Cardiac resynchronisation therapy (CRT) is an established therapy for patients with drug-refractory, highly symptomatic systolic heart failure (HF) and delayed ventricular conduction.1 CRT devices are designed to synchronise the mechanical activity within the left ventricle, between the left and the right ventricle and between the atria and ventricles. Resynchronisation induces reverse remodelling and reduces mitral regurgitation.2,3 This has been shown to improve cardiac function, resulting in an enhanced functional capacity and improved quality of life.4 Finally, large randomised trials have shown significant reductions in all-cause mortality and hospital admissions with CRT versus usual care.5,6 Thus, current guidelines recommend CRT combined with a pacemaker (PM; CRT-P) or implantable cardioverter–defibrillator (ICD; CRT-D) as a class I indication for patients with sinus rhythm (SR), New York Heart Association (NYHA) functional class III or ambulatory IV, a QRS duration ≥120ms, and left ventricular ejection fraction (LVEF) ≤35% despite optimal pharmacological therapy.7–9
Given the high morbidity and mortality in chronic HF, efforts are being made to explore new indications with the aim of extending the beneficial effects of CRT to selected subgroups of patients who do not fulfill current selection criteria. Nevertheless, the high number of patients not deriving any demonstrable benefit despite fulfilling current guidelines should be kept in mind, requiring careful assessment of evidence.
Data from multicentre clinical trials have recently been published and resulted in an update of the European Society of Cardiology (ESC) guidelines, further extending CRT indications to selected patients. The aim of this article is to give an overview of these recent trials and the respective changes to the guidelines. Furthermore, emerging indications and ongoing trials are reported and discussed.
New Cardiac Resynchronisation Therapy Indications within the European Society of Cardiology Guidelines
New evidence from clinical trials has recently led to an updated version (see Table 1) of the current ESC HF guidelines from 20087 and the cardiac pacing guidelines from 2007.9,10 In contrast to previous guidelines, these recommendations not only are based on the inclusion criteria of patients in randomised clinical trials, but also consider the characteristics and outcome of patients actually treated. With respect to evidence-based efficacy of CRT in patients with NYHA class III/IV, the recommendations have been adapted to no longer require left ventricular (LV) dilatation and to limit NYHA class IV to ambulatory patients. This had already been established in the US guidelines.8
Cardiac Resynchronisation Therapy in Patients with Mildly Symptomatic or Asymptomatic Heart Failure
The recent publication of two large randomised studies, the Multicentre automatic defibrillator implantation trial with cardiac resynchronisation therapy (MADIT-CRT)11,12 and the Resynchronisation reverses remodelling in systolic left ventricular dysfunction (REVERSE) trial,13,14 has led to the most substantial changes in current CRT guidelines. The results of these trials suggest that CRT reverses LV remodelling, resulting in a reduction of disease progression and HF-related adverse clinical events. While REVERSE failed to show a difference in the primary end-point (HF clinical composite response), the results indicate that CRT reverses LV remodelling and reduces the risk for HF hospitalisation.14 Results from the 24-month follow-up of the European cohort of REVERSE have shown less clinical disease progression and an improvement in LV function.13 With extended follow-up there was a significant reduction in the clinical composite end-point in the CRT group (19% worsened in the CRT-ON group compared with 34% in the CRT-OFF group; p<0.01), as well as a significant reduction in time to first HF hospitalisation or death in the CRT-ON group (hazard ratio [HR] 0.38; p<0.003). Correspondingly, over a follow-up of 2.5 years, MADIT-CRT showed that patients with CRT-D had a one-third lower risk for the primary end-point (death or other clinical HF events) compared with patients with regular ICDs.11 Separating the primary end-point, there was no significant difference in death, but CRT resulted in a 41% reduction in other clinical HF events.
Extending clinical follow-up of CRT patients with mildly symptomatic HF over five to 10 years seems likely to show an influence on overall mortality reduction according to our pathophysiological understanding and the current scientific data.15 That benefit, however, has yet to be proved. Therefore, CRT in mildly symptomatic HF patients (NYHA II, LVEF ≤35%) is currently recommended not to reduce mortality but rather to reduce morbidity and to prevent disease progression, and should be limited to patients with a high likelihood of CRT response (QRS ≥150ms; class I recommendation, level of evidence A).10
Patients with NYHA functional class I are excluded from this recommendation as such patients accounted for only a small proportion of the populations studied in MADIT-CRT (15%) and REVERSE (18%); in addition, these patients did not derive a demonstrable clinical benefit. For the same reason, the cut-off for QRS duration (≥150ms) was chosen because patients with wider QRS duration responded more favorably in MADIT-CRT and REVERSE.
In addition, women, especially those with left bundle branch block (LBBB), showed a better clinical response to device therapy in MADIT-CRT. Nevertheless, this is not considered in the guidelines as analysis by sex was not pre-specified. Contrary to the European guidelines, the US Food and Drug Administration (FDA) based its newly proposed recommendations solely on the MADIT-CRT trial, using the inclusion criteria rather than the actual treated patient cohorts. This led to extended indications also in ischemic NYHA I patients and in patients with LBBB QRS duration ≥130ms.
To further strengthen the evidence and to address the issue of potential mortality reduction, data from a large randomised study with longer follow-up of CRT in patients with mildly symptomatic HF are soon to be expected from the Resynchronisation/defibrillation in advanced heart failure trial (RAFT; ClinicalTrials.gov identifier NCT00251251).16 This study included 1,800 patients in SR or with atrial fibrillation (AF), LVEF ≤30%, and QRS ≥120ms, and randomised patients to either ICD or CRT-D in a 1:1 fashion. The follow-up is expected to cover 4.5 years.
Cardiac Resynchronisation Therapy in Patients with Atrial Fibrillation
The prevalence of AF in HF patients is high and increases with progressive heart disease to up to 50% in NYHA III/IV patients.17 Therefore, AF patients represent a relevant and important subgroup of patients. The main problem in patients with AF and CRT is the decreased percentage of effectively biventricular paced captured beats due to irregular spontaneous AF rhythm, which corresponds to increased mortality.18 Based on these observations, a consensus exists that complete ventricular capture is mandatory to maximise clinical benefit and improve the prognosis of patients with permanent AF and CRT.19 Besides the establishment of pharmacological rate control, complete atrioventricular (AV) block by AV nodal ablation may be required to assure adequate pacing. However, randomised studies on this issue are lacking and results from observational studies have yielded controversial results, with some studies proposing AV nodal ablation as a prerequisite for mortality reduction in AF patients undergoing CRT20,21 and others reporting similar prognostic and symptomatic benefits without AV nodal ablation.22,23
Data on catheter ablation in HF patients with AF are promising and indicate significant improvements in cardiac function, symptoms, exercise capacity, and quality of life.24 Furthermore, this approach has been shown to be superior to AV nodal ablation and consecutive CRT implantation with respect to LVEF, six-minute walking distance and quality of life.25 However, data are limited by small sample size, short follow-up, and generally controversial data on the benefit of rhythm control in HF patients.26
The updated ESC guidelines consider CRT for morbidity reduction in patients with permanent AF, highly impaired LVEF (≤35%), NYHA III/IV, and QRS ≥130ms, on the condition that complete ventricular capture (at least 95%)18 is facilitated, mostly through AV nodal ablation.10
Contrary to the CRT guidelines and based on expert consensus only (class IIb, level of evidence C), the new ESC guidelines on AF27 consider biventricular pacing after AV nodal ablation, even in patients with only mildly symptomatic HF, irrespective of QRS duration and with only mildly depressed LV function (LVEF <45%).
Further understanding of the effect of primary catheter ablation in CRT patients suffering from AF may result from currently ongoing trials such as the Atrial fibrillation management in congestive heart failure with ablation (AMICA; ClinicalTrials.gov identifier NCT00652522), which is comparing LV reverse remodelling after catheter ablation versus conventional medical therapy in AF patients with reduced LVEF (<35%) requiring an ICD or CRT-D.
Cardiac Resynchronisation Therapy in Patients with a Conventional Pacemaker Indication
Chronic right ventricular (RV) apical pacing has been shown to increase morbidity and mortality in patients with ICDs and impaired LV function.28 These adverse effects are attributed to the induction of delayed activation of lateral LV segments with consecutive intra- and inter-ventricular dyssynchrony. Additional placement of an LV lead might potentially correct this dyssynchrony and result in a better clinical outcome. In this way, biventricular pacing has been shown to be superior to conventional RV pacing with regard to LVEF, quality of life, and exercise capacity in patients with a conventional PM indication and pre-existing LV dysfunction.29,30
These results have recently been extended to patients with normal systolic function and a conventional PM indication.31 In a prospective, double-blind, multicentre trial, 177 patients were randomly assigned to biventricular or RV apical pacing.29 At 12 months, the mean LVEF was significantly lower in the RV pacing group than in the biventricular pacing group (55 versus 62%; p<0.001), whereas the LV end systolic volume was significantly higher in the RV pacing group than in the biventricular pacing group (36 versus 28ml; p<0.001). Correspondingly, studies of upgrading patients with RV apical pacing to biventricular pacing showed significant improvements in LV reverse remodelling, even after long-term RV apical pacing.32,33 Larger studies with a longer follow-up evaluating morbidity and mortality are needed before biventricular pacing can be recommended in patients with preserved LV function and an indication for permanent pacing.
Therefore, the updated ESC guidelines underline the importance of avoiding RV pacing and recommend consideration of CRT for morbidity reduction in patients with severely and mildly symptomatic HF with a permanent PM indication, irrespective of QRS duration, in the context of a significantly impaired LVEF (<35%; class IIb, level of evidence C).
In the future, two landmark trials will further address the effect of biventricular versus RV pacing on survival, quality of life, and functional capacity in patients with a bradycardia indication for ventricular pacing. The biventricular pacing for atrioventricular block to prevent cardiac desynchronisation (BioPace) trial34 and the Biventricular versus right ventricular pacing in heart failure patients with atrioventricular block (BLOCK HF; ClinicalTrials.gov identifier: NCT00267098) have included 1,800 and 1,600 patients, respectively, with each group having a wide range of LVEF values. Clinical follow-up is ongoing.
Emerging Indications
Cardiac Resynchronisation Therapy in Patients with Narrow QRS
Evidence for a clinical benefit of CRT has been proved in HF patients with prolonged QRS as a marker of ventricular dyssynchrony. The presence of mechanical dyssynchrony, however, has also been shown in about 30–40% of HF patients with a normal QRS duration.35–37 Therefore, several studies have explored the potential benefit of CRT in this subset of patients.
While small single-centre trials indicated improvements in LVEF and exercise capacity in HF patients with narrow QRS and baseline mechanical asynchrony,38–40 these results could not be replicated in two multicentre trials.41,42 Evaluation of screening techniques in electrically normal, mechanically dyssynchronous HF patients (ESTEEM-CRT) included 68 patients with NYHA III, narrow QRS (<120ms), LVEF ≤35%, and LV mechanical asynchrony as assessed by tissue Doppler imaging. Patients did not improve as measured by objective exercise performance or echocardiographic LV reverse remodelling. Subjective symptomatic improvement was attributed to bias or placebo.42 RethinQ, the first randomised study, enrolled 172 NYHA III patients with narrow QRS (<130ms), a standard ICD indication, and LV mechanical asynchrony assessed by tissue Doppler imaging. At six months, CRT did not result in an improvement of peak oxygen consumption as the primary end-point (CRT-ON group 46% versus CRT-OFF group 41%). In a pre-specified subgroup with a QRS ≥120ms, peak oxygen consumption increased in the CRT group (p=0.02) but was unchanged in the subgroup with QRS <120ms (p=0.45). HF events did not differ between treatment groups. Consequently, there is currently no evidence that CRT is indicated in HF patients with QRS <120ms. However, it should be emphasised that both multicentre trials included a small number of patients, had a limited follow-up (up to six months), and did not focus on rehospitalisation and long-term survival. From a principal pathophysiological viewpoint, the method of evaluation of mechanical asynchrony and its impact on optimal LV lead positioning represent further methodological limitations of these studies.
More data on these issues are expected from the ongoing Echocardiography-guided Cardiac resynchronisation therapy (EchoCRT) trial (ClinicalTrials.gov identifier: NCT00683696).43 This trial will randomise around 2,200 patients to CRT or no CRT. The presence of baseline LV mechanical dyssynchrony is being assessed by speckle tracking echocardiography and validated through a single echo core laboratory. The primary end-point is all-cause mortality or hospitalisation for cardiovascular events during long-term follow-up.
Conceptually, the optimal position of the LV lead may play a greater role in achieving a clinical CRT benefit in narrow-QRS patients. Further impact on the role of CRT among those patient cohorts may therefore be expected from therapy developments aiming at the description and achievement of individualised optimal LV pacing sites.
Cardiac Resynchronisation Therapy in Patients with New York Heart Association Class IV
The major experience with CRT derives from systolic HF patients with NYHA class III disease. Patients with NYHA class IV account for only 10% of all patients in clinical CRT trials and these patients were highly selected, ambulatory outpatients who were taking oral medications and had no history of recent hospitalisation.8,44 The two landmark trials, the Cardiac resynchronization–heart failure trial (CARE-HF) and the Comparison of medical therapy, pacing, and defibrillation in heart failure (COMPANION) trial, included only 50 and 217 NYHA IV patients, respectively. Patients requiring continuous intravenous therapy or admissions for HF during the last month were excluded. In the CARE-HF study, no significant differences in outcome between class III and IV patients were reported.6,45 Correspondingly, a recent post-hoc analysis of the COMPANION trial that examined the outcome of all enrolled patients in NYHA functional class IV demonstrated a significant improvement in time to all-cause mortality and hospitalisations in NYHA class IV patients, with a trend toward improved mortality. Total mortality rates at two years were 55 and 45% with CRT-D and CRT-P, respectively, compared with 62% in the control group.44
Based on these data, current US guidelines8 and the updated ESC guidelines10 recommend CRT to improve morbidity (but not mortality) in ambulatory class IV patients. Patients with more advanced HF receiving intravenous inotropic support with refractory fluid retention or progressive renal dysfunction are not included in this recommendation as these patients represent the highest-risk population for complications during any procedure and for early mortality after discharge.8 A few observational single-centre46,47 or retrospective studies48–50 have addressed these patients, and have controversial results. On the one hand, urgent CRT implantation in patients with severely decompensated HF seems to be feasible and associated with clincial improvement;46,50,51 on the other hand, a significant reduction of mortality has not been proved.47,48 In our experience, the concept of CRT represents a promising approach as a temporary treatment in patients with refractory cardiogenic shock and a lack of further treatment options. In our own series of 18 patients with cardiogenic shock and electrical (QRS ≥120 ms, LBBB) as well as mechanical dyssynchrony, temporary percutaneous LV stimulation was feasible and resulted in improved haemodynamics with weaning of catecholamines.52
Cardiac Resynchronisation Therapy in Patients with Mildly Impaired Ejection Fraction
Nearly 50% of hospitalised patients with HF have a preserved LVEF but reveal a similar mortality and rehospitalisation risk to patients with highly impaired LVEF.53,54 Current clincial guidelines recommending CRT implantation in patients with LVEF ≤35% are based on the entry criteria of landmark trials implementing this LVEF cut-off.5,6 However, two small single-centre studies55,56 as well as post-hoc analysis of the Predictors of response to CRT (PROSPECT) study57 suggest that patients with wide QRS and LVEF >35% might also benefit from CRT.
The only prospective study on this patient cohort included 15 patients with LVEF 35–45% and QRS >120ms. At three-month follow-up, significant LV reverse remodelling could be observed with reduced LV end systolic (86 to 70ml) and end diastolic (136 to 121ml) volumes and improved LVEF (39 to 44%).56 Correspondingly, significant clinical improvements (NYHA functional class, six-minute walking distance, and quality of life) were reported in 27 patients whose LVEF as assessed with cardiac magnetic resonance imaging was >35% (mean 44%).55 These observations were replicated in a recent post-hoc analysis of the PROSPECT study. Of 361 subjects with available data, 86 were found to have an LVEF >35% (mean 43%) when assessed by the core laboratory. These patients experienced similar clinical and structural benefits to patients with more impaired LVEF.57
The prevalence of electrical and mechanical dyssynchrony and the potential usefulness of conducting clinical trials on CRT in patients with HF and preserved LVEF is currently being investigated in the prospective, observational Karolinska-Rennes (KaRen) study.58 In general, current data are too limited to recommend CRT in this patient population.
Conclusion
CRT is an efficacious and cost-effective therapy in patients with highly symptomatic systolic HF and delayed ventricular conduction. However, one-third of patients undergoing CRT for current guideline-recommended indications do not derive a demonstrable benefit from CRT.59,60
This, as well as the higher complication rate of CRT-D implant procedures compared with single-chamber devices,61,62 shows that efforts should be made to better define patients who are likely to respond or to increase the responder rate by using better implantation strategies.
On the other hand, exploring new indications might extend the benefits of CRT to a broader HF population. The concept of an early CRT intervention to delay or prevent HF progression has driven recent trials in the mild HF population, ultimately resulting in an extension of the current recommendations. Whether this concept can be applied to patients with narrow QRS or mildly impaired LVEF still needs to be proved. The use of CRT in patients with severely advanced HF and recent cardiac decompensation represents a therapy extension on the other end of the clinical HF spectrum. Scientifc data and evaluation are limited by the naturally high mortality rate in this patient subgroup; therefore, clinical guidelines do not state this indication. However, the first clinical descriptions indicate a potential role of temporary CRT as an additional treatment option in pre-terminal HF patients.