Outcome of the Fidelis implantable cardioverter-defibrillator lead advisory: A report from the Canadian Heart Rhythm Society Device Advisory Committee
Article Outline
Background
The Medtronic Sprint Fidelis family of leads has recently been the subject of a widespread advisory. Lead failure rates are estimated at 2.3% at 30 months, 2.6 times the failure rate of the reference Medtronic 6947 lead.
Objective
The purpose of this study was to contact pediatric and adult implantable cardioverter-defibrillator (ICD) implant centers across Canada to determine the short-term response to the October 15, 2007 Medtronic Fidelis lead advisory.
Methods
All centers completed an 11-part survey to assess the frequency and presentation of lead failure, operator characteristics, and center's response.
Results
Lead failure was noted in 80 (1.29%) of 6,181 patients at 21.0 months, with inappropriate shocks experienced in 45 (56%) of the 80 patients (overall risk 0.73%). No deaths were attributed to lead failure. Sensing was the primary form of failure, seen in 60 leads (75%), with pacing failure in 10 (13%), and high-voltage failure in 15 (19%). Assessment of the previous routine ICD interrogation prior to the advisory or lead failure demonstrated evidence of altered lead performance in only 8 (10%) of the 80 leads. Inappropriate shocks typically were multiple (median 7, range 1–122), with a single shock seen in only 5 patients. Lead failure was noted in 18 of 23 centers, representing 89.8% of leads implanted, with at least one failure noted in 15 of 16 centers that implanted more than 200 leads. Forty-seven of the 135 operators in the 23 institutions implanted the 80 leads that subsequently failed. Only 16 operators were involved in more than a single lead that subsequently failed; seven operators participated in three or more leads that subsequently failed. Seven centers planned to replace leads in most pacing-dependent patients, and two centers planned to replace leads in patients unable to hear the alert tone.
Conclusion
This national experience suggests a Fidelis lead failure rate of 1.29% at 21 months, most often presenting with multiple inappropriate shocks without evidence of impending failure from routine lead follow-up. Lead failure did not appear to cluster around specific operators or around high-volume or low-volume implant centers.
Keywords: Defibrillator, Lead, Advisory, Recall, Lead failure
Introduction
The recent increased incidence and reporting of implantable-cardioverter defibrillator (ICD) device advisories has created the need for rigorous data upon which to make clinical decisions.1, 2, 3 The clinical implications of these advisories have caused management dilemmas for physicians. The Heart Rhythm Society's guidelines were published to provide physicians with clinical direction to enable appropriate management decisions in such circumstances.4 These guidelines called for collection of accurate data to guide decisions regarding competing risks for patients in the face of an advisory.5 Further multicenter data that can be used to substantiate current management strategies and guidelines are needed.6
On October 15, 2007, Medtronic announced a lead advisory on the Sprint Fidelis family of ICD leads in response to a growing concern regarding conductor fractures that were first detected by Medtronic in March 2007 and reported in Heart Rhythm in July 2007,7 leading to withdrawal of the product from further sales. The physician advisory letter indicated a 5 in 268,000 preliminary estimate of risk of death. The estimate of lead failure was based on data from a remote monitoring database (CareLink), returned product analysis, a small chronic lead study, and five reported deaths worldwide where the lead was considered a possible or likely contributing factor. We report the short-term Sprint Fidelis advisory experience from an independent group of investigators representing all ICD implant centers in Canada.
Methods
On October 15, 2007, ICD implant centers across Canada were contacted to collect preliminary data on the number of leads and the frequency of lead fracture for development of a Canadian Heart Rhythm Society position statement and member, physician, and patient statements.8 A formal survey was sent to all ICD implant centers in Canada on November 9, 2007, based on an 11-question survey focusing on frequency of lead failure and operator characteristics. At each center, an investigator collected data based on the urgent response of the arrhythmia management team to the advisory. The survey was approved by the University of Western Ontario Ethics Review Board. The study was unfunded and independently performed.
The study involved standardized responses to questions regarding failure frequency, presentation with shocks, lead component failure mechanism, number of involved operators, and anticipated response in patients with normal functioning leads who were pacing dependent or unable to hear the programmed alert tones. The survey covered all four models of Sprint Fidelis leads (models 6930, 6931, 6948 and 6949, Medtronic, Inc., Minneapolis, MN, USA). A standardized datasheet was used to obtain survey data. The diagnosis of lead failure was made by the local investigator. Lead failure was defined as nonphysiologic high-rate sensing with a high sensing impedance suggesting fracture, a change in sensing or pacing impedance suggesting imminent fracture, or a rise in high-voltage impedance suggesting coil fracture leading to a decision to perform lead replacement. Isolated poor R-wave sensing or high pacing threshold without impedance change was not considered lead failure related to the advisory. Investigators classified the primary mechanism of failure as sensing, pacing, or high voltage. Failure could be classified by more than one mechanism if there was clear failure of multiple components, typically both sensing and pacing.
Follow-up of patients was closed December 20, 2007. Follow-up of the cohort was calculated from annual sales numbers provided by the manufacturer based on market release in July 2004 until product withdrawal on October 15, 2007. Follow-up was calculated from time of sale to October 15, 2007, assuming linear sales volumes through the fiscal year.
All 23 centers that were contacted, representing 25 hospitals, agreed to participate and provided complete survey data. No center declined participation. The three pediatric centers were combined to create a single pediatric center experience for purpose of analysis. Data were collected, stored, and analyzed by the Arrhythmia Service at University Hospital in London, Ontario. Patient identity was not collected by the survey and was only used for clinical care purposes within each institution. Data are expressed as mean ± standard deviation or as a percentage unless otherwise specified.
Results
Twenty-one of the 23 centers contacted patients by telephone or registered mail within several weeks of the advisory release (Figure 1). In general, patients were offered assessment and counseling within 1 month, adhering to the manufacturer's recommended changes in programming with use of alert tones. Lead failure was noted in 80 (1.29%) of 6,181 patients at 21.0 months of follow-up. No deaths were attributed to lead failure.
Figure 1.
Volume of leads implanted in each of the 23 centers. Note that three pediatric centers have been combined for purpose of analysis.
Lead failure presented with inappropriate shocks in 45 (56%) of the 80 patients (overall risk 0.73%). Inappropriate shocks typically were multiple (mean 15.5 ± 15.0, median 7, range 1–122), with a single shock seen in only five patients (Figure 2). Sensing was the primary form of failure, seen in 60 leads (75%), with pacing failure in 10 (13%) and high-voltage failure in 15 (19%). Assessment of the previous routine ICD interrogation prior to the advisory or lead failure demonstrated evidence of altered lead performance in only 8 (10%) of the 80 leads.
Figure 2.
Number of inappropriate shocks experienced by patients presenting with lead fracture. Five patients experienced a single shock. Patient 44 experienced 122 shocks. One patient experienced inappropriate shocks while traveling in a foreign country, and device interrogation was not available.
Lead failure was noted in 18 of 23 centers, typically in large-volume centers (Figure 3). These 18 centers represented 89.8% of the total population (5,548 leads). At least one failure was noted in 15 of 16 centers that implanted more than 200 leads. Two fractures (6.3%) were noted in 32 pediatric patients from three centers.
Figure 3.
Individual center failure rates listed by centers from smallest to largest with respect to implant volumes, in the same order as Figure 1.Note the lack of relationship between implant volumes and failure rate, and that higher volume centers all experienced at least one failure. Center 2 experienced two failures (6.3%) among 32 pediatric patients.
Forty-seven of the 135 operators in the 23 institutions were involved in the 80 failed lead implants. Nine operators performed two implants in patients with leads that subsequently failed, and seven performed three or more implants in patients with leads that subsequently failed. All of the latter operators were in centers that implanted more than 300 leads. Remote system monitoring (Medtronic CareLink Network) was available in only 3 of 24 sites. Seven centers planned to replace leads in most pacing-dependent patients prior to obtaining results of the survey, and two centers planned to replace leads in most patients who were unable to hear the alert tone.
Discussion
This multicenter study demonstrated a Sprint Fidelis lead failure rate of 1.29% at 21 months. This estimate is in keeping with previous estimates from two data sources provided by the manufacturer of 1.6%+0.9/−1.8 from the chronic lead study and 0.7%+0.1/−0.2 from the Medtronic CareLink Network, both at 21-month follow-up (Personal Communication, Medtronic, Greenwood method). To date, Medtronic has identified 665 confirmed fractures in returned leads. Estimates from the chronic lead study based on 654 patients suggest a 97.7% lead survival at 30 months, compared to 99.1% for the 6947 lead. Of interest, the risk of failure in the current study is lower than the risk estimate from a combined lead models analysis by Kleemann et al,9 who reported a failure rate of 4.4% at 2 years in 990 patients, escalating to 15% by 5 years. The majority of leads in this study were implanted in the 1990s, so direct comparison to more recent leads may not be valid.
The duration of follow-up of the Sprint Fidelis family of leads is relatively short, making long-term estimates of the slope of the survival curve imprecise. The increased number of failures in the second year combined with the reported suspicious deaths in 5 of 268,000 patients presumably influenced the manufacturer to issue an advisory and recall the existing leads in this context. Of note, the absence of attributable deaths in the current study should not be considered particularly reassuring. Data were not collected on the extent of device interrogation available at individual centers in the context of patient deaths. Ongoing review of what often is scant information is underway at each institution.
The problem in question is a conductor fracture that typically occurs in either the distal portion of the lead or near the anchoring sleeve, resulting in lead fracture with artifactual high rate detection with nonphysiologic rates associated with inappropriate shocks and high lead impedance. The current study suggests that inappropriate shocks are the mode of presentation in about half of patients, lower than initial reports in the advisory but higher than the 33% reported by Kleemann et al9 for other leads. Activation of the alert feature may further reduce the risk of inappropriate shocks, although the time from impedance change to lead failure often is short. The current study also bears out initial reports that the major concern surrounds sensing with noise leading to shocks, and that pacing and high-voltage failure is uncommon. This likely will influence the decision to observe most patients who are pacemaker dependent. Investigators in this study have subsequently expressed reservations about replacing leads in dependent patients in light of the survey results suggesting that pacing failure is distinctly uncommon.
Investigation of operator characteristics did not demonstrate an obvious clustering of cases by operator or center. Individual operators may perform implants in a similar manner within each center, diluting the ability to detect individual operator practice influences. Nonetheless, individual center failure rates showed a range from 0% to 3.2%, if the pediatric experience is excluded. Lead failure combines a range of factors, including lead characteristics, implanting physician, and patient factors. Preliminary data suggest a higher risk of Fidelis facture in younger patients, consistent with previous observations that apply to a wide range of leads. This finding has been attributed to physical activity, which is less likely in an elderly ICD population.9, 10 The root cause of the increased fracture rate has not been established, but the current study supports that lead factors play a strong role. Further details regarding implant technique are crucial for clarification of implant and operator characteristics that may minimize risk of fracture in the future. Unfortunately, this level of detail was not available from the current survey.
The majority of investigators in this study decided not to replace leads in patients who were either pacemaker dependent or unable to hear the alert tones. Many of the investigators in the current study participated in a report that noted a higher than expected complication rate when advisory ICDs were replaced.11 This study reported a major complication rate of 5.8%, with lead extraction for infection in 1.9% of patients, which perhaps influenced the investigators' decision regarding the risk-to-benefit ratio in the current patient population. Despite concerns that this complication rate may have overestimated risk in some practice environments, there is a clear heightened awareness of the risks associated with “reflex” replacement in the context of a recall or advisory.
Conclusion
The current study demonstrates that Medtronic Fidelis leads have a failure rate of 1.3% at 21 months, most often presenting with inappropriate shocks. Lead failure does not cluster within operators or centers, suggesting that inherent lead vulnerability to fracture is the major contributing factor.
Appendix 1
Investigators and Institutional Affiliations
Andrew Krahn, London Health Sciences Center, London, Ontario
Jean Champagne, Quebec Heart Institute, Laval Hospital, Quebec City, Quebec
Jeffery Healey, Hamilton Health Sciences Center, Hamilton, Ontario
Doug Cameron, University Health Network, Toronto, Ontario
Christopher Simpson, Queen's University, Kingston, Ontario
Bernard Thibault, Montreal Heart Institute, Montreal Quebec
Iqwal Mangat, St. Michael's Hospital, Toronto, Ontario
Stanley Tung, St. Paul's Hospital, Vancouver, British Columbia
Laurence Sterns, Vancouver Island Health Authority, Victoria, British Columbia
David Birnie, University of Ottawa Heart Institute, Ottawa, Ontario
Derek Exner, Foothills Hospital, Calgary, Alberta
Ratika Parkash, QEII Health Sciences Centre, Halifax, Nova Scotia
Soori Sivakumaran, Royal Alexander Hospital, Edmonton, Alberta
Ted Davies, Scarborough Centenary Hospital, Scarborough, Ontario
Benoit Coutu, Centre Hospitalier Notre Dame, Montreal, Quebec
Eugene Crystal, Sunnybrook Hospital, Toronto, Ontario
Kevin Wolfe, St. Boniface Hospital, Winnipeg, Manitoba
Atul Verma, Southlake Regional Hospital, NewMarket, Ontario
Elizabeth A. Stephenson, The Hospital for Sick Children, Toronto, Ontario
Shubhayana Sanatani, BC Children's Hospital, Vancouver, British Columbia
Robert Gow, Children's Hospital of Eastern Ontario, Ottawa, Ontario
Sean Connors, Health Sciences Center, St John's Newfoundland
Felix Ayala Paredes, Centre Hospitalier Universitaire de Sherbrooke, Sherbrooke, Quebec
Vidal Essebag, McGill University Health Center and Hôpital Sacré-Coeur de Montréal, Montreal, Quebec
References
- Recalls and safety alerts involving pacemakers and implantable cardioverter-defibrillator generators. JAMA. 2001;286:793–799
- Pacemaker and ICD generator malfunctions: analysis of Food and Drug Administration annual reports. JAMA. 2006;295:1901–1906
- . Pacemaker and ICD generator reliability: meta-analysis of device registries. JAMA. 2006;295:1929–1934
- Recommendations from the Heart Rhythm Society Task Force on Device Performance Policies and Guidelines Endorsed by the American College of Cardiology Foundation (ACCF) and the American Heart Association (AHA) and the International Coalition of Pacing and Electrophysiology Organizations (COPE). Heart Rhythm. 2006;3:1250–1273
- . Lead failures: dealing with even less perfect. Heart Rhythm. 2007;4:897–899
- Multicenter experience with failed and recalled implantable cardioverter-defibrillator pulse generators. Heart Rhythm. 2006;3:640–644
- Early failure of a small-diameter high-voltage implantable cardioverter-defibrillator lead. Heart Rhythm. 2007;4:892–896
- . http://www.chrsonline.ca/patients/advisories.htm
- Annual rate of transvenous defibrillation lead defects in implantable cardioverter-defibrillators over a period of >10 years. Circulation. 2007;115:2474–2480
- . Managing device-related complications and transvenous lead extraction. In: Clinical Cardiac Pacing, Defibrillation, and Resynchronization Therapy. Third Edition. Philadelphia, PA: Saunders Elsevier; 2007;p. 855–930
- . Complications associated with implantable cardioverter-defibrillator replacement in response to device advisories. JAMA. 2006;295:1907–1911
Author affiliations are listed in Appendix 1.
PII: S1547-5271(08)00114-8
doi:10.1016/j.hrthm.2008.01.029
© 2008 Heart Rhythm Society. Published by Elsevier Inc. All rights reserved.
