Author + information
- Received November 16, 2015
- Accepted December 8, 2015
- Published online March 22, 2016.
- aCedars Sinai Heart Institute, Los Angeles, California
- bDivision of Cardiology, Virgina Commonwealth University (VCU) School of Medicine, Richmond, Virginia
- ↵∗Reprint requests and correspondence:
Dr. Kenneth A. Ellenbogen, Virginia Commonwealth University (VCU) School of Medicine, PO Box 980053, Richmond, Virginia 23298-0053.
The implantable-cardioverter defibrillator (ICD) lead is the most vulnerable component of the ICD system. Despite advanced engineering design, sophisticated manufacturing techniques, and extensive bench, pre-clinical, and clinical testing, lead failure (LF) remains the Achilles’ heel of the ICD system. ICD LF has a broad range of adverse outcomes, ranging from intermittent inappropriate pacing to proarrhythmia leading to patient mortality. ICD LF is often considered in the context of design or construction defects, but is more appropriately considered in the context of the finite service life of a mechanical component placed in chemically stressful environment and subjected to continuous mechanical stresses. This clinical review summarizes LF mechanisms, assessment, and differential diagnosis of LF, including lead diagnostics, recent prominent lead recalls, and management of LF and functioning, but recalled leads. Despite recent advances in lead technology, physicians will likely continue to need to understand how to manage patients with transvenous ICD leads.
Implantable-cardioverter defibrillator (ICD) lead failures (LFs) are important to the practicing cardiologist because of the serious consequences if not diagnosed and treated promptly. This review summarizes LF mechanisms, assessment, and differential diagnosis, as well as management of patients with recalled leads or lead advisories. In this review, mechanical LF is used to refer to structural failure of lead materials, which may or may not be identified clinically. By contrast, electrical LF refers to mechanical failures that result in failure of the lead to perform its clinical roles of sensing, pacing, or defibrillation. We recognize that all LFs do not fit neatly into this dichotomy.
Materials, Design, and Functional Considerations
Understanding the mechanisms of LF requires a basic familiarity with materials and structural design. The components of an ICD lead include the conductors, insulation materials, defibrillation coils, lead electrodes, fixation mechanism, yoke (branch point of individual conductor elements), and lead connector. Most newly implanted ICD systems use the DF-4 connection pin, rather than the previous, multicomponent yokes (DF-1 and IS-1). In the DF-4 design, the pace/sense conductor and defibrillation coil conductor(s) connect to a single, multi-interface connection pin. The advantages of this design are reduced pocket bulk and prevention of inadvertent reversal of high-voltage connections. To date, no systematic failures have been reported for DF-4 leads.
Although there are differences in details, most manufacturers use similar materials. All current ICD leads and most in service today are of the multilumen design (Figure 1). Cabled conductors are coated with PTFE (poly-tetrafluoroethylene) and ETFE (ethylene-tetrafluoroethylene) and placed in an insulating silicone cylinder with 3 to 6 lumens, which may be coated with polyurethanes or copolymer materials to accommodate conductors. Low-voltage conductors are typically composed of a multiphase alloy of nickel, cobalt, chromium silver, and molybdenum. High-voltage conductors contain a low-resistance core of silver or platinum.
A central coil conductor used for the pacing cathode (tip) allows for stylet insertion and facilitates extension or retraction of the fixation helix of active fixation leads. Conductors for the pacing anode (ring) and high voltage coils are arranged as parallel cables and distributed around the central coil. Differences in lead designs include symmetric versus asymmetric coil placement and location of cables, and location/inclusion of compression lumens (Figure 2).
All leads have at least 1 distal right ventricular (RV) shock coil. Dual-coil leads have a second shock coil, usually positioned in the superior vena cava (SVC). For left pectoral implants, there is no clinically significant difference in defibrillation efficacy between these 2 designs (1). Dual-coil leads may provide superior defibrillation for some right-sided implants, but they are associated with greater procedural extraction risk due to fibrotic tissue ingrowth into the proximal coil (2).
ICD leads have 2 types of sensing designs, both using the tip electrode as a cathode. The dedicated bipolar lead has a ring electrode as an anode dedicated to sensing. By contrast, the integrated bipolar lead uses the RV defibrillation coil, integrated with the shock circuit, as the anode. Therefore, a dedicated bipolar lead requires 1 more conductor than an integrated bipolar lead. The designs are equivalent with respect to sensing ventricular fibrillation (VF) (3). However, an integrated bipole creates a large “antenna” that is more susceptible to oversensing of diaphragmatic myopotentials, electromagnetic interference (EMI), and atrial far field signals. Dedicated bipolar leads are more susceptible to T-wave oversensing.
The need to implant multiple leads in an individual patient has motivated development of smaller diameter leads. Current ICD leads vary from 6.3 F to 8.6 F.
ICD LF Mechanisms
ICD LF can occur as the result of body/lead interaction, suboptimal implant technique, or intrinsic/random defects occurring anywhere along the lead. The recalled Medtronic Sprint Fidelis (Medtronic, Dublin, Ireland) and St. Jude Medical Riata (St. Jude Medical, Saint Paul, Minnesota) series of leads serve as examples of conductor and insulation failures, respectively.
Fidelis conductor fractures
The root cause of Fidelis conductor fractures is related to the lead’s extreme flexibility, which permitted bending with a short radius of curvature, applying high stress to metal components. More than 90% of Fidelis LFs were caused by fracture of 1 or more pace-sense conductors (4). In about 60% of Fidelis LFs, the cable to the ring electrode fractured distally, just proximal to the ring electrode. In this region, silicone medical adhesive was applied to the outside of the lead (“sealant zone”) to prevent ingress of body fluids, and constrained motion of the cable. This created a hinge point for flexion with each cardiac cycle. Cables provide great strength in extension to support suspension bridges, but they provide much less fatigue strength in flexion. Additional contributing factors are thought to include the length of the sealant zone and 2 factors related to the cable to the ring electrode: specific material characteristics of the MP35N alloy and the 1 × 19 design of the cable filars, which is weaker than the 7 × 7 design used in the more reliable Quattro lead. In about 40% of Fidelis LFs, the helix to the tip electrode fractured proximally, close to the anchor sleeve, because the helix was not stiff enough to prevent bending with a short radius of curvature during high-stress shoulder motion. High-voltage conductors fail in about 10% of Fidelis LFs, with or without pace-sense failures (5). The failure rate for Fidelis is approximately 20% at 10 years.
Riata-family insulation breaches
Silicone insulation is inert and stable in a biological environment; however, it is also soft and has a high coefficient of friction. Thus, it is susceptible to implant damage and to cold flow (“creep”), increasing deformation under a compressive load. Both insulation damage or creep can result in insulation breaches. Abrasions may be external (“outside-in”) abrasions (lead-to-lead, can-to-lead, and yoke-to-lead) from constant compressive loads, or internal (“inside-out”) abrasions caused by cyclical forces exerted by the lead cables, compressing silicone on the outer side of the lead’s curvature. Outside-in abrasions in-pocket may occur with any lead, but the risk probably is highest in silicone leads without an outer protective coating, such as the 8-F Riata and downsized 7-F Riata ST. Such outside-in breaches were the most common cause of electrical LFs in Riata leads.
By contrast, inside-out abrasions were first described in Riata-family leads. Approximately 80% occur between the shock coils and 10% under the coils (6). Presently, there is no consensus about the root cause of Riata’s susceptibility to this type of abrasion. However, recent reports of inside-out abrasions occurring in the Biotronik Linox S or SD lead indicate that the root cause is not limited to Riata-family leads (7). There are important design differences between the 2 leads, but both have a “symmetric” design, with the helical coil to the tip electrode in the center, as opposed to an “asymmetric” design, in which the pacing coil is offset. However, it is not certain whether this design similarity relates to the root cause.
In Riata-family leads without external copolymer tubing (Model 1500 and 7000 series), inside-out abrasions present most commonly as exteriorized cables that can be identified radiographically between the proximal and distal shock coils (8–11). At diagnosis, most leads with externalized cables have normal electrical function due to intact ETFE inner insulation. A meta-analysis of leads assessed by routine, passive monitoring reported externalized cables in 23.1% of leads, but electrical failure in only 6.3% (6). Riata ST Optim and Durata leads have an outer, abrasion-resistant copolymer tubing, originally intended to prevent outside-in abrasions. If inside-out abrasion occurs through the silicone, this tubing contains the cables and prevents exteriorization, providing it remains intact. It is not known whether this tubing exerts forces on the silicone that reduce the underlying process of inside-out abrasion of the silicone lead body. A multicenter study reported no externalized cables in 3-year routine follow-up, but such cases have been reported (11,12). Furthermore, most Optim-coated leads lack Optim between the silicone and shock coil. Thus, inside-out abrasions under the shock coil may result in contact of a conducting cable with a shock coil (13–15). Contact between the proximal coil and cable to the distal coil may short the high-voltage conductors, preventing shock delivery to the patient.
Incidence of LF
The incidence of LF is difficult to determine due to reporting bias. Monitoring of lead performance relies primarily on industry-based, post-market surveillance and voluntary reporting to the Food and Drug Administration (FDA). Most failed leads are not explanted and returned for analysis or reported to the FDA. Furthermore, systematic LF may not be recognized until enough leads have been in service for a sufficient period of time. For instance, a multicenter study, with short follow-up of 18 ± 16.7 months, reported a 0.21% incidence of insulation damage for Riata leads (non-Optim) (10), greatly underestimating the failure rate with longer follow-up. The MAUDE (Manufacturer and User Data Experience) database receives physician reports related to medical devices. In addition to underreporting and reporting bias, the utility of this database is limited by nonvalidated entries.
When recalled models are excluded from analysis, the incidence of LF for currently implanted leads ranges from 0.28% to 1.14%, showing that most leads demonstrate a high level of reliability (16). Lead survival rates for recalled versus nonrecalled leads separate at about 2 years after implantation.
Diagnosis of LF
LF may present with electrical malfunction of pace-sense components, with electrical malfunction of high-voltage components, or mechanical complications. Pace-sense malfunctions are diagnosed most frequently. Oversensing with normal pacing impedance is the initial electrical abnormality with either conductor fracture or insulation breach in ≥60% to 85% of LFs (4,17,18). One analysis of 84 LFs from multiple manufacturers, the initial presentation was oversensing with normal impedance in 70%, oversensing with abrupt change in impedance in 19%, and impedance changes without oversensing in 11%. Because LF-related oversensing is often rapid, it presents most commonly with inappropriate detection of VF in older ICDs. Today, it commonly presents with alerts for suspected oversensing (see the following section). It also presents as inhibition of bradycardia or resynchronization pacing (18). Pace-sense malfunctions may also present with loss of capture, undersensing, or abrupt decrease in R-wave amplitude.
Shock component malfunctions are most commonly identified on the basis of changes in shock impedance and, less commonly, as failed defibrillation shocks. Rarely, shorted high-voltage outputs due to insulation breaches may cause ICD generator failure. The incidence of shock electrogram (EGM) abnormalities is unknown because this is not monitored consistently.
Rarely, LF may present with mechanical complications. Exteriorized cables from inside-out insulation breaches may damage the tricuspid valve and serve as a nidus for formation of thrombus and vegetations in endocarditis.
Oversensing in the Diagnosis of Pace–Sense Component Malfunction
Oversensing refers to sensing of signals other than the QRS complex (19,20). Oversensing that varies with each cardiac cycle (cyclical oversensing) indicates an intracardiac source. Characteristics of the oversensed signals in pace-sense LF have been reported (18–20) and must be distinguished from other causes of oversensing (19) (Table 1).
In conductor fractures, 3 features almost always are present: 1) signals are intermittent and have a high dominant frequency; 2) they show 1 or more types of variability, including amplitude, morphology, or frequency; and 3) in dedicated bipolar leads, they are not recorded on the shock channel. Three additional features are typical: 1) although some signals may be cyclical, noncyclical signals are present in at least some recordings; 2) some intervals are too short to represent successive ventricular depolarizations (120 to 160 ms). These are referred to as nonphysiological short intervals, but they occur commonly as a result of unrelated, but sequential physiological signals and are nonspecific. Whether or not they are present, conductor fracture typically results in rapid oversensing; and 3) signal amplitude may exceed the range of the sensing amplifier, and thus appear truncated. Pacing, especially high-output, may precipitate these signals or exacerbate oversensing. Importantly, connection problems between DF-1 leads and headers result in identical EGM patterns.
Unlike conductor fractures, insulation breaches do not themselves generate abnormal signals. Instead, oversensed signals pass through the insulation breach to enter the intact conductor. Thus, EGM patterns vary, reflecting the source signal. Nonphysiological signals may be recorded from multiple conductors (e.g., inside-out abrasion) or multiple leads (lead–lead abrasion). Pectoral myopotentials recorded from shock EGMs that include the pectoral ICD can be a normal finding, but pectoral myopotentials recorded on the RV sensing channel usually indicate an in-pocket insulation breach. EGMs of inside-out breaches of Riata leads often have characteristic spikes on the sensing channel or on both sensing and shock channels, which may represent mechanical interactions (18). Like oversensing caused by conductor fractures, oversensing related to insulation breaches usually is intermittent and transient.
Impedance and Impedance Trends in the Diagnosis of LF
ICDs measure electrical resistance (impedance) periodically for both pace-sense and high-voltage conductors (17,19). An abrupt 50% to 75% relative increase in pace-sense impedance is a highly specific indicator of an ICD system problem, either conductor fracture or a connection problem. Conversely, a gradual impedance increase without oversensing usually occurs at the electrode–myocardial interface, and lead replacement is not indicated unless pacing or sensing is compromised. An abrupt decrease in impedance of a chronically implanted lead usually indicates an insulation breach, but overall impedance changes are rare in pace-sense insulation breaches.
Fractures of high-voltage conductors often present as abrupt increases in shock impedance, typically >75% (5). High shock impedance may also be measured if connections are faulty. Low shock impedance has been reported in shock component insulation breaches, but insulation may fail during high-voltage shocks, even if low-voltage impedance measurements are within the expected range. The Central Illustration summarizes the approach to the patient with suspected LF.
Remote Monitoring and LF Diagnostics
Remote monitoring is the accepted standard for follow-up of ICD patients and plays a central role in the management of patients with at-risk leads. Both patient-initiated and fully automated systems improve recognition of lead malfunction (21).
Two types of LF diagnostics are of greatest value when used with remote monitoring. The Medtronic lead integrity alert (LIA) algorithm and Boston Scientific latitude lead alert are designed to improve early detection of conductor fractures and to decrease inappropriate shocks by monitoring for nonphysiological short intervals, rapid oversensing, and abrupt changes in relative impedance (17). Once triggered, both initiate remote-monitoring notification in devices with wireless telemetry. Additionally, LIA initiates an audible patient alert and extends the detection duration for VF to reduce inappropriate shocks. Both types of alerts need to be reviewed for well-documented false positives unrelated to LF (18). Figure 3 shows an example of a LF diagnosed by LIA. Note that minor oversensing occurred, and impedance was normal, despite extensive insulation and conductor damage.
The Medtronic lead noise algorithm and St. Jude Medical SecureSense RV lead noise discrimination algorithm are designed to differentiate oversensing due to LF from ventricular tachycardia/VF and withhold therapies in the presence of oversensing. These algorithms compare signals on the sensing and shock channels. Sensing-channel signals that do not correspond to shock-channel signals indicate oversensing. Both are ineffective for oversensing related to the RV coil in integrated bipolar leads. The lead noise algorithm has been validated clinically. It has not been reported to withhold shocks for clinical VF; non–lead-related oversensing that would have resulted in inappropriate shocks triggers most of the false positives. SecureSense has not been validated clinically; false-positive triggers have occurred during normal rhythm (22).
There are fewer ICD diagnostics for shock conductors than for pace-sense conductors. However, an automatic shocking vector switching algorithm (Dynamic Tx, St. Jude Medical) can prevent pulse generator damage and ensure shock delivery in at least some high-voltage insulation breaches that cause shorting in the shock output circuit of dual-coil leads. If a short is present in 1 of the 2 dual-coil shock pathways (RV coil to can, RV coil to SVC coil) during a shock, it aborts the shocks and then delivers the next shock through the intact pathway (23).
Management of Patients with Functioning Recalled Leads
For a patient with a functioning recalled lead, management focuses around a comprehensive baseline assessment and prescribing an intensified monitoring plan. A baseline chest radiograph may identify incomplete lead pin insertion, which can mimic the clinical presentation of conductor fracture. Cinefluoroscopy is the best modality for identifying Riata exteriorized cables.
Recommended practices include programming the duration to detect VF to at least 6 to 8 s (30 to 40 beats) unless contraindicated, establishing a quarterly remote or in-office follow-up schedule, and patient and remote monitoring alerts for out-of-range pacing impedance, shock impedance, and specific algorithms that facilitate early diagnosis of LF on the basis of oversensing or that withhold inappropriate shocks caused by oversensing (21). Manufacturer-recommended specific programming changes should be made immediately. Table 2 provides specific recommendations for follow-up of Riata and Fidelis leads.
The decision to replace a functioning, recalled ICD (either at pulse generator change or electively) should be on the basis of multiple factors summarized in Table 3. There is no consensus on how to manage patients with externalized cables, but no electrical abnormalities. Once a decision is made to replace a lead, a further decision may be made regarding extraction or addition of a new ICD lead, also considering factors summarized in Table 3. Decision analysis models provide analytical, data-driven recommendations for making these complex decisions, but the recommendations depend critically on assumptions about risk (24).
Acute Management of Patients with Inappropriate Shocks
Patients with inappropriate shocks often present with multiple, closely spaced shocks without antecedent symptoms. Point-of-care (emergency department) express interrogation devices may advise the clinician of abnormal device parameters. If an inappropriate shock is suspected, a donut magnet should be applied until comprehensive interrogation and reprogramming with the manufacturer's programmer can be performed. In most cases, an audible tone emitted from the ICD indicates proper magnet application. When a magnet is applied to an ICD, detection will be disabled (as long as the magnet remains applied). Unlike pacemakers, ICDs do not alter bradycardia pacing parameters with magnet application.
Management of Patients with Confirmed LF
Acute management of patients presenting with inappropriate shocks in normal rhythm centers on prompt inactivation of ventricular tachycardia/VF detection using the ICD’s programmer. In an emergency, a magnet disables detection for all ICDs as long as it is applied. Point-of-care (emergency departments) express interrogation devices are available for some ICDs.
Long-term management of most patients with confirmed LF centers around the decision to extract or add a lead (Table 3). However, in a minority of cases, it may be possible to modify existing hardware or reprogram the device. For example, it may be possible to program the SVC coil out of the system if there is isolated failure of the proximal high-voltage conductor. Extraction is the most appropriate option in some patients, but morbidity is probably higher with extraction than with insertion of a new ICD lead alone (25). Removal of Riata leads is more complex and challenging than removal of Fidelis leads (26). However, adding a new lead to a system with a failed Riata lead is associated with a high incidence of future lead issues, up to 28% (27), because externalized conductors may interact with the newly implanted lead due to failure of the ePTFE coating on exteriorized cables (28). Regardless of the lead, only experienced operators in high-volume centers with surgical backup should perform extractions. With the availability of the totally subcutaneous ICD system, soon to be coupled with leadless pacing, abandoning the transvenous system is an option for some patients.
Despite development of leadless pacemakers and subcutaneous ICD systems, transvenous ICD systems retain important advantages, and development of more durable leads remains a clinically important challenge in both biomaterials and engineering design. The legacy of recalled leads casts a shadow over the development of new leads. To ensure reliability, the FDA has mandated clinical trials with larger sample sizes and longer follow-up. The risk is that this high standard may stifle research into new lead materials and designs, limiting clinical practice to older, large-diameter leads. In response to the Fidelis and Riata failures, engineers have developed improved methods, including enhanced finite-element computer modeling of lead performance and more rigorous pre-clinical testing. These LFs taught us the importance of remote monitoring networks and patient registries for recognizing LFs that appear late after implantation, and tracking failure rates, which may vary over time. They also motivated improved diagnostics and device-based algorithms for pace-sense LFs. Patients and clinicians need comparable features for earlier diagnosis of high-voltage LFs. More broadly, we need diagnostics that go beyond warning of impending LF and confirm that a lead is performing as expected, without early structural changes that will eventually lead to mechanical or electrical failure. Such diagnostics could play a role in bringing new lead designs and materials to market faster by providing assurance of expected lead performance (29). However, for the foreseeable future, physicians will require understanding of lead design, LF mechanisms, strengths and weaknesses of leads diagnostics, and management options for at-risk and failed leads to make the best management decisions for an individual patient.
Dr. Swerdlow has received honoraria for lectures from Medtronic, St. Jude Medical, and Boston Scientific; consulting fees from Medtronic; and has intellectual property with Medtronic. Dr. Ellenbogen has received honoraria for lectures from Medtronic, Biotronik, Boston Scientific, and St. Jude Medical; served on data safety monitoring boards for Medtronic, Boston Scientific, and St. Jude Medical; and received consulting fees from Medtronic, Boston Scientific, and St. Jude Medical. Dr. Kalahasty has reported that he has no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- electromagnetic interference
- Food and Drug Administration
- implantable cardioverter defibrillator
- lead integrity alert
- lead failure
- right ventricular
- superior vena cava
- ventricular fibrillation
- Received November 16, 2015.
- Accepted December 8, 2015.
- American College of Cardiology Foundation
- Epstein L.M.,
- Love C.J.,
- Wilkoff B.L.,
- et al.
- Swerdlow C.D.,
- Gunderson B.D.,
- Ousdigian K.T.,
- et al.
- Theuns D.A.,
- Elvan A.,
- de Voogt W.,
- et al.
- Deeprasertkul P.,
- Yunus A.,
- Thakur R.
- Hauser R.G.,
- Abdelhadi R.H.,
- McGriff D.M.,
- et al.
- Swerdlow C.D.,
- Gunderson B.D.,
- Ousdigian K.T.,
- et al.
- Ellenbogen K.A.,
- Gunderson B.D.,
- Stromberg K.D.,
- et al.
- Swerdlow C.D.,
- Asirvatham S.J.,
- Ellenbogen K.A.,
- et al.
- Mixobuchi M.,
- Enjoji Y.
- Mendenhall G.S.,
- Saba S.
- Mulpuru S.K.,
- Pretorius V.G.,
- Birgersdotter-Green U.M.
- Materials, Design, and Functional Considerations
- ICD LF Mechanisms
- Incidence of LF
- Diagnosis of LF
- Oversensing in the Diagnosis of Pace–Sense Component Malfunction
- Impedance and Impedance Trends in the Diagnosis of LF
- Remote Monitoring and LF Diagnostics
- Management of Patients with Functioning Recalled Leads
- Acute Management of Patients with Inappropriate Shocks
- Management of Patients with Confirmed LF