Author + information
- Maurizio Gasparini, MD* ( and )
- Paola Galimberti, MD
- ↵*Reprint requests and correspondence:
Dr. Maurizio Gasparini, Humanitas Clinical and Research Center, Electrophysiology and Pacing Unit, Via Manzoni, 56, Rozzano (MI) 20089, Italy
- cardiac resynchronization therapy
- implantable cardioverter-defibrillator
- heart failure
- left ventricular ejection fraction
In this issue of the Journal, Kutyifa et al. (1) present interesting and intriguing results derived from a post hoc analysis of the MADIT-CRT (Multicenter Automatic Defibrillator Implantation Trial With Cardiac Resynchronization Therapy) study. The correlation between baseline left ventricular ejection fraction (LVEF) and outcome after cardiac resynchronization therapy (CRT) was analyzed by dividing patients into 3 groups according to baseline LVEF: >30%, 26% to 30%, and ≤25%. The authors focused particularly on patients who, after an echo core laboratory re-evaluation, were found to have a baseline LVEF >30%, which was beyond the stated eligibility criteria of the MADIT-CRT trial.
Reverse remodeling and outcome were significantly better in patients with baseline LVEF >30% (1). In fact, the mean reduction in left ventricular end-diastolic volume with CRT, as well as reduction in hospitalization/deaths after CRT, was directly related to increasing LVEF. The paper is notable because these unequivocal results derived from such a large, prospective, multicenter trial (more than 1,800 patients with mild heart failure [HF]) could well lead guideline administrators to expand CRT indications. In addition to the general appreciation of this work, there are some major points that need to be discussed.
First, the main message of the paper is that in the mild HF population, patients with lower LVEF derive significantly less benefit from CRT in terms of reverse remodeling and HF hospitalizations with respect to patients with higher LVEF. Conversely, patients with somewhat better LVEF (in the range of 30% to 35%) may well be the best responders. Second, the widely accepted concept of the relevance of dyssynchrony to predict CRT response is contradicted by these data; in fact, in this large population affected with mild HF, the higher grade of dyssynchrony was inversely correlated with CRT response. Lastly, baseline “local” echocardiograms were re-analyzed by a centralized echo core laboratory that had not been used during the original study. The large percentage (38%) of patients with LVEF >30% (and therefore beyond the eligibility criteria) is a surprising but highly important finding. In addition, in some cases, there was remarkable discrepancy between local and centralized echo evaluation (enrolling patients re-evaluated with LVEF, 45%). This point truly questions the accuracy of baseline LVEF evaluation in the absence of centralized data collection and evaluation.
Patient selection in CRT and adequate follow-up: the “sweet spot” for CRT
Approximately one-third of CRT recipients fitting current CRT guidelines do not improve with this therapy. Conversely, once a therapy has been demonstrated to be effective in a selected population, obtaining striking results, clinical practice may drift away from evidence determined on the basis of clinical trials and try to extend these advantages to the largest portion of the HF population that could potentially benefit from such therapy.
From this perspective, there has been a meaningful evolution of CRT candidates over the years (Fig. 1). In the first era of CRT, patients with HF were enrolled in phase A (i.e., CRT as a “last-resort therapy”: at an advanced stage of the disease with significantly, and probably excessive, compromised left ventricular [LV] function). The hemodynamic “switch on/off” effect of CRT was evidently documented during hemodynamic acute studies (2,3), as well as in studies evaluating the acute effects of CRT on dyssynchrony. This kind of “the miraculous switch off-switch on effect” permitting the dramatic amelioration of volume-pressure curves and demonstration of unquestionably significant gains in systolic blood pressure, excited electrophysiologists and HF physicians. However, most of those compromised patients, unequivocally “responders” in the acute phase, nonetheless showed poor outcome at mid- to long-term follow-up. As long as the different CRT trials enrolled less-compromised patients, aiming to detect gain in reverse remodeling and mortality, a continuum of progressive response became evident, and patient enrollment shifted from phase A to B and C, with guidelines excluding nonambulatory patients with New York Heart Association (NYHA) class IV.
A preventive strategy has evolved, step by step, after the REVERSE (Resynchronization Reverses Remodeling in Systolic Left Ventricular Dysfunction) trial (4), the MADIT-CRT study (5), and the RAFT (Resynchronization/Defibrillation in Ambulatory Heart Failure Trial) study (6), thus expanding CRT guidelines to even NYHA II class patients (Phases D and E in Fig. 1).
The paper by Kutyifa et al. (1) adds new and interesting data supporting the notion that “moderately” compromised LV systolic function (the “sweet spot,” as described by the authors) may be associated with the greatest benefits from CRT. Moreover, these data confirmed our previous finding on HF remission observed after CRT (7). We sought to determine whether CRT induces an HF remission (NYHA I; LVEF ≥50%) and to define the incidence and predictors of such a process. At multivariable analysis, nonischemic etiology, LVEF 30% to 35%, and left ventricular end-diastolic volume <180 ml were strongly associated with HF recovery.
However, it is not really surprising that, at the beginning of device therapy in HF (when only simple implantable cardioverter-defibrillators [ICDs] were used), the best candidates and responders were “the sickest.” In these patients, use of the ICD led to a remarkable lowering of total mortality at short-term follow-up. With CRT in addition to ICD, the scenario has changed completely: from a very compromised HF patient saved only from sudden death (but subsequently dying of refractory HF), to a less-compromised HF patient saved from both sudden and HF death.
Conversely, less-compromised patients (with a lower baseline mortality rate and lower LV dysfunction) (5–7) obviously need a much longer follow-up to show CRT benefit, in terms of reverse remodeling and significant reduction of HF hospitalizations and total mortality. In these cases, the “miraculous switch-on effect” of CRT is probably difficult (if not impossible) to observe, and a much longer follow-up is likely needed to demonstrate significant effects on hard endpoints.
The concept of needing a much longer time to demonstrate mortality reduction has been indeed confirmed by Kutyifa et al. (1), who explained that this too short follow-up may be the reason why “CRT-D was not associated with a statistically significant effect on the risk of all-cause mortality.”
Of note, at baseline, the 3 groups presented significantly different end-diastolic volume (EDV) (being larger in LVEF <25%, as expected), and the EDV reduction observed after 12 months of CRT was significantly larger in patients with higher LVEF. Thus, at 1-year follow-up, the EDV differences became even more pronounced in the 3 groups (a mean of 123 indexed ml vs. 100 ml vs. 88 ml, respectively). Looking at such large discrepancies in LV volumes, it is possible that patients with higher baseline LVEF, presenting dramatically smaller EDV after 12 months of CRT, might show lower mortality rates at long-term follow-up.
Dyssynchrony and CRT
Until now, “the higher the dyssynchrony, the better the results from CRT” has been considered a paradigmatic concept. However, despite hundreds of studies stating that dyssynchrony identified by using echocardiography is a fundamental prerequisite for the success of CRT, no single predictive and reproducible dyssynchrony parameter has been confirmed as useful in randomized trials (3). Consequently, no single dyssynchrony criterion is included in recent guidelines (9).
Kutyifa et al. (1) now present extremely interesting data that simply contradict the “leitmotiv” concept that the higher the degree of LV dyssynchrony, the greater the response to CRT. From the data of the MADIT-CRT study, it seems reasonable to conclude that dyssynchrony seems to be correlated to large baseline LV volumes and low LVEF. More important, extreme degrees of dyssynchrony seem to be inversely correlated to CRT response, at least in patients with mild HF.
Did the MADIT-CRT LVEF criteria “truncate” patients who could benefit?
These new data from Kutyifa et al. (1) leads one to wonder about some important global implications of the MADIT-CRT results. The trial aimed at recruiting patients with LVEF ≤30%. Now, the echo core laboratory re-evaluation shows that many (38%) of the enrolled patients had LVEF >30%. This finding leads to 2 questions: 1) What would have been the results of MADIT-CRT if indeed the patients had all had LVEF <30%? 2) Had MADIT-CRT used the LVEF ≤35% criterion, already accepted for CRT indications by guidelines, how much larger a patient population would have been addressed by the trial and, consequently, would now be considered good candidates for CRT-D therapy?
In terms of the first question, because the study by Kutyifa et al. (1) clearly showed that the best results were in patients with mild HF with LVEF ≥30%, this finding can only mean that the overall results reported by MADIT-CRT would have been less positive (because many of the patients with the better results would not have been included.) To be perfectly clear, we have no grounds to suggest that the dramatic positive results of MADIT-CRT would have become negative or neutral; simply, there is a strong probability that they would have been somewhat less positive. In this regard, we must remember that both REVERSE (4) and RAFT (6), with similar patient populations (and LVEFs), also showed beneficial outcome for patients with mild HF treated with CRT.
However, perhaps the second question (What if the inclusion criteria had used the “natural” limit of LVEF ≤35%?) has even more important clinical consequences. Here, we may legitimately speculate that even more positive results (driven by patients with higher LVEF) would have been obtained if the enrollment criteria had included patients with LVEF of 30% to 35%. A clear message would then be reported by guidelines → NYHA II-ambulatory IV HF patients on OPT with QRS ≥120-130 ms and EF ≤35% should be eligible for CRT.
In the recent ESC CRT guidelines (9), prevalent up to the update decided literally in the past days (10), NYHA II patients with an LVEF of 30% to 35% had been excluded. Consequently, probably the best responders to CRT were in this “limbo” until guidelines committees took these new and important data into consideration. Taking into account the just-published 2012 AHA Focused Updated Guidelines for Device-Based Therapy of Cardiac Rhythm Abnormalities (10), patients with HF with LVEF ≤35% and NYHA class II are now considered as class I, level B in the presence of left bundle branch block and QRS duration ≥150 ms. This important decision is following the concept of a more “preventive therapy” for CRT in this population. Thus, the risk of leaving out patients with LVEF 30% to 35%, who are probably the best responders to CRT and who, per previous guidelines, had unfortunately been left “in limbo” (9), has been dramatically reduced.
Red traffic lights are still present, but the direction seems correct and the reverse path is from “the sickest patients benefit the most” to “the healthiest patients benefit the most.”
Both authors have reported that they have no relationships relevant to the contents of this paper to disclose.
↵* Editorials published in the Journal of the American College of Cardiology reflect the views of the authors and do not necessarily represent the views of JACC or the American College of Cardiology.
- American College of Cardiology Foundation
- Kutyifa V.,
- Kloppe A.,
- Zareba W.,
- et al.
- Auricchio A.,
- Stellbrink C.,
- Sack S.,
- et al.
- Nelson G.S.,
- Berger R.D.,
- Fetics B.J.,
- et al.
- Linde C.,
- Abraham W.T.,
- Gold M.R.,
- St. John Sutton M.,
- Ghio S.,
- Daubert C.
- Moss A.J.
- McMurray J.J.,
- Adamopoulos S.,
- Anker S.D.,
- et al.,
- ESC Committee for Practice Guidelines
- Tracy C.M.,
- Epstein A.E.,
- Darbar D.,
- et al.