Refocusing on the Primary Prevention of Heart Failure

Abstract

Purpose of review

The burden of heart failure (HF) is a significant national and global public health problem, with prevalence rates on the rise. Given the significant morbidity, mortality, and healthcare costs attributable to HF, it is of utmost importance to utilize preventive strategies to prevent the development of HF. Therefore, we sought to address how a multi-modal risk assessment approach can be used to stratify patients for HF risk and guide implementation of therapeutic strategies to prevent HF.

Recent findings

New externally validated, multivariate prediction models for incident HF can be applied in the general population and may be used to aide clinicians in assessing individualized HF risk and screening for HF. Recent clinical trial data suggest a natriuretic peptide biomarker-based screening approach coupled with team-based cardiovascular care to focus on optimization of guideline-directed medical therapy may help prevent new-onset HF. However, widespread implementation of clinical risk scores and/or biomarkers is needed.

Summary

In addition to promoting a heart healthy lifestyle, prevention and management of modifiable risk factors, including intensive blood pressure lowering and use of sodium-glucose cotransporter-2 inhibitors, can prevent incident HF.

Introduction

An estimated 6 million Americans have heart failure (HF), with prevalence rates on the rise [1]. By 2030, projections estimate the prevalence of HF will increase by 46%, resulting in greater than 8 million adults with HF [2]. HF is a global public health problem as well, with an estimated 26 million cases of HF worldwide [3]. The overall lifetime risk for heart failure is estimated to range between 20 and 45% [1]. The economic impact of HF in the United States (US) is significant, with an estimated total cost for HF of $30 billion and is projected to increase to an estimated $70 billion by 2030 [1]. As outlined by national practice guidelines from the American College of Cardiology (ACC) and the American Heart Association (AHA), screening is essential for early identification of those with HF or at increased risk for HF, as evidence supports that the onset of HF can be delayed or prevented by targeting modifiable risk factors [4, 5]. However several barriers to screening exist, including poor understanding of HF among the general population and lack of widespread adoption of a HF risk assessment and screening protocol [5]. In contrast with coronary artery disease, incorporation of a risk-based assessment for heart failure at an early stage is not part of routine clinical practice. Externally validated, multivariate prediction models for incident heart failure that can be applied in the general population may be used to aide clinicians in screening for HF and assessing individualized HF risk. Furthermore there are emerging strategies for the primary prevention of heart failure, including natriuretic peptide screening protocols, evidence-based intensive blood pressure lowering, and utilization of novel therapies, such as sodium-glucose cotransporter-2 (SGLT2) inhibitors. This review will discuss how a multi-modal risk assessment approach can be used to stratify patients for HF risk to then guide implementation of therapeutic strategies to prevent HF (Fig. 1).

Fig 1.

Multi-modal risk assessment approach to estimate and reclassify heart failure risk in the general population based on known risk factors within the context of the American College of Cardiology and American Heart Association heart failure staging construct.

Risk assessment of and screening for heart failure

Assessment for modifiable heart failure risk factors

HF is largely attributable to cumulative exposure to modifiable risk factors over the life course. The ACC/AHA guidelines define 4 stages as a progression of those at risk for HF (Stage A and B), which may result in an asymptomatic period that may last years to decades. Progression to symptomatic HF is labeled Stage C and advanced end-stage HF, D [4]. Given increases in mortality with transition from Stage A to Stage B and Stage B to Stage C, intervention to prevent progression to symptomatic heart failure is critical.

The AHA defines ideals cardiovascular health (CVH) as both ideal health behaviors (nonsmoking, body mass index < 25 kg/m², physical activity at goal levels, and pursuit of a diet consistent with current guideline recommendations) and ideal health factors (untreated blood pressure < 120/< 80 mmHg, fasting blood glucose < 100 mg/dL, and untreated total cholesterol < 200 mg/dL). However, the prevelance of ideal cardiovascular in the US is very low [6, 7]. Traditional risk factors for HF are common, with up to one-third of the US adult population having at least one HF risk factor, including hypertension, elevated body mass index (BMI), physical inactivity, diabetes mellitus (DM), tobacco use, and coronary artery disease (CAD) [8]. When approaching HF risk assessment, it is essential to characterize CVH and assess for these modifiable HF risk factors.

Hypertension is highly prevalent and one of the most important modifiable risk factors for HF [9]. The Framingham Heart Study (FHS) showed that hypertension antedated development of HF in 91% of patients. The risk of developing HF in hypertensive patients in comparison with normotensive individuals was almost 2-fold greater in men and 3-fold greater in women [10]. Neurohormonal blockade with anti-hypertensive agents that block the renin-angiotensin-aldosterone system (RAAS) in particular may be an appealing strategy for risk reduction of HF compared with other blood pressure lowering agents. However, robust data supports the greatest impact on prevention results from the magnitude of BP lowering, independent of anti-hypertensive class. Similarly, DM increases the relative risk of HF by almost 2-fold in men and 4-fold in women [11,12,13,14]. Furthermore, poor glycemic control independently increases the risk of HF; with each 1% increase in hemoglobin A1c (HbA1c), there is an 8–36% increased risk of incident HF [15,16,17,18].

The relationship both of elevated BMI and physical inactivity with an increased risk of HF is also well-established [19,20,21,22,23]. Those with an elevated BMI have an earlier onset of HF than those with a normal BMI [19,20,21]. In the FHS, each increment of 1 kg/m² in BMI was associated with an increased risk of HF by 5% for men and 7% for women [23]. Furthermore, increasing BMI has been shown to have a dose-response linear relationship with HF with preserved ejection fraction (EF) [19]. Analogously, physical inactivity has an inverse dose-response relationship with HF risk [22].

Cigarette smoking increases the risk of incident HF, independent of coronary artery disease (CAD) status [24, 25]. Cigarette smoking has been linked with oxidative stress and deleterious inflammatory effects on the cardiovascular system [26, 27]. Additionally, CAD is a well-described risk factor for HF and shares a pathophysiologic relationship with the aforementioned risk factors in addition to dyslipidemia [28].

Heart failure risk assessment models

Several multi-variable models exist to estimate future risk for HF. Ideally these models allow for individualized HF risk assessments in patients at risk for HF (stage A) and to better identify those with asymptomatic structural heart disease (stage B) to implement targeted prevention strategies. These tools also play an essential role in investigational strategies for HF prevention.

The Health, Aging and Body Composition (Health ABC) Heart Failure Risk score is an externally validated risk prediction model for incident risk of HF in the elderly. This 9-variable model was developed in a cohort of 2935 people aged 70–79 years participating in the Health ABC Study in the US. The scoring system classifies patients into four groups of 5-year HF risk (5%, 5–10%, 10–20%, 20%) using the following variables: age, history of CAD, history of smoking, systolic blood pressure, heart rate, serum glucose, serum creatinine, serum albumin, and electrocardiographic left ventricular hypertrophy (LVH). While this model had good calibration (χ² 6.24, p = 0.621) and discrimination performance (C index of 0.73) across sex and race overall, the risk was underestimated in white men [29, 30].

The Heart “OMics” in AGEing (HOMAGE) study developed a 5 year incident HF risk prediction model in the elderly using a meta-analysis of 10,236 people from three cohorts (Health ABC Study, Valutazione della PREvalenza di DIsfunzione Cardiaca asinTOmatica e di scompenso cardiaco (PREDICTOR), and Prospective Study of Pravastatin in the Elderly at Risk (PROSPER)). The PREDICTOR trial included a population-based registry of 2001 people over the age of 65 years in four Italian cities. The PROSPER trial cohort included 5804 people over age 70 years with a history of vascular disease or high risk of developing vascular disease in Scotland, Ireland, and The Netherlands. The HOMAGE risk prediction model was externally validated using the Anglo-Scandinavian Cardiac Outcomes Trial (ASCOT), including a cohort of 13,121 high-risk hypertensive patients aged 40 to 80 years, in Scandinavian countries, the United Kingdom, and Ireland [31,32,33]. This study validated 10 variables to predict a 5-year risk of incident heart failure hospitalization (including age, sex, BMI, smoking, diabetes mellitus, history of CAD, antihypertensive medication use, systolic blood pressure, heart rate, and serum creatinine), and showed good discrimination (C index 0.71) and calibration (χ² 7.9, p = 0.54) [33].

Limitations of the Health ABC and HOMAGE models include non-generalizability to middle-aged and young adults. While several other multi-variable models exist to estimate future risk for heart failure (Table 1), incorporation of these models into clinical practice has not been adopted given these prior models are derived from non-contemporary cohorts with limited ethnic diversity and lack of extensive external validation in a population without prevalent cardiovascular disease [34, 35•].

Table 1.

Characteristics, Risk Factors, and Endpoints of HF Risk Scores

Risk score information

Risk factors/covariates/endpoint

Risk equation

Cohorts and countries

Participants’ race/ethnic Background

Participants’ Hx of CVD

Participant age range

Data collection years

Risk window (years)

Age

Sex

Race

Hx of CHD

SBP

BP Tx

PCP-HF

FOF, CHS, ARIC, CARDIA, MESA

European-American, African-American

Free of CHD and stroke at baseline

30–80 years

1985–2002

10

X

X

X

X

X

WATCH-DM

ACCORD, ALLHAT USA

European-American, Hispanic, African-American,

Included participants with hx of CHD

40–79 years

2001, 2003–2005

5

X

X

X

TIMI - HF

SAVOR-TIMI 53, DECLARE-TIMI 58

European, European-American, African-American, Hispanic

Included participants with hx of CHD

Over age 40 years

2010–2011

2.1

x

Risk score information

Risk factors/covariates/endpoint

Risk equation

Total Chol

LDL Chol

HDL Chol

DM

Fasting glucose

DM Tx

Smoking

BMI

Heart Rate

ECG

Biomarkers (e.g., albumin, BNP)

Endpoint

PCP-HF

X

X

X

X

X

X

X (QRS)

HF

WATCH-DM

X

X

X

X (QRS)

HF

TIMI - HF

The Pooled Cohort equations to Prevent Heart Failure (PCP-HF) is a race- and sex-specific 10-year risk equation for developing incident HF. This contemporary analysis utilized 33,010 people, aged 30 to 79 years from pooled individual-level data from 7 population-based cohorts with long-term follow-up. These included 5 large, racially and geographically diverse, National Heart, Lung, and Blood Institute–sponsored cohort studies for derivation, including the ARIC (Atherosclerosis Risk In Communities) study, CARDIA (Coronary Artery Risk Development in Young Adults) study, CHS (Cardiovascular Health Study), FOF (Framingham Heart Study Offspring Cohort), and MESA (Multi-Ethnic Study of Atherosclerosis). The study design for these cohorts has been previously described [36,37,38,39,40,41,42]. Age, BMI, systolic blood pressure, fasting blood glucose levels, cholesterol levels, smoking status, and QRS duration were identified as predictors of the risk of developing HF over 10 years. A web-based tool exists of this prediction model (available at pcphf.org). This model was further externally validated in the Prevention of Renal and Vascular End-stage Disease cohort and Jackson Heart Study where it demonstrated good-to-excellent discrimation (C index ranging from 0.71 to 0.88 in white and black men and women), and strong calibration (χ² < 20 for all sex-race groups). Overall, the PCP-HF tool has excellent discrimination and calibration among a diverse group of US and European adults enhancing its broad generalizability and may be a valuable tool in risk assessment of HF on both an individual and population-based level [35•].

To focus on participants with diabetes who are at heightened risk, the WATCH-DM risk score (Weight [BMI], Age, hyperTension, Creatinine, HDL-C, Diabetes control [fasting plasma glucose], QRS Duration, MI, and CABG) leveraged machine learning to develop a ten clinical variable score to predict risk of incident HF hospitalization using clinical trial data. This risk score was developed using a cohort 8756 participants from the Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial, and was externally validated in a cohort of individuals with diabetes mellitus in the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT). The study design for these trials has been previously described [43, 44]. This risk score utilizes ten variables routinely obtained for patients with DM including: age, BMI, systolic and diastolic blood pressure, fasting plasma glucose, high-density lipoprotein level, serum creatinine, QRS duration, and history of prior myocardial infarction or CABG. The WATCH-DM Risk score provides a 5-year risk assessment for incident heart failure in patients with diabetes and also provides an integer-based risk score (very low, low, average, high, or very high risk) for ease of integration in clinical use, with good discrimination (C index 0.72) and calibration (χ² 10.58, p = 0.23) [45•].

The TIMI Risk Score for Heart Failure in Diabetes (TRS-HF DM) is an externally validated risk score to predict hospitalization for heart failure (HHF) in patients with DM. This tool was derived using 8212 patients from the placebo arm of the SAVOR-TIMI 53 trial (Saxagliptin Assessment of Vascular Outcomes Recorded in Patients With Diabetes Mellitus–Thrombolysis in Myocardial Infarction 53) and externally validated in the placebo-treated patients in the DECLARE-TIMI 58 trial (Dapagliflozin Effect on Cardiovascular Events– Thrombolysis in Myocardial Infarction 58). The study design of these trials has been previously described [46, 47]. Five clinical variables were indentified to predict the risk of HHF, including prior heart failure, history of atrial fibrillation, coronary artery disease, estimated glomerular filtration rate, and urine albumin-to-creatinine ratio. The TRS-HF DM score provides an integer-based for HHF (low, intermediate, high, and very high risk) with good discrimination (C index 0.78 to 0.81) and calibration (χ² 4.64, p = 0.20) [47].

Biomarker screening

The current ACC/AHA HF Guidelines list a class IIa recommendation to obtain a screening natriuretic peptide in those at risk for HF [4]. This recommendation was based on evidence from the St Vincent’s Screening to Prevent Heart Failure (STOP-HF) and PONTIAC trials, which showed natriuretic peptide screening paired with early intervention prevents the development of HF. The STOP-HF Trial was a single-center, unblinded study of 1374 people at risk for HF in Ireland (with hypertension, diabetes mellitus, or known vascular disease) without know structural heart disease or symptoms of HF who were randomized to B-type natriuretic peptide (BNP) screening or standard of care. Groups in the intervention arm with a BNP level > 50 pg/ml underwent echocardiography and referral to a team-based specialist cardiovascular service. All patients received coaching by a specialist nurse who emphasized individual risk status and the importance of adherence to medication and healthy lifestyle behaviors. The BNP-based screening group had reduction in the primary endpoint of asymptomatic LV dysfunction with or without newly diagnosed heart failure [48•]. In PONTIAC, 300 patients with diabetes mellitus type 2 without a history of cardiac disease and a N-terminal pro b-type natriuretic peptide (NTproBNP) > 125 pg/ml were randomized to individualized visits for the up-titration of renin-angiotensin system (RAS) antagonists and beta-blockers versus usual care. The treatment group had a reduction in the composite endpoint of hospitalization or death due to cardiac disease [49].

Electrocardiography (ECG) screening

Abnormal baseline ECG abnormalities have been shown to be predictive of future HF events. In a middle-aged adult cohort without cardiovascular disease (MESA), a QRS duration greater than 100 ms was significantly associated with incident HF suggesting that ECG abnormalities may identify left ventricular structural changes in asymptomatic patients [50]. An analysis of 2915 older adults participating in the Health ABC study showed baseline and new ECG abnormalities are independently associated with increased risk of incident HF after adjusting for clinical HF risk factors. Minor abnormalities minor were ST segment or T-wave abnormalities. Criteria for major ECG abnormalities included Q-QS wave abnormalities, major ST-T abnormalities, LVH, atrial fibrillation or atrial flutter, Wolff-Parkinson-White, complete bundle-branch block, or intraventricular block. The risk of heart failure increased significantly with the increasing severity of ECG abnormalities [51]. While increased QRS duration has been incorporated into contemporary pooled-cohort heart failure risk calculators, identification of more severe ECG abnormalities may prompt consideration of assessment with an echocardiogram to evaluate for asymptomatic structural heart disease.

Echocardiogram screening

An abnormal ECG or elevated natriuretic peptide level should prompt consideration for an echocardiogram to evaluate for asymptomatic left ventricular dysfunction. While evidence to support community-wide echocardiographic screening for asymptomatic left ventricular dysfunction is inadequate [52, 53], some evidence suggests a screening echocardiogram may benefit individuals at increased risk for HF.

In a community-based screening study in 438 asymptomatic adults over age 65 with at least one HF risk factor used clinical characteristics and functional capacity, followed by ECG abnormalities to define low- and high-risk groups for new HF. Those identified to be high risk for HF were more likely to have an abnormal echocardiographic findings, and therefore may potentially benefit from cardioprotective medications [54]. Furthermore, a screening echocardiogram protocol in 5227 adults over age 60 recruited from primary care clinics found increased detection of both Stage B and Stage C HF, with subsequent increase in use of cardioprotective medications (including ace-inhibitors, beta-blockers) in these groups [55]. However, data supporting that screening echocardiographic assessments improve HF and cardiovascular outcomes is lacking.

Assessment for heart failure risk enhancers

One limitation of relying wholly on multi-variable HF risk assessment tools is lack of inclusion of some known risk enhancers for HF in these tools. It is important to consider these risk-enhancing comorbidities to personalize HF risk screening assessment, as these risk prediction tools may underestimate risk. In addition to traditional risk factors, conditions to consider include adverse pregnancy complications, systemic inflammatory conditions, atrial fibrillation, history of treatment for cancer, chronic kidney disease, liver disease, and sleep disorders.

Complications of pregnancy including pre-eclampsia, gestational hypertension, gestational diabetes, and premature delivery portend an increased risk of future heart failure [56, 57]. Pre-eclampsia has been linked with a four-fold increased risk of incident heart failure [58•]. Systemic inflammatory conditions including connective tissue disorders and human immunodeficiency virus (HIV) are also associated with an increased risk for HF patients with rheumatoid arthritis (RA) are estimated to be at twice the increased risk of incident HF [59, 60]. RA enhances the risk of HF from both ischemic and non-ischemic causes, with associations between increased disease activity and incident non-ischemic HF [61]. Patients with lupus are at an increased risk of HF, attributed to accelerated atherosclerosis and higher cardiovascular risk factor disease burden (i.e., hypertension) in comparison with those without lupus [60, 62, 63]. HIV is an independently associated risk factor for heart failure [64]. In comparing people living with HIV to non-HIV infected controls, the reported relative risk of heart failure and diastolic dysfunction in those with HIV was 1.7 and 3.0, respectively [65]. Atrial fibrillation (AF) is a known risk factor for heart failure, with some reports of a 3-to-4 times higher incidence of heart failure in those who have AF in comparison with those without AF [66,67,68,69]. Furthermore, permanent AF has been shown to portend higher risk of incident HF in comparison with paroxysmal AF [70]. The inter-relationship between AF and HF is complex, particularly given these conditions often co-exist with shared risk factors and pathophysiology.

Cancer treatments including anthracyclines, human epidermal growth factor receptor 2 antagonists and chest radiation portended a risk for both early-onset and late-onset HF [71,72,73,74]. Other disease states linked with increased incidence of HF include chronic kidney disease [75,76,77], non-alcoholic fatty liver disease [78], and obstructive sleep apnea [79]. An important component of a HF risk assessment is family history. Obtaining a 3-generation family history of HF is recommended [80]. If the family history is concerning for a genetic cardiomyopathy, the patient’s individual HF risk should not be assessed only using pooled cohort data. Instead, consider referral to genetic counseling, genetic testing, and further cardiac testing [80, 81].

Treatments aimed at heart failure prevention

Once a HF risk assessment is complete, risk stratification should help drive the management for that patient. In all patients, including those without HF risk factors, the cornerstone of the discussion should be counseling on the importance of a heart healthy lifestyle to optimize cardiovascular health to reduce the lifetime risk of incident HF. In those who are at increased risk for HF, additional preventive strategies should be targeted at optimizing modifiable risk factors.

Lifestyle

The AHA has defined ideal cardiovascular health based on seven metrics that compose Life’s Simple 7, which include not smoking, maintaining an optimal BMI, physical activity at goal levels, healthy diet, optimize cholesterol, reduce blood glucose levels, and control blood pressure [73]. Greater adherence to Life’s Simple 7 has been associated with a lower lifetime risk of heart failure and greater preservation of cardiac structure and function [82, 83].

Evidence suggests adherence to a Mediterranean or DASH-type (Dietary Approaches to Stop Hypertension) diet may protect against new onset of HF [84, 85•]. Data from a large, diverse, and contemporary US population without known CAD found adherence to a plant-based dietary pattern was associated with a 41% lower risk of incident HF hospitalization; in contrast, a southern-based diet was associated with a 72% higher risk of HF hospitalization [86].

Increased physical activity is associated with lower incidence of HF. Data from the Physicians Healthy Study [87] and the Women’s Health initiative [88], both large-cohort community-based studies demonstrated relative high levels of physical activity are associated with a lower risk of HF [89]. A prospective study of 137,303 women aged 50 to 79 years showed that higher levels of self-reported total recreational activity and higher levels of walking are associated with a lower risk of developing HF [90]. Furthermore, data from the California Men’s Health Study, a cohort of 82,695 men over age 45 years without HF followed for 10 years showed increased physical activity and lower sedentary time were associated with lower HF risk [91].

Obesity prevention and management

In patients with obesity, weight loss has been shown to reduce the incidence of HF. Clinicians should integrate weight loss treatment in patient’s care plans, including lifestyle, pharmacologic, and surgical weight loss strategies. In a study of nearly 40,000 obese people without heart failure in a Scandinavian registry, gastric bypass surgery was associated with approximately half the incidence of HF in comparison with intensive lifestyle treatment. These effects were largely mediated by weight loss, with a 22.6 kg more weight loss in the surgery group at 2 years, and appeared to be partially mediated by effects of treating atrial fibrillation, diabetes mellitus, and hypertension [92].

While overweight and obesity are associated with greater incidence of HF, the findings of an “obesity parodox” in patients with HF, whereby the association of obesity is associated with longer survival, has raised concern about intentional weight loss following disease onset. It is important to clarify the other potential drivers of lower BMI that may occur due to frailty, increased catabolic state, or lack of metabolic reserve resulting in weight loss. Therefore, it is not that obesity is protective, but the severity of illness may contribute to the so-called “obesity paradox” [21, 93,94,95]. Ultimately, focus on maintenance of a normal body weight would allow for promotion of CVH and promotion of a healthier, longer life free of HF.

Hypertension management

Controlling hypertension is associated with a lower risk of incident HF [8]. Current ACC/AHA guidelines recommend targeting a blood pressure of < 130/80 mmHg in patients at increased risk of HF (stage A) [10]. Secondary analysis of the Systolic Blood Pressure Intervention Trial (SPRINT) demonstrated targeting a systolic blood pressure of less than 120 mmHg is associated with reduced incidence of HF. In this population of patients with hypertension at increased risk of cardiovascular disease but without diabetes, intensive blood pressure lowering by targeting a systolic blood pressure of < 120 mmHg in comparison with targeting < 140 mmHg was associated with a significant reduction in new HF. However the rates of adverse events (AF) were higher in the intensive blood pressure group, including hypotension, syncope, electrolyte abnormalities and acute kidney injury [96, 97, 98•]. One approach to better inform who would benefit from more intensive blood pressure lowering would be the use of risk stratification using multivariate risk modeling. A recent post hoc analysis using SPRINT data applied the PCP-HF prediction model to stratify the SPRINT cohort into three tertiles of 10-year incident HF risk (low, intermediate, or high). In the high predicted HF risk group, there was a significant reduction in incident HF in the intensive BP lowering arm. However, this benefit of an intensive blood pressure target was not observed within low and intermediate heart failure risk group. In the intensive versus standard BP treatment arms, serious adverse events occurred in 40.9 vs. 38.7% in the high predicted HF risk group, respectively. This data suggests intensive BP lowering for HF prevention is most beneficial to individuals at high predicted risk of HF and minimizes unnecessary AE in those individuals at lower risk of HF [99].

Glucose-lowering therapy

Currently, limited data are available to guide targeted hemoglobin A1c goals in patients with type-2 diabetes mellitus to reduce the incidence of HF [100]. Sodium-glucose cotransporter-2 (SGLT2) inhibitors are the first class of glucose-lowering therapies that have been shown to reduce the risk of HF hospitalization in patients with DM [47, 100,101,102]. SGLT2 inhibitors have been shown to have a similar relative risk reduction of hospitalization for heart failure (HHF) across a wide spectrum of baseline cardiovascular risk [48, 103]. Applying the Health ABC HF Risk Score to stratify patients with DM and without HF by baseline 5-year HF risk demonstrated the use of a SGLT2 inhibitor reduced cardiovascular mortality and HF hospitalizations in patients at both high and lower risk of HF [104]. Futhermore, utilizing the TRS-HF DM to assess baseline risk for HHF in patients with DM, those with the highest baseline risk for HHF derive the greatest absolute risk reduction of HHF from SGLT2 inhibitors [48•].

Notably, the majority of patients in cardiovascular outcome trials for SLGT2 inhibitors did not have HF at baseline. Therefore, the reductions in heart failure events provide compelling evidence for the benefit of SLGT2 inhibitors in the primary prevention of incident HF among patients with diabetes, irrespective of glycemic control [103]. The Dapagliflozin and Prevention of Adverse Outcomes in Heart Failure (DAPA-HF) Trial randomized patients with NYHA Class II-IV HF with an EF < 40% HFrEF to dapaglifozin or placebo, in those with and without T2DM. The risk of worsening HF or death from cardiovascular causes was lower in those who received dapagliflozin than among those who received placebo, regardless of the presence or absence of diabetes [105]. These data suggest SGLT2 inhibitors may have a therapeutic role to prevent incident HF in both those with and without diabetes.

Conclusion

Externally validated, multivariate risk assessment of HF may be incorporated into routine clinical practice for personalized risk assessment and targeted prevention strategies. Using a step-wise approach, once a patient is identified at increased risk, practitioners may consider a biomarker assessment (i.e., natriuretic peptide) or cardiac testing (i.e., echocardiogram). Furthermore, data suggests a natriuretic peptide biomarker-based screening coupled with optimization of guideline-directed medical therapy can help prevent new-onset HF. Utilization of these tools may improve identification of those (a) with stage B HF to implement guideline-directed medical therapy or (b) moderate-to-high HF risk individuals who would benefit from targeted primary prevention strategies (i.e., lower blood pressure targets or, initiation of SGLT2 inhibitors, or neurohormonal blockade). Further studies are need to develop a cost-effective strategy for further cardiac testing in those identified at increased risk of HF. Despite the promising accumulating data on risk prediction for HF, categorizing HF as a single entity is an important limitation that must be acknowledged. HF is a heterogenous syndrome with different subtypes that may require consideration of different predictors and further personalization with improved phenotyping of HF risk.

In the future, more widespread utilization of genetic testing may be incorporated for a more precision medicine-focused approach to HF risk assessment. Advances in knowledge about the genetic impact on HF risk are growing rapidly [106•]. Increased accessibility to and reduced cost of an individual’s genomics data may lead to identification of new risk-enhancing genes in asymptomatic patients. For example, truncating variants in the Titin gene (TTNtvs), a known pathologic variant in familial and sporadic dilated cardiomyopathies [107], have been recently associated with risk for several other non-ischemic cardiomyopathies, including peripartum cardiomyopathy [108], alcohol-induced cardiomyoatphy [109], and cancer therapy–induced cardiomyopathy [110]. In the general population, in comparsion with noncarriers, asymptomic carriers of TTNtvs are at a substantially increased relative risk of having a subnormal ejection fraction and developing HF in the future [111]. However, not all carriers of TTNtvs go on to develop HF; therefore, further data is needed understand the determinants of disease expression to guide HF preventation strategies in asymptomatic TTNtvs carriers [111, 112]. In addition to genomics, other rapidly expanding HF “omics” data (i.e., metabolomics, pharmacogenomics, epigenomics, proteomics, microbiomics) are promising to enhance understanding of HF pathophysiology, characterization of HF sub-phenotypes, and development novel HF therapeutics, but the additional clinical predictive value of this data is not yet clear [113]. These data will need to be combined with current models for a more personalized risk assessment and consideration of therapeutic strategies.

Furthermore, novel systems of care approach for HF risk assessment and screening are needed to allow more widespread delivery of effective, efficient strategies focused on HF prevention. Get With the Guidelines – Heart Failure (GWG-HF) is an voluntary, hospital-based quality initiative to improve adherence to guideline driven in-hospital treatment of heart failure. Data has shown since implementation of GWG-HF, hospitals participating in the GWG-HF program have had improvements in their processes of care for patients with HF [114]. Further development of systems of care–based programs targeted at heart failure prevention are urgently needed to improve implementation of HF prevention in clinical practice.

Ultimately, the most important way to prevent HF is to promote a heart healthy lifestyle from birth to older adulthood. A comprehensive patient-centered lifecourse approach that accounts for maintenance of optimal CVH (including healthy diet, 150 min of moderate to vigorous physical activity, avoidance of tobacco exposure, maintenance of a normal BMI) is needed to reduce individual-level and population-level morbidity and mortality related to HF [115].