INTRODUCTION — Cardiac resynchronization therapy (CRT) is a treatment for selected patients with chronic heart failure with reduced ejection fraction (HFrEF) and bundle branch block; CRT involves biventricular pacing, or pacing of only the left ventricle [1-6]. CRT can be achieved with a device designed only for pacing (CRT-P) or with the added capability for defibrillation (CRT-D) (image 1).
The rationale for CRT is that ventricular dyssynchrony impairs the function of a failing ventricle. Resynchronization may improve performance and reverse the deleterious process of adverse ventricular remodeling, improve quality of life, reduce heart failure hospitalizations, and improve survival. CRT does not obviate medical therapy but is incremental therapy used in conjunction with guideline directed medical therapy. (See "Cardiac resynchronization therapy in heart failure: Indications and choice of system", section on 'Rationale for CRT'.)
The implantation technique for CRT and initial programming considerations will be reviewed here.
Indications for CRT and outcomes of CRT in patients in sinus rhythm or with atrial fibrillation and cardiac pacing in patients with heart failure are discussed separately. (See "Cardiac resynchronization therapy in heart failure: Indications and choice of system" and "Overview of pacemakers in heart failure" and "Cardiac resynchronization therapy in heart failure: Indications and choice of system", section on 'Rationale for CRT'.)
SYSTEM COMPONENTS — The most common CRT pacing configuration involves three leads (right atrial, right ventricular [RV], and left ventricular [LV]) with a three-lead pulse generator (image 1). Compatibility of the connector pins of all pacing leads into the device header across manufacturers is facilitated by the International Standardization Organization, whereby connections for bipolar and unipolar leads (IS-1), quadripolar leads (IS-4), and defibrillation leads (DF-4) are uniform and interchangeable. Thus, contemporary lead and generator technology can be freely mixed among vendors, including epicardial and transvenous leads.
Right atrial lead — An atrial lead allows sensing and tracking of sinus rhythm to synchronize ventricular activation to follow either a spontaneously occurring sinus beat or a pacemaker-delivered atrial beat.
Patients with permanent atrial fibrillation or flutter are typically implanted without an atrial lead given the absence of functional atrial contraction. CRT in patients with atrial fibrillation is discussed separately. (See "Cardiac resynchronization therapy in atrial fibrillation".)
Right ventricular lead — The RV lead is an essential component for biventricular pacing systems. Most RV leads are placed transvenously. The RV lead tip is usually placed apically to provide the greatest RV to LV tip separation; in the case of CRT-D, this position also provides an optimal shocking vector. The transvenous RV lead is typically secured with either an active fixation helix or tined mechanism into the RV apex, apical septum, mid septum, or outflow tract. His bundle pacing is an emerging alternative to CRT and is discussed separately (waveform 1).
Epicardial RV lead placement is generally performed only in patients with access problems (eg, venous thrombosis) and in patients requiring concomitant cardiac surgery (eg, coronary artery bypass grafting) with which epicardial lead placement can be easily combined. When an epicardial approach is selected, the lead is typically secured anteriorly to the right of the interventricular groove. (See 'Surgical LV epicardial lead placement' below.)
A key role of the RV lead is sensing intrinsic rhythm, as measured by the bipolar (or integrated bipolar) or unipolar R wave amplitude measured in millivolts. Such sensing is useful to inhibit pacing during ventricular arrhythmias and for tachyarrhythmia therapies (ie, shocks or antitachycardia pacing) in the case of a CRT-D system. (See "Implantable cardioverter-defibrillators: Overview of indications, components, and functions".)
In the majority of implanted CRT systems, the RV lead paces. As discussed below, the RV lead can be programmed to deliver pacing that is either offset from or simultaneous with LV stimulation. For some innovative CRT methods (eg, adaptive CRT or true LV only CRT), pacing can be entirely deferred. (See 'Initial programming considerations' below.)
Left ventricular lead — All CRT candidates have evidence of native or paced electrical dyssynchrony, usually due to electrical delay or block in the left-sided His-Purkinje conduction system. Outcomes and treatment effects following CRT are best in patients with true left bundle branch block. Outcomes in patients with nonspecific interventricular conduction delay and right bundle branch block (RBBB) are more controversial without definitive evidence of benefit. Pacing stimulation delivered from the LV lead is therefore essential to compensate for the delayed LV activation in order to restore electrical and mechanical synchrony of the ventricular contractions. While CRT pacing can be accomplished using LV leads with unipolar, bipolar, or quadripolar electrodes placed transvenously or with unipolar or bipolar electrodes placed epicardially, contemporary implantations are now most commonly performed transvenously using leads with quadripolar electrodes. These quadripolar leads give the implanting and treating physician more pacing vectors to pace the ventricle and therefore more options for tailoring therapy and avoiding potential complications (eg, phrenic nerve/diaphragmatic capture) (see 'Multi-site pacing' below). The final pacing configuration is chosen based on the absence of phrenic nerve stimulation near the programmed output, favorable electrical parameters to minimize battery depletion (ie, pacing threshold, impedance), and desired electrical findings (ie, Q-LV [site of greatest electrical delay], paced QRS [shorter duration and optimal morphology]).
IMPLANTATION TECHNIQUE — CRT can be initiated via a de novo implant or as an upgrade from an existing single or dual chamber system. General criteria for placement of cardiac implantable electronic devices apply, including identification of candidacy for moderate sedation and absence of contraindications to device implantation (such as active bloodstream infection). In most centers, contrast dye is administered to guide lead implantation and there is a risk of contrast nephropathy in patients with advanced chronic kidney disease. (See "Anesthetic considerations for electrophysiology procedures", section on 'Procedures for cardiac implantable electronic devices'.)
De novo implantation — Most de novo CRT implantations are performed via a transvenous approach. A pocket for the device is surgically created within the plane of prepectoral fascia 2 to 3 cm below the clavicle. The leads are extended via venous access obtained in either the axillary, subclavian, or cephalic veins and anchored to the pectoral muscle with nonabsorbable suture to tie down sleeves.
Left ventricular (LV) lead implantation via the coronary sinus (CS) requires cannulation with a sheath or catheter, and should be performed by experienced operators. Cannulation tools include specially designed curved long-sheaths, and in cases of difficult access deflectable catheters with bipolar electrodes to identify the characteristic electrogram of the CS ostium involving approximately equally sized atrial and ventricular components. Some operators employ contrast injections, while others use only fluoroscopic landmarks. Soft-tipped guidewires and subselecting catheters can be used to both engage the ostium and target specific LV venous branches of interest. Anatomic and functional barriers to CS cannulation include an eccentric ostium due to anatomic variation, or a consequence of right atrial dilatation. Other barriers include the presence of venous valves (Thebesian valve, Valve of Vieussens) or coronary sinus strictures.
In a contemporary study of de novo transvenous CRT implants among 2014 patients from 114 centers, the overall success rate was 97.1 percent using a median 17 minutes of fluoroscopy, 25 cc of contrast dye, and an a 2.6 percent adverse event rate [7]. These outcomes are better than those reported by a similarly sized study approximately 10 years earlier [8], suggesting that technical procedural advancements (ie, cannulation tools, CS venoplasty, quadripolar LV pacing leads), and/or operator experience have increased success rates while reducing complications. The collective experience of an implanting center also appears to be important. In an analysis of the National Inpatient Sample Database involving 410,104 de novo CRT implants between 2003 and 2011, lower hospital volume was associated with worse outcomes, including higher complication rates including increased in-hospital all-cause mortality [9].
Surgical LV epicardial lead placement is discussed below. (See 'Surgical LV epicardial lead placement' below.)
Upgrade from a pre-existing pacing system — CRT upgrade from a single or dual chamber transvenous device involves adding an LV pacing lead. The additional lead is most commonly placed transvenously. However, approximately 5 to 10 percent of leads cannot be implanted transvenously via the coronary sinus. Therefore, a small minority of patients require surgical epicardial LV lead placement. Upgrade to a CRT system is appropriate for the patient with a pre-existing pacemaker who develops heart failure with electrical disease substrate that meets criteria for CRT (eg, left bundle branch block). This includes patients with progressive cardiomyopathy induced by a high percentage of right ventricular pacing (pacing-induced cardiomyopathy). (See "Cardiac resynchronization therapy in heart failure: Indications and choice of system".)
The benefits of CRT upgrade should be weighed against the procedural risk and complexity of adding a lead, particularly in the presence of venous occlusive disease [10]. In the REPLACE registry, the complication rates for procedures involving lead additions were greater than those for generator replacement alone [11]. In an analysis comparing 19,546 CRT upgrades with 464,246 de novo procedures in the United States between 2003 and 2013, the upgrades were independently associated with increased mortality (odds ratio [OR] 1.91), cardiac perforation (OR 3.2) and need for lead revision (OR 2.09) [12]. Factors to consider when deciding whether to proceed with CRT upgrade and in choosing an approach for upgrade include the patient’s age, comorbidities, presence of venous occlusive disease, accessibility to an experienced lead extraction center, and accessibility to an experienced cardiac surgery center.
Contrast venography is recommended prior to creating incisional access for any CRT upgrade procedure. The presence of venous occlusive disease complicates the addition of an ipsilateral transvenous LV lead. Venous occlusive disease (predominantly axillary or subclavian occlusion) occurs in 10 to 30 percent of patients with implantable pacing devices [13-17]. Established risk factors for venous occlusive disease include the number of leads traversing the vein [15] and the presence of larger diameter implantable cardioverter-defibrillator leads, particularly those with superior vena cava (SVC) shocking coils [17]. However, since less than 10 percent of patients with venous occlusive disease are symptomatic or have physical exam findings such as asymmetric peripheral edema [16], venography is recommended prior to creating incisional access for CRT upgrade procedures.
Approaches to CRT upgrade in the presence of venous occlusive disease include the following:
●One common approach is transvenous extraction of one of the pre-existing leads to allow guidewire-retained access for replacement, plus the addition of a new LV lead [18]. In the LEXICON study of 2405 extracted leads, the indication of venous occlusion represented 4.5 percent of the total [19]. In the total study population, clinically successful lead extraction was accomplished in 97.7 percent and the major adverse event rate was 1.4 percent, including procedural death in 0.28 percent. Risk factors for complications include body mass index <25 kg/m2 and procedures performed in low volume extraction centers.
●Surgically placed LV epicardial lead, as detailed in the next section [20-22].
●"Deep" transvenous access medial (ie, central) to the point of obstruction. Disadvantages include heightened risk of vascular or thoracic trauma, including pneumothorax and increased risk of lead fracture on follow-up [23,24]. Some centers utilize smaller bore needles and guidewires (ie, micropuncture kits) to mitigate the risk of creating a large pneumothorax.
●Contralateral implant of a de novo system with full or partial abandonment of the pre-existing system. Disadvantages include increased risk of venous occlusive disease involving the SVC with multiple traversing leads, device-device and lead-lead interactions, and long-term management of redundant leads, particularly in young patients, or those at risk for future bloodstream infection [25]. This approach is less desirable given its limitations, especially in younger patients.
●Parasternal or paraclavicular lead tunneling to an alternate point of venous entry, including the ipsilateral internal jugular vein if patent, or the contralateral axillary/subclavian venous system [26]. Concerns include risk for skin erosion over bony sternal or clavicular prominences where there is typically minimal subcutaneous tissue, and creating long-term lead management problems in the event of a need for subsequent transvenous extraction. For these reasons, this approach also seems to be less commonly utilized.
●Percutaneous subclavian venoplasty using high-pressure dilating balloons is an alternative approach to vein obstruction [14].
Surgical LV epicardial lead placement — Epicardial lead placement entails an increased risk of adverse events and provides no greater benefit compared with transvenous placement, so it is generally regarded as a second-line option in patients who have had unsuccessful transvenous placement or have venous occlusive disease. The exception is for patients requiring concomitant cardiac surgery (ie, valve surgery or coronary artery bypass grafting) with which LV epicardial lead placement can be easily combined. In a series of 42 patients with dedicated "stand alone" LV epicardial lead placement using a mini-thoracotomy approach, the mean length of stay was 3.4 days and 30-day adverse event rate was 17.5 percent, including 4.8 percent mortality, 7.5 percent LV lead noncapture, and 5 percent infection [21]. A similar study of 30 stand-alone LV lead placements via mini-thorocatomy following failed transvenous placement reported successful implant in all patients without any procedural or 30 day mortality, and a lower mean length of stay of 1.3 days. However, this study did not report electrical lead testing data nor bleeding or infectious complication rates [27]. A study comparing the clinical outcomes of 96 patients with surgically placed epicardial LV leads to transvenous implants found no difference over a mean five years of follow-up in which the overall clinical response rate to CRT pacing was 65 percent [28] In this study, major adverse events occurred in 5 percent of patients undergoing stand-alone LV lead placement and 14 percent of patients with concomitant LV lead placement.
The earliest CRT systems involved surgically placed epicardial leads via sternotomy or full thoracotomy approach. Contemporary techniques for dedicated epicardial LV lead placement include a mini-thoracotomy approach involving 3 to 5 cm incisional access in the fourth or fifth intercostal space just anterior to the mid axillary line and single lung ventilation to allow dissection into the pericardial space for implantation of active fixation LV leads [20]. A key advantage of surgical epicardial lead placement is that lead placement is not confined to the anatomic branches of the LV venous circulation as is the case with transvenous placement. The lead is fixed directly to the cardiac tissue using a specifically designed instrument, and electrically tested in the same manner as transvenously placed leads. Surgical leads are connected to the pacing device using an IS-1 connector in the case of a bipolar, or a Y-adapter lead extender allowing two unipolar leads to be configured to an IS-1 connector.
LEFT VENTRICULAR LEAD POSITION — The optimum left ventricular (LV) lead location paces viable myocardium (ie, avoids scar) from the site of greatest electrical and mechanical delay while avoiding phrenic nerve stimulation.
The available evidence suggests that the optimal position for the LV lead is generally lateral or posterolateral. Because the posterolateral wall is often the latest segment to contract in a dyssynchronous LV in the presence of left bundle branch block (LBBB), it is generally the targeted location. In the follow-up results of the MADIT CRT study, only patients with lateral and posterior LV lead locations derived long-term mortality benefit from CRT over ICD alone [29]. Additionally, a more basal (versus apical placement) has also been associated with better rates of CRT response.
Transvenous passage of the lead via the coronary sinus into a coronary vein constrains lead location to available branches. Classic LV venous anatomy features an anterior vein coursing in the interventricular groove, an anterolateral branch coursing diagonally from LV apex to base, a midlateral and/or posterolateral vein that often shares a common trunk, and the middle cardiac vein coursing posteriorly to the apex. The lateral and posterolateral branches usually represent the optimal targets for LV pacing leads (image 2).
A key advantage of surgical epicardial lead placement is that lead placement is not confined to the anatomic branches of the LV venous circulation as is the case with transvenous placement. The key challenge with epicardial LV lead implantation is achieving a sufficiently lateral location. Sometimes this can be difficult if the surgical exposure is not optimal.
Right to left interelectrical delay — Evidence-based recommendations for CRT are strongest for patients with LBBB and QRS duration >150 msec (see "Cardiac resynchronization therapy in heart failure: Indications and choice of system", section on 'Indications for referral for CRT'). The presence of LBBB causes overall delayed LV activation compared to the right ventricle, leading some operators to measure the interlead electrical delay (IED) time between sensing of native impulse at the RV and LV implanted leads. In a study of 68 patients, IEDs were independently associated with CRT response (reverse remodeling), even when accounting for the presence of myocardial scar (OR 3.99; 95% CI 1.02-15.7) [30]. In another study involving 160 patients, an IED ≥100 ms was associated with more pronounced LV reverse remodeling when lead implantation site was guided by identification of the latest mechanical delay in a non-scarred myocardial segment [31]. Further studies are needed to determine the clinical utility of IED measurement.
Site of greatest electrical delay — Targeting the site of greatest LV electrical delay for pacing stimulation to optimize electrical resynchronization has been associated with favorable CRT outcomes [32-35], and is therefore recommended. Quantitatively, the site of LV activation can be measured from the onset of the Q wave recorded on the surface electrocardiograph to the timing of the local LV electrogram (Q-LV). At the time of implant, operators have an opportunity to measure this interval (in milliseconds) as one of the potential parameters to consider in selecting the site for LV pacing stimulation.
The following studies illustrate the importance of LV electrical delay as measured by Q-LV as a predictor of improvement in hemodynamic improvement:
●In separate prospective acute hemodynamic studies of 31 and 32 patients undergoing CRT implant, Q-LV was strongly associated with hemodynamic improvement as measured by dP/dt max independent of the pacing mode [32,33].
●In a prospective study of 156 patients, Q-LV was the only independent predictor of improvement in LV ejection fraction and decrease in LV end systolic volume in follow-up [34].
●In analysis of the SMART AV trial involving 426 patients, Q-LV was associated with greater improvement in mitral regurgitation, suggesting an additional physiologic mechanism for targeting the site of greatest electrical delay [35].
Site of greatest mechanical delay — Some experts consider the site of greatest mechanical delay as a factor ancillary to the site of greatest electrical delay when identifying the site for LV pacing. Targeting the site of greatest mechanical delay for pacing stimulation has been associated with improved CRT outcomes in some studies.
Echocardiographic-based imaging and magnetic resonance imaging [36-38] have been utilized to evaluate the value of placing the LV pacing lead at or near the area of maximal mechanical delay. Tissue Doppler imaging (TDI) was initially assessed in a series of 54 patients [39] and a greater reduction in end-systolic volume was associated with LV pacing at the site of maximal delay. TDI has been the most widely studied and utilized method in clinical practice [40-47]. Additional echocardiographic-based methods include myocardial strain imaging [48,49]. (See "Tissue Doppler echocardiography", section on 'Use in heart failure and resynchronization therapy'.)
Although direct measures of mechanical dyssynchrony have been investigated as a means of identifying responders to CRT [50-54], the clinical utility of such assessment has not been established.
Left ventricular scar — The LV lead is optimally positioned in the vicinity of viable myocardium and not in or immediately adjacent to scar. Pacing in regions of LV scar has been associated with lower response rates. In addition, pacing in a densely scarred region may require higher pacing outputs, which can lead to premature battery drain.
The following observations suggest that the presence, location, and/or extent of LV scar may impact response to CRT:
●In a series of 40 patients with indications for CRT who underwent cardiac magnetic resonance imaging, 14 had a transmural posterolateral scar [36]. These patients had a lower response rate to CRT (14 versus 81 percent of patients without posterolateral scar).
●In a series of 50 patients with indications for CRT who underwent single-photon emission computed tomography imaging, global scar burden, number of severely scarred segments, and scar burden near the LV lead were all inversely correlated with increase in LV ejection fraction after CRT [55].
●In a feasibility study of 15 patients, multi-modality imaging including cardiac magnetic resonance (CMR) and CT were integrated with fluoroscopy to guide LV lead placement by defining the venous anatomy, avoiding scar, and identifying the phrenic nerve [56]. The use of real-time image-guidance (involving overlaying the CMR and CT datasets onto live fluoroscopy) for LV lead placement in six patients was associated with pacing closer to the target area than in the nine patients who received LV lead placement without such overlays. Reverse remodeling was nominally but not significantly greater in the real-time image guidance group (LV end-systolic volume change -30 versus -19 percent)
Electrocardiographic measures of scar were associated with a reduced odds of reverse remodeling [57].
COMPLICATIONS
Early complications — Early CRT implantation complication rates have varied among studies but appear to be higher with CRT upgrades as compared with de novo implantation [12]and lower among high volume centers [9]. In a contemporary study of 2014 patients from 114 centers undergoing de novo transvenous CRT implant, the composite complication rate was 2.6 percent at three months, most commonly involving LV lead dislodgement in 1.7 percent of patients and phrenic nerve stimulation not amenable to reprogramming in 0.5 percent of patients [7]. A meta-analysis of clinical trials, albeit among older studies, reported an overall acute complication rate of 14 percent that is largely driven by lead-related complications but also includes 0.8 percent perioperative mortality [58]. Of note, one study looking exclusively at CRT upgrades found an acute complication rate of 11 percent [59].
Higher in-hospital mortality rates were observed in a cohort of 26,887 patients undergoing implantable cardioverter-defibrillator (ICD) and/or CRT implantation that included older adults with rates ranging from 0.7 to 1.2 to 2.2 percent in patients aged <80, 80 to 85, and >85 years [60]. While patients over 80 years old undergoing CRT implantation may face higher complication rates than younger patients, longer-term data suggest that mortality rates for these patients are only slightly higher than in the general octogenarian population. A single-center study of 95 consecutive patients over age 80 who underwent CRT (86 percent with a defibrillator) survived a mean 4.1 years (CI 3.7-4.5) and demonstrated only modestly worse overall survival experience when compared with the general octogenarian population with the Kaplan-Meier curves only beginning to diverge at two years post-CRT follow-up [61].
The contemporary availability of quadripolar LV pacing leads has markedly reduced phrenic nerve stimulation by affording multiple different pacing vectors in order to avoid stimulating the left hemi-diaphragm [62]. Additional acute complications include coronary sinus or coronary vein trauma, pneumothorax, pocket hematoma, and infection [8,63-65]. There are also specific concerns with LV lead placement such as prolonged radiation exposure due to the complexity of the transvenous implantation procedure, which may have acute skin effects and contribute to long-term radiation-related risks [66].
Late complications — The incidence of late complications is illustrated by the following results from a review of implantation success rates and safety outcomes in 89 studies of patients undergoing CRT or CRT-ICD (CRT-D) implantation [65]:
●In 54 studies (6123 patients) of CRT-alone devices, 5 percent of CRT devices malfunctioned and 2 percent of patients were hospitalized for infections in the implant site over six months of follow-up. During a median follow-up of 11 months, lead problems occurred in 7 percent of CRT devices.
●While there is a theoretical risk that pacing from an LV lead may be proarrhythmic due to alterations in depolarization and repolarization sequences [67], pooled analysis from 14 randomized controlled trials did not demonstrate any excess risk of sudden death or noncardiac death in CRT device recipients.
●In 36 studies (5199 patients) of combined CRT-ICD devices, 5 percent of CRT-ICD devices malfunctioned, 1 percent of patients developed site infection, and lead problems were detected in 7 percent of patients over 12 months of follow-up.
INITIAL PROGRAMMING CONSIDERATIONS
Approach to initial programming — The optimal pacing strategy at the time of implant has not been fully defined, and may be patient specific.
●A key goal for all patients is to achieve 100 percent CRT pacing. (See 'AV timing' below.)
●The following are additional recommended strategies to optimize CRT pacing (figure 1):
•To program left ventricular (LV) pacing stimulation from the site of greatest electrical delay as measured by Q-LV, and to critically evaluate the surface electrocardiographic (ECG)-paced QRS duration and morphology in response to any chosen pacing configuration and to avoid wider QRS morphologies, those resulting in a left bundle branch block, or an apical QRS morphology.
•To apply fusion pacing and/or multi-site pacing algorithms at the time of implant for qualifying patients. (See 'Fusion pacing' below and 'Multi-site pacing' below.)
•To program CRT devices to promote battery longevity by choosing stimulation sites with the lowest possible electrical thresholds and greater impedance measurements to reduce current drain while utilizing the shortest possible pulse width within the range of acceptability that also does not result in phrenic nerve stimulation. Withholding right ventricular (RV) pacing stimulation, as often the case with fusion pacing, is also expected to extend battery longevity.
•Patient complaints of phrenic nerve stimulation should be evaluated promptly, and addressed with prompt reprogramming once detected.
AV timing — For CRT pacing, a shortened atrioventricular (AV) delay (ie, 90 to 120 ms) is generally applied to prevent native AV conduction, and thus increase the percentage of biventricular pacing. (This contrasts with the strategy for standard pacemakers, for which the AV interval is often extended beyond 200 ms to promote native AV conduction.) In a post-hoc analysis of the MADIT CRT study involving 1235 patients, sensed AV delays less than 120 ms demonstrated overall better clinical outcomes when compared with longer sensed AV delays, likely attributable to a greater percentage of biventricular pacing in place of intrinsic activation via the native diseased His-Purkinje system [68].
In patients who fail to experience improvement in symptoms or cardiac function following CRT implantation, postprocedural adjustment in the AV delay ("AV optimization") performed in conjunction with Doppler echocardiographic assessment may be helpful as a means of improving LV systolic function and reducing presystolic mitral regurgitation. The AV delay can be programmed so that the end of atrial contraction (marked by the end of the A wave) is timed to coincide with the onset of ventricular contraction (marked by the onset of systolic mitral regurgitant flow) [69]. However, the clinical efficacy of AV optimization has not been established as randomized clinical trials have failed to demonstrate benefit [70].
VV timing — With CRT devices, the timing between the right and left ventricular pacing stimulus (VV timing) can be adjusted, but the optimal programming strategies for VV timing have not been fully defined and may be patient specific. There is limited evidence to support routine adjustment of VV timing using echocardiographic assessments.
Optimizing electrical synchrony — The surface ECG-paced QRS duration and morphology are carefully evaluated with each LV pacing configuration. Pacing vectors are often adjusted until the narrowest QRS duration is achieved. In general, pacing vectors are avoided if they result in wider QRS morphologies (less electrical synchrony), morphologies similar to left bundle branch block (delayed LV activation), or an apical QRS morphology (less basal pacing). The paced QRS duration and morphology at the time of implant must be judged in the context of anatomic lead location and LV substrate. Maximizing electrical resynchronization using the surface QRS duration as a surrogate marker is beneficial in many instances (as discussed in the fusion pacing section below). However, achievement of a narrow QRS does not always translate into improved mechanical synchrony. Intrinsic ventricular scar patterns can create "balanced delay" in which the overall right and left ventricular activation are relatively simultaneous (ie, paced QRS appears narrow) but remain mechanically dyssynchronous due to nonhomogenous conduction through and around intrinsic LV scar.
Mismatch between electrical resynchronization and mechanical resynchronization likely explains why an CRT implant strategy focused on achieving the shortest possible paced QRS duration has yielded mixed echocardiographic and clinical results:
●Among patients with little or no intrinsic conduction (ie, complete heart block), achieving a more narrowly paced QRS duration was associated with reduction of LV end-systolic volume of ≥15 percent from baseline [71].
●In a substudy of the PROSPECT trial, greater shortening of the paced QRS over the intrinsic QRS duration was also associated with significantly increased likelihood of clinical and echocardiographic response (odds ratio [OR] 0.89) [72].
●In an analysis of the MADIT CRT trial, shortened paced QRS duration was not qualitatively associated with better outcomes, while apical pacing demonstrated worse outcomes [73].
An apical pacing morphology is always to be avoided and typically characterized by a left bundle branch block appearance in V1 with a superior axis that is negative in limbs leads II, III, and aVF with a widened QRS >200 ms.
Optimizing mechanical synchrony — Some experts favor adjusting the timing between the right and left ventricular pacing stimulus (VV timing) to favor earlier (eg, by 20 to 50 ms) right or left ventricular activation with the degree of interventricular synchrony assessed by echocardiography. Among patients who undergo VV optimization, cardiac performance is generally optimal when the LV pacing impulse is pre-excited relative to the RV impulse. While VV optimization is often performed in nonresponders to CRT therapy, the evidence for its use is mixed (see 'VV timing' above). It remains an area of continued investigation [74].
Emerging techniques
Fusion pacing — An increasing number of clinicians test the application of a fusion pacing algorithm at the time of implant for patients demonstrating intrinsic AV nodal conduction, and who have a device with an applicable algorithm. For patients with intrinsic left bundle branch block and normal AV conduction, CRT can be achieved by fusion pacing as an alternative to biventricular pacing; the concept is to create fusion between intrinsic conduction occurring antegrade over the intact right bundle and LV pacing ipsilateral to the left bundle branch block. Fusion pacing entails timing the delivery of the LV pacing impulse such that the intrinsic RV activation is "fused" with the paced LV activation. The required delay is evaluated by measuring the time interval for intrinsic conduction from the right atrial lead to the RV lead.
Thus, fusion pacing should not be equated with "LV only" CRT pacing at largely shortened AV intervals without incorporation of sensed atrial-RV conduction time interval. Indeed, early randomized trials including PATH-CHF and DECREASE-HF, B-LEFT HF, and GREATER EARTH yielded similar outcomes when comparing LV only pacing in patients with left bundle branch block with biventricular pacing at shortened AV intervals [75-78].
Proprietary algorithms have emerged to synchronize ventricular contraction by delaying the timing of LV pacing stimulation to synchronize with intrinsic RV conduction by measuring the sensed atrial-RV conduction timing [79]. Another algorithm allows dynamic adjustment of the AV delay to accommodate each patient's changing need as the ventricle remodels while allowing the option of using varying combinations of LV pacing, RV pacing, and intrinsic conduction (ie, so-called "triple fusion") to be characterized at the time of procedural implant [80,81]. Ventricular activation associated with the native His-Purkinje system is faster than paced ventricular activation wave fronts spreading centrifugally from the stimulus. Thus, fusion-pacing algorithms commonly result in a more narrowly paced QRS than either traditional biventricular pacing or LV only pacing as utilizing an intact right bundle and/or septal His-Purkinje allows more rapid electrical depolarization (figure 2).
In the Adaptive CRT trial, 478 patients were randomized in a 2:1 ratio to fusion pacing (LV only pacing synchronized to RV contraction using the proprietary algorithm) versus standard CRT with echocardiographic-guided optimization. At mean 20-month follow-up, the fusion pacing group demonstrated significantly fewer subsequent all-cause hospitalizations (OR 0.54, 95% CI 0.31-0.94) [82].
Multi-site pacing — There is limited data on the use of multi-site pacing and the utility of this approach is still under investigation. Some experts test the application of a multi-site pacing algorithm at the time of implant for patients undergoing quadripolar LV placement, and for whom have a device with an applicable algorithm. Quadripolar LV pacing leads have markedly reduced the incidence of phrenic nerve stimulation and improved the success rates of transvenous LV lead implant via the coronary sinus by opening up greater pacing programmability beyond the limited vectors offered by a bipolar lead [62]. This has led to a programming strategy using two or more LV pacing configurations simultaneously to overcome transmyocardial conduction delay and/or improve intraventricular dyssynchrony. In an observational study of 110 patients comparing standard with multi-site pacing, the one-year echocardiographic response rate with optimally positioned leads was 72 percent with standard programming and 90 percent with multi-site pacing [83].
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Arrhythmias in adults" and "Society guideline links: Cardiac implantable electronic devices".)
INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5th to 6th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10th to 12th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.
Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)
●Basics topic (see "Patient education: Cardiac resynchronization therapy (The Basics)")
SUMMARY AND RECOMMENDATIONS
●In most patients, cardiac resynchronization therapy (CRT) is delivered though three leads (right atrial, right ventricular, and left ventricular [LV]). The positioning of the LV lead is critical. Optimal LV lead placement is generally lateral or posterolateral. (See 'System components' above and 'Left ventricular lead position' above.)
●CRT can be initiated via a de novo implant or as an upgrade from an existing single or dual chamber system. (See 'Implantation technique' above.)
●The LV lead for CRT is most commonly accomplished via a transvenous coronary sinus route. (See 'Left ventricular lead' above and 'Implantation technique' above.)
●Epicardial lead placement entails an increased risk of adverse events and provides no greater benefit compared with transvenous placement so it is a second-line option in patients with unsuccessful transvenous placement, venous occlusive disease. The exception is for patients with planned cardiac surgery (ie, coronary artery bypass grafting) for whom LV epicardial lead placement can be easily combined. (See 'Surgical LV epicardial lead placement' above.)
●Complications of transvenous CRT implantation include unsuccessful placement, coronary sinus or coronary vein dissection/trauma, pericardial effusion/tamponade, pneumothorax, diaphragmatic/phrenic nerve pacing, inability to place a coronary sinus lead, and infection. Complication rates appear to be higher for upgrades than de novo implants, and lower at high volume centers. (See 'Complications' above.)
●A key goal for all patients is to achieve 100 percent CRT pacing. For CRT pacing, a shortened atrioventricular (AV) delay (ie, 90 to 120ms) is generally applied to prevent native AV conduction, and thus increase the percentage of biventricular pacing. (See 'AV timing' above.)
●The optimal programming strategies for timing between the right and left ventricular pacing stimulus (VV timing) for CRT devices has not been fully defined. At implant, it is important to evaluate the surface electrocardiograph-paced QRS duration and morphology in response to each chosen pacing configuration and seek the narrowest QRS duration while avoiding a left bundle branch block or an apical QRS morphology. (See 'Optimizing electrical synchrony' above.)
●As alternatives to traditional biventricular pacing at shortened AV intervals, fusion pacing and multi-site pacing algorithms have emerged as potentially useful tools to improve resynchronization therapy for qualifying patients. (See 'Emerging techniques' above.)
ACKNOWLEDGMENT — The UpToDate editorial staff thank Dr. Leslie A. Saxon and Dr. Teresa DeMarco for their past contributions as authors and Dr. Wilson Colucci for his past contributions as section editor to prior versions of this topic review.
