This editorial refers to ‘Response to preventive cardiac resynchronization therapy in patients with ischaemic and nonischaemic cardiomyopathy in MADIT-CRT’†, by A. Barsheshet et al., on page 1622,‘Time-dependent benefit of preventive cardiac resynchronization therapy after myocardial infarction’‡, by A. Barsheshet et al., on page 1614,and ‘Cost-effectiveness of cardiac resyncronization therapy in patients with asymptomatic to mild heart failure: insights from the European cohort of the REVERSE (Resynchronization Reverses remodeling in Systolic Left Ventricular Dysfunction)’§, by C. Linde et al., on page 1631
After the publication of the COMPANION and the CARE-HF trials,1,2 cardiac resynchronization therapy (CRT) has become accepted as an efficient therapeutic modality, over and above optimal drug treatment.3 CRT not only improves symptoms and quality of life but also reduces morbidity and mortality in patients with New York Hear Association (NYHA) class III and IV, depressed left ventricular ejection fraction (LVEF), and a wide QRS.3 More recently, the guidelines on device therapy in heart failure from the European Society of Cardiology recommended CRT also in patients with NYHA class II despite optimal drug treatment, with an LVEF <35% and very wide QRS ≥150 ms, to reduce morbidity and/or to prevent disease progression.4 This important commendation to prevent disease progression by CRT was based on the results of the REVERSE and MADIT-CRT trials.5,6 The recent publication of the RAFT trial confirmed the benefits of CRT in preventing clinical disease progression in patients with mild heart failure.7 Importantly, in contrast to REVERSE and MADIT-CRT, the RAFT study showed that CRT reduced mortality in this mild heart failure cohort.5–7 This significant effect was demonstrable in the RAFT trial, probably on account of the longer clinical follow-up of 5 years as compared with 2 years in the REVERSE and MADIT-CRT trials.5–7
As the patient population eligible for CRT continues to expand, together with the fact that a substantial proportion of patients remain non-responsive to CRT, it is imperative that we attempt to better define responders and non-responders to CRT, as well as address the cost-effectiveness of this therapeutic intervention. The results from the REVERSE, MADIT-CRT, and RAFT trials showed some uniformity in that a QRS duration of >150 ms was predictive of a favourable response to CRT.5,6 The MADIT-CRT trial identified women and the presence of a left bundle branch block (LBBB) as two important determinants of response to CRT. There has been much debate regarding the influence of the aetiology of the myopathic process on clinical outcome. A variance in the pathophysiology accompanying the ischaemic (ICM) and non-ischaemic (NICM) substrates serve to fuel this controversy even further.
Barsheshet et al. have compared (i) the efficacy of CRT in ischaemic and non-ischaemic patients;8 and (ii) the time-dependent benefit of CRT after myocardial infarction (MI) in the MADIT-CRT trial.9 In their study, both ischaemic and non-ischaemic patients were observed to benefit from CRT. There was a significant reduction in the combined endpoint of heart failure and death (mainly driven by heart failure events) in both ICM (34%) and NICM (44%), without any significant treatment–aetiology interaction.8 In comparison, within the REVERSE trial, the percentage of worsened patients (as measured by the clinical composite score) was significantly reduced at the 1-year follow-up in NICM subjects as compared with the ICM group.10 In contrast, the time to first heart failure hospitalizations was significantly reduced in the ICM subset but not in the NICM subset. The absence of a significant reduction in the NICM group is probably a consequence of the low event rate in the NICM group.10 Interestingly, the lower event rate in the NICM patients observed in the REVERSE trial was comparable with that demonstrated by Barsheshet et al in the ICD (implantable cardioverter defibrillator) group (control group) of the MADIT-CRT study.8,10 There was a significantly higher rate of death or heart failure at 3 years follow-up in ICM patients as compared with NICM patients (34% vs. 24% respectively, P = 0.007), implying differences in the underlying pathophysiology of these two substrates.
In their recent work, Barsheshet et al. have shown that within patients with ischaemic and non-ischaemic cardiomyopathy, there are specific subgroups of patients who may respond better to CRT. In ICM, patients with a wide QRS (>150 ms) or a systolic blood pressure <115 mmHg did significantly better, with a reduction in the 3-year heart failure/death event rate.8 However, in NICM patients, CRT was associated with a pronounced benefit among females and in patients with diabetes mellitus concomitant with an attenuated efficacy among men and non-diabetic patients.8 The heightened response in the particular subset of the ICM patients appears to ferret out a sicker subgroup with a higher risk for death or heart failure and thereby with an elevated likelihood of response to CRT. The clinical benefit of CRT was also superior in patients with LBBB in both ICM and NICM subgroups.8 A part of this beneficial response in patients with LBBB could be attributed to the fact that the left ventricular activation sequence and the pattern of mechanical dyssynchrony may be more predictable in the setting of an LBBB, as opposed to the non-LBBB morphology. Standard left ventricular lead placement strategies, targeting the left ventricular free wall, may consequently prove more useful in this substrate as opposed to the non-LBBB morphology. Nevertheless, this QRS morphology-driven response has raised the awareness of the importance not only of a wide QRS, but also of the LBBB morphology in the mild heart failure population eligible for CRT. These important observations might have an important clinical impact for enhancing patient selection for CRT, save for the fact that this analysis should be interpreted cautiously, especially due to the inherent limitations associated with subgroup analysis and potential subgroup interactions.
Another significant finding is that the magnitude of CRT-induced left ventricular reverse remodelling was significantly higher in the NICM group as compared with the ICM group.8 In this subanalysis of the MADIT-CRT trial, the left ventricular end-systolic volume decreased by a mean value of 37 ± 16% in the NICM group in contrast to a reduction of 29 ± 14% in the ICM group at 1 year follow-up. Importantly, similar reductions in left ventricular end-diastolic volume and left atrial volume, accompanied by an improvement in the LVEF were observed in NICM patients. A recent report from the REVERSE trial reinforces these results, with NICM being an independent predictor for left ventricular reverse remodelling.10 This study showed that there was a >3-fold reduction in left ventricular end-diastolic and end-systolic volume indexes and a 3-fold increase in LVEF in patients with NICM. It would be of interest to note if there was an aetiology-specific relationship between echocardiographic remodelling and clinical outcome in this group of patients with mild heart failure.
In a second study, Barsheshet et al. studied the time-dependent benefits of preventive CRT after MI in a subset of the MADIT-CRT trial.9 This substudy included 704 ICM patients with documented MI and a subset of ICM patients without prior MI. The median value of the time between the MI and the randomization was 8 years. The primary finding was that in patients with ICD therapy (control group), there was a time-dependent relationship of the occurrence of heart failure or death. Left ventricular remodelling is a continuous complex process leading to progressive left ventricular dilatation with larger left ventricular volumes in patients with remote MI, and consequently more clinical events, as shown in this study. The patients without documented MI exhibited a prognosis identical to patients with remote MI beyond the median time. The benefit of CRT to prevent death and heart failure in the population with prior MI became significant only in patients implanted ≥8 years after the MI with a 58% reduction in the risk of heart failure or death. In patients with ‘recent MI’, <8 years, CRT did not provide significant reduction in death or heart failure events. Patients with ICM, but without a prior MI also benefit from CRT, with a reduction of death or heart failure by 57%. Interestingly, greater benefit from CRT was seen in patients without prior MI, where the magnitude of left ventricular reverse remodelling was significantly more pronounced than in the population with documented MI. Unfortunately, there is no clear comparison for the effect of CRT on the left ventricular reverse remodelling between patients enrolled before and beyond the median time interval. Since the disease process in heart failure is a continuum, it appears unlikely that the results of this study will influence our selection process, dictated by the time interval from the previous MI. It is noteworthy, that CRT was much more efficient in preventing death or heart failure events in patients without a documented MI or in those patients with a prior MI, but with significantly larger volumes. Again, it is imperative that we note that the inferences being derived here are from non-specified subgroups analysis, the subgroups themselves being limited by their small sizes, with <200 patients in each quartile. Importantly, although these findings are interesting, they need to be validated prospectively.
Finally, Linde et al. have studied the cost-effectiveness of CRT in patients with mild heart failure, within the REVERSE trial.11 Although the previous guidelines for resynchronization therapy had recommended CRT for patients with refractory severe heart failure, a low LVEF, and wide QRS, the recently published RAFT trial will tilt the balance towards prophylactic CRTs in the mild heart failure patient population. As the indications for resynchronization therapy expand, it becomes more imperative that the cost-effectiveness of CRT is addressed. The authors showed that the CRT group gained 0.94 life years or 0.8 quality-adjusted life years (QALYs) compared with the CRT off group. An additional upfront cost of €11 455, yielded an incremental cost-effectiveness of €14.278 per QALY gained, with CRT being cost-effective after 4.5 years. This is consistent with the cost-effectiveness of CRT in the population with moderate to severe heart failure as demonstrated within the COMPANION and CARE-HF trials.12,13 The cost-effectiveness of CRT in the mildly symptomatic will probably be reinforced with the demonstration of the efficacy of CRT on mortality by the RAFT trial.7
CRT has substantially altered the natural course of heart failure and exerts its physiological impact through favourable ventricular remodelling, which in turn improves clinical outcomes. In the setting of finite health care resources, it is imperative to demonstrate the cost-benefit advantages of new technological approaches. As the population eligible for resynchronization therapy expands, the need to be more cognisant of its cost-effectiveness and to refine the selection criteria will become more important. It is important to underline that the recent European guidelines included the pre-specified cut-off value of 150 ms as the ECG criterion for the NYHA II patients and not as usual the inclusion criteria of the REVERSE and MADIT-CRT trials. Recent substudies, such as those discussed here, will help us better understand the differences in the pathophysiology and response of different cardiac substrates. As these data become replicated via robust, prospective studies, it will help us better select patients likely to respond, making preventative CRT an even more cost-effective endeavour.
Conflict of interest: none declared.