Introduction

Heart failure (HF) causes a significant global health burden and has become increasingly prevalent1. The trajectory of heart failure varies from patient to patient and is often determined by the etiology, genetic factors, and treatment received2. It has been observed that a subset of heart failure with reduced ejection (HFrEF) patients experience improvement or recovery of their ejection fraction, spontaneously or after therapy3. However, the natural history, mechanism, and determinant factors of heart failure with improved ejection fraction (HFimpEF) have not been well described. Further, the lack of a unified definition of this subset of heart failure patients who experienced ejection fraction improvement hinders future clinical studies that seek to guide the management of such patients.

Definitions

The definition of HF with improved ejection fraction has evolved4. The 2013 ACC/AHA guidelines were the first to appreciate patients with HFpEF who previously had HFrEF as clinically unique from those with persistently preserved or reduced EF; this cohort was classified as “Heart failure with preserved ejection fraction (HFpEF), improved.”5 HFrEF patients with improved EF have been increasingly recognized as a distinct clinical entity different from those with persistent HFrEF or HFpEF due to differences in outcomes and trajectories. It has been observed that although these patients have improvement in ejection fraction (EF), some patients are still at risk for relapse of heart failure, especially when guideline-directed medical therapy (GDMT) is terminated6. To unify the nomenclature, the Universal Definition defined HFimpEF as HF with a baseline LVEF of ≤40%, a ≥10-point increase from baseline LVEF, and a second measurement of LVEF of >40%. The European Society of Cardiology (ESC) classified HF with improved EF as patients with a history of overtly reduced LVEF (≤40%), who later present with LVEF ≥ 50%. The 2022 ACC/AHA guidelines defined HFimpEF as HF with a baseline LVEF of ≤40% and a follow-up measurement of >40%3. Such definitions emphasize the importance of longitudinal measurement of EF to differentiate the HFimpEF from HFmrEF and HFpEF, as patients who demonstrate a progressive decline in LVEF or who have persistently low LVEF generally face a worse prognosis than those with an improved LVEF7,8,9. This also highlights the necessity of continued monitoring and individualized management of patients based on their trajectory. However, discrepancies among these definitions—particularly regarding the degree of LVEF improvement required and the target threshold for “improvement”—create ambiguity in clinical categorization. These differences underscore the difficulty in defining HFimpEF, complicating efforts to standardize diagnosis, assess prognosis, and guide therapy across different healthcare settings. Despite best efforts with the Universal Definition, slight variance in definition may influence collection of outcome and clinical data. In this review, we aim to provide a comprehensive overview of the epidemiology, mechanisms, diagnosis, and management of HFimpEF.

Prevalence

The prevalence of HFimpEF has been described in the literature but varies significantly due to the differences in population and definition, ranging from 9 to 52%1,2,3,4. In the Val-HeFT trial, 9.1% of patients with a baseline EF < 35% achieved an EF > 40% after one year1. In the large prospective IMPROVE HF study, which enrolled 3994 patients, 28.6% of HFrEF patients showed an improvement of over 10% in EF at 24 months10. The CHAMP-HF registry, involving 2092 patients with a baseline EF of ≤40%, demonstrated that 33% of these patients experienced a ≥10% improvement in EF, with approximately one-third achieving an EF of more than 50%11. Data from the CARDIOCHUS-CHOP registry indicated that 52% of patients with baseline HFrEF showed an EF improvement exceeding 40%12. The Heart Failure Optimization study found that in patients de novo HFrEF patients who were prescribed a cardioverter-defibrillator, 46% had an LVEF improvement >35% at 90 days and 77% had an LVEF improvement >35% at 265 days13. Using the updated definition of HFimpEF, recent studies suggest that the prevalence of HFimpEF within the general heart failure population is estimated to be around 20–30%.

Global registries demonstrate variability across the continents and different geographies. A heart failure registry in Hungary reported that 19.5% of 833 patients with HFrEF at baseline achieved HFimpEF at one-year follow-up14. A cohort study from a Korean registry involving 1509 hospitalized HFrEF patients found that 31.3% had HFimpEF one year after initial hospitalization15. Similarly, two large registries from Asia indicated that 20–30% of patients with baseline HFrEF achieved HFimpEF within a year16. Regardless of ethnicity or country of origin, the prevalence of HFimpEF has been shown to be ~20–30%.

Prognosis

Patients with HFimpEF and resolution of HF symptoms generally exhibit better prognoses than those with persistent HFrEF or HFpEF. In the Val-HeFT trial, patients with HFimpEF experienced nearly a 50% reduction in mortality risk (rate ratio (RR) 0.49 [0.31–0.76]) compared to those with persistent HFrEF after a median follow-up of two years. Additionally, in the CARDIOCHUS-CHOP registry, patients with recovered LVEF had significantly lower mortality and hospitalization rates compared to both HFrEF (mortality RR 0.71 [0.54–0.92], hospitalization RR 0.89 [0.75–1.0]) and HFpEF patients (mortality RR 0.70 [0.54–0.89], hospitalization RR 0.86 [0.76–1.0])12. A large retrospective cohort study of 2166 patients also showed that those with HFimpEF had the lowest rates of mortality and hospitalization compared to patients with persistent HFrEF and HFpEF8. A meta-analysis pooling data from nine studies involving 9491 heart failure patients found a 56% reduction in all-cause mortality and a 60% reduction in hospitalization rates for HFimpEF compared to HFrEF patients, alongside a 58% reduction in all-cause mortality and a 27% reduction in hospitalization rates9. Importantly, although HFimpEF patients have the best prognosis in terms of death, cardiac transplantation, and hospitalization rate compared to other cohorts of heart failure patients, HFimpEF patients have elevated biomarkers and a higher risk of heart failure hospitalization and mortality compared to healthy controls4,13,17. This is especially true in symptomatic HFimpEF patients, who exhibit worse prognoses compared to asymptomatic HFimpEF patients and have been shown to have comparable hospitalization rates and outcomes to HFpEF patients8.

Despite the improvements, HFimpEF patients remain at risk for a future decline in EF. Data from the Trieste Heart Muscle Disease Registry, which included 800 patients with nonischemic cardiomyopathy (NICM), found that 57% of patients had HFimpEF with a median time of 13 months. However, of this cohort with improved EF, 41% had a recurrent decline in LVEF (EF ≤ 40%) over a median follow-up of 11 years18. Among 174 HFrEF patients who achieved an EF of ≥45% after beta-blocker therapy, 26% experienced a deterioration to below 45% at a median follow-up of 9.2 years13. In the TRED-HF trial, 36% of DCM patients with improved EF and resolution of symptoms experienced a relapse of DCM within six months after withdrawal of GDMT. In another study involving 408 DCM patients, those with persistently improved EF showed significantly better transplant-free survival compared to those who relapsed after 8.5 years of follow-up18.

In summary, HFimpEF represents a significant and unique subset of heart failure patients compared to those with persistent HFrEF. Although associated with a better prognosis than HFrEF and HFpEF, HFimpEF patients, especially if symptomatic, still face an elevated risk of mortality and hospitalization, and are susceptible to relapses, particularly with the withdrawal of GDMT.

Predictors

Data from landmark clinical trials suggest that GDMT, including beta-blockers and angiotensin-converting enzyme inhibitors/angiotensin receptor blockers (ACEI/ARB), is associated with improvements in EF and generally correlates with better outcomes19,20,21,22. Real-world data further illuminate additional clinical features that may help predict improvements in LVEF beyond the effects of GDMT. For instance, the IMPROVE HF study, which included 15,381 HFrEF patients, identified several factors—female sex, no prior myocardial infarction, nonischemic heart failure etiology, and absence of digoxin use—as being associated with a greater than 10% improvement in EF23. Similarly, the SwedeHF study, comprising 4942 patients, found that female sex, nonischemic causes, higher blood pressure, and shorter heart failure duration were predictors of HFimpEF24. Additionally, the absence of LBBB was associated with better odds of EF improvement compared to those with LBBB25. Genetic factors, cardiac imaging features, and biomarkers have also been linked to HFimpEF, which will be discussed further.

Factors associated with worsening EF in patients who achieved HFimpEF have also been studied. In addition to the aforementioned withdrawal from GDMT, prolonged QRS interval, presence of LBBB, older age, lower heart rate, and larger LVEDD have been identified as independent predictors of recurrent LVEF deterioration26. DM and ischemic etiology were also found to be associated with a higher likelihood of worsening EF27. A study from Cleveland Clinic found that male sex, non-White race, atrial fibrillation, and the presence of an ICD were negatively associated with maintaining improved EF17. There appear to be no reliable predictors of persistent recovery of EF among 408 patients with DCM, who initially experienced improvement, according to a long-term follow-up study27. (Table 1). While these demographics offer a general overview of the populations and comorbidities studied to date, patients with HFimpEF likely represent a much more complex cohort with unexplored variables that warrant further investigation.

Table 1 Summarize predictors of LVEF improvement in HFimpEF
Full size table

Mechanisms

The underlying biological process of HFimpEF is known as reverse remodeling. In cardiac remodeling, cardiomyocytes in a failing heart undergo maladaptive changes including abnormalities in gene expression, mechanical properties, metabolic pathways, and the extracellular matrix (ECM)28. During myocardial remodeling, cardiomyocytes, ECM, and the myocardium experience a series of maladaptive changes. Myocyte alterations include hypertrophy, re-expression of fetal genes, abnormal excitation-contraction coupling, beta-adrenergic desensitization, and altered metabolism. Myocardial changes encompass myocyte necrosis, apoptosis, and autophagy, as well as myocardial fibrosis driven by fibroblast proliferation, MMP activation, and inflammation28. These changes induce alterations in ventricular geometry, resulting in either concentric remodeling (wall thickening with reduced compliance) or eccentric remodeling (wall thinning with chamber dilation)29. These alterations ultimately shift the volume-pressure relationship in the left ventricle, contributing to clinical heart failure. First described in 1995, reverse remodeling was observed both in patients with ischemic and nonischemic cardiomyopathy among patients with HFrEF treated with beta-blockers. Other than medical treatment strategies, changes in EF and LV volumes were observed in device therapies with CRT and other ventricular geometry and ventricular unloading interventions30,31 (Fig. 1).

Fig. 1: Mechanism of HFimpEF.
figure 1

LVEF left ventricular ejection fraction, HFimpEF heart failure with improved ejection fraction. Legend: This figure illustrates the various treatments for heart failure with improved ejection fraction that depend on the etiology of heart failure. These treatments include revascularization, circulatory support, anti-arrhythmic devices and guideline directed medical therapy. Such therapies can contribute to improvement in cellular and molecular cardiac remodeling to potentially improve ejection fraction.

Full size image

Studies have shown that treatment with LVAD implantation, CRT, RAAS inhibition, and beta-blockade is associated with reduced cardiac hypertrophy, decreased fetal gene expression, improved cell metabolism and energetics, increased beta-adrenergic responsiveness, and reversal of cardiac fibrosis and ECM protein deposition32. Some of these studies demonstrated accompanying improvements at the cellular level with improvement in contractility and increased beta-adrenergic responsiveness in cardiac myocytes33.

An in vitro study of failing human hearts showed a 50% reduction in β-receptor density and a 45% reduction in isoproterenol-mediated adenylate cyclase stimulation compared to normal functioning hearts. It is hypothesized that the decreased β-receptor density leads to subsensitivity of the β -adrenergic pathway and therefore decreased muscle contraction34. The geometric changes to the heart appear to be involved in its impairment. Research into the molecular and cellular mechanisms of reverse remodeling suggests that most heart failure therapies promote this process by reversing defects, remodeling cellular defects at the cardiac myocyte level, accompanied by improvement in cardiac mechanics and hemodynamics resulting in reverse remodeling of the myocardium, and reducing left ventricular dilation28,35.

Importantly, while recovery from heart failure is possible, reverse remodeling often does not equate to complete restoration of myocardial function. Studies of gene expression during LVAD therapy have shown that most abnormally upregulated genes in cardiomyocytes do not normalize36. In inflammation-triggered dilated heart failure models, the normalization of LV structure and function was only accompanied by partial normalization (~60%) of gene expression associated with heart failure36. Persistently abnormal genes affect all aspects of cardiac remodeling, including ECM, cytoskeletal elements, excitation-contraction coupling, metabolism, and sarcomeric protein families. Moreover, when subjected to adverse stimuli again, exaggerated hypertrophic response and increased mortality were noted36. Observations from LVAD patients also suggest that while many experience normalization of LV chamber geometry, changes in the ECM and the ratio of LV wall thickness to LV wall radius may not return to normal37.

These observations have led to the theory that true myocardial recovery which may be seen in a small proportion of patients, may represent a reversible process called “elastic deformation,” where cardiomyocytes may be able to return to their original form after the removal of adverse stimuli. In contrast, “plastic deformation” terminology is used when cardiomyocytes do not fully revert to a normal state due to irreversible damage and remodeling28. Therefore, reverse remodeling without complete recovery is better described as a “remission” rather than “recovery,” given the persistent risk of future heart failure4. This viewpoint also supports clinical observations that the recovery from heart failure is largely dependent on etiology. While myocardial recovery from transient injuries—such as toxin exposure, inflammation, peripartum cardiomyopathy, tachycardia or stress-induced cardiomyopathy, or viral infections—is possible, most heart failure cases resulting from infarction, long-term hemodynamic alterations, diabetes, or genetic factors are less likely to achieve complete recovery due to underlying irreversible remodeling3,38,39.

Clinical assessment

Symptoms and physical exam

Patients with HFimpEF often experience similar typically milder than compared to those seen in HFrEF. Common symptoms include dyspnea, fatigue, and exercise intolerance, although the severity of these symptoms tends to be less pronounced in HFimpEF patients. However, even with improvement in ejection fraction patients may still experience symptoms. Specifically, a small proportion of patients may have a resolution of symptoms (NYHA Class I) or signs but may have a relapse with re-exposure to cardiac stress or injury3,4. Acknowledging symptom severity in HFimpEF does not always correlate directly with ejection fraction, it is important to recognize that distinguishing HFrEF from HFimpEF based on symptoms is unreliable, as some patients with HFrEF can remain asymptomatic despite a low LVEF. Furthermore, in those with HFimpEF who continue to experience significant symptoms and physical exam findings, such as jugular venous distention, peripheral edema, expiratory crackles, and cool or pale skin, other factors outside of LVEF may be contributing to their condition40. Cardiac conditions affecting intravascular filling pressures, left and right interdependence, arrhythmias, valvular dysfunction, and cardiac constrictive or restrictive processes. Additionally, non-cardiac factors, such as pulmonary or metabolic factors, can also contribute to symptom burden.

Given the overlap in symptoms between HFrEF and HFimpEF, relying solely on physical exams and symptom reviews to evaluate improvements in cardiac function is insufficient to guide treatment decisions. A comprehensive review of biomarkers, recent imaging, medication adherence, physical exam findings, and detailed symptom review is essential to properly manage HFimpEF and optimize patient care.

Biomarkers

Despite the recovery in ejection fraction and milder symptoms in patients with HFimpEF, biomarkers can provide a clearer picture of diagnosis and prognosis. Key biomarkers include natriuretic peptides (NP) and high-sensitivity cardiac troponins (hs-cTn)41. According to US HF clinical guidelines published in 2022, measurement of brain natriuretic peptides (BNP)or N-terminal prohormone of BNP (NT-proBNP) is recommended for risk stratification in chronic heart failure and can aid in the diagnosis of HF42. The European Society of Cardiology also emphasizes the utility of NPs as reliable indicators of intracardiac volumes and filling pressures with higher levels increasing the likelihood that patients’ dyspnea is due to HF42. Clinicians can use NPs in symptomatic patients (dyspnea and fatigue) to help in early diagnosis between acute and chronic HF. In addition, NPs have strong prognostic value and are a great tool for predicting outcomes like death and HF-related hospitalizations, even when patients have other cardiac conditions, such as myocardial infarction or atrial fibrillation42. NP levels serve as a biochemical marker of cardiorenal impairment providing additional information beyond imaging in the evaluation of HFimpEF. Elevated levels must always be interpreted alongside cardiac imaging to identify the underlying cause of HF, as GDMT effectively lowers NP levels, which are linked to better outcomes.

Genetic and proteomic signatures specific to HFimpEF can help distinguish it from other heart failure types43,44. A study by Adamo et al.45 identified proteomic profiles in HFimpEF patients using high-content aptamer-based proteomic technology, revealing small variabilities compared to other HF subtypes. While there were overlaps with HFpEF, distinct differences were observed with HFrEF45. These differences may inform medical management and device interventions for HFimpEF patients. However, applying these proteomic signatures in therapy management faces limitations due to the complexity and cost of assays, as well as a lack of superiority over NT-proBNP in predicting adverse outcomes3.

Biomarkers such as soluble suppression of tumorigenicity 2 (sST2), Galectin-3 (Gal-3), and Growth differentiation factor-15 (GDF-15) all provide additional insights into the prognosis of heart failure. These biomarkers are implicated in cardiac stress and fibrosis46,47.However, there are no current studies on HFimpEF and sST2. Yan et al. found an inverse relationship between ejection fraction and sST2 levels in HFpEF and HFmrEF47. Gal-3, associated with fibrosis and inflammation, can provide information in risk stratification for HFimpEF patients48. While there is limited evidence, some studies suggest there is a relationship between elevated Gal-3 and prognosis48. Further research is necessary todetermine Gal-3 fluctuations in patients improving ejection fraction from HFrEF. Lastly, GDF-15, related to inflammation and oxidative stress, has been implicated in heart failure prognosis49 Otaki et al.50 found that elevated GDF-15 levels were associated with increased cardiac-related events and mortality in a Japanese cohort50. Biomarkers such as sST2, Gal-3, and GDF-15may provide potential applications for HFimpEF patients. While current studies are limited, findings suggest that these markers can help with risk stratification. Further research is essential to better understand the specific roles of these biomarkers in HFimpEF and their potential for guiding clinical decisions.

Imaging modalities

Imaging plays a critical role in the diagnosis and management of HFimpEF by providing valuable information on left ventricular ejection fraction, diastolic dysfunction, structural abnormalities, and hemodynamic parameters3. Imaging tools that are particularly useful across all HF subtypes are 2D echocardiography and cardiac magnetic resonance imaging (CMR). Each modality offers clinicians additional information, especially in cases of HFimpEF. Echocardiography has been used to evaluate left ventricular ejection fraction since 1962, facilitating the assessment of heart failure severity51,52. Echocardiography allows clinicians to assess cardiac function and gain insights into potential EF recovery. One study found that higher tricuspid annular plane systolic excursion (TAPSE), smaller end-diastolic volume (162.51 mL vs. 208.54 mL, P < 0.001), and smaller left ventricular end-systolic volume (119.81 mL vs. 157.13 mL, P < 0.001) are associated with ejection fraction recovery, helping to identify patients likely to recover53. Speckle-tracking echocardiography (STE) further enhances myocardial function assessment by providing strain analysis. In patients with HFimpEF, STE can measure global longitudinal strain (GLS), identifying persistent subclinical systolic dysfunction that may not be apparent on a traditional 2D echocardiogram54. STE imaging would also allow for monitoring improvements and decreased strain that is associated with HFimpEF54.

Several limitations complicate the diagnosis of HFimpEF. LVEF is measured via echocardiography, which can be affected by image quality, intra- and inter-observer variability, and physiological fluctuations. Measurement errors may be influenced by human factors, including bias or knowledge of the patient’s condition. Furthermore, the diagnosis of HFimpEF requires longitudinal follow-up, and inadequate serial assessments may lead to missed opportunities for proper reclassification as a patient’s condition evolves. Such inconsistencies can result in the misclassification of patients, potentially categorizing them as having heart failure with mildly reduced ejection fraction (HFmrEF) or HFpEF instead of recognizing HFimpEF. Finally, symptom improvement, which is the primary goal for patients, may not always correlate with improvements in LVEF. Conversely, patients may experience symptomatic relief even without a significant change in LVEF55,56. While imperfect, echocardiogram is great at detecting structural and functional parameters of the heart, which allows clinicians to follow myocardial recovery and guide GDMT for patients with HFimpEF.

Cardiac magnetic resonance imaging (CMR) provides even more detailed insights into cardiac morphology and function57. Like other imaging techniques, CMR assesses both systolic and diastolic function but also provides structural changes in detail. For instance, CMR enables clinicians to detect fibrosis, inflammation, and other tissue damage around the heart through late gadolinium enhancement and T1 and T2 mapping, which are adverse components of the remodeling process, thereby adding prognostic value for HFimpEF patients58. CMR is also useful for assessing cardiac volumes, mass, and ejection fraction for both ventricles, especially when echocardiograms do not provide sufficient information. While not specifically tracking HFimpEF in studies, CMR is a valuable tool for assessing etiology, scar burden, and progression of heart failure52.

Management

Pharmacologic intervention

The pharmacological management of HF, particularly through GDMT, of HF has been extensively studied and has demonstrated improvements in patient quality of life and mortality (Table 2)59,60. However, while current data provides a wealth of information on either HFrEF or HFpEF, the current data are limited in clinical guidance of patients with HFimpEF. The European Society of Cardiology (ESC) Guidelines offer limited information on how to manage patients with improvement in HF61, while the AHA Guidelines provide only a single Class I, Level B recommendation for the maintenance of GDMT in patients with HFimpEF3. One might ask, if EF improves, is there any benefit to continuing GDMT? The answer lies in examining the continued advantages of GDMT and additional therapies used in HF, which often overlap in benefits for patients with HFimpEF.

Table 2 Pharmacologic treatment in HFimpEF
Full size table

Beta-blockers (BBs) are important in the reversal of cardiac remodeling in HFrEF patients and likely prevent adverse remodeling in HFimpEF patients. BBs work by inhibiting chronic sympathetic activation, which counters the harmful effects of catecholamines on the heart, thus preventing adverse remodeling62. Moreover, while reducing the potential for unfavorable remodeling, BB has been shown to improve both right and left ventricular EF, suggesting that patients with HFrEF can achieve significant recovery of EF63. The importance of continuing BB therapy inHFimpEF is underscored by studies like Enzan et al.64, which found that patients who discontinued BB after EF recovery had a higher incidence of left ventricular EF decline during a two-year follow-up period compared to those who continued BB therapy (24.0% vs. 19.6%; odds ratio [OR], 0.77; 95% CI, 0.63–0.95; p = 0.013). This finding highlights the importance of maintaining BB therapy, even after EF recovery64. To maximize benefits, such as reducing adverse remodeling and improving EF, BBs should be titrated to their maximum or maximally tolerated doses in HFimpEF patients15. BB are essential for preventing adverse cardiac remodeling and improving EF in HFimpEF patients, and their continued use after EF recovery significantly reduces the risk of left ventricular decline.

Similar to BBs, angiotensin-converting enzyme inhibitors (ACEi), angiotensin receptor blockers (ARBs), and angiotensin receptor neprilysin inhibitors (ARNIs)), have a role in the treatment of HFimpEF65,66. Improving and maintaining left ventricular function is crucial for patients with HFrEF on the path to recovery, as well as those who have already transitioned to HFimpEF. A study by McElderry et al.17 demonstrated that the continued use of ACEi, ARBs, and ARNI after initial EF improvement, as observed via echocardiogram, was associated with sustained enhancement of cardiac function17. In addition to improving EF, these medications have been linked to better survival rates and reduced HF-related hospital admissions17. All GDMT indicated for treatment of HFrEF play a crucial role in GDMT for HFimpEF, leading to better patient outcomes.

Mineralocorticoid Receptor Antagonists (MRA) are unique in GDMT due to their multifaceted biological effects and clinical benefits in HFimpEF. A study examining the deletion of the mineralocorticoid receptor from smooth muscle cells in mice suggests that MRAs are beneficial in heart failure by inhibiting aldosterone binding to its receptor. This action reduces collagen deposition and fibrosis in the myocardium by limiting the expression of pro-fibrotic genes, thereby preventing harmful cardiac remodeling67. Additionally, MRA therapy has been shown to modestly improve LVEF, making it a valuable therapeutic option for patients with HFimpEF68. The importance of continuing MRA therapy after EF improvement has been highlighted in studies like that of Chen et al. (2023), which demonstrated that patients who discontinued spironolactone had a higher incidence of heart failure relapse (HR: 2.3%, 95% CI, 1.2–4.5; p = 0.01)68. Therefore, incorporating MRAs into the long-term management of HFimpEF is crucial for maintaining the structural and functional integrity of the myocardium, preventing fibrosis, and ensuring long-term optimal ejection fraction.

Despite current guidelines recommending the continuation of MRAs to prevent adverse cardiac events, MRA therapy is often discontinued due to unfavorable side effect profile. Second-generation MRAs, such as eplerenone, have reduced affinity for androgen, progesterone, and glucocorticoid receptors, resulting in a more favorable side effect profile while demonstrating similar benefits in patients with HF. More recently, third-generation MRAs such as finerenone are in development for use in patients with HF. These third-generation, nonsteroidal mineralocorticoid receptor agonists exert highly specific inhibition of mineralocorticoid receptor hyperactivation, thereaby preventing myocardial fibrosis and inflammation without the systemic side effects associated with first-generation MRAs69. Ongoing studies are evaluating the utility of third-generation MRAs in HF, which will help guide more precise and tolerable therapeutic choices in HF management70.

Sodium-glucose cotransporter-2 (SGLT2) inhibitors are newer medications that have significantly reduced HF hospitalizations and cardiac events since their introduction into clinical practice71. The mechanism of action of SGLT2 inhibitors involves inhibiting the reabsorption of sodium and glucose in the kidneys, resulting in reduced volume overload and improved hemodynamics. Studies have shown that SGLT2 inhibitors also reduce oxidative stress, cardiomyocyte death, and inflammation—all of which contribute to the prevention of heart failure progression72,73. SGLT2 inhibitors have further demonstrated benefits in cardiac remodeling, as evidenced by a reduction in left ventricular end-diastolic volume (−1.5 ± 25.4 ml) and left ventricular end-systolic volume (−26.6 ± 20.5) compared to placebo (p < 0.001)73. Moreover, a pre-specified analysis of the DELIVER trial showed that HFimpEF patients who took dapagliflozin experienced a reduction in worsening heart failure events (HR: 0.78, 95% CI, 0.61–1.14), cardiovascular death (HR: 0.62, 95% CI, 0.41–0.96), and total worsening heart failure events (rate ratio: 0.68, 95% CI, 0.50–0.94)74. This study suggests that incorporating SGLT2 inhibitors alongside GDMT can effectively maintain improvements in cardiac function and prevent adverse outcomes in HFimpEF patients. Thus, the inclusion of SGLT2 inhibitors into GDMT not only enhances heart failure management in patients with improved ejection fraction but also plays a vital role in maintaining cardiac function, preventing negative outcomes, and reducing the risk of hospitalization and cardiovascular events.

Beyond the traditional and newer medications included in GDMT for heart failure, GLP-1 receptor agonists (GLP1RA) have recently been shown to have significant benefits in HFpEF patients with obesity and are now indicated in this population. In comparison to placebo, HFpEF patients on GLP1RA were shown to have a reduction in heart failure symptoms, as measured by the Kansas City Cardiomyopathy Questionnaire (estimated difference: 7.8 points; 95% CI: 4.8–10.9; P < 0.001), a reduction in mean body weight (estimated difference: −10.7 percentage points; 95% CI: −11.9 to −9.4; P < 0.001), and improved exercise capacity, as assessed by the 6-minute walk test (estimated difference: 20.3 m; 95% CI: 8.6–32.1; P < 0.001)75,76. GLP1RA have shown varying results on improvements in LVD function and ejection fraction77,78. While GLP1RA have demonstrated benefits in HFpEF patients with obesity and HFrEF patients with diabetes, their role in HFimpEF remains understudied and warrants further investigation (Fig. 2).

Fig. 2
figure 2

This figure illustrates the optimal medical therapy and device therapy of HFimpEF such as Mitraclip and beta blockers.

Full size image

The question of how quickly to uptitrate GDMT therapy is addressed by the STRONG-HF trial, in which patients demonstrated more robust response with intensive, rapid up-titration of GDMT. In this study, patients were randomized to receive high-intensity or usual care; the high-intensity ground received full doses of prescribed therapy with MRA, BB, and renin-angiotensin blockers within 2 weeks of discharge after hospitalization with acute heart failure. Patients in the high-intensity care group saw significantly reduced symptom burden, improved quality of life, and a lower risk of all-cause death or heart failure readmission compared with usual care79. Therefore, physicians should prioritize rapid GDMT uptitration following acute heart failure admissions.

As discussed, GDMT plays a crucial role in maintaining the health of patients with improved ejection fraction. The TRED-HF trial confirms this, showing that patients with recovered ejection fraction from dilated cardiomyopathy who stopped their GDMT experienced significantly higher relapse rates compared to those who continued their treatment ([95% CI 28.5–67.2]; P = 0.0001). Additionally, those who discontinued therapy showed adverse remodeling, evidenced by an increase in left ventricular mass and a reduction in left ventricular global longitudinal strain6,80. A long-term follow-up of TRED-HF showed that over 65% of participants in the trial with recovered DCM met the criteria for relapse over more than 6 years. Most of these patients were started on low doses of GDMT immediately following the trial, but this still resulted in relapse80. The findings of TRED-HF and its long-term follow-up reinforce the notion that HFimpEF represents a transient remission of systolic function rather than a full recovery in most patients. The risk of relapse is more likely than complete myocardial recovery6,28,81. Therefore, one of the critical issues in managing HFimpEF is ensuring continued adherence to GDMT. It is important to note that while continued therapy is crucial for sustained LVEF improvement, TRED-HF also demonstrated the therapeutic benefit of long-term management and monitoring. Therefore, patient and physician education is essential to emphasize the importance of maintaining GDMT and long-term monitoring even after LVEF improvement.

Device-based interventions

Implantable Cardioverter-Defibrillator (ICDs) and Cardiac Resynchronization Therapy (CRT) device-based interventions are both integral in HF management3.To reduce the risk of sudden cardiac death (SCD), guidelines recommend the implantation of an ICD in patients with an EF ≤ 35%, following three months of optimal GDMT and with a life expectancy of over one year82. Studies, such as the one conducted by Al-Sadawi et al. have found that the risk of ventricular arrhythmias (VA) is still present for HFimpEF (4.1%), therefore there is still utility for an ICD in patients with improved ejection fractions83. In addition, a secondary analysis of the SCD-HeFT trial also, emphasized the utility of ICD therapy in patients with improved heart failure lower rates of all-cause mortality in comparison to the placebo group (2.6 vs 4.5 per 100 person-year follow-up)84. While an improvement in EF appears to reduce the risk of VA, the risk of SCD persists. While an improvement in EF appears to reduce the risk of VA, the risk of SCD persists. LVEF itself, even in patients post-MI, is a poor predictor of SCD85. The available data supports ICD generator replacement for most HFimpEF patients, particularly those with a history of shock therapy, high arrhythmia risk, or abnormal ECG findings. Future studies should focus on determining the timeline when the risk of arrhythmic events is low enough for patients with improved ejection fraction to consider having their ICDs removed, as well as identifying which subpopulations of HFimpEF may not require continued ICD therapy. The risk of removal or deactivation of ICD will need to be weighed in the context of continued ICD therapy and individualized to each patient. Most patients with HFimpEF do not require removal unless they have an ICD-related infection/endocarditis, electrical or rhythm, or valvular-related complication.

Current guidelines recommend the use of CRT in patients with HFrEF, NYHA class symptoms of II or greater, and a QRS duration of 150 ms or greater3. The main goal of CRT is to improve the coordination of cardiac contractions via the pacing of both ventricles. Historic and new studies have proven CRT alone and CRT-D has significantly reduced the all-cause mortality rate, heart failure hospitalizations, and heart failure events, but these studies were primarily focused on patients with HFrEF (</ = 30–45%)86,87. Current research from a prospective study following REVERSE study participants showed that individuals with CRT developed a significant increase in LVEF over 5 years (6.0 ± 10.8%, (P < 0.000)) suggesting CRT benefit in improving EF in conjunction with optimal GDMT and the potential for continued improvement of EF and LV remodeling for patient in the HFimpEF population35. The growing evidence that CRT may also beadvantageous for those with HFimpEF, indicates a broader potential in heart failure management with this subpopulation. The STOP-CRT study found that 7.5% of patients who underwent CRT and discontinued medications that block neurohormonal activity (BB and ACEi) experienced worsening heart structure or function. While CRT demonstrated benefits in supporting recovery and showed low rates of serious heart problems following the discontinuation of neurohormonal therapy, some risks remain for patients with other heart-related conditions. Therefore, discontinuing pharmacological therapy after CRT may not be suitable for everyone88.

The trajectory, prognosis, and risk of each patient may be different, and the continuation of therapies will need to be considered according to each patient’s unique profile. For example, it is common for patients to become pacer-dependent for maintaining a heart rate, and in those individuals, continuation of CRT would be important to prevent pacemaker or LBBB-induced cardiomyopathy even in the context of HFimpEF.

Prognosis of heart failure with improved ejection fraction (HFimpEF) according to specific etiologies

Despite advancements in novel therapies, diagnosis, and regular follow-ups, the prognosis of HF remains poor1,32,89,90. Some patients still experience severe outcomes shortly after diagnosis, often marked by LV remodeling32,91. Studies show that patients with improved LVEF have a lower risk of death and reduced hospitalization rates compared to those with persistent HFrEF or worsening EF27,92,93,94. In a meta-analysis by He et al., which included 9 studies and 9491 participants, it was found that patients with HFimpEF had a significantly lower risk of cardiac hospitalization (HR: 0.40, 95% CI: 0.20–0.82) and mortality (adjusted HR: 0.44, 95% CI: 0.33–0.60) compared to those with HFrEF. Furthermore, compared to heart failure with preserved ejection fraction (HFpEF), HFimpEF was associated with a moderately lower risk of mortality (HR: 0.42, 95% CI: 0.32–0.55) and hospitalization (HR: 0.73, 95% CI: 0.58–0.92)95.

The course of a patient’s HF depends largely upon the etiology, genetic and pathophysiological factors, presence of comorbidities, and the timeliness, appropriateness, intensity, and responsiveness to HF therapy4. Patients with non-ischemic etiologies of HF have a more favorable prognosis and greater likelihood of reversal of systolic dysfunction94,96.

Ischemic cardiomyopathy

Ischemic cardiomyopathy (ICM) is generally defined by the presence of LV dysfunction <35% coexisting with—and thought to be caused by—severe coronary artery disease97. The ischemic injury results in pathological remodeling and disorganization, causing impaired contractility. The presence of ICM portends a poor prognosis, with data showing a 30% worse survival compared to nonischemic LV dysfunction98. Increased scar burden and recurrent myocardial infarctions are associated with a low likelihood of recovery and improvement in LVEF97. Current guidelines recommend “complete” myocardial revascularization in ICM patients, but do not specify by what means to accomplish the revascularization99. Trials show that the mechanism by which revascularization improves LV function and overall prognosis appears to differ between coronary artery bypass graft (CABG) and percutaneous coronary intervention (PCI), with data favoring CABG.

Early surgical revascularization via CABG is associated with better long-term mortality reduction compared to delayed intervention100. Velazquez et al performed a randomized control trial of 1212 patients with LVEF < 35% randomly assigned to undergo either CABG plus optimal medical therapy (OMT) compared with OMT alone. The study found that rates of all-cause mortality were lower over 10 years in the cohort assigned to CABG therapy compared to patients receiving medical therapy alone. Further, CABG has been shown to improve LVEF, which is directly correlated with all-cause mortality and prognosis. In a retrospective study examining outcomes post-CABG, it was found that patients with an LVEF improvement of ≥10% exhibited a better prognosis than those without LVEF improvement at 7 years (85.9% vs 63.5%, P = 0.001)101.

In contrast to CABG patients, patients with ischemic cardiomyopathy undergoing percutaneous revascularization (PCI) may not experience the same mortality benefit. The REVIVE trial examined whether patients with LV dysfunction <35% saw improvement in outcomes after PCI with OMT compared with OMT alone. This study found that revascularization by PCI did not result in a lower incidence of death, heart failure hospitalizations, or improvement in EF102. A retrospective cohort study examining over 10,000 patients who underwent PCI found that LVEF improved slightly at 6–8 months. LVEF improved the most in patients with severely reduced (<40%) baseline EF. Most notably, change in LVEF was independently associated with the risk of death even after adjusting for baseline LVEF. For every 5% increment in the change in LVEF, the adjusted risk for 5-year all-cause and cardiac death was reduced by 9% and 14%, respectively. Change in LVEF appears to have prognostic value in patients with impaired (LVEF < 50%) but not in patients with preserved baseline LV systolic function (LVEF ≥ 50%)103.

There is ongoing debate regarding the appropriate medical regimen for patients with improved EF shortly after revascularization therapy. Certain pillars of GDMT, such as BB therapy, may be withdrawn after recovery of EF above 40%. Discontinuation of therapy was found to be non-inferior to continued therapy in a French cohort of 3698 patients. In this study, the median time between the last MI and randomization was 2.9 years, meaning the patients had already received several years of BB therapy before stopping. More analysis is needed to evaluate the length of therapy required before discontinuation104. Other data suggests that withdrawal of beta-blockade leads to a reduction of ejection fraction105. More studies are needed to further evaluate ideal medical therapy for patients with improved EF after ischemic injury.

Takotsubo cardiomyopathy

Takotsubo cardiomyopathy (TC) is described as transient LV dysfunction triggered by emotional distress that commonly returns to baseline EF. A meta-analysis of 28 case series found that 96% of TC patients had improved EF from 20–50% to 59–76% over 7–37 days106. However, despite normal LVEF and serum biomarkers, some patients with TC may have long-lasting symptomatic and functional impairment, including reduced apical circumferential and global longitudinal strain107. Furthermore, some studies suggest that Takotsubo patients may have long-term outcomes similar to patients with acute coronary syndrome.

Longer-term studies have been mostly retrospective with small sample sizes and have shown mixed results. Elesber et al. and Valbusa et al. found that after acute complications of TC are addressed, patients with TC have improved EF to baseline with no significant difference in mortality compared to gender-matched controls108,109. In contrast, Lonesco et al., found that 52% of 27 TC patients reached a primary end-point of death, cardiogenic shock, SCD, or cardiac re-hospitalization over a mean follow-up of 27 ± 16 months, significantly higher than matched controls110. The largest prospective study of TC patients followed 136 patients over 2.3 ± 2 years. They found that 95% of patients ultimately improved their EF values > 50%, and 5% of TC patients had delayed improvement >2.5 months. Compared to age- and sex-matched controls, TC patients had increased significantly increased mortality (p = 0.016). Furthermore, BBs were not shown to be absolutely protective against either initial or recurrent SC events and seem to have an unproven benefit111. While it appears that most TC patients have improved ejection fraction and normal mortality, the data suggests that a subset of TC has increased mortality compared to the normal population. Larger studies over longer periods need to be done to further examine the outcomes and utility of medical therapy in TC patients.

Alcohol and drug-induced cardiomyopathy

Alcoholic cardiomyopathy (ACM) is a leading cause of non-ischemic DCM in the United States, and it is estimated that ~21–36% of all non-ischemic DCMs are attributed to alcohol112. Acute ACM occurs when a large volume of alcohol instigates myocardial inflammation, leading to myocardial damage and tachyarrhythmias including AF and ventricular fibrillation. The key diagnostic element of chronic ACM is LV dysfunction in the setting of longstanding alcohol use and the absence of CAD. It is postulated that chronic ACM results from an increase in reactive oxygen species levels in myocytes leading to cardiac dysfunction113. Long-term outcomes of alcohol-induced cardiomyopathy show a better prognosis than ischemic cardiomyopathy. Beyond GDMT, the mainstay treatment of ACM involves cessation of alcohol. Several studies show that complete abstinence results in an improvement in LVEF114. Moreover, ACM patients who reduce their consumption to moderate levels (<60 g/d) show similar LVEF improvements to those who withdraw from alcohol completely115. ACM patients whose EF improved had better outcomes (CV death or heart transplant) over 82 months compared to those whose EF did not improve116.

However, patients with high or prolonged duration of alcohol intake may have significant persistent LV dysfunction even after cessation of alcohol and initiation of medical therapy117. These results suggest that the earlier patients with ACM abstain or moderate their alcohol consumption, the more likely they are to improve their LVEF and their overall outcomes. Patients with cocaine-induced cardiomyopathy generally develop systolic dysfunction due to multiple pathophysiological mechanisms, including increased myocardial demand, coronary spasm, and myocardial injury and infarction118,119. First-line treatment for cocaine-induced cardiomyopathy includes cessation of cocaine use, which leads to significant improvement in systolic function. Additional medical therapy is consistent with GDMT, although special consideration is given to the use of BBs120. The literature is generally divided about the use of BBs. Traditionally, selective BBs are avoided due to the risk of unopposed alpha-adrenergic receptor stimulation. However, studies have shown that beta-blockade therapy is safe and may be effective in treating cocaine-induced heart failure121,122. In fact, the use of BB therapy in chronic cocaine use is associated with reduced hospital admissions and reduced 30-day mortality121. The data is limited and insufficient to draw generalizable conclusions, so further research is necessary123,124. In most substance abuse-related cardiomyopathies, cessation of illicit drug and substance use is the first step along with initiation of GDMT for HFrEF. Given the high incidence of recurrent injury, we recommend that GDMT should be continued125.

Cancer therapy-related cardiomyopathy

Given the significant advancement in cancer therapeutics, cancer therapy-related cardiomyopathy and cardiotoxicity are becoming a more prevalent health concern. Anthracyclines are a widely used chemotherapy drug that treats hematologic, soft-tissue, and solid malignancies and carries cardiotoxic properties in a dose-dependent manner. The mechanism is not completely understood but is proposed to be the formation of anthracycline-iron complexes that stimulate free-radical formation126,127,128. Prevention of anthracycline-related cardiomyopathy includes limitation of peak plasma concentration, administration of iron-binding compounds like dexrazoxane, and therapy with ACE inhibitors, BBs, and MRAs127,129. There exists some evidence that pre-treatment with spironolactone can preserve LV systolic and diastolic function, but further study is needed to support this conclusion130. Trastuzumab is a HER2 inhibitor that is used in the treatment of HER2+ breast cancer. However, cardiac dysfunction is a well-documented adverse effect; 13% of patients treated with paclitaxel and trastuzumab went on to develop cardiac dysfunction of NYHA class III or IV compared with 1% in a group treated with paclitaxel alone. Furthermore, 27% of patients in a group treated with anthracycline, cyclophosphamide, and trastuzumab developed cardiac dysfunction of NYHA class III or IV127,131. Fortunately, symptoms generally improved with standard medical management131. While the LV dysfunction associated with trastuzumab is sometimes reversible, relapse can occur upon re-exposure. It is generally not considered safe to remove medical therapy after improvement in ejection fraction6, especially if additional treatment with trastuzumab is required. However, more specific research regarding reversible exposure-related cardiomyopathies is warranted. In HFimpEF secondary to chemotherapy exposure, most GDMT is continued due to recurrence with re-exposure to chemotherapy or other injuries or cardiac stress. In patients who experience a recurrence of cancer and require re-exposure to chemotherapy, cardiac evaluation, and GDMT should be initiated early to help support cardiac function during treatment.

Peripartum cardiomyopathy

Peripartum cardiomyopathy (PPCM) is a pregnancy-related form of systolic heart failure that presents during pregnancy or soon after delivery. Exact mechanisms are unknown but immunologic, inflammatory, and genetic factors are thought to play a role117,132. One hypothesis of PPCM pathology involves increased oxidative stress during pregnancy, which generates a cardiotoxic sub-fragment of prolactin133. A National Institute of Health-sponsored study is ongoing to explore the safety and efficacy of bromocriptine in blocking prolactin breakdown to prevent the development of PPCM134. Management of PCCM is similar to other forms of nonischemic cardiomyopathy and includes GDMT with considerations for fetal safety during pregnancy and lactation. Fortunately, greater than 50% of patients with PPCM have recovery of their LVEF135,136, but relapses are common in certain patients, especially in subsequent pregnancies. Approximately 20% of women with recovered systolic function will have a decline in LVEF in subsequent pregnancies137. A study examining genetic factors in PPCM found that the presence of TTN truncating variants was significantly correlated with lower EF at 1-year follow up138. Evidence for continued treatment with GDMT after normalization of LVEF is currently insufficient, and discontinuation of medications should be done under careful observation. Therefore, long-term follow-up and cardiac care are required for good clinical practices and further data gathering. Moreover, patients should be counseled regarding relapse risk in future pregnancies.

Genetic cardiomyopathy (LMNA, TTN mutations)

It has been proposed that recovery from genetic cardiomyopathy may not be expected due to the irreversibility of the underlying cause. However, it is now known that the prognosis of genetic cardiomyopathy directly depends on the underlying gene mutation. Mutations in the TTN gene are the commonest genetic subtype of DCM and likely account for up to 25% of idiopathic DCM cases139. The TTN gene encodes for the sarcomere protein titin, and mutations resulting in TTN truncation are a common cause of dilated cardiomyopathy140. Importantly, TTN-related cardiomyopathy tends to have a favorable remodeling response to GDMT. A retrospective study examining genetic cardiomyopathies in The Netherlands found that compared to LMNA and idiopathic DCM, those with TTN mutations presented at older ages, less often developed LVEF < 35%, and had better composition outcomes of death, ventricular arrhythmia, and heart transplantation. Furthermore, close to 50% of TTN subjects saw an LVEF improvement of >10% after the initiation of GDMT141. Another study examined different outcomes of LV assistance devices (LVAD) implantation in end-stage DCM in TTN-positive vs TTN-negative patients. Of the TTN-positive cohort, 6 out of 10 recovered enough LV function to explant their LVADs. The overall results also suggest no difference in cardiac improvement between TTN-positive and negative patients, indicating the promise of recovery in the genetic DCM cohort142.

LMNA-related dilated cardiomyopathy (DCM) is a separate subtype of genetic cardiomyopathy that is associated with early conduction system disease and arrhythmias, which often precede LV systolic dysfunction and chamber dilation143,144. Patients with LMNA mutations demonstrate a lower likelihood of LV improvement compared to those with TTN truncating variants, even when treated with optimal GDMT145. Consequently, LMNA cardiomyopathy is associated with a more refractory disease course, and cardiac transplantation may be considered in advanced cases. Due to the high risk of sudden cardiac death, ICDs are recommended in LMNA-related DCM when clinical indications are met. Emerging research is exploring novel biomarkers and targeted therapies for LMNA-related cardiomyopathy. Preclinical models suggest that MAPK pathway inhibition, including ERK and JNK inhibitors, may improve EF and LVESD. However, these findings are currently limited to animal studies146.

It is important to highlight the cross-section of exposure-related and genetic cardiomyopathies. There is not a straightforward dose-dependent relationship between toxin exposure and degree of cardiomyopathy, particularly in ACM. Ware et al examined the relationship between alcohol intake and dilated cardiomyopathy, finding the prevalence of truncated variant TTN to be 13.5% in ACM patients compared with 2.9% in control patients125. While there is a more direct association between cumulative drug exposure and cancer therapy-induced cardiomyopathy (CCM), there is still substantial variability in individual response. Indeed, patients with CCM have more rare protein-altering variants than comparative cohorts, and patients with truncated variant TTN experienced worse more heart failure and impaired myocardial recovery than those without147. Finally, a study found that in genetic evaluation of 172 women with peripartum cardiomyopathy, the prevalence of TTNtv was significantly higher than that in control population (15% compared with 4.7%) but similar to that of DCM patients (17%). In this population, the presence of TTNtv was significantly correlated with a lower EF at 1-year follow up138. This suggests a genetic predisposition that may underly individual risk of toxin- and exposure-related cardiomyopathy. Genetic testing should therefore be considered inpatient in patients with ACM, CCM, and other acquired cardiomyopathies to inform risk stratification, prognosis, and long-term management.

These findings highlight the potential for HFimpEF in genetic cardiomyopathy, particularly TTN-related disease, and underscore the importance of genetic testing in risk stratification and long-term management planning. As precision medicine advances, future guidelines will benefit from integrating genetic profiles to personalize treatment decisions and identify patients most likely to sustain EF improvement.

Tachycardia-induced cardiomyopathy

Tachycardia-mediated cardiomyopathy (TMC) is a reversible form of HF that can be both a cause and effect of HF148. The advent of both medical rate/rhythm control and ablation has been shown to yield an improved EF in TMC. While TMC can be caused by any tachyarrhythmia, the most common cause of TMC is atrial fibrillation, which causes reduced diastolic filling, reduced stroke volume, and elevated diastolic pressure149. A single-center randomized trial in Germany investigated the effects of catheter ablation versus medical therapy in 363 HFrEF patients with AF. After a median follow-up of 38 months, the ablation group had significantly less all-cause mortality (13.4% vs 25%, p = 0.01) and hospitalizations for heart failure (20.7%, vs 25.9%, 0 = 0.004). Furthermore, an improvement in LVEF ≥ 35% was observed in 68% of ablation patients and 50% of the 18 patients who crossed over to the ablation group. Approximately a third of patients with frequent PVCs develop cardiomyopathy. Several studies have documented an improvement in LVEF following PVC ablation in >80% of patients150,151,152. In one study, early improvement in LVEF after ablation (>25% increase at 1 week) was shown to predict complete recovery of LV systolic function151. Emerging data suggest that the presence of fibrosis on CMR may identify patients with TMC who are less likely to recover their LV function. These patients may be at elevated risk of recurrence of TMC as well as sudden cardiac death151. Anywhere from 1 to 3% of patients with thyrotoxicosis present with HFrEF, thought to be due to TMC153,154. In a study of 34 patients with HFrEF secondary to hyperthyroidism, those who achieved a euthyroid state had a significantly improved LVEF (55% vs 30%, p < 0.001) at 3 months compared to those who remained hyperthyroid154. Notably, 15% of the original 34 patients had persistent HFrEF despite euthyroid state. TMC patients have an overall improved prognosis compared to ischemic causes of HF. While treatment of the underlying arrhythmia in TMC appears to successfully improve LVEF, these patients remain at an elevated risk for recurrence and sudden cardiac death.

Myocarditis

Myocarditis is defined as inflammation in the myocardium and manifests with diverse clinical presentations ranging from chest pain or mild dyspnea to acute cardiogenic shock. The most frequently identified causes of myocarditis are viruses155, but myocarditis can also stem from other infections, such as bacteria, protozoa, or fungi156; autoimmune disorders, such as systemic lupus erythematosus, sarcoidosis, or Sjogren’s syndrome157; or immune checkpoint inhibitors158. Up to 20% of myocarditis. In most myocarditis cases, cardiac abnormalities usually resolve without clinically relevant residual damage159. However, up to 20% of myocarditis patients result in persistent DCM160. While the overall incidence of myocarditis is unknown, retrospective studies report that 9–16% of unexplained nonischemic DCM have histological evidence of myocarditis, which suggests an inflammatory component that persists beyond the initial course155,161.

The initial step in evaluating for myocarditis involves recognizing a scenario that is clinically consistent with the diagnosis, such as a patient with a prior respiratory infection, a previous diagnosis of autoimmune disease or myocarditis, a family history or an unexplained death, or an exposure to immune checkpoint inhibitors. In addition to an EKG, troponin, and echocardiography (class 1, level C recommendation)117,155, the cornerstone of myocarditis diagnosis involves CMR and endomyocardial biopsy (EMB). In clinically stable patients with acute onset of symptoms within 2–3 weeks from presentation, CMR is the most accurate means of diagnosis of myocarditis and can be used to stratify risk according to scar pattern162,163. In patients with myocarditis presenting as cardiogenic shock, high-grade heart block, or symptomatic ventricular tachycardia, diagnosis can be made through EMB (AHA and ESC class 1,level B recommendation). The use of EMB allows clinicians to characterize inflammation and evaluate persistence of viral presence, which helps guide treatment in certain circumstances. Certain rare forms of myocarditis, such as giant cell myocarditis (GCM) and eosinophilic myocarditis (EM), can only be identified by EMB164,165.

It is recommended that all patients with acute myocarditis receive GDMT for the management of HF and arrhythmias5,166. Furthermore, expert consensus recommends 3–6 months of abstinence from competitive sports to decrease the risk of remodeling and sudden death167. The treatment for GCM and EM involves steroids sometimes in combination with azathioprine, although the data is limited to anecdotal evidence168,169 Beyond these rare causes of acute myocarditis, the use of immunosuppression is not beneficial170.

As opposed to acute myocarditis, those with EMB-proven non-viral chronic myocarditis seem to benefit from immunosuppressive therapy with prednisone and azathioprine with significantly improved EF compared to controls5,166. These studies were limited by the small number of patients and limited follow-up duration. A larger retrospective case series demonstrated an association between immunosuppressive therapy results and heart transplantation-free survival as compared with standard heart failure therapy alone171. In clinical trials in patients with chronic viral myocarditis, the use of antivirals (for human herpesvirus 6) and interferon-B (for enterovirus and adenovirus) has yet to show a clinically significant difference in EF improvement or CV death172,173. There are many ongoing treatment trials based on EBM findings. Until larger scale trials have been published, the use of shared-decision making should be used given the knowledge gap and lack of robust data for myocarditis management.

Conclusion

In conclusion, HFimpEF represents a distinct and important heart failure population, different from patients with persistent HFrEF. While HFimpEF is associated with a better prognosis compared to HFrEF, patients with HFimpEF still face an elevated risk of mortality and hospitalization, particularly if GDMT is discontinued. Until more evidence emerges, current guideline recommendations suggest the continuation of GDMT in patients with HFimpEF. However, as reviewed, specific considerations may be given based on the underlying etiology of heart failure. Further research is needed to better stratify which subgroups of HFimpEF patients require long-term therapy and to more thoroughly evaluate their long-term outcomes.