Device therapy is an important adjunct to medical heart failure therapy. In a group of 31 consecutive HF patients who underwent CCM implant at our hospital, we observed improvement in a meaningful proportion of patients on several fronts: NYHA class, hospitalizations and hospitalized days, systolic and diastolic function, and weight reduction. Based on our single-center experience, CCM may provide an opportunity to augment the care of certain heart failure patients.

NYHA class improvement of at least one class occurred in over 65% of the patients in this study. This is relatively consistent with prior studies, although our proportion improved is slightly less than what was observed in clinical trials and large registries [3,4,5]. That the proportion with NYHA class improvement was slightly lower than previously published may represent simple variation across populations, may be related to differences in comorbidities, or a limitation of our small sample size. However, as it is well documented that impoverished communities have worse HF outcomes, it is worth noting that this study represents data from one of the nation’s most destitute communities, at what is the hospital providing the most charity care in the USA [22, 26, 30]. Nonetheless, improved NYHA functional class is associated with lower risk of hospitalization, lower risk of all-cause mortality, as well as improvements in overall quality of life, and has been used as a benchmark endpoint for clinical improvement in HF [15].

Heart failure is the most common reason for hospitalization among Medicare patients and the second most common overall, resulting in over 1 million hospitalizations per year in the USA [24]. Mortality rates for Medicare patients is between 15 and 20% within 90 days of hospitalization, and all patients with depressed EF, regardless of age, have a 5-year mortality of 69%, and median survival of 6 months [7, 17]. Therefore, reduction in hospitalizations is its own indication of reduction in morbidity and may be a surrogate measure of other untoward outcomes. CCM therapy has previously been shown to reduce hospitalizations, and our findings are consistent with previously demonstrated reductions in HF hospitalizations [1, 3,4,5, 18]. Although the point-estimates were not significantly different, we observed lower annual hospitalizations and annual days hospitalized in patients who had CCM placed more proximate to HF diagnosis, which aligns with CRT literature describing improved outcomes if device placement occurs prior to initial HF hospitalization, with incrementally worse outcomes as time passes from index hospitalization to ultimate device implant [18, 20]. This suggests that it is not just the HF intervention itself (device or otherwise) that is critical, but also its timing that may have important implications regarding outcomes. Preventing hospitalizations is a central tenet in the therapy of HF patients, as well as being a focus of improvement to reduce healthcare costs [14, 30].

The EF improvement demonstrated in this observational study was predictable based on prior literature, including in both ICM and NICM patients [3, 9]. A potential implication of the EF improvement with CCM is whether CCM-therapy could render primary prevention ICD indication moot in some patients, who with CCM, experience improvement of EF to > 35%. Nonetheless, in CRT and non-CRT heart failure populations, recovery of EF does not normalize the risk of sudden death [32, 34]. There is no data that EF improvement in CCM patients reduces the risk of sudden death, and in fact, there was early concern that the increased intracellular myocardial calcium observed with CCM therapy could increase arrhythmogenicity. Ultimately, this hypothesis was not borne out in safety and then subsequent clinical studies, perhaps because CCM increases systolic, rather than diastolic calcium stores [1, 2, 6, 16, 28, 29]. It is important to emphasize, though, that just as patients with recovery of EF after primary prevention ICD remain at higher risk for ventricular arrhythmias than the general population, there is no data to suggest that those with EF recovery related to CCM would fair differently [23]. Certainly, the current study is too small to evaluate ICD therapies or ventricular arrhythmias as an outcome—although none of the patients experienced ICD therapy for ventricular arrhythmias, yet one patient without an ICD did die in his sleep 1 month after CCM implant. Our observation that 40% of ICD candidates (EF ≤ 35% after 3 months of GDMT) experienced an improvement in systolic function to EF > 35% is provocative, although the clinical application of these findings requires rigorous study before suggesting that ICDs be withheld in this subgroup.

CCM therapy has a lusitropic effect at the cellular level, and in an initial safety trial in patients with heart failure with preserved EF, was shown to be safe and to improve diastolic filling index (E/E’) [21]. Certainly, improvements in diastolic function are important, as improvements are associated with decreases in hospitalizations and mortality among those with HF [8]. In our study, more than half of those with DD had improvement in diastolic filling index, suggesting CCM therapy could have a positive effect on diastolic function in the clinical setting. Improved diastolic function did not correlate with improvements in EF nor with type of cardiomyopathy. Notably, in patients with improved NYHA class, 78% had improvement in EF, 44% had improvement in DD, and 28% had improvement in both. Among a small group of patients, it is inappropriate to draw firm conclusions, but it may be that CCM affects myocytes in multiple ways, which may manifest differently, depending on the heart. For example, as provocative conjecture, perhaps it is that diastolic improvements are at play in the increased benefit CCM has been shown to provide in those with LVEF 35–45% [2]. Bigger studies with similar granular data on the relationship between systolic and diastolic function may be clarifying.

Although weight loss can be a poor prognostic indicator for worsening HF (cardiac cachexia), which was evident in the outlying patient, in this study, most of the studied patients showed weight loss to go along with other clinical improvements [31, 33]. A significant portion of patients had substantial weight loss after CCM, and unsurprisingly, the five patients who experienced improvement in NYHA class, hospitalizations, systolic, and diastolic function, all had meaningful weight loss. Weight loss, in the setting of improving functional capacity and overall cardiac function, may be a product of increased activity and/or improvements in volume status after CCM placement—and may portend better outcomes [25].

Despite large proportions of patients demonstrating improvement in several endpoints, it is important to highlight that several patients did not improve. In the CRT population, so-called “non-responders” comprise approximately 30% of those treated, with some proportion of failures to respond related to insufficient GDMT, limited coronary sinus branch anatomy, and failures in device programming [11, 27]. But in CCM, where anatomy may not be as relevant and programming far more simplistic, it interesting to consider why some benefit, while others do not. One patient in our experience did profoundly poorly—spending more time in the hospital than out of it, and ultimately died of progressive heart failure, all after CCM implant. There were no signs prior to CCM to suggest this patient would not be a good candidate. The patient was adherent to all medications, had a BMI of 33.8 kg/m2, an EF of 35%, a left ventricular end diastolic dimension of 50 mm, and an estimated glomerular filtration rate of 50 mL/min/1.73 m2. He had suffered an out of hospital ventricular fibrillation arrest, aborted by therapy from his ICD, approximately 2 weeks prior to CCM implant. A consideration for his progressive disease may be the chronicity of his cardiomyopathy (first diagnosed 4 years prior to CCM implant); however, others had similar lead-times prior to CCM implant who fared better. Another patient with EF decline was our single dialysis-dependent patient. Dialysis patients may be a specific group of patients in whom the risk/benefit must be very carefully considered, given their increased risk of infection with implanted cardiac devices [12]. Alternative studies may have been performed as a matter of protocol, such as cardiopulmonary exercise testing or 6-min walk test, to help identify which patients are more likely to benefit from CCM.

The limitations of this study relate to its small, uncontrolled, and unblinded nature. All measured outcomes were unblinded, and vulnerable to bias. Improvements in measured outcomes may have been related to CCM, or other unmeasured changes, like improvement in GDMT or improved vigilance of care in patients who had undergone what was a novel procedure at our institution. The lack of uniformity in follow-up time, as well broad period over which implants took place (2020–2023), may have resulted in differences in management, including that SGLT2 inhibitors and sacubitril/valsartan to replace ACEi/ARB in NYHA class III HF patients did not become part of societal guideline recommendations until 2022 [13]. Echocardiograms were performed at varying time intervals, and therefore, potential fluctuations of EF, in either direction, may not have been identified. Additionally, echocardiogram interpretations were performed by variable attending readers. Patients did not uniformly undergo quantitative testing of functional capacity before and after, such as a 6-min walk test and cardiopulmonary exercise testing. Weight data was collected from varying appointments, different scales, and at non-uniform intervals post-CCM.