Characteristics of Sprint Fidelis lead failure

Abstract

Background. The Medtronic Sprint Fidelis ICD lead is prone to failure and the rate of failure seems to be increasing. The aim of this study was to investigate the rate of Sprint Fidelis lead failure, the characteristics, the mode of presentation and possible predictors of lead failure.

Methods and Results. The rate, characteristics and presentation of Sprint Fidelis lead failure was assessed in this single-centre survey. 619 Sprint Fidelis ICD leads were implanted at our centre between December 2004 and August 2007. The mean follow-up was 32±10 (range 22–60) months; 35 patients (5.7%) required a lead re-implantation because of failure of the pace-sense conductor. Mean duration of lead survival was 23±12 (2–46) months and the rate of failure did not stabilise during follow-up. The mode of presentation was inappropriate shocks in 16 patients (45.7%), alarm alert in 12 patients (34.3%), and detection at routine follow-up in seven patients (20%). In 31 patients (89%), interrogation data revealed a sudden rise in impedance and/or frequent short VV intervals prior to lead failure and in five patients an isolated decrease of R wave (<2.5 mV). The interrogation data were not different from patients with shocks compared with patients without shocks. The interrogation data at routine follow-up in the first three months after implant were normal and stable.

Conclusion. The rate of Sprint Fidelis lead failure reaches 5.7% at a mean follow-up duration of 32 months. The rate of failure does not seem to stabilise. Routine follow-up can not predict lead failure or prevent inappropriate shocks. (Neth Heart J 2010;18:12-7.)

The number of implantable cardioverter defibrillator (ICD) implants is increasing every year. ICD lead failure is fortunately a rare but a potentially very serious complication. It might result in failure to deliver therapy or result in inappropriate shock delivery, therefore ICD lead failure is associated with morbidity, mortality, increased health care costs and psychological distress.^1,2 In April 2007 first reports regarding Sprint Fidelis pace-sense lead malfunction were published and the manufacturer decided to suspend further distribution in October 2007.^2,3

A few studies have been published describing failure rates and trying to discover predictors for failure. The aim of this study was to describe patients presenting with a Sprint Fidelis lead failure and particularly the mode of presentation and implications of this lead issue on patient management. This paper emphasises that close monitoring with remote care and frequent outpatient hospital visits is still mandatory.

Methods

A total of 619 patients undergoing a Sprint Fidelis ICD lead implantation at the Isala Clinics in Zwolle, the Netherlands were included in this study. Patients with lead failure were identified and clinical data were collected from the patient records. ICD implantation reports and device interrogation data were retrieved from technical reports. We evaluated device interrogation details at the moment of implantation, the following day, three months after implantation, the last visit before lead dysfunction and at the time the patient presented with lead failure. A sudden nonphysiological change in impedance or sensed R wave or frequent short VV intervals was regarded as a lead dysfunction. Different modes of presentation were compared: sudden increase in impedance, frequent short VV intervals, sudden decrease of R wave and inappropriate ICD shocks.

Statistical analysis

Continuous variables were expressed as a mean ± standard deviation and were compared using an unpaired Student's T test. Categorical variables were compared using the Χ² test. A Kaplan-Meier curve was created to demonstrate lead survival. A p value <0.05 was considered to be statistically significant. Analyses were performed using SPSS 15.0.

Results

Patients

Between December 2004 and August 2007, 619 Sprint Fidelis ICD leads were implanted at the Isala Clinics. The mean follow-up was 32±10 months. Thirty-five patients (5.7%) required a lead re-implantation because of failure of the pace-sense conductor. There were no failures of the shock conductor. The mean duration of lead survival was 23±12 months. The survival of the Sprint Fidelis leads implanted at our institution is shown in figure 1. Clinical characteristics of patients with lead failure are shown in table 1. Most patients suffered from ischaemic cardiomyopathy (57.1%). The favourite position for the tip of the shock lead was the right ventricular apex (91.4%) and in the majority of patients the subclavian vein was used for entrance (88.6% vs. 11.4% cephalic vein approaches).

Figure 1.

Sprint Fidelis lead survival. Number of patients at risk at 0, 20, 40 and 60 months

Table 1.

Patient characteristics.

Age	66±14 (21-81)
Gender (M/F)	25/10
Cardiomyopathy
- Ischaemic	20 (57.1%)
- Dilating	9 (25.7%)
- Hypertrophic	3 (8.6%)
- No cardiomyopathy	3 (8.6%)
Ejection fraction (%)	31±13 (range 10-60)
QRS >120msec	10 (28.6%)
BMI	25.2 ± 4.3 (range 20-38)
Implanted system
- Biventricular device	7 (20%)
- Single chamber	11 (31.4%)
- Dual chamber	17 (48.6%)
Lead tip position
- Apex	32 (91.4%)
- RVOT	2 (5.7%)
- Septum	1 (2.9%)
Vein used for insertion
- Subclavian	31 (88.6%)
- Cephalic	4 (11.4%)
Time to revision (months)	23 ±15 (range 2-46)

BMI=Body Mass Index; VOT=right ventricular outflow tract.

Mode of presentation of Sprint Fidelis lead failure

As shown in figure 2, in 16 patients (45.7%) the mode of presentation was inappropriate shocks. Twelve patients had multiple shocks ranging from two to 16 shocks at the time of presentation. Ten patients presenting with shocks had a sudden non-physiological increase of impedance. Fourteen patients had frequent short VV intervals and in two patients the only finding was an isolated pronounced decrease in sensed R wave (R wave <2.5 mV). Both patients had a dislodgement of the leadtip; one was found in the right atrium and the other was retracted in the caval vein. In one case of multiple inappropriate shocks a transoesophageal echocardiography showed that the tip perforated the intraventricular septum. Interrogation data in this patient detected frequent short VV intervals but a normal impedance and sensed R wave. There was one patient with inappropriate shocks who had a lead integrity algorithm (LIA) installed a few months before presentation. At the moment of data analysis, three of the 35 patients with lead failure had the LIA software installed.

Figure 2.

This diagram shows the mode of lead failure presentation. See text for details.

Nineteen patients with lead failure (45.3%) did not present with (inappropriate) ICD shocks. In 12 of these patients the alarm alert was triggered because of frequent short VV intervals or a sudden increase in impedance. In six patients lead failure was discovered at their routine follow-up: two patients had an increase in impedance and four patients a pronounced decrease of sensed R wave. There was one patient who presented in the outpatient clinic with chest pain after trauma (a closet fell on his chest) and appeared to have a sudden loss of sensed R wave.

There were no statistical differences in baseline characteristics and interrogation data between patients with shocks compared with patients without shocks and no difference in patients presenting with a sudden increase in impedance compared with patients with a decrease in sensed R wave. The combination of interrogation data abnormalities are illustrated in figure 3. Figure 4 shows a typical example of intracardiac electrograms of a failing Sprint Fidelis lead.

Figure 3.

Venn diagram showing number of patients with characteristic lead performance abnormalities.

Figure 4.

A representative example of stored intracardiac electrogram from a patient with a Sprint Fidelis lead who presents after having received 16 inappropriate ICD shocks. All shocks were inappropriate due to oversensing (diagram A). Interrogation revealed >50,000 short VV intervals and a sudden impedance rise to >3000 Ω. Diagram B shows the sudden onset of ventricular oversensing.

Discussion

Main findings

Sprint Fidelis pace-sense ICD lead failure is a major concern. This slender lead has a vulnerable pace-sense unit that is prone to fracture and can cause inappropriate ICD shocks. Several studies have been published with a varying incidence ranging from 2 to 3.3%.^1-3 We found a failure rate of 5.7%. It seems that the incidence of lead failure will increase over time (figure 1). The manufacturer withdrew the Sprint Fidelis leads from the market after observing that its 30-month failure rate had reached 2.3%. In late 2008 Medtronic announced that the lead's failure rate had climbed to as high as 6.3% at 42 months.

Comparison with previous studies

Recently Hauser and Hayes found a Sprint Fidelis failure rate of 3.75% per year during an average followup of 27 months.⁴ Furthermore, the rate of failure appeared to accelerate after the first implant year whereas the failure rate of other lead models was low (0.58% per year) and remained stable. The authors support the idea that in selected patients a prophylactic lead replacement should be considered during ICD generator replacement.

The mechanism and predictors of lead failure remain unexplained. Farwell and colleagues identified two independent predictors of lead failure: 1) left ventricular ejection fraction (LVEF), i.e. an increased LVEF increases the risk of failure, and 2) a cephalic venous access was associated with a reduced risk of failure compared with subclavian vein access.⁵ The reason that we found a higher failure rate compared with previous reports might be related to a longer follow-up. As mentioned before, lead failure is not a phenomenon that occurs primarily in the first period postimplantation; in fact the hazard of fracture increases with time.⁵ Another explanation of the higher than expected failure rate in our patients might be the relatively high rate of subclavian vein approach: 88.6 vs. 11.4% cephalic approaches. Farwell et al. had 76% subclavian approaches.⁵ A third reason for the higher failure rate could be the experience of the implantor. It has been suggested that the risk of lead fracture is related to the duration of the implantation procedure and experience of the implantor.⁶ The Isala Clinics is a teaching hospital where part of the procedures are performed by a cardiologist in training under the supervision of an experienced electrophysiologist. On the other hand, Krahn et al. did not find evidence that lead failure is operator dependent in their multicentre study.⁷

Importance of remote care and lead integrity alert software

The majority of patients with lead failure, with or without inappropriate shocks, have a sudden rise of pacing impedance. Kallinen et al. emphasise that a large proportion of patients presenting with inappropriate shocks had a rise in impedance within 24 hours before ICD shock delivery.² A more frequent follow-up would, therefore, not prevent inappropriate shocks. The audible alarm triggered on impedance change seems a more reliable tool in early detection of lead failure. However, the alarm is often not heard and in some cases the shocks are delivered just a few hours after onset of impedance change. Remote monitoring with a CareLink system is an instrument to perform device interrogation at home, usually when the patient is asleep. As a daily interrogation is performed, a potential lead failure is detected in an early phase. Therefore, the majority of inappropriate shocks could be prevented.⁸ Another important method for early identification of lead failure and to reduce the number of inappropriate shocks is the Lead Integrity algorithm (LIA) software developed by Swerdlow et al.⁹ The algorithm can be downloaded in ICD generators and initiates an alert when triggered by either oversensing or an excessive increase in impedance. Once the LIA is triggered it increases the programmed number of intervals to detect (NID) to reduce the likelihood of an inappropriate shock. According to the authors the LIA could prewarn 95% of patients with a lead fracture and subsequently could prevent inappropriate shocks. Recently a modification was developed to the LIA. Both of the LIA and the modified LIA systems have three components: high impedance and two oversensing components, i.e. short VV intervals (RR ≤130 msec) and ≥2 rapid nonsustained VTs (mean RR ≤220 msec). Whereas the original LIA triggers on high impedance or both oversensing criteria, the modified LIA requires two of the three components. The modified LIA appeared to have a significantly higher positive predictive value for ICD lead fracture (85.7 vs. 73.2%) and less false positives.¹⁰ However, the LIA system is only available for Medtronic ICDs. In our study all patients who required lead replacement had a Medtronic ICD generator. Only three patients had the LIA software installed when admitted for lead revision. One of these three patients presented with inappropriate shocks.

Sprint Fidelis lead failure mainly affects the pace-sense conductor as manifested by sudden increase in pacing impedance and oversensing of non-physiological lead signals. However, Kenigsberg et al. found that 50% of lead failures were detected by a R-wave diminution to <5 mV. A multivariate regression analysis revealed that the mean R-wave amplitude during the first four days postimplantation was an independent predictor of future R-wave diminution.¹ We found a loss of R wave in nine patients with lead failure (25.7%). Five of these patients had a pronounced decrease in R wave with a normal impedance and no frequent sensed short VV intervals. Although these patients required a lead replacement, it is unclear if this phenomenon (isolated loss of r wave) is related to the increased vulnerability of the Sprint Fidelis lead. We did not find a relation between the R wave amplitude in the early days after implantation and the loss of R wave when patients presented with a lead fracture.

Prophylactic lead replacement in certain patient categories?

Often lead failure is a very sudden process and inappropriate shocks are arrhythmogenic, exposing patients to an increased risk of fatal arrhythmias rather than preventing SCD. Although identifying risk factors, adjusting ICD programming and intensifying followup may prewarn patients with a failing Sprint Fidelis lead and may reduce inappropriate shocks, one should consider prophylactic lead replacement in patients with a high risk of ventricular arrhythmias as mentioned earlier. The question is which patient should have a prophylactic lead replacement and which patient can be ‘safely’ managed with intensive and adjusted followup. Priori et al. introduced a model for this decisionmaking process for the prophylactic replacement of ICD generators that have been listed on the advisory letter.¹¹ The model is based on the risk/benefit ratio of a replacement procedure. The ‘number needed to replace’ (NNR) provides an estimate of the number of lives saved when replacing an ICD that has been listed on the advisory letter. Four parameters are used for the NNR formula: expected annual SCD rate, residual device life, difference in failure rate between the device listed on the advisory letter and the replacement device and the placement procedure mortality risk. Although this model is developed for ICD generator failures a modified formula for lead failure could guide the clinician in the decision of prophylactic lead replacement. Priori et al. point out that the NNR methodology for lead issues would be more complicated for several reasons. First the consequences of lead failure such as undersensing or oversensing resulting in inappropriate shocks and non-delivery of shocks are in general more complicated than a dysfunctional generator. Secondly, the risks associated with lead replacement are considerably greater for replacing ICD leads than pulse generators.¹¹ At the moment the benefit of prophylactic lead replacement does not seem to outweigh the risks associated with the procedure. Therefore it is only recommended in exceptional cases. The manufacturer's independent physician quality panel advises on Sprint Fidelis issues, the patient management recommendations are shown in table 2.

Table 2.

Manufacturer's Sprint Fidelis patient management recommendations.

• When a lead fracture is suspected or confirmed, prompt patient attention is recommended. Patients should contact their physician without delay if they experience unexpected shocks

• The Lead Integrity Alert (LIA) is expected to provide three days advance notice prior to inappropriate therapy to 76% of the patients with lead fractures. As a result, we strongly recommend that all Sprint Fidelis patients who have the ability to upgrade to Lead Integrity Alert do so promptly

• The risk of prophylactic intervention appears to be greater than the risk of serious injury resulting from lead fracture even for pacemaker-dependent patients, except in select individual patient circumstances as determined by the physician

• Special circumstances may apply to device change-out or upgrade procedures when a lead fracture has not occurred. At least four options are available, each of which carries risks and benefits that should be taken into consideration:

◦ Leave a properly performing lead intact; this is likely to be the best choice for the majority of patients

◦ Place a new ICD lead without extraction of the existing lead

◦ Place a pace-sense lead without the extraction of the existing lead. This option reflects the observation that approximately 90% of Fidelis failures are related to fractures in the pace-sense circuit. It is unknown what the failure rate of the high voltage conductor would be should a pace-sense conductor failure occur in the existing Sprint Fidelis lead

◦ Unusual patient circumstances may warrant extracting and implanting a new ICD lead. Factors to consider when making this decision include patient life expectancy, age and co-morbidities, number of implanted leads and duration of implant, and patient preference. Medtronic's Independent Physician Quality Panel recommends that if a lead requires removal, the procedure be performed by a physician with extensive lead extraction experience

Conclusions

Previous reports on Sprint Fidelis lead failure vary in incidence and mode of presentation, which can be explained by the fact that different definitions of lead failure are used and that most of the studies are relatively small sized, retrospective, single-centre studies. Although one cannot prevent lead fracture, the number of patients presenting with inappropriate shocks can be reduced. Identifying the mechanism and risk factors for lead failure is crucial for developing an adequate system to predict lead failure, to prevent inappropriate shocks or failure to deliver therapy and for the decision-making process of prophylactic lead replacement. Therefore, more research preferably multicentre national analysis is needed.