Introduction

Cardiac contractility modulation (CCM) is a relatively new device-based (the Optimizer™ system by Impulse Dynamics, Orangeburg, NY, USA) therapy for heart failure with reduced ejection fraction (HFrEF). It can be recommended for patients with optimal medical therapy (OMT)-persistent symptoms of HF. CCM involves the administration of a biphasic non-excitatory electrical impulse to the interventricular septum during the ventricular absolute refractory period. CCM affects myocardial cell biology: in vitro studies showed that it enhances myocardial contraction via calcium circulation regulation without increasing myocardial oxygen consumption [1]. The results of clinical studies supported these data: the administration of CCM was associated with clinical improvement and lower hospitalization rates in patients with HFrEF (II-IV NYHA class, left ventricular ejection fraction (LVEF) < 40%) [2,3,4,5,6]. The meta-analysis of randomized controlled trials conducted by Giallauria et al. [7] revealed a moderate effect on peak oxygen consumption (VO2 peak) and exercise tolerance, thus improving the quality of life. Still, it demonstrated no effect on LVEF and mortality.

The impact of HF etiology on CCM therapy efficacy remains controversial. The etiology of heart failure (nonischemic vs. ischemic) was not associated with improvement in the etiology-dependent subgroup analysis of the FIX-HF-5 study [8]. On the contrary, the single-center study conducted by Fastner et al. reported a significant increase in LVEF and LV reverse modeling in NICM patients in the mid-term perspective [9]. A year later, Fastner et al [10] presented their findings in a larger cohort of patients. The authors did not reveal any significant difference in 12-month LVEF between ICM and NICM patients. These changes achieved the level of significance only in patients completing the 5-year follow-up.

Here, we present the 12-month results of our multicenter HF etiology-specified analysis of CCM therapy efficacy and safety in real-world practice.

Methods

This multicenter retrospective analysis included data from 166 patients with OMT-resistant HFrEF and NYHA class II-IV who underwent CCM therapy system implantation using Optimizer™ IV or Optimizer™ Smart (Impulse Dynamics, Orangeburg, NY, USA) from 2013 to 2019 in four medical centers in Russia within the framework of the national CCM therapy device implantation program. The federal program protocol was initially approved by the National Ethics Committee and, after that, by each centers’ Independent Ethics Committee. All patients signed a written informed consent before recruitment.

The inclusion criteria were age ≥ 18 years, OMT-resistant HFrEF, and NYHA class ≥ II. The exclusion criteria were patients on the heart transplant waiting list, a history of myocardial infarction, PCI, cardiac surgeries within three months, acute myocarditis, hypertrophic cardiomyopathy, reversible causes for HF, mechanical tricuspid valve, severe comorbidities (acute decompensation, injury, or failure in other organ systems).

The implantation procedures were performed under local anesthesia with lidocaine hydrochloride (10 mg/ml). An infraclavicular incision was made, and a pocket was created in the right subclavian region. The leads via the subclavian approach were advanced to the heart. Patients implanted with Optimizer™ IV got three screw-in leads - one atrial (Boston Scientific 7741 Ingevity IS-1 52 cm) and two ventricular leads (Boston Scientific 7742 Ingevity IS-1 59 cm, St Jude Tendril STS IS-1 59 cm); patients with Optimizer™ Smart - only two ventricular leads (Boston Scientific 7742 Ingevity IS-1 59 cm, St Jude Tendril STS IS-1 59 cm). The atrial lead was fixed into the right atrial appendage; the two ventricular leads were implanted in the right ventricular aspect of the interventricular septum. The tip-to-tip distance between ventricular leads was ≥ 2 cm.

Follow-up

The clinical follow-up duration was 12 months, with follow-up visits at 2, 6, and 12 months. Devices’ interrogation (Optimizer™ and CRT-D) and data analysis were performed at each follow-up visit. Data regarding clinical events and healthcare utilization were also collected. We thoroughly screened the available medical records of all 166 patients. The following parameters were collected at baseline and 12 months after implantation: demographics, medical history, physical examination data, laboratory examination data, NYHA class, six-minute walking distance (6MWD), transthoracic echocardiography (TTE), NTproBNP levels. Since February 2020, the follow-up protocol has been changed due to the COVID-19 pandemic. In some patients, the 12-month follow-up data was acquired remotely.

Statistical analysis

Continuous variables were presented as median (Me), interquartile range (IQR), mean (M), and standard deviation (SD) depending on the distribution. Categorical variables were presented as frequencies (percentages). Differences in the patients’ continuous data results were checked for significance with the Wilcoxon signed-rank test or the Student’s T-test, respectively. The chi-square test was used to compare categorical variables. The correlation between metrics was evaluated by calculating the Spearman correlation coefficient. For survival analyses, Kaplan-Maier curves were done. A two-tailed p-value ≤ 0.05 was regarded to be significant. Data were analyzed with the R programming language.

Results

Baseline characteristics

Data from 166 patients with OMT-resistant HFrEF who underwent Optimizer device implantation from 2013 to 2019 in four medical centers in Russia were analyzed in this study. The median age of the total study participants was 59 [54.0;66.0] years; the majority were males (83.7%). The median LVEF was 29% [24.1–33.0], with a median NYHA class of 2 [2;3]. The ischemic etiology of HF was revealed in 100 patients (61.5%) (ICM group); the non-ischemic group (NICM) evolved 66 patients (38.5%). Patients in the ICM group were significantly older (61[57–69] vs. 55 [42.8–61], p < 0.001), more frequently had hypertension (79% vs. 42.4%, p < 0.001), and chronic kidney disease (43% vs. 22.7%, p = 0.012). Patients in the NICM group had significantly more often atrial fibrillation (AF) (58% vs. 74%, p = 0.048), larger enddiastolic volume (EDV) (249 [208;309] vs. 220 [192;271], p = 0.019) and endsystolic volume (ESV) (183 [147;230] vs. 154 [128;199], p = 0.003). They also more frequently received angiotensin receptor/neprilysin inhibitors [28.8% vs. 9%, p = 0.002] and anticoagulants [77.3% vs. 60%, p = 0.032]. Only a third of the patients (29.5%) were previously implanted with ICD/CRT-D due to the limited quantity of state-funded available devices. Baseline characteristics of patients in detail, as well as the number of available records for each parameter, are presented in Table 1.

Table 1 Baseline clinical, echocardiography characteristics and NTproBNP
Full size table

Successful device implantation was carried out in all patients. During the 12-month follow-up, five patients developed major pectoral muscle pacing, requiring device lead replacement in three patients. In two patients, we set the device parameters so that the CCM pacing was continued with one lead, and the other was used only for sensing.

CCM device pacing rate

The CCM device pacing rate data was available for 134 patients (80.7%). The CCM pacing rate > 75% was achieved in 112 patients (83.6%); 86 of them (76.8%) have also reached the level of > 90% with a median rate of 95.2% [88.2;99.0]. The subanalysis of the CCM pacing rate in patients with AF showed a high Optimizer pacing rate regardless of AF type (Fig. 1).

Fig. 1
figure 1

CCM pacing rates according to the rhythm

Full size image

Survival

21 patients (13%) died during the 12-month follow-up. Cardiovascular death was confirmed in 12 patients. 9 patients died due to acute decompensation of chronic heart failure; 3 patients developed sudden cardiac death (only one had a previously implanted ICD). No mortality differences existed between patients with ICM and NICM (14.4 and 10.8% p = 0.51) (Fig. 2).

Fig. 2
figure 2

Survival of patients in ICM and NICM groups

Full size image

Changes in transthoracic echocardiography, six-minute walking distance, and NTproBNP parameters

The thorough analysis of changes in TTE parameters, 6MWD test, and NTproBNP levels is presented in Table 2. We missed 12-month follow-up data in some patients due to remote follow-up during the COVID-19 pandemic.

Table 2 Echocardiography parameters, NYHA, 6-minute walking distance, and NTproBNP in 12-month follow-up
Full size table

The overall 12-month LVEF was improved in most of the patients (Fig. 3). Moreover, 12-month LVEF in 16 patients (24%) in the NICM group increased up by ≥ 40%. The analysis of 12-month TTE data revealed a significant increase in LVEF (+ 2.0 [2.0;6.0] vs. +7.7[2.0;12.0], p < 0.001), and a significant decrease in end-systolic volume (-3.0 [-18.0;10.0] vs. -24.0[-74.5;0.0], p < 0.001) and end-diastolic volume (-3.0[-22.75;21.8] vs. -24[-69.5;6.0], p = 0.007) in the NICM group.

Fig. 3
figure 3

LVEF changes at 12-month follow-up

Full size image

We detected an improvement in the 6MWT, but these changes did not reach the level of significance (+ 75 [22;108] vs. +80 [10;160], p = 0.851).

The NTproBNP level decreased in both groups, which was more pronounced in the NICM groups. However, this decrease did not reach statistical significance (-262.55 vs. -591.0, p = 0.091).

We revealed a modest NYHA class improvement in both groups, which was more significant in the NICM group (p = 0.041). Almost half of the patients completing 12 months of follow-up showed ≥ 1 NYHA class improvement (Fig. 4). Patients with improving/stable NYHA class significantly less frequently suffered from COPD (17.5% vs. 37%, p = 0.047), AF (56% vs. 89%, p = 0.010), and lower extremity edema (22.4% vs. 50%, p = 0.008). They also had lower baseline systolic blood pressure (110.0 [105.0;120.0] vs. 120.0[110.0;129.0], p = 0.032), baseline pulmonary artery systolic pressure (34.5[26.2;44.8] vs. 40.0[31.5;55.0], p = 0.030), ESD (57.0[52.0;63.0] vs. 61.0[56.0;68.0], p = 0.036), baseline NTproBNP (1279.0[665.0;2280.0] vs. 2472.0[1208.0; 6492.0, p = 0.001). The detailed comparative analysis of patients with improving/stable and worsening NYHA classes/fatal outcomes is presented in Table 3.

Fig. 4
figure 4

NYHA class at baseline and 12-month follow-up (a) baseline NYHA in ICM group, (b) 12-month NYHA in ICM group; (c) baseline NYHA in NICM group; (d) 12-month NYHA in NICM group

Full size image
Table 3 Baseline clinical, echocardiography characteristics and NTproBNP in groups with and without clinical impairment
Full size table

Discussion

The impact of HF etiology on CCM therapy is poorly studied and remains controversial. Such an important HF marker as LVEF was not assessed in RCTs as the primary endpoint of CCM therapy; most of the RCTs assessed exercise tolerance [2,3,4,5]. So, the interpretation of LVEF changes following Optimizer system implantation is understudied. Our analysis revealed that CCM therapy in NICM patients led to significant LVEF improvement with no statistically significant effect on 12-month survival. We also detected a significant improvement in the NICM group in other TTE data, such as EDV and ESV, reflecting the LV reverse remodeling. Our findings correlated well with the study results conducted by Fastner et al. [9], showing not only a remarkable LVEF improvement in NICM patients but also an LV reverse remodeling process reflected in LVEDD reduction.

On the one hand, these results could have been predictable. Although the SCD rates were similar for ICM and NICM patients [11], an older age, significantly more frequent comorbidities, worse response to HF drug therapy, a higher risk of cardiovascular death (especially due to myocardial infarction), and worse survival in patients with ischemic cardiomyopathy [12,13,14,15] suggest a poorer outcome of CCM therapy in those patients. However, Abraham et al. [8] did not reveal any association between HF etiology and patients’ improvement. LVEF improvement was not dependent on HF etiology in the study conducted by Müller et al., too [16]. On the other hand, the last study evolved patients with > 35% LVEF, which can be a possible explanation for such discrepancy in study results.

Our study also confirmed the tendency for the NTproBNP level to decrease, which was more pronounced in NICM patients. It is interesting to notice that NTproBNP level reduction tendency reached the level of statistical significance in the study conducted by Kuschyk et al. [17] evaluating the long-term efficacy (34.2 ± 28 months) of CCM therapy. This may reflect the long-term beneficial effects of CCM therapy on cardiac myocytes. However, Kuschyk et al. did not specify the HF etiology.

In our study, the 12-month 6MWD improvement did not reach the significance level. We also did not reveal any differences in 6MWD changes depending on HF etiology. Most of the studies, including RCTs, showed a significant increase in exercise tolerance, particularly in 6MWD [7, 18,19,20,21,22]. The involvement of more patients with advanced heart failure in our study can be the possible explanation for not reaching the significance level for 6MWD improvement during 12 months of follow-up. However, the 12 months of follow-up confirmed the more significant improvement of the NYHA in NICM patients.

Our subanalysis of patients’ characteristics with and without clinical impairment revealed, that patients without clinical improvement more frequently suffered from COPD, AF, and edema, and had higher baseline systolic blood pressure, baseline pulmonary artery systolic pressure, ESD, and baseline NTproBNP. Our results indicate the crucial role of patients’ thorough selection for effective CMM therapy, as patients with progressive congestive heart failure, edema, and frequent hospitalizations may not benefit from CCM therapy. Thus, we could recommend, that these patients should immediately be placed on the waiting list for heart transplantation skipping the CCM therapy device implantation. However, further larger, multicenter studies with a control group and optimal medical therapy are necessary to clarify these findings.

Limitations

There are some possible limitations of this study: First, this is a retrospective analysis of the prospective collected data. These data were collected within the framework of a state-funded CCM therapy implementation program in Russia. The state-funded character of the program predefined the number of participants and the duration of the clinical follow-up visits. Second, optimal medical therapy has changed over the years; most patients were implanted with Optimizer devices before ARNI and SGLT2 inhibitors became available in Russia. Third, the Optimizer device provides information about the device pacing rate, which is not always the same as the CCM therapy rate. Fourth, due to the COVID-19 pandemic, some follow-up visits were done remotely, and we don’t have complete follow-up data for some patients.

Conclusion

Patients with non-ischemic HFrEF showed a significantly higher improvement in LVEF and LV reverse remodeling following CCM therapy device implantation. There was no significant association between HF etiology and survival in drug-resistant HFrEF patients following CCM therapy.