38.20 Pacing for bradycardia: Device features and algorithms and clinical impact

Cardiac pacing is the treatment of choice for patients with bradycardia. Several existing algorithms and features facilitate individualization and optimization of device programming on an individual patient basis. Individually based optimal choice of device and pacing mode should be the rule. Correct programming of available pacemaker algorithms and features is a prerequisite for optimal patient treatment.

A pacemaker system for permanent cardiac pacing consists of a pacemaker and one lead (single-chamber pacemaker) or two leads (dual-chamber pacemaker) implanted in the right atrium or right ventricle, or both. The pacemaker contains a battery, the power source, and a pulse generator, the electronic unit controlling the behaviour of the pacemaker (see Chapter 38.17 for more details).

Most commonly, the power source is a constant-voltage lithium–iodine battery that has predictable voltage behaviour over time. The battery voltage remains relatively constant throughout most of its discharge. Expected battery longevity for cardiac pacemakers is 6–12 years. Battery current drain is highly dependent on pacemaker programming. A pacemaker with 6 years of battery longevity under nominal pacing parameters may reach its replacement time at 2 years at one extreme or more than 10 years at the other extreme. The clinical indicators of battery depletion vary between manufacturers and models, the most common being a gradual decrease in telemetered battery voltage. Furthermore, estimated remaining battery longevity is indicated during follow-up and an elective replacement indicator is activated when battery depletion approaches and the pacemaker should be replaced.

Pacemaker leads are bipolar or unipolar. Bipolar systems have both electrodes near the distal end of the lead, the tip electrode acting as the cathode and the more proximal ring electrode as the anode. In a unipolar system, a single electrode (the cathode) is located at the lead tip, and the pacemaker can act as the other electrode (the anode). Bipolar leads reduce the risk of myopotential over-sensing, far-field sensing, crosstalk, and local skeletal muscle stimulation, and they allow programmable switching between bipolar and unipolar configurations. Therefore, bipolar leads should be standard in both the atrium and ventricle. Most bipolar leads have a coaxial design with the inner coil terminating at the tip electrode and the outer coil at the ring electrode.

Active endocardial fixation of pacing leads by screwing a fixed or retractable tip-screw into the atrial or ventricular myocardium is most commonly used. Active fixation reduces the frequency of atrial lead dislodgements and allows for straightforward lead implantation in different locations of both right atrium and ventricle, such as the right atrial septum and right ventricular outflow tract. Today, passive fixation by tines or wings trapped in the trabeculae of the right atrium or ventricle is rarely used. For paediatric patients and patients without venous access to the right side of the heart, epicardial leads are the choice. Modern leads have a small tip surface area with a porous surface, promoting low thresholds, low current drain, and good sensing. Steroid-eluting leads with a small reservoir of glucocorticoids at the lead tip are recommended to diminish the inflammatory reaction at the electrode–endocardial interface and improve acute and chronic pacing thresholds and sensing.

In a pacemaker system, the weak link is the pacing lead, most commonly causing complications. Recently, the concept of leadless pacing was introduced. Two leadless pacing systems were tested in observational feasibility cohort studies and found to meet pre-specified safety and efficacy endpoints.^1,2 Using specially designed delivery tools, the leadless pacemaker is introduced via the femoral vein and implanted in the right ventricle where it is anchored either by a helical screw-in fixation electrode¹ or by electrically inactive tines.² So far, leadless pacing is used for single-lead ventricular pacing only. To date, no comparative studies have been conducted that document non-inferiority of leadless pacing to transvenous pacing. Therefore, leadless pacing should be reserved for selected patients in whom traditional transvenous pacing is judged unsuitable or associated with a significantly higher complication rate, such as, for example, no venous access from the upper extremities or haemodialysis with an extremely high infection risk.

Bradycardia caused by atrioventricular (AV) block requires pacing in the ventricle. Dual-chamber pacing is superior to single-lead ventricular pacing in reducing pacemaker syndrome and improving exercise capacity. However, it is well documented that dual-chamber pacing does not prolong survival or reduce morbidity as compared with single-lead ventricular pacing.^3,4,5 The current recommendation is to use dual-chamber pacing as the first choice for patients with AV block^6,7 (Figure 38.20.1). The choice should be made on an individual basis, taking into consideration complication risk, costs, and individual patient factors. Rate-adaptive single-lead ventricular pacing is a reasonable alternative for inactive, elderly patients with AV block.⁵ It is well documented that left ventricular ejection fraction is lower with ventricular pacing as compared to intrinsic activation because of the abnormal electrical activation resulting in an abnormal and less synchronous contraction. Therefore, in case of severely reduced left ventricular ejection fraction and clinical heart failure, cardiac resynchronization therapy should be considered at initial implantation if a high percentage of ventricular pacing is expected.⁸ For patients with persistent atrial fibrillation and AV block, rate-adaptive single-lead ventricular pacing is the mode of choice.⁶

For patients with sinus node disease, single-lead ventricular pacing is associated with a higher risk of atrial fibrillation and thromboembolism than single-lead atrial or dual-chamber pacing.^4,9 AV block with the need for a ventricular pacing lead occurs in 1.5–1.8% of patients with sinus node disease per year,^10,11 and atrial fibrillation is more common with single-lead atrial than with dual-chamber pacing.^12,13 Therefore, dual-chamber pacing is the pacing mode of first choice.⁶ Ventricular pacing should be minimized by AV delay management, either by means of manual optimization of the AV interval or by programming of AV hysteresis (Figure 38.20.1).

All currently available pacemakers possess rate response algorithms intended to imitate physiological increases in heart rate by means of different sensors incorporated into the devices. These sensors convert to pacing rate the strength of an input signal from a surrogate parameter related to physical or emotional exercise.

A range of different sensors have been in use. Currently the most prevalent are the accelerometer (sensing of body acceleration), minute ventilation, and closed loop-based sensors (the latter two are impedance-based between ventricular lead and pacemaker casing). In some devices, sensor blending is used to benefit from the different sensitivity and proportionality in rate response in the individual sensors, for example, an accelerometer sensor is sensitive to initiation and termination of exercise, and a minute ventilation sensor has better proportionality to intensity of exercise.¹⁴

While there is good documentation for sensors to deliver rate responses with various and modest sensitivity and proportionality, there is only weak evidence for the clinical benefit.^15,16

There are recommendations for using rate response pacing in sick sinus syndrome with clear chronotropic incompetence,¹⁷ atrial fibrillation with AV block, and when choosing single-lead ventricular (VVI) pacing in pacing-dependent patients whether in sinus rhythm or atrial fibrillation.^17,18

Haemodynamically, the optimal AV delay in pacemaker patients differs between individuals, for most around 125–200 ms at rest.^19,20,21,22 In dual-chamber pacemakers, sensed AV delay is typically programmed 20–50 ms shorter than paced AV delay to compensate for atrial sensing taking place after the start of atrial activation. In patients with persistent AV block, the most common practice is to use a rate-adaptive AV delay that shortens with higher heart rate. Small crossover studies indicate better patient well-being and exercise tolerance with rate-adaptive AV delay,^21,23 while no larger randomized trials have been conducted.

For patients without persistent AV block, ventricular pacing should be minimized by programming. For this purpose, either an individual optimization of the AV delay or programming of AV hysteresis is recommended.⁶ AV hysteresis automatically increases the AV delay to search for intrinsic AV conduction thereby promoting intrinsic activation. It is well documented that AV hysteresis can reduce ventricular pacing.^24,25 Allowing an AV delay up to 250–270 ms is common. Beyond this length of the AV delay, it is not likely that the benefits of avoiding ventricular pacing exceed the benefits of preserving the AV synchrony.^12,13 Haemodynamically, a very prolonged AV delay is disadvantageous, and prolonged AV conduction is associated with increased risk of atrial fibrillation.²⁶

Two manufacturers have developed more advanced algorithms to minimize unnecessary ventricular pacing, the ‘managed ventricular pacing’ algorithm in Medtronic devices, and the ‘SafeR’ algorithm in Sorin devices. Both algorithms act as single-lead atrial pacing modes switching to dual-chamber pacing modes in case of defined higher levels of AV block. These algorithms are the most effective in reducing ventricular pacing.^27,28,29,30 As compared with dual-chamber pacing with a very high percentage of ventricular stimulation, managed ventricular pacing reduces atrial fibrillation in patients with sinus node disease.³¹ However, as compared with dual-chamber pacing programmed with a normal or slightly prolonged AV delay, neither of these two algorithms provide any clinical benefit with respect to preventing atrial fibrillation or other important clinical endpoints.^32,33,34 Furthermore, several cases have been published of ventricular fibrillation induced by short–long–short sequences occurring with managed ventricular pacing.^35,36,37,38 These algorithms allowing intermittent AV block therefore should not be used for patients with AV block.

Tracking modes in pacemakers (e.g. DDD and VDD) have the inherent risk of tracking atrial tachyarrhythmia to the maximum tracking rate.³⁹ To avoid this, mode switching to a non-tracking mode is necessary when atrial tachyarrhythmia occurs, usually to DDI(R) or VDI(R) (uninterrupted atrial sensing to decide on termination of the arrhythmia). The algorithms governing mode switching differ between device companies.^40,41 A prerequisite for reliable mode switching is good atrial sensing (e.g. of low-amplitude atrial fibrillation). The various algorithms differ with regard to degree of programmability, number of fast atrial intervals required, consecutive interval counting, mean atrial interval, and so on, before mode switching occurs.⁴¹ The atrial tachycardia detection rate is usually programmable. Basic pacemaker intervals are an important consideration in evaluating mode-switching performance. While atrial tachycardia sensing can occur during the post-ventricular atrial refractory period, the post-ventricular atrial blanking period does not allow atrial sensing. Potentially, this may lead to a lack of mode switching in regular atrial arrhythmias such as atrial flutter,⁴² where every other atrial event may coincide with the post-ventricular atrial blanking period. For this reason, many devices include an additional mode-switching-related algorithm, atrial flutter response, to reveal underlying atrial flutter in such situations.⁴³ Mode switching should be activated whenever a tracking mode is programmed.¹⁷

The most common pacemaker-mediated tachycardia is endless loop tachycardia (ELT) which can occur in dual-chamber pacing modes with atrial tracking.³⁹ The refractory interval post-ventricular atrial refractory period prevents ELT in most circumstances. Commonly, ELT is initiated by premature ventricular contractions, giving rise to retrograde conduction to the atrium beyond the post-ventricular atrial refractory period, leading to tracking of the sensed retrograde atrial event, constituting the loop. Preventive algorithms differ between devices, but usually include extension of the post-ventricular atrial refractory period after premature ventricular contractions and algorithms to terminate the ELT by interrupting the loop (e.g. by withholding one ventricular paced beat, extending the post-ventricular atrial refractory period, or by delivering atrial pacing when atrial sensing and ventricular pacing occurs at maximal tracking rate, or at a programmable ELT rate). Detection algorithms differ between devices (AV interval modulation to detect VA linking indicative of ELT, ELT rate programmability, etc.).

In general, it is recommended to programme these features on when dual-chamber pacing in an atrial tracking mode is operative.

Most contemporary devices include automatic threshold determination as a programmable option for atrial and right ventricular leads. Algorithms differ between companies mainly with regard to whether there is beat-to-beat capture verification or not.³⁹ They depend on detecting the evoked response in the myocardium from the pace stimulus and discerning this from the polarization voltage on the lead tip. The output is adjusted automatically with a programmable safety margin above the measured threshold. Safety in pacing is provided when threshold changes are reliably detected. These automatic features are most beneficial when high pacing percentages or high thresholds are present and increased battery longevity can be anticipated, or when remote monitoring is planned.^44,45 Pacing thresholds should still be tested at in-clinic visits to verify correspondence with automatic measured values.

Rate hysteresis allows the rate to go below the nominal lower rate limit. When intrinsic beats are sensed, pacing will occur when reaching the hysteresis rate, while paced beats resume at the programmed lower rate if intrinsic events are not registered. The related feature, night rate, in effect allows a programmable lower rate limit at night-time below the nominal lower rate limit and mimics physiological diurnal rate variation.

These features are recommended on an individual basis. No evidence for beneficial effect is present. By reducing pacing, they can be battery conserving.

Algorithms to terminate or prevent atrial arrhythmias have been applied with equivocal evidence for effect. Atrial overdrive and preference pacing to suppress premature atrial beats supposed to initiate atrial tachyarrhythmia as well as non-competitive atrial pacing to avoid pacing in the atrial vulnerable phase are present in different variations in many devices. While seemingly physiologically rational, the evidence is not strong for clinical effect.⁴⁶

Active pacing termination features exist in a few devices with the assumption that even more disorganized atrial rhythms (i.e. atrial fibrillation episodes) are initiated by short bursts of more organized atrial arrhythmia amenable to pace termination. The evidence is controversial.^32,47,48 However, in the light of ample evidence that automatic focus firing from the pulmonary veins initiates atrial fibrillation episodes, this assumption is controversial, lending more support to the preventive features.

These algorithms should be programmed on an individual basis.¹⁷

Modern pacemakers sample a vast amount of diagnostically useful follow-up information³⁹ (for details, refer to Chapter 38.22). Importantly, some of this information is only sampled if the pacemaker is programmed to do so, for example, high ventricular rate episodes, pre-episode electrogram storage, and automatic capture threshold information over time. Information on the performance of some of the algorithms described previously in the ‘Atrioventricular delay programming’ section is included in follow-up data (e.g. percentage of time spent in mode switch, sensor-indicated rate distribution, episodes with ELT, managed ventricular pacing function, and AV interval) giving opportunity for refining the function of these algorithms in the individual patient at follow-up.

Magnetic resonance imaging (MRI) conditional pacemakers contain a MRI safe mode algorithm to be programmed on during MRI, but currently only a minority of implanted systems will be MRI safe. However, most pacing systems can undergo MRI (1.5 T MRI) at minor risk,⁴⁹ provided that appropriate reprogramming is performed according to management guidelines.¹⁷ An important point in relation to programming is whether the patient is pacemaker dependent, that is, whether an asynchronous or an inhibited mode should be programmed during MRI. Features that could elicit inappropriate pacing, such as rate smoothing, trigger pacing, and conducted atrial fibrillation response, must be programmed off. A relative contraindication for MRI is the presence of abandoned leads.⁵⁰

Remote follow-up is becoming the standard of care in the implantable cardioverter defibrillator population. In pacemaker patients, the cost:benefit ratio is less favourable. Most contemporary pacemakers have the option to be followed in remote monitoring systems. Presently, the choice to do so is based on an individual assessment (e.g. in older non-ambulatory patients or other situations where specific reasons favour this option). According to the most recent ESC GL on cardiac pacing, remote device management is recommended in patients who have difficulties attending in-office follow-up and in patients who have a device component that is recalled or on advisory.⁵¹ Remote device management may allow spacing in-office follow-up visits up to 24 months. Automatic threshold measurement and other options should be activated to derive most benefit (see Chapter 38.22 for more details).