The ‘diamond' approach to personalized drug treatment of heart failure with reduced ejection fraction

Heart failure (HF) is a complex clinical syndrome with symptoms and signs due to cardiac dysfunction, leading to high hospitalization and morbidity. HF treatment has rapidly developed in recent decades, and breakthroughs have been made. Although conventional neurohormonal blockade therapies, including β-blockers, angiotensinconverting enzyme (ACE) inhibitors, angiotensin receptor blockers (ARBs), and mineralocorticoid receptor antagonists (MRAs), significantly improve the prognosis of patients with heart failure with reduced ejection fraction (HFrEF), mortality and rehospitalization remain high. Therefore, new therapies are needed. Previous studies demonstrated that ivabradine, angiotensin receptor-neprilysin inhibitor (ARNI), sodium-glucose cotransporter 2 (SGLT2) inhibitor, vericiguat, and omecamtiv mecarbil (OM) are beneficial for HFrEF. However, there is a lack of systematic review of the most optimal manner to use under various clinical conditions. This review summarizes the current knowledge regarding these therapies to give suggestions regarding clinical use timing, application scope, and optimal therapies under various conditions. Most importantly, we propose the HF diamond approach to express the necessity of conjunction of therapies. Different from the current guidelines, we suggest to use the diamond approach in an early and comprehensive manner at the beginning of ventricular remodeling in HFrEF to prevent further deterioration of HF and maximize the prognosis of patients.


Introduction
Heart failure (HF) is the biggest challenge to cardiovascular disease treatment in the 21st century, with substantial morbidity, rehospitalization, and mortality, which affected estimated 23 million people worldwide [1]. The incidence of HF in American adults increased from 5.7 million to 6.2 million in the past few years [2]. The data from GWTG®-HF (Get With The Guidelines-Heart Failure, Dallas, Texas, USA) showed that the rates for 5-year mortality and readmission in patients hospitalized for HF exceeded 80% and 75%, respectively [3]. According to the 2021 ESC (European society of cardiology, Brussels, Belgium) heart fail-ure guideline, HF is now subdivided as either reduced ejection fraction (rEF) (HFrEF: left ventricular ejection fraction (LVEF) ≤40%), mildly reduced EF (HFmrEF: LVEF of 41-49%), or preserved EF (HFpEF: LVEF ≥50%) [4].
Although conventional neurohormonal blockade therapies significantly improve the prognosis of patients with HFrEF, mortality and rehospitalization remain high. A study enrolled 11,064 HFrEF patients showed that under neurohormonal blockade therapies, the 2-year mortality for patients with worsening HF events was as high as 22.5%, and the hospitalization for recurring malignant HF within 30 days was 56% [5].
A large amount of data showed that conventional therapies cannot meet the current needs, and HF treatment urgently needs to be updated. New therapy for HFrEF has been rapidly developing in recent years. Therapies such as ivabradine, angiotensin receptor-neprilysin inhibitor (ARNI), sodium-glucose cotransporter 2 (SGLT2) inhibitor, vericiguat, and omecamtiv mecarbil, etc., which have been proved to improve the prognosis of HFrEF patients. This article mainly clarified the timing, scope of application, contraindications for the clinical use of each agent, and optimal therapies under various clinical conditions. Multiple factors are involved in the occurrence of HF and progressive treatment, and therefore, it is necessary to combine drugs with different mechanisms for the most successful treatment of HFrEF. The effective combination of angiotensin-converting enzyme (ACE) inhibitors (ACEI)/angiotensin receptor blockers (ARBs), mineralocorticoid receptor antagonists (MRAs), and β-blockers is widely recognized.
We proposed the diamond approach and it included new drugs and provided possible combinations based on the mechanism and clinical trials so that the most suitable drug regimen was used for an individual patient, and most importantly, to propose the necessity of the conjunction of medicines (Fig. 1). Different from the current guidelines, we proposed to use the diamond approach in an early and comprehensive manner at the beginning of ventricular remodeling in HFrEF to prevent further deterioration of HF and maximize the prognosis of patients (Fig. 2).
Bronchospasm was one of the adverse reactions caused by β-blockers, but the use of β-blocker therapy was not contraindicated by chronic obstructive pulmonary disease (COPD) [18,19] or bronchial asthma [20]. The use of cardioselective β-blockers was relatively safe in asthma patients [20]. β-blockers should be gradually titrated to the maximum tolerated dose or the target dose recommended by guidelines [14][15][16][17]. Study found that patients who reached the target dose and target heart rate had the lowest mortality, and those who only met the target dose attained higher survival if they achieved the target heart rate [21].

MRAs
Aldosterone was the terminal hormone of the RAAS and played a role in myocardial remodeling. In addition to diuresis and potassium preservation, aldosterone caused myocardial interstitial fibrosis [6]. MRAs blocked the action of aldosterone on mineralocorticoid receptors and reversed cardiac remodeling. Spironolactone [31] and eplerenone [32] significantly reduced the mortality and hospitalization of HFrEF (LVEF <35%). The 'escape phenomenon' of aldosterone occured after long-term use of ACEIs/ARBs. ACEIs/ARBs alone was not ideal for reducing aldosterone in this setting, and MRA therapy was necessary. For patients who still has symptoms of HF even under the treatment of ACEIs/ARBs and β-blockers, MRAs were recommended according to guidelines [14][15][16][17]. MRA should be avoided in the following situations: patients with kidney failure (estimated glomerular filtration rate (eGFR) >30 mL/min/1.73m 2 ), hyperkalemia (>5.0 mmol/L), or patients who are pregnant [14][15][16][17].

Ivabradine
The correlation between adverse cardiovascular events and rapid heart rhythm has been confirmed in patients with cardiovascular disease, and lowering the heart rate reduced cardiovascular risk [33][34][35]. A retrospective cohort study found that high resting heart rates often occured in HFrEF and were always associated with adverse outcomes [36]. Ivabradine slowed down the heart rate by selectively inhibit- NYHA, New York Heart Association; SGLT2i, sodium-glucose cotransporter 2 inhibitor; ACEI, angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blocker; ARNI, angiotensin receptor-neprilysin inhibitor; HFrEF, heart failure with reduced ejection fraction; HF, heart failure.
ing the funny ion current (I f ) channel of the sinoatrial node cells. It dose-dependently reduced heart rate and became more effective when the heart rate was faster [37]. Ivabradine also significantly reversed left ventricular remodeling [38].
In the SHIFT trial, ivabradine significantly reduced HF hospitalization but not cardiovascular or all-cause mortality compared with placebo [34]. A prospective study showed that ivabradine is beneficial for HFrEF by effectively improving the symptoms and quality of life [39]. Ivabradine was added to therapy when patients have received the maximum dose of β-blockers but still had a resting sinus heart rate ≥70 beats/min, or if symptoms remained after optimized treatment [14,16,17].
Ivabradine should be avoided in patients with sinus bradycardia, sinoatrial block, or second degree or above atrioventricular block [14,16,17]. Due to its mechanism of action, it had no effect on cardiac inotropy or systemic vascular resistance [33]. However, cornerstone research did not include patients with hypotension (blood pressure (BP) <90/50 mmHg), and therefore, the specific impact of ivabradine on blood pressure is still unclear [34]. Ivabradine might increase atrial fibrillation by blocking the I f channel in sinus node tissue [40,41]. However, several individual cases [42,43] and a small sample study [44] indicated that ivabradine slowed the ventricular rate and improved heart function in patients with atrial fibrillation. It is necessary to perform further clinical trials to confirm whether ivabradine is an effective agent in this condition.

ARNI
As a multi-compound drug composed of the ARB valsartan and the neprilysin inhibitor sacubitril, ARNI showed an extraordinary effect on the reversal of cardiac remodeling. The PARADIGM-HF trial demonstrated that ARNI exceeded enalapril in reducing cardiovascular mortality (hazard ratio (HR), 0.80 [95% CI, 0.71-0.89]) and hospitalization for HFrEF (HR, 0.79 [95% CI, 0.71-0.89]) patients [45]. The superiority of ARNI compared with enalapril in improving the quality of life has also been proven [46].
In the EVALUATE-HF trial, a significant reversal in cardiac remodeling was observed after 3 months of ARNI treatment for HFrEF patients [47]. The PROVE-HF study further explored the association between improvement of ven-tricular remodeling and the reduced level of N-terminalpro-brain natriuretic peptide (NT-ProBNP) [48]. For patients hospitalized with acute HF, the TRANSITION study demonstrated the safety and efficacy of early in-hospital initiation of ARNI after hemodynamic stability [49], and the PIONEER-HF trial confirmed the superiority of initial treatment with ARNI compared to enalapril through the 8 weeks of follow-up [50]. Similar conclusions were made in several studies where ARNI exceeded ACE inhibitors/ARBs in reducing hospitalization [51] and improving the quality of life for HFrEF patients during the 12 months of follow-up in real world practice [52]. The cardio-renal benefit provided by natriuretic peptides indicates that ARNI was superior to other traditional RAAS inhibitors for the treatment of HF and CKD.
Despite causing a modest increase in the urine albuminto-creatinine ratio (UACR), ARNI was found to slow the rate of decrease in the eGFR more effectively compared to enalapril [53]. Prospective researches have proved that patients with HFrEF and CKD can benefit from ARNI [54,55]. ARNI has become a new cornerstone and first-line therapy for HFrEF in guidelines [14,16,30]. Common side effects of ARNI were related to hypotension, renal insufficiency, and rare angioedema. It should be further noted that the contraindications to ACE inhibitors/ARBs also apply to ARNI.

SGLT2 inhibitors
By inhibiting sodium-glucose cotransporter 2 (SGLT2) which presented at the early proximal tubule, SGLT2 inhibitors prevented the reabsorption of the majority of filtered urinary glucose, and lowered blood glucose levels. Except as glucose-lowering agents [56], SGLT2 inhibitors showed beneficial effects on hospitalization for HF, and cardiovascular and total mortality in patients with diabetes [57][58][59][60][61]. The specific mechanisms of SGLT2 inhibitors that confer cardiac benefits remained unknown, and might be related to lowering of blood pressure, diuresis, weight loss, amelioration of myocardial metabolism and fibrosis, and reduction of the excessive activation of the sympathetic nervous system (SNS) and RAAS [62][63][64].
A subgroup analysis of the DAPA-HF trial [71] showed that the combination of SGLT2 inhibitors and ARNI was efficacious and safe. Caution was advised in patients with genital and urinary tract infections [72,73], ketoacidosis [74], hypovolemia (e.g., hypotension, dehydration, and cerebral infarction), or hypoglycemia (when combined with insulin or an insulin secretagogue) [75]. Ketoacidosis was a serious but extremely rare clinical condition in patients on SGLT2 inhibitors. The rate reportedly was less than 0.76/1000 patientyears in patients receiving canagliflozin [76] and 1/1000 for empagliflozin [57]. How SGLT2 inhibitors might be contributing to ketoacidosis has not been fully understood, but major illness, prolonged starvation, heavy alcohol use, and lower insulin doses were potential ketoacidosis triggers.

Vericiguat
The nitric oxide-soluble guanylate cyclase-cyclic guanosine monophosphate (NO-sGC-cGMP) pathway was the main regulator in myocardial metabolism and endothelial function, and it was a potential target for the treatment of chronic HF [77]. As a new oral sGC stimulator, vericiguat was cardioprotective by increasing the activity of cyclic guanosine monophosphate (cGMP) [78,79].
In the SOCRATES-REDUCED trial [80], vericiguat was proven to be well-tolerated and safe in HFrEF during the 12 weeks of follow-up. The VICTORIA trial [81] enrolled HFrEF patients with more severe heart failre than those in other contemporary clinical trials, and found that even under guideline-directed medical therapy (GDMT), vericiguat still reduced the composite primary outcome of cardiovascular death or first HF hospitalization (P = 0.02) over a median follow-up of 10.8 months. This result was mostly driven by reduced HF hospitalization, with a statistically nonsignificant reduction in cardiovascular death. A secondary analysis of the vericiguat trial showed that there were time-dependent risks of events according to index hospitalization subgroups, in which the worsening outpatient subgroup had the lowest risk compared to those at <3 months or 3 to 6 months after HF hospitalization, no significant difference was found in the risk reduction between these subgroups [82]. Based on those positive clinical results, vericiguat is considered for symptomatic HFrEF (even under GDMT) according to the 2021 ESC heart failure guideline [4].
OM was a selective, small-molecule cardiac myosin activator (also referred to as a cardiac myotrope) that binded to the catalytic domain of myosin and increased cardiac contractility without affecting cardiac myocyte intracellular calcium concentrations or myocardial oxygen consumption [86,87]. Different from conventional inotropic agents, OM was a potential treatment for HFrEF patients due to its properties [88].
The recent large randomized controlled trial GALACTIC-HF was conducted to demonstrate this hypothesis [89]. This trial enrolled HFrEF patients with EF ≤35%, and most of these patients were under standard HF therapy. The results showed that OM still reduced the composite primary outcome of the time to cardiovascular death or first HF event (P = 0.03). However, no statistical differences were found for secondary endpoint events including cardiovascular death and the first HF hospitalization. The benefit of OM for patients with HF was moderate compared to neurohormonal blockade therapies. However, OM remained safe and effective even under standard medical care for HF, which maked it a promising treatment. However, it is worth noting that OM can induce ischemia if plasma concentrations that prevent complete relaxation of the heart are achieved [90].

'Diamond' approach to treatment
Many patients with heart failure have not received the optimal treatment due to clinicians' insufficient awareness of the importance of drug combinations and excessive caution of adverse reactions it brings. Therefore, we proposed the diamond approach in this review to express the necessity of conjunction of therapies and exhibit possible combinations based on the latest clinical researches.
HF frequently have comorbidities such as hypotension, arrhythmias, and kidney disfunction. Considering that it may take months to prescribe all therapies recommended in the diamond approach, we believe the choice of optimal agents should vary along with different conditions and is worthy of attention. The following summarize the related studies and give some suggestions about personalized treatment in this regard.
Unless contraindicated or not tolerated, usage of the diamond approach in an early and comprehensive manner is crucial. In clinical practice, commonly step-up therapy of HF may reduce the benefit in patients with heart failure. Therefore, we combined the diamond approach and proposed a more aggressive treatment for HF (Fig. 2).

Special considerations 3.1.1 Hypotension
Patients with HF often had low BP. HF therapies such as ACEIs/ARBs, MRAs, and β-blockers lowered BP. ARNI had a stronger antihypertensive effect [91]. SGLT2 inhibitors also lowered BP through significant diuresis. In recent studies, it was found that SGLT2 inhibitors had a direct na-triuretic effect rather than promoting osmotic diuresis [92] and urine volume increased without an increase in urinary sodium over time [93]. A post-mortem analysis from the EMPA-REG OUTCOME trial found that empagliflozin reduced SBP (≥5 mmHg) in the fourth week of use [94]. Although ivabradine had little effect on hemodynamics, related studies did not included patients with hypotension. The use of ivabradine was not recommended in this setting [14].
The VICTORIA trial found that patients receiving vericiguat therapy had a higher incidence of hypotension and syncope, which was related to the drug's mechanism [81]. The GALACTIC-HF trial excluded patients with low blood pressure (SBP <85 mmHg), and found no deleterious effects of OM on BP [89]. In patients with HFrEF and hypotension, it was necessary to distinguish whether low perfusion exists. It should be noted whether patients are tolerant of the above HF drugs before initiating therapy if there is the pre-existing condition of low perfusion (e.g., dizziness, fatigue, cold limbs, oliguria).
For patients with hypotension but lack of evidence of low perfusion, usage of these disease-modifying drugs should be considered. For these patients, it is necessary to titrate from a small dose and strictly monitor BP and heart rate. There are often improvements in heart function and hypotension for HFrEF patients after a period of treatment. When hypotension occurs, the most important action to be taken is reducing unnecessary vasodilators and diuretics and then adjusting the above medications.

HF after acute myocardial infarction (AMI)
Myocardial ischemia, infarction, and scar formation caused by coronary heart disease were the most common causes of HF. A randomized controlled retrospective (PARADISE-SWEDEHEART) study performed in Sweden found that the incidence of HF in patients after MI was as high as 13-32%, and these HF patients were associated with higher morbidity and rehospitalization. Early, comprehensive, and standardized drug treatment largely determined the prognosis for these patients. ACEIs [95,96]/ARBs [28], MRAs [97], and β-blockers [98] reduced the mortality and hospitalization in patients after AMI. Unless contraindicated or not tolerated, the early use of ACEIs/ARBs [99] and MRAs [97,99] were universally recognized. Because the early use of β-blockers (<24 hours) increased cardiogenic shock and/or death, they should be initiated after hemodynamic stability in these patients [14,16,100]. A short-acting plain tablet of a β-blocker was preferred in this setting. By antagonizing the RAAS and strengthening the natriuretic peptides, ARNI had a stronger effect on the reversal of cardiac remodeling compared with the use of ACEIs/ARBs [47,48].
In the recent PARADISE-MI trial, it was found that when compared with an ACEI, ARNI reduced the primary endpoints (cardiovascular death, HF hospitalization, or outpatient development of HF) by 10% in patients with AMI, although statistical difference was not reached (P = 0.17). This study also observed that ARNI improved heart function more effectively because of its gradual action. Whether OM is benefical for AMI was not clear, since the GALACTIC-HF trial did not include relevant patients [89]. Because OM effectively increased cardiac contractility without increasing myocardial oxygen consumption and arrhythmia, it benefited patients after AMI. The specific effects of early use of ivabradine, SGLT2 inhibitors, and vericiguat on AMI remained unknown due to a lack of data.

CKD
CKD and HF often coexist, with CKD being present in 40-50% of chronic HF patients [101]. In addition to hemodynamic disturbances, the continuous activation of the SNS and RAAS also played a crucial role in CKD [102,103].
The key to treatment lied in breaking the vicious cycle between the neuroendocrine system and hemodynamic disorder. ACEIs/ARBs [26,[104][105][106][107]/ARNI [53,108] and MRAs [32,109] improved the prognosis of HFrEF with CKD. ARNI was a more optimal choice than ACEIs/ARBs for CKD because its superior cardiorenal effect has been proven. In the Kidney Disease Improving Global Outcomes (KDIGO) Controversies Conference, it was suggested that ARNI is an essential medicine for patients with HFrEF and CKD [110].
A meta-analysis including several randomized trials [111] found that β-blockers offer consistent benefits for patients with HFrEF and moderate CKD (eGFR: 30-60 mL/min/1.73 m 2 ). Although a study on ivabradine in CKD patients was lacking, ivabradine is considered safe in moderate CKD as a result of low renal clearance rate of 20% [112]. SGLT2 inhibitors caused a mild, short-term decrease in eGFR in the early stage, but it did not affect the long-term renal protective effect [57,60,61,113,114]. Vericiguat was beneficial for CKD because it promoted the production of cGMP. After a post-hoc analysis of the VICTORIA trial was performed, it was found that the beneficial effects of vericiguat on the primary endpoint were consistent across the full range of eGFR (>15 mL/min/1.73 m 2 ), irrespective of worsening renal function [115].
Out of consideration for renal insufficiency, diseasemodifying drugs including ACE inhibitors, ARBs, ARNI, and MRAs are usually underused. Stopping these drugs or administering a low-dose of these drugs always leads to poor outcomes. Unless contraindicated or not tolerated, all patients with HFrEF and CKD should be routinely and comprehensively receiving these therapies, and drugs need to be titrated to the maximum dosage under strict monitoring.

End-stage renal disease (ESRD)
Due to most clinical trials excluding patients with advanced CKD (eGFR <30 mL/min/1.73 m 2 ), there was a lack of strong evidence of safety and effectiveness for medical therapies that were used with these patients. However, some small non-randomized controlled studies have been conducted with these patients, and we can learn from them. β-blockers were found to be associated with lower morbidity and mortality in HFrEF patients with advanced CKD [116], and among β-blockers, carvedilol was the first recommendation for dialysis patients with HFrEF [117,118].
A study showed that ARNI increased the LVEF and was well tolerated in HFrEF with ESRD on dialysis [119]. It was found that in the DAPA-CKD [120] trial, dapagliflozin can delay the deterioration of renal function and reduce cardiovascular death in patients with CKD (eGFR: 25-75 mL/min/1.73 m 2 ). It is important to note the dialysis rate of drugs in patients with advanced CKD. Additional clinical studies are needed to guide the medical treatment of these patients.

Arrhythmia-induced cardiomyopathy (AiCM)
Known as reversible non-ischemic dilated cardiomyopathy, AiCM was caused by tachycardias, atrial fibrillation (AF), and premature ventricular contractions. The elimination of arrhythmia was the optimal treatment for AiCM, and it reversed cardiomyopathy. Because AiCM often occured with HF, therapies for HF should also be considered for AiCM [121]. Thus far, guidelines only recommend HF treatments for tachycardia-induced cardiomyopathy (T-CM). We suggested that the same therapy should be used for atrial fibrillation-induced cardiomyopathy (AF-CM) and premature ventricular contraction-induced cardiomyopathy (PVC-CM), because these patients were very likely to benefit from these disease-modifying drugs.
There were double benefits from the use of β-blockers because they prevented arrhythmia and antagonize the sympathetic (beta adrenergic) nervous system, and their use should be placed first in the medical treatment of HF. ARNI further reduced the occurrence of arrhythmias compared with ACEIs/ARBs [122]. In addition to decreasing cardiac remodeling, the extra anti-arrhythmic effect of ARNI might be derived from enkephalinase inhibitors [123]. Dapagliflozin was found to effectively reduced the incidence of atrial fibrillation/atrial flutter events by up to 19%, which may have resulted due to the improvement of cardiomyocyte metabolism by SGLT2 inhibitors [124]. The use of ivabradine should be avoided in non-sinus tachyarrhythmia. Although a few case reports and small studies found that ivabradine may be useful for treatment for AF, the associated clinical data are still lacking.

HF in different stages
The two HF classification schemes that are widely used are the American College of Cardiology Foundation/American Heart Association (ACCF, West End, Washington, D.C., USA/AHA, Dallas, Texas, USA) staging system and the New York Heart Association (NYHA, Dallas, Texas, USA) functional classification. The stage system mentioned by ACCF/AHA considers the development and progression of HF, whereas the NYHA classes focused on exercise capacity and the severity of symptoms of HF. According to the different classifications, the treatments for HF were also dynamic and subject to change. Different from the current guidelines [16,125,126], we proposed a much more aggressive treatment for HF.
Stage A HF exhibited no structural heart disease or HF symptoms. However, due to combined risk factors, these patients are vulnerable to heart failure. Thus, we held the same point of view as the guidelines in terms of stage A treatment, that was, risk factors that may lead to or contribute to HF should be modified [15].
Stage B HF was characterized by asymptomatic cardiac dysfunction, and the neuroendocrine system was activated at this stage [127]. It was previously demonstrated that ACEIs [128] and β-blockers [129] reduce the risk of HF and reverse ventricular remodeling in asymptomatic patients with reduced EF. The 2013 ACCF/AHA guidelines recommended the use of ACEIs/ARBs and β-blockers in stage B HF [15]. Although other disease-modifying drugs mentioned for use in the diamond program were not recommended in guideline [15], we believe that the early and comprehensive use of these drugs is reasonable and necessary to prevent ventricular remodeling and improve outcomes.
Patients in stage C have symptoms of heart failure based on structural heart disease. In this stage, recommended therapies include ACEIs/ARBs/ARNI, β-blockers, MRAs, SGLT2 inhibitors, and ivabradine [17]. It is more important to use disease-modifying drugs for these patients to prevent further deterioration of heart function, and therefore, increasingly aggressive use of vericiguat and OM is warranted.
Stage D patients were associated with worsening heart function and often required inotrope or device therapy. However, therapies that improve prognosis should be considered if patients can tolerate them, or even tolerate only a small dose. Drugs should be initiated at very low doses, and patients should be closely monitored for signs or symptoms of intolerance. Vericiguat and OM may exert a satisfactory effect on these patients due to their mechanisms of action.

Conclusions
The continuous update of HF drugs enables the use of numerous effective targeted therapies for patients with HFrEF, and brings hopes to further improve the outcome of these patients. How to use these drugs to maximize the benefits in these patients is what clinicians must consider. Based on evidence-based clinical data, the diamond approach tries to give the suggestions on most appropriate drugs for individual HF patient. Furthermore, we believe the key of HF treatment is to effectively prevent deterioration of HF. The 'diamond' approach proposed in this review did not focus only on the derivation of individualized and optimized treatments, but also conveyed the perspective of aggressive treatment for HFrEF.

Author contributions
HBG and HT wrote the manuscript with support from PP and ZZ. HBG, HT, YJH and DW revised the manuscript under PP and ZZ guidance. All authors read and approved the final manuscript.

Ethics approval and consent to participate
Not applicable.

Acknowledgment
Not applicable.