Historical Milestones in Development

Permanently implantable pacing leads evolved from the temporary pacing wires that were first used to provide bradycardia support.³ The initial permanent transvenous leads were unipolar and consisted of a basic conductor, an insulator, and a connector pin. The electrodes were large and polished with high-polarization properties, low electrode-tissue impedance, and excessive current drain. No lumen was present in these leads to allow for stylet insertion, so lead implantation was a long, difficult task. Also, no fixation mechanism was present, so lead displacement rates were high. The development of bipolar pacing leads minimized far-field oversensing,⁴ requiring a major reconsideration of lead body design and structure, because two longitudinal conductors needed to coexist inside the lead, separated only by a thin layer of insulation. This and subsequent increases in complexity required advances in materials science so as not to compromise reliability.

The development of passive-fixation tines was a huge step forward and dramatically reduced the rate of reoperation for lead dislodgement compared with existing flange-tipped pacing leads.^5–7 Active-fixation helices were introduced subsequently and further improved the stability of implanted leads. Modern electrodes were developed with small geometric surface area but large effective area as a result of porous surfaces, and these optimized the trade-off between high current density and low polarization. Chronic threshold rises and late exit block were still major problems in cardiac pacing at that stage, however, and remained so until the revolutionary incorporation of an elutable dexamethasone reservoir into a porous-tip electrode, essentially eliminating this complication.⁸ Although this likely resulted from the glucocorticoid and not the redesigned porous titanium-tip electrode, this remained unproved until a pioneering randomized double-blind trial compared two otherwise identical leads and electrodes, with and without steroid elution.⁹ This unequivocally demonstrated the benefit was the result of the dexamethasone. Subsequent advances have included the universal standardization of connector pins (and consequently device headers), improving lead longevity through the use of better component materials, and improving sensing through the use of narrower bipoles.

Lead Structure And Polarity

The basic functions of any cardiac device lead—pacing and sensing—require conductors for current flow between the pulse generator and myocardium and insulators to prevent short-circuiting of this current. These elements may be configured in two ways. First, a pace/sense circuit may be composed of a lead containing one conductor that connects to a negatively charged tip electrode, the cathode, from which electrons flow through the myocardium and thoracic cavity back to the active, positively charged pulse generator (IPG), the anode (Fig. 4-1, A). Such a lead, with only one conductor and electrode, is called a unipolar lead because only one electrode is in contact with the heart (although it is part of a bipolar circuit). It has been established that pacing thresholds are lower at most pulse widths when the intracardiac pacing electrode is configured as the cathode rather than as the anode¹⁰ (see Chapter 1). Although once the only option, unipolar leads have largely been replaced by bipolar leads. The alternative system consists of a lead body containing two conductors (separated and surrounded by insulation) that connect to a cathode-tip electrode and an anode ring electrode located several millimeters more proximally (Fig. 4-1, B). The IPG is not an active electrode in a bipolar circuit, and such leads are said to have bipolar configuration.

Figure 4-1 A, Unipolar pacing circuit, with an intracardiac cathode located on the lead tip in the right ventricle (RV). The circuit is completed by the active pulse generator, which forms the anode. B, In a bipolar circuit, both cathode and anode are located on the lead, with the anode generally formed by a ring electrode proximal to the tip cathode.

Unipolar systems have a large interelectrode distance from the tip of the intracardiac lead to the IPG, and this large field of view makes them vulnerable to the oversensing of myocardial signals (T waves or electrograms from another chamber, known as “crosstalk”) and far-field nonmyocardial signals such as pacing artifacts (another form of crosstalk), skeletal myopotentials, and nonphysiologic electrical “noise” in up to 38% of patients.¹¹ Electrical signal oversensing can have significant consequences on pacing behavior, including ventricular output inhibition and inappropriate mode switching caused by far-field ventricular oversensing on the atrial channel.^12,13 Because of the additional consequence of inappropriate shock delivery if oversensing were to occur in an ICD, it is particularly important for ICD leads to have bipolar sensing circuitry. Since the IPG itself serves as one of the electrodes in a unipolar system, skeletal muscle capture and pectoral muscle stimulation may result in the setting of a higher programmed pacing output. Bipolar leads have been shown to have similar but slightly higher, acute and chronic stimulation thresholds to unipolar leads.⁴ Bipolar leads also carry an element of redundancy in that they can be used in either the bipolar or the unipolar configuration, as in isolated failure of the proximal, anodal conductor, with no demonstrated difference in performance between the two pacing configurations.¹⁴ In addition, the much lower amplitude of the bipolar pacing artifact greatly reduces the likelihood of crosstalk and oversensing. The great advantage of bipolar leads is related to their superior sensing properties. These factors have led to the near-universal adoption of the bipolar lead as the configuration of choice in cardiac device systems.¹⁵ The only potential disadvantage of bipolar leads is that reliability is lower than with the less complicated unipolar lead,¹⁶ although some studies have not supported this contention.^15,17

However, insulation failure usually has negligible effects on unipolar lead performance, a clear reliability advantage. In terms of the structure of the lead body, the simple, single-conductor design of unipolar leads had to be modified to accommodate the greater number of components and increased complexity of the bipolar lead body. There are two main bipolar lead designs, coaxial and coradial (Fig 4-2). Coaxial leads have an inner conductor that extends down the length of the lead to the tip electrode, the cathode, arranged in a coil configuration with a central lumen to allow for passage of a stylet at implantation. This coil is covered by a cylindrical length of inner insulation, which in turn is wrapped by another coil conductor that also runs down the lead to the ring electrode, the anode. A second, outer layer of insulation and lead covering protect the ring conductor from the outside environment, thereby completing the design. Although this coaxial model has been the industry standard for pacing leads for many years, the resulting bulk and stiffness of this four-layer design are significant compared with the simpler unipolar leads. Coradial bipolar leads addressed some of these concerns with new conductor and insulator technology that was subsequently used in ICD lead design. In coradial leads, a single coil extends down the length of the lead (again with a central lumen to allow for stylet insertion) and consists of two parallel, alternating conductor strands, one of which connects to the cathode and the other to the anode. Each conductor strand is individually coated with a bonded layer of ethylene tetrafluoroethylene (ETFE) fluoropolymer insulation (originally by DuPont, Wilmington, Del), which serves to insulate each strand from the other, despite being intertwined. The single, two-component coil is surrounded by a single, outer insulation covering.¹⁸ These leads have comparable bulk (~5F diameter) and flexibility to unipolar leads, often with tines or a fixed, nonretractable helix to minimize size; electrical parameters and reliability have not been demonstrably inferior to coaxial leads.^18–21

Figure 4-2 A, Coaxial lead body design consists of two nested coil conductors to the tip and ring electrodes, each wrapped in a separate layer of insulation. The inner coil connects to the tip cathode and contains a central lumen for stylet passage; the outer coil connects to the ring anode. B, In coradial design, a single coil contains the parallel conductor strands to the tip and ring electrodes, with each strand covered by a fluoropolymer insulator. A surrounding layer of outer insulation covers the lead body.

Defibrillator leads have even greater complexity, with two, three, or four conductors that connect to the pacing/sensing electrodes as well as to the high-voltage shocking coil(s), depending on the specific type of lead. Because of the prohibitive bulk that would result from a coaxial design with more than two conductors, ICD leads generally have a different type of structure, known as a multilumen design. This type of ICD lead consists of a long cylinder of insulating material, with separate internal channels running down its length (Fig. 4-3). Each conductor runs down an individual channel, usually with its own additional covering tube of insulation. The configuration and number of channels differ, depending on the number of conductors, the specific manufacturer, and whether air-filled channels are also incorporated into the design to improve lead handling and protect the conductors from flexion-related damage in the dynamic intracardiac and intravascular environment.

Figure 4-3 Multilumen ICD lead design consists of a single body of insulation with several internal channels that contain the conductor elements (central pace/sense coil and outer high-voltage cables), each covered by its own layer of fluoropolymer insulation. A surrounding layer of outer polyurethane-based insulation is present. HP, High-performance.

(Courtesy Medtronic, Minneapolis.)

One other lead body structure worth mentioning are small (4.1F diameter), isodiametric, lumenless pacing leads (Model 3830 SelectSecure; Fig. 4-4). The lack of a central lumen and the use of a conductor cable (rather than coil) allows for increased insulation redundancy, high tensile strength, and reduced bulk. It also means that the active-fixation helix is nonextendable because there is no central coil conductor to transmit torque from the connector pin to extend or retract the helix. Fixation of such leads to the myocardium is achieved by turning the entire lead body. To enable implantation without a stylet, the lead is deployed through an 8.5F steerable catheter delivery system (SelectSite C304). Short-term results from this lead have been reasonable, with no signs of increased lead displacement or fracture risk.^22,23 The small external diameter of this lead has been considered to be of some benefit in pediatric patients²⁴ and patients with congenital heart disease,²⁵ but the suggestion of a resulting improvement in chronic venous patency from the smaller size is yet to be substantiated in long-term follow-up. In addition, selective site pacing, in both the atria and the ventricles, is facilitated by the deflectable sheath, although steerable and preformed stylets can also be used with conventional leads to achieve this. No published experience with extraction of chronically placed lumenless leads exists, and the lack of a lumen to accommodate a lead-locking device may be a significant disadvantage of this design.

Figure 4-4 Isodiametric, lumenless, 4.1F pacing lead (Model 3830 SelectSecure). There is no central lumen for stylet passage, so the lead is deployed through an 8.5F steerable catheter delivery system (SelectSite C304) at implantation.

(Courtesy Medtronic, Minneapolis.)

Conductors

Both the coaxial and the coradial bipolar pacing lead body designs have been in widespread use and proved to be relatively fracture resistant over time. This durability is at least partially related to the materials and design of the conductor elements used in pacing leads. One study found conductor fracture over a 17-year follow-up in 19 of 561 right ventricular pacing leads (3.4%) and then generally at points of high stress, such as at the lead anchoring sleeve and the costoclavicular ligament (subclavian crush).²⁶ The basic material used for most conductors over the last 3 to 4 decades has been MP-35N (SPS Technologies, Cleveland), an alloy of nickel, cobalt, chromium, and molybdenum. The main advantage of MP-35N is its high strength and resistance to corrosion.²⁷ Its main disadvantage is its high electrical resistance, but this has been overcome with the development of composite-wire conductors that incorporate low-resistance metals such as silver and stainless steel with high-strength materials such as titanium, platinum, and platinum-iridium alloy.

In pacing leads, these materials are generally incorporated into a drawn filled tube (DFT) composite-wire conductor strand consisting of a thick, strong body of MP-35N or titanium that has been filled with a central core of softer, low-resistance metal such as silver, often encased in a further outer shell of platinum alloy (Fig. 4-5, A). The DFT conductor strands are then formed into multifilar (i.e., multiple strands or filaments) coils with a central lumen in coaxial leads, as described earlier, or further coated in a layer of ETFE fluoropolymer before being arranged in bifilar coradial coils,²⁸ again with a central lumen. The alternative way of forming a composite-wire conductor is the drawn brazed strand (DBS) method, generally consisting of six strands of a high-resistance material (e.g., MP-35N) that are tightly molded over a central strand of silver, such that the inner, low-resistance metal is forced between and around the strong outer strands (Fig. 4-5, B). This results in a low-resistance, durable composite conductor wire that can once again be formed into a multifilar coil for use in pacing leads.²⁹ However, it is also used to form strong, multifilar, braided cables that are employed as the conductors for the high-voltage circuitry of ICD leads. DBS cables are discussed later.

Figure 4-5 Composite-wire conductor design.

A, Drawn filled tube (DFT) composite wire; thick body of MP-35N is filled with softer, low-resistance silver. B, Drawn brazed strand (DBS) wire; six strands of MP-35N are tightly molded over a central strand of silver. DFT and DBS composite wire can then be further fashioned into multifilar coil and cables.

(Courtesy St Jude Medical, Sylmar, Calif.)

Insulation

Pacing lead insulation is a critical and vulnerable component that is often the cause of lead failure.^30,31 One of the problems with the major materials used for lead insulation (polyurethane, silicone rubber, fluoropolymers) is that they were not originally designed specifically for this purpose, and each has disadvantages when used as part of a biologic pacing system. They were incorporated by early lead engineers into their designs for lack of any purpose-built alternatives, and it is only recently that any viable, tailored alternatives have been developed.

Poly(ether)urethane is a synthetic, segmented polymer with very high tensile strength and resistance to mechanical abrasion. From a physical perspective, this means that a thinner layer of insulation can be used to cover the lead conductors, thereby reducing the overall lead diameter. Also, polyurethane has excellent lubricity and handling characteristics, with a low frictional coefficient, which can facilitate implantation of two or three lead systems by decreasing the physical interactions between the leads. However, polyurethane leads are stiffer and not fully biostable, being subject to in vivo polymer degradation that can cause insulation and late lead failure.³² The two main types of polyurethane are Pellethane 80A and Pellethane 55D (Upjohn, CPR Division, Torrance, Calif), with 80A particularly prone to biodegradation.

Two mechanisms of in vivo, biologically mediated degradation of polyurethane predominate: environmental stress cracking (ESC) and metal ion oxidation (MIO)³³ (Fig. 4-6). When polyurethane is subjected to repeated mechanical stresses in the presence of surface-active agents provided by macrophages and α₂-macroglobulin, parallel polymer molecules are disordered and interchain bonds weakened. This is the mechanism of ESC,³⁴ manifesting as a deep surface cracking visible with electron microscopy, particularly at the anchoring sleeve and venous entry portions of the lead. By comparison, no mechanical stress is required for MIO to develop because this is a chemical reaction that occurs particularly at the conductor-insulator interface.³⁵ Biologic oxidants, such as peroxide from surrounding inflammatory cells,³⁶ can result in the release of metal cations from the conductor, especially cobalt and nickel ions. These cations cause oxidation of the soft zones in the polymer, starting at the α-carbon of the ether bond and resulting in chain break. The harder Pellethane 55D is much less susceptible to both ESC and MIO than 80A but is significantly stiffer. Nevertheless, the long-term performance of Pellethane 55D has been much better than Pellethane 80A, a material notoriously prone to failure.

Figure 4-6 Mechanisms of polyurethane insulation failure on electron microscopy.

A, Environmental stress cracking (ESC) results from mechanical stress, leading to disruption of polymer molecule arrays and interchain bonds. B, Metal ion oxidation (MIO) of the polymer soft segments causes polymer chain break.

(Courtesy St Jude Medical, Sylmar, Calif.)

Silicone rubber (polysiloxane) has none of the disadvantages of polyurethane. It is completely inert and biostable over extended periods and thus not vulnerable to MIO or ESC. Silicone is more flexible, which may decrease the risk of damage to cardiac structures, including lead perforation. It is also more thermally resistant to the effects of electrocautery, which can be a significant advantage during pulse generator replacement or system revision.³⁷ Initially, a softer peroxide-catalyzed silicone rubber, MDX4-4515-50A, was used for pacing leads, but new versions with higher tensile strength and abrasion resistance have been developed. These include high-performance (HP) silicone, extra-tear-resistant (ETR) silicone, and Novus (Med-4719, Nusil Technologies, Carpinteria, Calif), produced by hybridizing HP and MDX4 silicone. Unfortunately, no variety of silicone rubber has the positive characteristics of polyurethane, including surface lubriciousness and ease of handling. Although these friction-related problems of silicone can be addressed by an outer layering of polyurethane, the main problem with silicone relates to its lower tensile strength and susceptibility to abrasion and tears (Fig. 4-7, A). This means that silicone insulation is more prone to damage during implantation and subsequent interactions in the device pocket, and that a thicker insulation layer must be used to maintain lead reliability, thereby increasing lead bulk. Also, as with glass, silicone is a fluid material and thus prone to a gradual flowing process away from pressure points, known as “cold flow,” which can lead to thinned, denuded, and abraded troughs or depressions adjacent to areas of thickened rubber polymer³² (Fig. 4-7, B).

Figure 4-7 Mechanisms of silicone rubber insulation failure.

A, Weaker molecular structure of silicone makes it more prone to abrasions and tears at points of higher mechanical stress. B, Glasslike fluid composition of silicone renders it susceptible to cold flow away from areas of higher pressure.

(Courtesy St Jude Medical, Sylmar, Calif.)

The fluoropolymers polytetrafluoroethylene (PTFE) and ETFE (DuPont) have some of the advantages of both silicone and polyurethane, with good abrasion resistance and biostability.³⁸ Unfortunately, their great stiffness excludes them from functioning as the primary insulating material in any lead, but they can be used, as outlined earlier, as a thin coat on conductor strands, particularly in coradial leads. This inner insulating layer can prevent electrical communication between conductor strands and can also protect them from interacting with an adjacent outer layer of polyurethane, thereby reducing MIO.

Recently, the development of a specific insulation material for cardiac leads has suggested that the deficiencies of the standard insulators in pacing leads may finally be addressed. The material is a copolymer hybrid composed of 48% silicone, 40% polyurethane hard segment, and 12% polyhexamethylene oxide soft segment; it is called Elast-Eon (Aortech Biomaterials, Clayton, Victoria, Australia). This blending of insulating materials results in the product having the strength, lubricity and abrasion resistance of polyurethane, but the flexibility and biostability of silicone.³⁹ Early bench and clinical testing have been promising to date, with little evidence of cold flow, ESC, or MIO phenomena, the main weaknesses of silicone and polyurethane. The material has been used in the St. Jude Medical Optim family of pacing (Models 1888T and 2088T) and ICD leads (Durata), but no long-term data on its chronic performance are yet available.

Electrodes

Electrodes represent the final common electrical and mechanical interface between the lead and myocardium. The design and construction of the electrodes influence the acute and chronic electrical performance of the lead. Because the electrode type also determines the fixation mechanisms employed and the lead tip pressure generated, however, mechanical problems such as lead dislodgement and cardiac perforation can also be affected by the tip electrode.

The materials used for lead electrodes have evolved over the past 40 years and have included titanium, anodized platinum, platinum-iridium alloys, Elgiloy (an alloy of cobalt, iron, chromium, molybdenum, nickel, and manganese), and vitreous carbon. All these display relative inertness in vivo, which minimizes inflammatory reactions and corrosion over the long term. Minor corrosion is seen, however, with Elgiloy and platinum alloys when in situ for extended periods, although this is of uncertain clinical significance.⁴⁰ While attempting to control for other differences, several human and animal studies have compared these materials, but the results are conflicting.^41–44 In general, applications have been found for these materials in electrode composition, with most having reasonable in vivo performance and longevity, although vitreous carbon cathodes have particular strength, inertness, and long-term electrical reliability.⁴⁵

More important than the precise material composition of electrodes seems to be their geometry, particularly the surface area. A large cathodal surface area in contact with the endocardium results in a low current density at the electrode-tissue interface and consequently a higher capture threshold. As tip electrodes were reduced in surface area, apart from better thresholds, higher tissue-electrode impedances were also seen, and this increased pulse generator longevity.^46,47 Against this beneficial effect of a smaller electrode is the adverse effect on tissue polarization seen with a smaller electrode surface area. Polarization refers to the electrical afterpotential generated at the electrode-tissue interface by ionic movement induced by the pacing stimulus. At pacing pulse delivery, a layer of positively charged cations is attracted to and surrounds the negatively charged cathode (and a larger layer of negatively charged anions surrounds the cation layer), which opposes current flow from the pacing electrode. This polarization potential is recorded as an afterpotential following the stimulus artifact. To maximize the advantages of small-radius electrodes, yet minimize the disadvantage of the polarization effect more prominent with smaller electrode surface area, current electrodes are small in size but have a complex, porous microscopic surface (Fig. 4-8). Regardless of the method of achieving a complex, highly textured surface, such as coating the electrode with microspheres or a woven mesh of microscopic metallic fibers, the resulting large surface area has proved beneficial with regard to pacing and sensing performance.

Figure 4-8 Sintered porous electrode (left) and laser porous electrode (right).

(From Hirshorn MS, Holley LK, Skalsky M, et al: Characteristics of advanced porous and textured surface pacemaker electrodes. PACE 6:525-536, 1983.)

Titanium nitride–coated or iridium-coated cathodes with fractal surface structure achieve the goal of enhanced surface area with a negligible polarization signal⁴⁸ (Fig. 4-9). Such electrodes have permitted the development of beat-by-beat automatic capture threshold and pacing output management systems that measure the evoked myocardial response following every pacing stimulus. By verifying myocardial capture for each paced beat, with the immediate delivery of a backup higher-output pulse if capture is lost (Fig. 4-10), the pacemaker can maintain a pacing output only marginally above the capture threshold, thereby prolonging pulse generator battery life.⁴⁹ Another type of automatic threshold algorithm does not verify beat-to-beat capture, but performs an automatic threshold test daily and adjusts the pacing output accordingly (with larger safety margin above capture threshold). These types of automatic capture-detection algorithms can be unreliable in the setting of a significant polarization afterpotential, because a large polarization signal cannot be distinguished from the local myocardial evoked response.⁵⁰

Figure 4-9 Magnified view (A) and electron micrograph (B) of fractally coated, titanium nitride electrode tip, which optimizes surface area and minimizes polarization resistance.

(Courtesy St Jude Medical, Sylmar, Calif.)

Figure 4-10 Automatic loss of capture algorithm (Autocapture). The atrioventricular (AV) delay is shortened to ensure ventricular pacing (A). The pacing output is automatically decreased until loss of capture is detected by failure to sense a myocardial-evoked response at B2. A backup 5-V pulse is delivered at B3, followed by an automatic gradual increase in pacing output that restores myocardial capture (C), and then resumption of dual-chamber pacing is seen, with the programmed AV delay (D). Modern leads produce negligible polarization afterpotential and thus allow for high-fidelity sensing of the myocardial-evoked response.

(Courtesy St Jude Medical, Sylmar, Calif.)

Electrodes implanted in myocardium elicit a strong foreign body granulomatous reaction that results in the formation of a local fibrous capsule at the lead tip, which can have a profound effect on electrical performance. Smooth-surfaced electrodes elicit the formation of a thick, electrically insulating capsule of fibrous and granulation tissue that may impede electrode fixation to the endocardium.⁵¹ Complex, porous electrodes, in contrast, are manufactured to have extremely small (20-50 µ in diameter) surface pores that allow for endocardial attachment by way of tissue ingrowth into these regions, which yields a mechanical advantage in addition to the electrical benefits mentioned earlier. A fibrous reaction also occurs with highly textured electrodes, but a lesser insulating effect is observed because the electrically active myocardium is closer to the electrode,⁵¹ resulting in improved electrical parameters over time.⁵²

To curtail further the fibrous reaction elicited by the electrode and to decrease the incidence and magnitude of chronic threshold rises, a pharmacologic anti-inflammatory strategy was pursued with glucocorticoid-eluting porous electrodes (Fig. 4-11). These leads showed excellent long-term capture and sensing parameters, with a dramatic reduction in late exit block (i.e., loss of myocardial capture).⁸ In an elegant and seminal double-blind randomized trial, Mond et al.⁹ conclusively proved the beneficial effect of steroid elution by comparing two otherwise identical porous titanium electrodes, only one of which incorporated dexamethasone in its tip. There was no difference in capture threshold seen at implant. Although there was an expected late rise in threshold in the standard leads starting at 2 weeks, this was completely abolished in the steroid-eluting lead, both during the trial period and in long-term follow-up.⁵³ Animal studies confirmed that the biologic basis of this electrophysiologic effect was through the glucocorticoid-induced attenuation of the inflammatory reaction, with thinner, less cellular fibrous capsules.⁵⁴

Figure 4-11 Dexamethasone-eluting reservoirs in passive-fixation (A) and active-fixation (B) leads.

(Courtesy St Jude Medical, Sylmar, Calif.)

In bipolar leads, another factor affecting electrical performance, particularly far-field sensing, is the interelectrode distance between the tip cathode and the ring anode. Narrow-spaced bipolar electrodes in atrial leads have been shown to reduce far-field ventricular electrogram sensing without compromising atrial sensing.⁵⁵ In a randomized trial that compared an atrial lead with a 1.1-mm tip-to-ring spacing (Model 1699T Optisense, St. Jude Medical) with a standard 10-mm spacing, the close-spaced bipole showed minimal far-field ventricular sensing at a sensitivity of 0.3 mV, compared to 30% 1-year incidence in the standard-bipole group.⁵⁶ This translated into a large reduction in inappropriate mode switch events, from 23% in the control group to 4% in the narrow-bipole group.

Fixation Mechanisms

Initial pacemaker leads had no fixation mechanisms, so lead dislodgement rates were high. The first solutions to this problem were wedge-shaped flanges and fins, but these had limitations. The development of scalloped tines was the first robust solution to the problem of lead displacement,⁵ and comparisons with earlier methods of passive fixation were favorable.⁵⁷ Tines are usually constructed from the same material as the covering insulation of the lead, typically silicone rubber, and protrude backward from the base of the tip electrode. Their orientation is designed to allow advancement of the lead for initial implantation but to prevent retraction and dislodgement by engaging the myocardial trabeculae of the right atrial appendage and right ventricular apex (Fig. 4-12, A).

Figure 4-12 Lead fixation mechanisms.

A, Passive-fixation tined lead. The tines are constructed of outer insulation material and facilitate advancement of the lead but prevent its retraction by engaging myocardial trabeculae in the right ventricle or right atrial appendage. B, Extendable-retractable helix is deployed to screw directly into the myocardium in an active-fixation lead.

(Courtesy St Jude Medical, Sylmar, Calif.)

Although passive-fixation tined leads perform well, with low rates of dislodgement, they are less reliable for pacing at nontraditional sites, and the development of active-fixation leads with a helix at the tip improved the options for pacing site selection (Fig. 4-12, B). These leads generally have an extendable-retractable screw made of platinum-iridium or similar alloy that is usually electrically active as part of the cathode. This terminal helix is deployed (i.e., extended out of lead) with clockwise rotation of the connector pin, which transmits torque via the central coil conductor to the helix mechanism at the other end of the lead. A small number of leads use a stylet-driven extension-retraction mechanism. Some smaller-diameter leads have a fixed helix that requires rotation of the entire lead body around the stylet for deployment into the myocardium. Steroid elution from a reservoir at the base of the helix helps to dampen the inflammatory reaction produced by traumatic engagement of the myocardium by the screw. Active- and passive-fixation leads perform similarly overall,⁵⁸ with some differences related to their construction. Active leads tend to be easier to extract, whereas passive leads tend to have lower chronic thresholds with higher impedances that prolong pulse generator longevity. The clinician must always consider the potential for perforation⁵⁹ and trauma to surrounding structures⁶⁰ when using active-fixation leads.⁶¹

Lead Anchoring Sleeves

Securing the lead body to the fascia or muscle at the floor of the device pocket at implantation is an important step to prevent lead dislodgement. The sutures used must be fastened firmly but, given the relative fragility of lead insulation as previously outlined, must not be allowed to compromise the lead structure or integrity (Fig. 4-13). Although the anchoring sleeve provides a mechanism to achieve this balance, there has been little development in this area. After being tied down tightly, sleeves made of MDX silicone were found to produce much lower lead deformation than ETR silicone,⁶² but no long-term in vivo data are available. Suture sleeves that grasp the lead by snapping into place (rather than relying on circumferential deformation of sleeve by encircling suture) have been proposed but are not in current use.

Figure 4-13 Lead anchoring sleeves.

Each sleeve contains up to three grooves to allow for suturing the sleeve to the lead body, then to the underlying deep fascia in the pocket.

(Courtesy St Jude Medical, Sylmar, Calif.)

Connector Terminal Pins

The connector terminal pin forms the interface of the lead with the header of the pulse generator and has gone through significant evolution. With successive generations of pulse generator and lead technology, various standards for connector terminal pins for pacing leads have been used,^63,64 as well as two different pin-fixation mechanisms within the device header, the side lock, and the now-universal set screw. Reliability has been high with both pin-header fixation interfaces, although the more forgiving tolerance of the set-screw platform has led to its widespread adoption.⁶⁵ This design requires a self-sealing grommet to cover the head of the set screw and prevent corrosion of the pin or header connections because pocket fluid enters the header. Sealing fins are required around the lead body or within the header channel to prevent fluid leaking into the header at that site, regardless of the fixation mechanism. Different patterns of connector pins include the 5- and 6-mm-long pins, the 3.2-mm low-profile pin, the VS-1 pin, and the current standard IS-1 connector pin (Fig. 4-14). Compatibility between terminal pin type and header channel type is still of major concern when changing pulse generators in patients with older leads. Modern pulse generator headers conform to IS-1 specifications (Fig. 4-15) and are backwards-compatible with VS-1 pins. Occasionally, for example, adapters may be required to downsize a long pin connector to an IS-1 header.

Figure 4-14 Evolution of pace/sense connector pins, culminating in current IS-1 standard (bottom).

Figure 4-15 A, IS-1 pace/sense connector pin terminal. B, Cross section of IS-1 header.

Lead Placement and Fixation Mechanisms

For accessing the stimulation site, CRT lead systems come with a range of outer guide introducers to facilitate coronary sinus cannulation and provide support during lead placement. Although the initial transvenous CRT leads were stylet driven, over-the-wire leads were rapidly introduced because of difficulty negotiating venous tortuosities and acute takeoff angles. The development of an over-the-wire system was a major advance,⁶⁸ and such leads are routinely required to select and place leads in the best possible target venous branches, usually over the posterolateral left ventricle. These leads allow for front or back loading of a 0.014-inch angioplasty wire into the stylet lumen. With or without the assistance of inner guiding catheters, the wire may be used to subselect target venous branches (including those with acute takeoff angles) and the lead may then be tracked over the pre-positioned wire. Obtaining distal vascular access with a wire also allows for venoplasty to facilitate lead placement in the setting of stenotic vascular segments.

Once at the target site, and with retraction of the wire or stylet back into the lead body, most lead tips have some degree of memory and assume the form of a preshaped curvature, which may be a smooth curve, an angled cant, or a spiral (Fig. 4-17). These distal lead shapes provide some degree of stability in the target vein by exerting modest lateral pressure on the inner vein surface, thereby increasing friction and a resistance to forward or backward migration. LV lead models that have no distal curvature use small, distal tines to engage the venous wall, but these tines likely have a reduced ability to fix the lead passively in place, particularly in larger vein branches. Thus, stability of tined leads probably relies more on lead flexibility and the cumulative friction along the lead length wherever it is in contact with the venous wall. These straight leads seem to have a higher risk of dislodgement, and subsequent generations of CRT leads have not incorporated tines.

Figure 4-17 Representative variety of currently available resynchronization leads with differing passive-fixation curvatures.

(Courtesy Medtronic, Minneapolis; St Jude Medical, Sylmar, Calif; and Boston Scientific, St Paul, Minn.)

In most cases the major contributor to LV lead fixation in the coronary venous system has been the extent to which the outer lead diameter is circumferentially opposed to the inside wall of a venous tributary of given caliber. If the lead has been advanced snugly into a vein branch, significant fixation friction forces may be generated, but lead stability remains variable. The risk of dislodgement also depends on upstream factors, including the number and size of redundant loops and tortuosities in the proximal course of the lead body, cardiac motion, tricuspid valve regurgitation, and interactions with other intracardiac hardware, which can each generate traction forces on the LV lead. Consequently, there is a significant rate of CRT lead displacement of up to 1% to 5% that may require surgical lead revision if there is a significant change in electrode position, ventricular capture threshold, or phrenic capture.^66,69,70

The concept of active-fixation leads in the coronary venous system has been brought to clinical practice with the Attain StarFix lead (Model 4195, Medtronic, Minneapolis), which has three sets of deployable lobes that fold out and protrude from the lead body just proximal to the tip (Fig. 4-18). These 12 circumferential lobes are extended by an external sliding-push tubing mechanism, and radiopaque markers are used to indicate whether one, two, or all three sets of lobes are extended or retracted. In a recent multicenter study of 408 implants, Crossley et al.⁷¹ found a very low lead dislodgment rate of 0.7% with stable electrical parameters over a mean follow-up of 23 months. Phrenic nerve capture with this unipolar lead required lead revision in 11 patients (2.5%). This reasonable performance must be weighed, however, against potential difficulties of extraction, which fortunately has not been a major issue in the other LV lead models to date. In this study of the active-fixation LV lead, 26 patients (6.4%) underwent an attempt at lead revision. In two patients, no attempt to remove the lead was made, and in the remaining 24 revisions, five attempts failed (21%). These patients had maximum implant duration of only 941 days. Of the seven patients who had a lead in situ for more than 1 year, four (57%) were unable to be extracted.

Figure 4-18 Active-fixation resynchronization lead (Attain StarFix, Model 4195) with deployable lobes to grip the lumen of the coronary venous tributary in which it is implanted.

(Courtesy Medtronic, Minneapolis.)

This study is similar to the experience of Nagele et al.,⁷² who encountered extraction difficulty caused by immobile lobes in two patients. The ability to un-deploy the lobes seems to be independent of the duration of implant and may be related to local branch vein thrombosis or an ingrowth of fibrous tissue through the lobe loops. Although both these studies were small and represented early experience with this lead, the difficulties with extraction of the StarFix lead contrast with the relative ease of extraction of chronic LV leads, even after extended periods since implantation.⁷³

Design of CRT Leads

The original CRT leads were exclusively unipolar because of size considerations, given that LV leads had to be advanced through a long, smaller-caliber sheath that was used to engage the coronary sinus selectively. An early example was the EasyTrak lead family (Models 4510/11/12/13, Guidant, St. Paul, Minn.), which had a 6F lead body diameter and an LV-1 connector terminal pin compatible only with LV-1 headers on Guidant CRT pulse generators. Since then, numerous developments in lead size, polarity, and CRT device programmability have greatly increased the options available to the implanting operator, including lead and device options to manage the problem of phrenic nerve capture, as well as to find suitable pacing configurations in the setting of myocardial scar. Most significantly, bipolar CRT leads were developed that were no larger in diameter than the previous generation of unipolar leads and that could therefore be deployed through the same sheath delivery systems and into venous branches of the same caliber. Newer LV lead versions include 4F to 5F leads that can be advanced through smaller venous subselection sheaths, into more diminutive vein branches, which increases the number of lead positioning options.

The interelectrode distance in bipolar CRT leads varies from 8 to 21 mm. The great advantage of bipolar leads is the option of programming between multiple different pacing electrode permutations, including true bipolar and extended bipolar configurations, with the option of selecting whether to use the LV tip or the LV ring electrode as the pacing cathode (Fig. 4-19). These programming options may be invaluable in the patient with posterolateral LV fibrosis or a large field of phrenic nerve capture, where an alternate pacing polarity may obviate the need for lead revision. True bipolar LV pacing also eliminates the problem of simultaneous right ventricular (RV) anodal capture, which can be seen in the setting of unipolar LV lead use and can sometimes partially undermine the benefit of LV pacing.⁷⁴ Anodal capture from the RV ring or coil electrode can hinder proper CRT delivery by eliminating the ability to stimulate the left before the right ventricle, and occasionally by leading to absent LV pacing if the pacing output is inadvertently programmed below the true LV capture threshold, because RV anodal capture is mistaken for LV cathodal capture during threshold testing.⁷⁵

Figure 4-19 Multiple different pacing polarity configurations are possible with modern bipolar resynchronization leads and generators. Each arrow originates at the cathode and ends at the anode of current left ventricular (LV) lead pacing configurations. RV, Right ventricular.

No significant differences have been noted in the short-term and long-term complication rates after CRT implantation among the various leads and CRT systems commercially available.⁷⁶

Historical Evolution of ICD Leads

Epicardial patches placed through thoracotomy were the leads first used in human ICD systems.⁷⁷ The pulse generators in these systems were large and required placement in an abdominal pocket. Shock energy was delivered through a free-floating 12F nonpacing endovascular titanium spring coil that served as the anode, with the patch as the cathode. Sensing was accomplished initially by epicardial patches (or apical cup electrode in the first implant), but this sensing mechanism proved to be unreliable, and dedicated epicardial and endocardial leads were adopted early. These first-generation ICD systems had a high risk of complications, particularly in the large endovascular coil, including superior vena cava thrombosis, coil fracture, insulation failure, and coil retraction into the subclavian vein caused by lack of a lead anchoring sleeve.

Consequently, the second-generation ICD devices in the late 1980s were fully epicardial systems, with shock delivery between two epicardial patches and tachyarrhythmia sensing through an epicardial lead. Epicardial patches also had problems, however, including an ongoing infection and pericarditis risk,⁸⁵ a possible adverse effect on transthoracic external defibrillation threshold, and a perioperative mortality of up to 4% at implantation in these patients, who were generally at high risk for thoracotomy or median sternotomy.^86,87 These patch electrodes also displayed poor in vivo durability over time, with a failure rate of up to 28% at 4 years. More than half of patients with obvious lead failure or fracture were asymptomatic. The epicardial pace/sense leads incorporated in these systems were also suspect, with a failure rate of 14.1% at 8.5 years in one study, with lead fracture accounting for half these failures.⁸⁸ In contrast to patch failure, patients with epicardial pace/sense lead failure were likely to develop multiple inappropriate ICD shocks from oversensing.⁸⁹

By the early 1990s, advances in pulse generator design and technology allowed miniaturization for infraclavicular pectoral implantation, although this would require an entirely endocardial defibrillation, sensing, and pacing lead system. The initial transvenous leads were implanted with the pulse generator in an abdominal pocket, and the leads were tunneled subcutaneously from their pectoral vascular insertion site down to the abdomen. Such leads, including the Transvene family (Medtronic), were initially based on coaxial pacing lead platforms. Their development added ATP capability to ICD systems, but also new lead-related problems. Apart from the large bulk of these leads (up to 12F), predisposing to venous occlusion and subclavian crush,⁹⁰ lead failure was frequent, demonstrating a progressive late incidence related to both conductor fracture and insulation defects.⁹¹ Addressing these problems required paradigm shifts in lead body design as well as advancements in materials technology, which ushered in the modern era of multilumen, narrow-caliber transvenous ICD leads.

ICD Lead Structure

The basic elements in pacing leads are also found in transvenous ICD leads: conductors, insulation, electrodes, fixation mechanisms, and connector terminals. In addition, however, ICD leads contain the high-voltage circuitry that links the pulse generator with one or two high-voltage shock coils located on the lead body (Fig. 4-20). One of these coils is located in the right ventricle, a variable distance proximal to the tip pace/sense electrode, whereas the other coil, if present, is positioned on the part of the lead located in the superior vena cava.

Figure 4-20 A, Dual-coil ICD lead with second high-voltage shock coil residing in the superior vena cava. B, Single-coil ICD lead with one shock coil in the right ventricle only. The active ICD pulse generator generally serves as the cathode or anode for the first phase of biphasic shocks.

Two basic lead body architectures can be used to incorporate all these elements into an ICD lead. First, the coaxial structure, a design essentially borrowed from the pacing platform of the same name, consists of concentric coiled conductors (to tip electrode, ring electrode, and shock coils) layered inside one another and separated by intervening layers of insulation, with the whole assembly housed in an outer covering layer of insulation. The innermost coil connected down to the tip conductor contained a lumen for stylet insertion that assisted with endocardial lead implantation. The best known examples of this design were the Medtronic Transvene and the Ventritex TVL families of ICD leads, first used in the early 1990s. Although initial reports were positive,⁹² subsequent longitudinal data showed that these leads were prone to early and late failure, caused by a combination of conductor and insulation problems.^83,91 Whether these failures were mainly related to the limited durability of the conductor and insulation (polyurethane 80A) materials, or resulted from an inherent weakness of the bulky coaxial architecture in situations of increased complexity as found in ICD leads, will never be known; this design was no longer in use by the late 1990s.

The second main paradigm for ICD lead structure, and still the dominant architecture for modern ICD leads, is the multilumen design, in which parallel cable and coil conductors run to the tip, with ring and shock electrodes in separate channel lumens, all within a single frame of insulation; some models also have additional, empty decompression lumens. The conductor to the tip electrode is conventionally coiled and contains the central lumen for stylet insertion. The coil also permits the transfer of torque when the implanting physician rotates the pace/sense connector terminal pin to deploy an active-fixation helix at the distal lead tip. Each conductor, both cable and coil, is invested with an inner layer of insulation, typically of fluoropolymer type, and the entire lead body is encased in a further outer insulation layer (see Fig. 4-3).

ICD Lead Sensing Design

Sensing function is critical in ICD systems to allow for accurate rhythm discrimination and to prevent inappropriate shocks from oversensing as well as failure of tachyarrhythmia treatment caused by undersensing. Therefore, unipolar ICD leads were not developed because of the prohibitive risk of oversensing noncardiac potentials, such as pectoral and diaphragmatic myopotentials. Bipolar ICD lead sensing can be arranged according to one of two paradigms: integrated bipolar and dedicated bipolar (also called true bipolar) designs. Sensing in integrated bipolar leads occurs between the tip electrode and the RV shock coil, which doubles as the sensing anode, the distal extent of which is located 6 to 15 mm proximal to the tip. By comparison, dedicated bipolar leads additionally include a conventional ring electrode (the anode) 10 to 15 mm back from the tip, similar to a bipolar pacing lead (Fig. 4-21). The additional electrode means that an extra conductor is required in dedicated bipolar leads, but the smaller anode in these leads confers superior sensing performance with reduced far-field oversensing.^93,94

Figure 4-21 Design pullback.

“Pullback” refers to the distance between the lower active edge of the right ventricular (RV) defibrillation electrode and the RV apex. Design pullback should be minimized for lower defibrillation thresholds. The diagrams show the impact of sensing methods on catheter design and the magnitude of the RV electrode pullback. A, Leads using integrated bipolar sensing/pacing methods (from tip to RV defibrillation electrode) typically have a shorter pullback distance (e.g., 12 mm here). B, Leads using dedicated bipolar sensing/pacing methods have larger distances from catheter tip to active RV defibrillation electrode because of sensing ring (e.g., 25 mm here).

The larger sensing anode (the distal shock coil) in integrated bipolar leads may be of particular concern in patients with small right ventricles, where the proximal end of the coil may extend across the tricuspid anulus and thereby pose a risk of far-field atrial oversensing.⁹³ Integrated bipolar leads may be more prone to oversensing high-frequency respirophasic noise transients (i.e., diaphragmatic myopotentials), which may cause pacing inhibition and inappropriate shocks.⁹⁵ The increased current density at the smaller anode of a dedicated bipolar lead does, however, increase the likelihood of anodal myocardial stimulation in CRT systems in which LV pacing is programmed in an “extended bipolar” fashion, with the RV ring doubling as the LV pacing anode,⁷⁵ although in the current era of bipolar LV leads and pacing vector programmability, this is less of a concern.

The main theoretical advantage of integrated bipolar leads, apart from their greater simplicity (which may reduce the chance for component failure), relates to the smaller distance between the RV shock coil in these leads and the RV apex, because there is no intervening ring electrode between the shock coil and the lead tip. This is known as the pullback distance, and data suggest a modestly improved defibrillation threshold with shorter pullback distance,^96,97 although this was more important before the current era of high-output, biphasic devices and greater defibrillation safety margins. Theoretical concerns about differences in R-wave amplitudes between the two lead designs have not been borne out by comparative studies,⁹⁸ but some studies suggest longer sensing latency⁹⁸ with integrated bipolar leads, as well as the potential for undersensing of postshock ventricular fibrillation, presumably from local myocardial stunning around the shock coil and its proximity to the tip cathode.^99,100 This phenomenon was not seen when the shock coil was more than 6 mm from the tip electrode. However, in terms of sensing de novo ventricular fibrillation, no significant differences have been demonstrated between the two lead platforms,¹⁰¹ and both are in clinical use today, although dedicated bipolar systems are implanted more frequently.

Conductors

As with pacing leads, similar materials are used in ICD lead conductor construction, with MP-35N being the dominant metal alloy employed. As outlined earlier, its advantages include its corrosion and fracture resistance, but MP-35N–based conductors can still fracture at points of high stress (e.g., costoclavicular ligament). Recent testing suggests that this risk may be related to the presence of contaminant titanium inclusions, although this is controversial.¹⁰² In ICD systems, however, low-resistance conductors are essential to minimize voltage drops across the lead during high-voltage shock delivery. This requirement mandates the use of composite wire technology in which individual conductor strands combine higher-resistance MP-35N with a low-resistance metal such as silver in one of two patterns. The DFT pattern (see Fig. 4-5, A) produces conductor strands that are generally formed into coils, whereas in the DBS pattern (see Fig. 4-5, B), strands are twisted into multifilar braided cables (Fig. 4-22). The latter are particularly important in ICD leads because they have very high tensile strength and are more fracture resistant than coils.

Figure 4-22 Filar composition of unraveled high-voltage conductor cable alongside multifilar central pace/sense conductor coil in dissected ICD lead.

Failed ICD leads frequently demonstrate intact high-voltage cables with fractured pace/sense coil conductors. One conductor coil strand is generally required, as outlined earlier, to create a lumen for stylet insertion and to permit active-fixation helix deployment. However, the use of multiple parallel cables in a multilumen ICD lead design produces a stronger lead of smaller diameter than would be possible if only coiled conductors were used (as in the now-obsolete coaxial configurations). Each cable generally is housed in its own lumen of the multilumen lead by coating it with stiff, fluoropolymer insulation, preventing abrasion and damage to the soft, silicone-based insulation of the lead body (see Fig. 4-3). Decompression spaces, either within the lumen that houses a conductor or as separate channels in the housing insulation, may reduce the stress on the conductors, particularly in the setting of repeated lead flexion.

Electrodes and High-Voltage Coils

Electrode composition and design in ICD leads conform to the same basic principles as outlined earlier for pacing leads, with steroid elution, current density, and polarization considerations again dictating the material composition of the tip cathode and ring anode in dedicated bipolar leads. Fixation mechanisms are also similar, with either electrically active extendable-retractable helices or flexible tines being employed.

The high-voltage defibrillation shock coil is the defining feature of the ICD lead. All models include at least one shock coil, wrapped around the distal lead body, 5 to 6 cm in length and positioned in the right ventricle, whereas dual-coil leads have a second shock coil located either 17 cm or 21 cm proximal to the shaft end of the RV coil, with this superior vena cava (SVC) coil being slightly longer, 7 to 8 cm (see Fig. 4-20). The RV coil serves as the defibrillation anode or cathode during phase 1 of a biphasic shock, depending on how the ICD is programmed to deliver high-voltage therapy, with the SVC coil and active pulse generator casing together serving as the other shock cathode or anode. With single-coil ICD leads, the pulse generator solely functions as the second high-voltage electrode. Dual-coil leads may assist in lowering the defibrillation threshold (DFT), although the improvement in DFT with the addition of an SVC coil may be modest and of greater benefit when the electrical impedance during single-coil shock delivery is higher.¹⁰³ Because the high-voltage shock vector plays a role in DFT, some ICDs permit programming changes to eliminate either the SVC coil or the pulse generator itself from the shocking circuit, thereby improving the DFT in certain patients. Shocking coil–only leads are also available for implantation in the coronary sinus or azygous vein, or subcutaneously around the left thorax to improve the shock vector in patients with an elevated DFT and unreliable defibrillation with standard shock configurations.

The materials used in shock coil construction are generally platinum-iridium alloy or platinum alloy–clad tantalum, because titanium coils produce metallic oxides during high-voltage shocks that prevent their use in ICD leads. Polarization also affects the design of shock electrodes, as it does for pacing electrodes, but unlike the situation described earlier with pacing leads, fractally coating the shock coils does not improve defibrillation efficacy.¹⁰⁴ However, some data suggest that iridium oxide coating may improve lead porosity and reduce polarization, even to the point of seeing a clinically meaningful reduction in DFT.^105,106

Although an essential part of the ICD lead, the presence of endovascular defibrillation coils poses distinct physical problems, especially related to fibrous tissue ingrowth, which can make lead extraction difficult and potentially hazardous, as discussed later.

Insulation

Essentially the same materials used in the manufacture of pacing lead insulation are also used in the context of ICD leads, and their respective advantages and disadvantages are discussed earlier. The multilumen design of ICD leads relies on a single, large-caliber body of insulation that contains the conductor lumens. Other insulation elements may include the jackets investing the conductor cables or an additional outer insulation covering, but the mechanical characteristics of the lead will largely be determined by the material used to construct the lead body. Because ICD leads are larger and more complex than pacemaker leads, silicone rubber is the dominant option for ICD lead body insulation; polyurethane 55D and fluoropolymers are too stiff for this purpose, and polyurethane 80A is too susceptible to biodegradation. The relative structural advantages of polyurethane 55D, lubriciousness and high tear strength, can be used to good advantage instead as an outer insulation layer surrounding the lead body. In this location it is far away from the metal conductors within the thick silicone rubber casing and a fluoropolymer jacket. This barrier prevents exposure to metal ions from the MP-35N conductor that can trigger the oxidative degradation of the soft, polyurethane segment. Likewise, although the major structural weaknesses of fluoropolymers (related especially to rigidity) preclude their use as the primary insulator, inner insulation of the coil and cable conductors is well suited to these highly biostable, abrasion-resistant materials. Generally, ETFE is used for coating of the cable conductors, and PTFE is used to line the central lumen lead body that houses the coil conductor.

Although silicone rubber may be biostable and flexible enough to be used as the dominant material for ICD lead body insulation, its softness and high frictional coefficient make lead implantation difficult if the outermost surface of the ICD lead is silicone. Outer coverings of polyurethane may address this disadvantage to some extent, but the long-term mechanical risks related to silicone’s softness remain, as well as its susceptibility to abrasion and cold flow, and require the use of thicker silicone layers and thus increased lead caliber. This weakness is particularly marked at high-stress points such as at the costoclavicular ligament or in the pocket next to the pulse generator, where harsh local mechanical forces can lead to progressive insulation abrasion and local silicone thinning at pressure points. If lead insulation is sufficiently thinned or denuded that a low-impedance dielectric is created, particularly where a high-voltage conductor strand closely opposes the metal casing of the pulse generator in the device pocket, a short circuit with arcing can result during shock delivery, leading to ineffective defibrillation and permanent damage to the lead, the ICD circuitry, or both.¹⁰⁷

As mentioned in the section on pacing leads, newer insulation materials are being developed for specific application in cardiac device leads, including one already in clinical use, Elast-Eon (Aortech Biomaterials, Clayton, Victoria, Australia). This substance is a hybrid copolymer composed of near-equal amounts of silicone and polyurethane and, in extensive laboratory experience, combines advantages of each.³⁹ A family of pacing and ICD leads uses this material as an outer insulator (Optim, St. Jude Medical), but as yet no leads use this material for the primary lead body insulation.

ICD Lead Connector Terminals

As with the initial CRT leads, early ICD leads had high-voltage circuitry that terminated proximally in connector pins compatible only with pulse generators made by the same vendor. This was a major problem, particularly at generator change procedures, when lack of hardware cross-compatibility could limit device selection options and increase procedural complexity. The first solutions involved developing a wide range of adapters and extenders that could be used to switch connector pin design. However, this added further complexity to the system, increased the number of electrical connection points, and introduced another component that was subject to similar modes of failure as the rest of the lead.

Eventually, industry-wide standardization occurred in ICD systems, such that the pace/sense coil conductor terminates in a standard IS-1 pacing pin that is placed in the IS-1 port of the ICD header, and each high-voltage shock coil terminates in a standard DF-1 (ISO-11318) connector pin. In a dual-coil ICD lead, therefore, three tails ending in one IS-1 pin and two DF-1 pins emerge from a trifurcation yoke attached to the main lead body (Fig. 4-23). Three separate connector ports are found on the ICD header to accommodate the trifurcated end of a dual-coil ICD lead (one IS-1 port and two DF-1 ports). The yoke and lead tails of the ICD lead are subject to insulation and conductor failure, just as in the rest of the lead system. They also increase pocket bulk and may contribute to erosion risk, particularly in the setting of a dual-chamber or biventricular device system, with atrial and LV leads also present. Additional abandoned lead ends might also be present in the device pocket, if nonfunctional or redundant leads remain in place in the setting of lead failure or device upgrade. Besides increasing pocket bulk and interaction between device system components, the greater number of connector pins in a standard ICD lead allows inadvertent connector pin reversal in the device header, and the thick fibrous tissue that frequently encapsulates the lead trifurcation yoke can make a generator change or system revision more challenging. The development of the new IS-4/DF-4 standard connector pin has allowed the integration of two DF-1 high-voltage terminals and one bipolar IS-1 pace/sense terminal, all into one connector pin (Fig. 4-24). This new design addresses the mechanical issues associated with current multiterminal ICD leads, but the long-term reliability of these connector pins is currently unknown.¹⁰⁸ In addition, this all-in-one connector pin design precludes the ability to replace a single component of the ICD lead, such as the pace/sense or a shocking coil component, in the setting of pace/sense malfunction or elevated DFT.

Figure 4-23 ICD lead connector terminals.

A, Standard IS-1 pace/sense connector terminal. B, High-voltage DF-1 pin connects shock coil(s) to generator header. C, Proximal end of dual-coil ICD lead, with central IS-1 pace/sense pin, two high-voltage DF-1 pins, and trifurcation yoke to interface the three terminal pins with lead body.

Figure 4-24 Schematic representation and photograph of the IS-4/DF-4 connector pin incorporating high-voltage and low-voltage elements in the one pin and thus obviating the need for an external trifurcation yoke.

(Courtesy St Jude Medical, Sylmar, Calif.)

Extraction Considerations

Increasingly, ICD systems are being implanted in younger patients with evolving indications outside the context of acquired cardiomyopathy. Channelopathies, arrhythmogenic RV dysplasia, congenital cardiac abnormalities, and sarcoidosis are indications for implant that affect younger patients, often children, who are expected to live with their ICD leads (since generators are changed at regular intervals) over their lifetime. Because ICD lead failure rates continue to accrue over time with current lead and materials technology,⁸¹ and lead infection can occur at any time without warning,¹⁰⁹ lead manufacturers and physicians must give due consideration to this aspect of device therapy at implant. In addition, abandoned sterile leads are increasingly being removed to allow for reduced intravascular hardware burden, avoidance of inappropriate oversensing,¹¹⁰ and potentially increased long-term venous patency.¹¹¹

Lead extraction, even with the modern tools and expertise,¹¹² remains a potentially hazardous endeavor,^113,114 particularly in young patients with long-standing dual-coil ICD leads, in which tissue ingrowth into the interstices between the helical turns of the shock coil wire may be causing dense, fibrous adhesions and tissue encapsulation of the lead body (Fig. 4-25). With some data indicating that defibrillation efficacy may not be significantly improved with the addition of the SVC coil,¹¹⁵ one approach is to consider implanting only single-coil ICD leads in patients expected to have longer implant duration.¹¹⁶ However, the real solution lies in the manufacture of shock coils that are easier to extract.

Figure 4-25 A, Chronically implanted ICD lead shows extensive adhesions binding the lead tip and shock coil to the trabeculated ventricular endocardium. B, After lead extraction, extent of shock coil–associated fibrosis is evident. C and D, Cross-sectional photomicrographs of a chronically implanted lead show fibrous tissue invasion into interstices between helical turns of the shock coil.

(C and D courtesy St Jude Medical, Sylmar, Calif.)

All newer ICD lead bodies are isodiametric, with the same or a smaller maximal external diameter as proceeding from the proximal to the distal end of the lead (Fig. 4-26), and are 8-Fr size or smaller, which already improves the ease of extraction over previous transvenous ICD lead systems. Manufacturers also have taken different approaches to the problem of shock coil adhesion to the venous and cardiac walls¹⁰² (Fig. 4-27). Silicone backfilling of the interstices between the coils has been used in Medtronic leads to reduce fibrous tissue ingrowth. This approach may be effective¹¹⁷ but is limited by the surface area of the outer curvature of the round coil strands; the interstices cannot be fully obliterated because a large, exposed coil surface area is required for effective defibrillation. In Biotronik and St. Jude Medical leads, use of flatwire technology allows complete backfilling of the interhelix turns with silicone, allowing for a completely isodiametric shock coil segment while still reducing fibrous tissue ingrowth, as demonstrated in animal testing.¹⁰² A third strategy, employed in Boston Scientific leads, is to cover the defibrillation coils with a thin, encapsulating jacket of semiporous expanded polytetrafluoroethylene (ePTFE; Gore-Tex, Gore and Associates, Flagstaff, Ariz). As with all fluoropolymers, this material is abrasion and tear resistant as well as biostable, but it has been molecularly modified to make it semiporous. This allows for free fluid and ion flow so as not to significantly alter high-voltage electrical conductance during shocks from the underlying defibrillation coils. Consequently, no significant difference in clinical defibrillation efficacy has been demonstrated with the use of ePTFE coils.¹¹⁸ In the setting of an integrated bipolar sensing ICD lead and additional abandoned leads in the right ventricle, mechanical interactions between the large RV coil and adjacent lead materials may occasionally result in sensed electrical artifacts. Oversensing from this type of lead-on-lead interaction may be reduced or eliminated if the metal strands of the RV coil are covered by an ePTFE sleeve.¹¹⁰ The real advantage of the semiporous sleeve, however, is that cellular migration cannot occur into the covered underlying coil interstices, thus reducing tissue ingrowth into, and local adherence of, the shocking coils over time. This has made ICD lead extraction easier in animal models, compared with an identical lead without covering sleeves.¹¹⁷ Recent clinical data have corroborated the potential benefit of modified ICD lead shock coils to improve the ease of extraction.¹¹⁹

Figure 4-26 Lead body diameter may not be uniform throughout (top) or may be isodiametric (bottom), which improves the ability to extract chronically implanted leads.

(Courtesy Medtronic, Minneapolis.)

Figure 4-27 Solutions to the problem of shock coil fibrous tissue ingrowth.

Top panel shows the shock coil interstices backfilled with silicone. Middle two panels demonstrate effective backfilling between flatwire helix turns. Bottom two panels show an ePTFE covering sleeve over the shock coil.

(Courtesy Medtronic, Minneapolis; St Jude Medical, Sylmar, Calif; and Boston Scientific, St Paul, Minn.)

ICD Lead Failure

The ICD lead is one of the most complex and vulnerable parts of the ICD system, and lead failure is a significant problem (Box 4-1). ICD lead failure, perhaps more so than with pacing lead failure, can lead to sudden death. This can occur not only by loss of pacing support in pacemaker-dependent patients, but as most often seen, also by causing recurrent inappropriate shocks due to oversensing of lead noise. These shocks can induce ventricular fibrillation, which the system may then not be able to defibrillate successfully.¹²⁰ Early data from the transvenous lead experience held great promise,¹²¹ but late lead failure was seen at longer-term follow-up across all lead designs and manufacturers.⁸¹ Patient-related factors, such as excessive upper body activity, can contribute to ICD lead failure, as can procedural factors at implant, including subclavian (rather than cephalic) access,^122,123 rough handling, traumatic transvenous insertion, overtorquing of active-fixation helices, and kinking of lead loops in the pocket. However, none of these factors is positively identified in most cases; the difficulty in extracting ICD leads (possibly with additional lead damage during its extraction) means that the failure mechanism usually remains obscure. A systematic solution to this problem must involve better ICD lead design and manufacture, with advances in materials technology.

The Transvene family (Medtronic) of coaxial ICD leads had a structural failure rate of 38% at 8 years,⁸³ which was more often caused by fracture of the outer high-voltage coil conductor than disruption of the polyurethane 80A insulation. Despite changes in design and materials, however, newer generations of ICD leads are also prone to late failure. Kleemann et al.⁸¹ found that cumulative survival of 990 ICD leads implanted between 1992 and 2005 was only 60% at 8 years, mainly from insulation defects, including in silicone-based leads. One in five leads implanted for a decade or longer failed. This study was weighted toward the older lead models, however, and 95% of implants were through subclavian access, which may have affected failure rates and mechanisms.

In an industry-wide push for increasing miniaturization, small-caliber ICD leads were introduced in the mid-2000s. Again, although early data gave no cause for alarm,¹²⁴ more recently¹²⁵ the small-caliber (6.6F) Sprint Fidelis lead (Models 6930/6931/6949/6948, Medtronic) has been identified as much more prone than other contemporary lead models to early and late failure, usually presenting with conductor fracture, noise oversensing, and inappropriate shocks. Initial failure rate was 1.29% at 21 months,¹²⁶ but risk appears to increase with time, up to 5.7% to 12.1% at 3 years,^127–129 a failure rate of 3.75% per year, versus 0.58% per year or less for other contemporary ICD lead models.^129,130 The mechanism in most cases is a fracture of the pace/sense hexafilar cathode conductor, either proximally, near the lead anchoring sleeve, or distally, near the anode ring electrode.^102,131 Although no longer being implanted, the problem of failure with this lead is ongoing because of the large number of patients with this model in place.