracture rate was 3.6% versus <0.5%/year for the other ICD leads implanted at our center (Table 2). Similarly, the non-Fidelis fracture rate observed by Hauser and Hayes1 was 0.582% compared to 3.75%/year for Fidelis leads in their study.
Third, we observed a trend, although not statistically significant, that patients who developed Fidelis lead fractures had a higher follow-up EF (37% vs 30%, p = NS) than patients who did not exhibit fractures. In our patients, LVEF was found to be predictive of lead fracture based on analysis of LVEF in its continuous form in the Cox proportional hazard model. For every 10% increase in LVEF, the risk of lead fracture increased by 30% (hazard ratio 1.30, 95% CI 1.05 to 1.61, p = 0.013; Figure 4). If fracture is more common in patients with more preserved EF, it is not known whether this is due to a greater local stress placed on the lead in the heart or whether it relates to greater upper extremity activity that could explain the difference. In our study neither age nor gender was found to be predictive of lead fracture.
Figure 4.
Cumulative probability of Sprint Fidelis implantable defibrillator lead failure based on LVEF groups (<20%, 20% to 45%, >45%).
The reason for the discrepant rates reported for lead fracture remains unclear. Lead fractures in general have been ascribed to lead design, implantation technique, and patient factors.1, 8 Different datasets also could differ in completeness of follow-up. At least 1 report has found that the cephalic approach may carry a lower rate of lead fracture.1 Nearly all of our implants used the extrathoracic axillary approach versus the intrathoracic subclavian vein or cephalic approaches. The higher fracture rate in our study could be due in part to the noncephalic approach, although the very low fracture rate observed for other leads in our center and the favorable characteristics of this approach argue against this explanation.
We concur with the advocacy that detection of a lead fracture mandates a rapid response to deactivate the device as rapidly as possible, i.e., within hours of diagnosis. In addition, use of sophisticated algorithms to minimize inappropriate shock is advisable. Increasing fracture rates with time tends to make a stronger case for lead replacement in a patient already undergoing ICD system operation, although such decisions should be individualized, weighing age, venous patency, pacemaker dependency, other morbidities, and patient wishes. Overall, we believe further data showing a higher fracture rate would be needed before such a recommendation would be appropriate.
Of the 38 fractured leads at our center, 11 were extracted. This decision was approached on a case-by-case basis and the extraction procedures were uncomplicated and all patients have not reported further lead problems. Furthermore, we feel it is imperative that with the significant risk of inappropriate shocks, even with the Lead Integrity Alert software (Minneapolis, Minnesota) programmed into the device, the patient should be evaluated quickly and appropriate changes in device function should be made to avoid inappropriate shocks. To date, data from Medtronic, this study, and others do not reveal a trend toward increased mortality associated with lead fracture or the occurrence of an inappropriate shock or shocks in the setting of a Fidelis lead fracture. However, repetitive ICD shocks are known to lead to detectable troponin increases consistent with myocardial injury and/or activation of the neurohormonal cascade leading to worsening heart failure and ultimately cardiovascular mortality.9, 10, 11, 12
The limitations of this study include a relatively small sample and retrospective method. There were no data on the site of lead fracture in our individual cases. Mode of death was not determined in patients without Fidelis lead fracture, and in the absence of postmortem interrogation, we cannot completely exclude that these patients had developed a lead fracture or that such a potential fracture could have proved fatal. Despite these limitations, we believe that these data significantly add to the survival analysis of these leads and substantiate the increasing hazard of Fidelis lead fracture in a time-dependent manner.
References
- . The increasing hazard of Sprint Fidelis implantable cardioverter-defibrillator lead failure. Heart Rhythm. 2009;6:605–610
- . Accelerating risk of Fidelis lead fracture. Heart Rhythm. 2008;5:1375–1379
- . Canadian Heart Rhythm Society Device Advisory Committee: outcome of the Fidelis implantable cardioverter-defibrillator lead advisory: a report from the Canadian Heart Rhythm Society Device Advisory Committee. Heart Rhythm. 2008;5:639–642
- . Survival Analysis. In: New York: Springer; 1997;p. 536
- . Inappropriate implantable cardioverter-defibrillator shocks in MADIT II: frequency, mechanisms, predictors, and survival impact. J Am Coll Cardiol. 2008;51:1357–1365
- . Inappropriate therapy and fatal proarrhythmia by an implantable cardioverter-defibrillator. Heart Rhythm. 2005;2:307–309
- . Gross and microscopic pathological changes associated with nonthoracotomy implantable defibrillator leads. Circulation. 1998;98:1517–1524
- . Lead and device related complications in the Antiarrhythmic Versus Implantable Defibrillators (AVID) trial. Am Heart J. 2001;141:92–98
- . Electrical storm in patients with an implantable defibrillator: A matter of definition. Ann Noninvasive Electrocardiol. 2007;12:375–382
- . Electrical storm presages nonsudden death: the Antiarrhythmics Versus Implantable Defibrillators (AVID) trial. Circulation. 2001;103:2066–2071
- . Recurrent ventricular arrhythmia storms in the age of implantable cardioverter defibrillator therapy: a comprehensive review. Prog Cardiovasc Dis. 2008;51:229–236
- . Effects of repeated electrical defibrillations on cardiac troponin I levels. Am J Cardiol. 1999;83:270–272
PII: S0002-9149(09)02220-6
doi:10.1016/j.amjcard.2009.08.655
© 2010 Elsevier Inc. All rights reserved.
