Cardiac resynchronization therapy (CRT) achieved by implantation of a biventricular pacemaker is recommended for symptomatic patients with heart failure (HF) in sinus rhythm with a QRS duration≥130ms and left bundle branch block (LBBB) QRS morphology and with left ventricle ejection fraction (LVEF)≤35% despite optimal medical therapy to improve symptoms and reduce morbidity and mortality. In patients with atrial fibrillation (AF), it is essential to provide a strategy to ensure biventricular capture.1 The response to CRT when there is electrical asynchrony is known. However, identifying patients who will be responders to this therapy is challenging. The initial studies performed to demonstrate mechanical asynchrony by echocardiography did not reveal its usefulness in predicting response to CRT, usually defined as left ventricle reverse remodeling. Therefore, echocardiography assessment of left ventricular asynchrony has no additional benefit over electrocardiogram alone.2 Although rather infrequent, CRT implantation is associated with short-term and long-term complications. Indeed, a European survey showed that the peri-procedural complication rate was 5.1%, and major adverse events occurred in 5.4% of upgraded and 4.4% in de novo CRT recipients.3 Identifying patients with a high likelihood of being responders is crucial, so CRT is to be offered to this group, maximizing the risk-benefit profile of this therapy.

The early inward motion and thickening/thinning of the ventricular septum associated with LBBB, called septal flash (SF),4 and the apical rocking (ApR) are mechanical consequences of LBBB. Both can be evaluated visually using M-mode echocardiography and results from an imbalance of myocardial function due to intraventricular conduction delays, regional damage, or both.5 Some data suggest that their presence, unlike classical echocardiographic parameters, can indicate effectiveness after CRT, and it seems to be associated with a better prognosis.6 Therefore, this study aimed to analyze the relationship between the ApR and SF presence before the biventricular pacemaker implantation and the outcomes after CRT in a non-selected chronic heart failure population.

Baseline characteristics of the 96 patients included in the analysis are summarized in Table 1. De novo CRT implantation was performed in 91 patients, 5 patients (5.2%) underwent upgrade to CRT. Briefly, ApR or SF was present in 55% of the patients before CRT, but both were only present in 34.4% of patients. In the ApR/SF group, 88.7% had SF, 71.7% had ApR, and 62.3% had both. The baseline characteristics and outcomes of SF and ApR are described in Table 1 of the supplementary data.

Patients with ApR/SF at baseline were younger, had more frequently a QRS≥150m (although the mean QRS was not statistically different), and had lower hemoglobin levels. There were no other significant differences at baseline. There were no differences in the prevalence of SF, ApR, or both between men and women (the supplementary data). Most of the patients were Caucasian.

The median follow-up was 42 (25–65) months. During follow-up, 31 (32.5%) patients died, and 30 (31.3%) patients were admitted to the hospital due to worsening of HF. The composite endpoint outcome of death and HF hospitalization occurred in 45% of the patients. All-cause mortality was not significantly lower in the ApR/SF group; however, the composite outcome of HF hospitalization or all-cause death was significantly lower in patients with ApR/SF than in patients without ApR/SF (32% vs 60%; P<.001). The same result was observed in terms of HF hospitalization (17% vs 49%; P=.001) (Table 2; Fig. 2). Patients with ApR/SF significantly improved the number of meters walked at the 6MWT (median [percentile 25-percentile 75) 61m (0–101) vs 6 (−11–37)m; P=.054; and had an increase of at least 50m more frequently 60% vs 19%; P=005), and there was a significantly higher percentage of CRT echocardiographic super-responders (defined as LVEF>50%) (Table 2).

Fig. 2.

Kaplan–Meier curves for long-term outcome divided by the present of septal flash or apical rocking.

(0.11MB).

As expected, patients with the composite endpoint had a significantly worse renal function, lower hemoglobin level, and higher NT-proBNP level at baseline. Furthermore, fewer patients improved by >50 meters in the 6MWT and reached LVEF>50% at 1-year follow-up compared with patients without the composite endpoint (Table 3). The presence of SF and the presence of ApR/SF between the groups with and without the composite endpoint were significantly different (31% vs 60%; P=.02; and 40% vs 68%; P=.005; respectively) (Table 3).

Results of both univariate and multivariate analyses of the association between ApR or SF and the composite endpoint are summarized in Table 4. In univariate analyses, the composite endpoint was associated with the absence of ApR/SF, higher renal impairment, QRS duration>150ms, NT-proBNP level, and lower hemoglobin. In multivariate analysis, the presence of ApR or SF (HR, 0.36; 95%CI, 0.15–0.86]; P=.019), female sex, hemoglobin and logNT-proBNp were associated with the composite endpoint. When analyzed separately (SF/ApR and non-SF/ApR), the only baseline characteristics associated with the endpoint were hemoglobin and NT-proBNP levels in both groups and estimated glomerular filtration rate in the SF/ApR group.

ApR or SF was present in 55% of the patients in our series. Ghani et al. described the presence of ApR in 45% of the patients undergoing CRT,11 while Gąsior et al. documented 28% of patients with SF.12 In the most extensive study, published by Stankovic et al., ApR and SF were observed in 64% and 63% of patients, respectively, and these patients were younger, more frequently female, had more often LBBB morphology, longer QRS duration, less ischemic disease, and less atrial fibrillation.4 With 55% of patients with asynchrony parameters, our study does not differ so much from these results. Several studies have also shown that SF or ApR and its relationship with remodeling are associated with the etiology of HF, with lower prevalence and LVEF improvement in ischemic patients.11,12 We did not see this difference in our series; the only significant differences we found between the groups with and without ApR/SF were age and hemoglobin level. However, our relatively small sample size might have modulated these results.

ApR/SF before CRT in this study was independently associated with a lower incidence of all-cause death or HF hospitalization during a follow-up of more than 3 years. This better prognosis is consistent with the studies by Stankovic et al. and Ghani et al.6,11 that showed a more favorable long-term survival after CRT when ApR or SF were present at baseline. Both studies had a follow-up similar to ours, with a 46 to 5.2 years median period. However, HF hospitalization was only included in the composite outcome in the study by Ghani et al.11 Interestingly, the population included in that study was comparable to ours and showed a 21% mortality rate and a 21% HF hospitalization.

The CRT benefit identified in patients with ApR/SF was observed in terms of survival and hospitalizations and in terms of exercise capacity measured by the 6MWT. In patients with a 6MWT available at baseline and at 1-year follow-up, 60% of patients with ApR/SF improved more than 50 meters, whereas it was only true in 19% of patients without ApR or SF (P=.005). CRT response is usually also evaluated by the presence of reverse LV remodeling, defined by changes in LV end-systolic volume, LVEF at 3–6 months after CRT implantation or both. The presence of LVEF>50% at follow-up has been proposed as a criterion for super-response,9 and the long-term outcome of super-responders of CRT is excellent.13 We observed super-response to CRT in the present study in 26% of the patients. The presence of ApR/SF before CRT implantation was associated with a super-response defined as LVEF recovery (LVEF>50%). This result is consistent with other studies that showed that ApR was independently associated with super-response to CRT.14

Finally, the early CRT studies suggested that CRT might not be effective for AF patients.15 However, the Certify trial showed the importance of achieving a high biventricular pacing percentage, preferably by atrioventricular junction ablation. The clinical outcome after CRT for patients with AF treated with atrioventricular junction ablation is similar to that observed in sinus rhythm.16 In our study, 32% of patients had AF as a baseline rhythm without differences between both groups. The presence of AF did not limit the echocardiographic assessment of SF and ApR. Interestingly, the better outcomes observed in the group with baseline ApR/SF after CRT were also present in patients with AF. These findings are similar to those in the study by Gabrielli et al., where they describe that the presence of SF at baseline is associated with a higher likelihood of CRT response in HF patients with long-standing AF.17

More than half of the patients with CRT indication who underwent biventricular pacemaker implantation had ApR or SF at baseline. Still, demographic and clinical characteristics could not allow to identify them. The presence of ApR or SF was associated with lower all-cause mortality or HF hospitalization and improved functional status and LVEF at follow-up. These results highlight the importance of carefully assessing the presence of ApR and SF in all patients irrespective of clinical characteristics to identify patients who will present a better response to CRT.

This study received no funding.

Conceptualization, methodology, validation, review draft and editing, project administration, S. Ruiz Bustillo and N. Farré; software, A. García Durán and F. Escalante; formal analysis, N. Farré; investigation, S. Ruiz Bustillo, A. García Durán, A. Mas-Stachurska, E. Vallès, L. Belarte Tornero and N. Farré; resources, N. Farré and A. Mas-Stachurska; data curation, A. García Durán, F. Escalante, A. Calvo Fernández, J. Jimenez López, E. Solé González, R. Morales Murillo, S. Ruiz Bustillo and N. Farré; writing original draft preparation, S. Ruiz Bustillo; visualization, A. García Durán, A. Calvo Fernández, A. Mas-Stachurska, F. Escalante, E. Vallès, J. Jimenez López, L. Belarte Tornero, Sandra Valdivielso More, E. Solé González, R. Morales Murillo, J. Martí Almor; supervision, N. Farré. All authors have read and agreed to the published version of the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.