Heart failure after myocardial infarction: incidence and predictors

Abstract

Aims

The aim of the present paper was to provide an up‐to‐date view on epidemiology and risk factors of heart failure (HF) development after myocardial infarction.

Methods and results

Based on literature review, several clinical risk factors and biochemical, genetic, and imaging biomarkers were identified to predict the risk of HF development after myocardial infarction.

Conclusions

Heart failure is still a frequent complication of myocardial infarction. Timely identification of subjects at risk for HF development using a multimodality approach, and early initiation of guideline‐directed HF therapy in these patients, can decrease the HF burden.

Introduction

Despite the remarkable advances in the treatment of coronary artery disease and acute myocardial infarction (MI) over the past two decades, MI remains the most common cause of heart failure (HF). ¹ According to the time sequence of MI occurrence and HF development, three clinical presentations differing in pathophysiology, clinical characteristics, and outcomes can be identified: (i) HF onset at the time of MI presentation, (ii) HF developing during hospitalization for MI, and (iii) HF onset after discharge from the index hospitalization.

Heart failure developing at the time of myocardial infarction hospitalization

The factors that contribute to the pathogenesis of HF development at the time of the MI hospitalization include myocardial compromise due to myocardial necrosis, myocardial stunning, and mechanical complications such as papillary muscle rupture, ventricular septal defect, and ventricular free wall rupture. Within 30 min of ischaemia, cardiomyocyte structural changes and oedema develop, leading to progressive myocyte death after 3 h of ischaemia. Reperfusion itself causes a second wave of injury through the production of reactive oxygen species. Despite successful epicardial reperfusion, the embolization of thrombotic debris leads to ongoing microvascular dysfunction and myocardial ischaemia. ² The inflammatory response to myocyte death also contributes to HF development. Furthermore, HF at this stage can be also triggered by exacerbation of pre‐existing HF and co‐morbidities, for example, anaemia, chronic kidney disease (CKD), or chronic obstructive pulmonary disease.

Opposing trajectories of HF incidence presenting at MI admission and during a hospital stay have been observed in the last decades. While the proportion of HF cases at MI admission has increased (from 4% in 1992–1996 ³ to 12–13% in 2001–2011 ⁴ , ⁵ ), the proportion of HF cases developing during the hospital stay has decreased (from 39% ³ to 4–28% ⁴ , ⁵ , ⁶ ). The increase in HF at MI presentation may be explained by recent improvements in pre‐hospital care, which led to a decrease in out of hospital mortality. ⁷ , ⁸ On the other hand, the decrease in in‐hospital HF may be caused by the introduction of percutaneous coronary intervention (PCI), which leads to more substantial myocardial salvage as compared with thrombolysis. In the nationwide SWEDEHEART registry, ⁶ the incidence of in‐hospital HF complicating MI has fallen from 46% in the thrombolytic era (the year 1996) to 28% in PCI era (the year 2008) (Figure 1 ). Higher myocardial salvage by PCI can also explain the increase in the proportion of patients with HF with preserved ejection fraction, which increased from 18% in 1998 to 30% in 2008. The second explanation for the in‐hospital HF decrease may be the change in MI diagnosis, which is currently based on troponin level and allows the detection of less severe MI cases with a lower risk of HF development.

Figure 1.

Percentage of patients with index myocardial infarction undergoing percutaneous coronary intervention (PCI) (orange line) and with in‐hospital heart failure (HF) (blue line)—adapted from SWEDEHEART study. ⁶

Heart failure developing after myocardial infarction hospitalization

Heart failure developing after MI hospitalization is a consequence of cardiomyocyte death and scar formation, which triggers chronic neurohumoral activation (renin–angiotensin–aldosterone and sympathetic nervous system up‐regulation) and ventricular remodelling. Left ventricular (LV) remodelling is more pronounced in men, patients with larger infarct size, and late or unsuccessful reperfusion of epicardial or microvascular bed. ⁹ Ventricular remodelling changes ventricular geometry and leads to wall thinning, ischaemic mitral regurgitation, and further cardiomyocyte loss.

Heart failure development after hospital discharge is very prevalent. It is diagnosed in approximately 13% of patients at 30 days and 20–30% at 1 year after discharge for MI. ¹⁰ , ¹¹ The incidence of HF after MI discharge is highest in the first months, and then it drops and remains stable at a rate of 1.3–2.2% per year afterwards. ¹¹

Clinical impact of heart failure after myocardial infarction

The development of HF after MI has a significant impact on outcomes, regardless of the HF type. ¹² Among patients with a history of MI, HF development increases total mortality risk three‐fold and cardiovascular mortality four‐fold. The timing of HF development also has an impact on adverse events. HF developing more than 3 days after MI is associated with a 43% higher mortality risk as compared with patients with HF developing in the first 3 days after MI. ¹² This may be explained by different risk factors and mechanisms leading to HF at different time points.

The need for heart failure screening and prevention

As recognized by both the European Society of Cardiology ¹³ and the American Heart Association, ¹⁴ HF prevention is an urgent public health need. The population of patients after MI represent a high‐risk group for HF development, in which HF screening and prevention is of particular importance. Missed or delayed diagnosis of HF compromises patient prognosis and increases therapy costs. This underscores the need for close follow‐up of MI patients at risk for HF development, which has been shown to result in improved patient adherence, higher prescription of recommended therapy, and lower cardiovascular hospitalizations. ¹⁵ , ¹⁶ , ¹⁷

The present paper reviews risk factors and biomarkers associated with HF development after MI and suggests a multimodality approach for screening (Figure 2 ). This information is intended to assist clinicians in identifying patients at particular risk of HF development after MI and early initiation of guideline‐directed HF therapy in these patients at high risk of adverse clinical events.

Figure 2.

Multimodal approach for prediction of heart failure after myocardial infarction (MI).

Clinical risk factors

The impact of different clinical risk factors on the HF risk after MI is shown in Table 1 .

Table 1.

Clinical risk factors for HF

Clinical risk factors	Increase in risk of post‐MI HF
Age, increase by 10 years	20–50%
Female sex	15–34%
History of previous MI	21–89%
Hypertension	7–70%
Diabetes	30–42%
Glomerular filtration, decrease by 10 mL/min/1.73 m2	10%
Heart rate, increase by 10 b.p.m.	7–23%
Atrial fibrillation	20–51%

HF, heart failure; MI, myocardial infarction.

Age

The incidence of in‐hospital HF is three times higher in patients 75–85 years old as compared with those 25–54 years old. After hospital discharge, HF incidence is six times higher in the older age group. ¹¹ After multivariate adjustment, the in‐hospital HF risk increases by approximately 50% ⁵ and post‐discharge HF by 20–50% ¹⁸ , ¹⁹ , ²⁰ for every 10 years of age.

Gender

Female sex was found to be independently associated with increased HF risk after MI in some studies, but not in others. ¹⁸ , ¹⁹ , ²¹ , ²² , ²³ In the studies that reported higher risk associated with female sex, the excess HF risk ranged from 15% to 34%. ⁶ , ²⁰ , ²⁴

Several reasons may explain higher HF risk in women. Compared with men, female patients presenting with MI are older and have a higher prevalence of co‐morbidities and worse functional status. ²⁵ The impact of co‐morbidities such as diabetes, hypertriglyceridemia, and metabolic syndrome on cardiovascular risk appears to be higher in women than men. Furthermore, gender disparities in MI presentation ²⁶ , ²⁷ and less aggressive hospital care of female patients, ²⁷ including the underuse of revascularization, may further contribute to the higher HF risk in women. ²⁸

Number and location of infarct‐related artery

Multi‐vessel disease (MVD) reflects the high atherosclerotic burden with more prominent endothelial dysfunction and systemic inflammation. ²⁹ Patients with MVD are generally older and have diabetes and renal impairment as common co‐morbidities. MVD is associated with lower ejection fraction ²² , ²⁹ and increased risk of major adverse cardiovascular events (MACE), including HF, by 80%. ²⁹ Anterior MI is associated with a higher risk of adverse remodelling ³⁰ and HF. ³¹ The higher risk of HF associated with anterior MI is caused by the greater magnitude of irreversible LV damage, as compared with other MI locations. ³²

Prior myocardial infarction

A history of MI increases the risk of HF by 21–89%. ⁵ , ⁶ , ¹⁸ , ²¹ , ²⁴ , ³³ , ³⁴ Excess risk may be explained by pre‐existing systolic and/or diastolic dysfunction.

Arterial hypertension

Many studies reported that arterial hypertension increases the risk of HF. The excess risk associated with arterial hypertension ranged from 7% to 70%. ⁶ , ¹⁸ , ²⁰ , ²¹ More common microvascular injury and myocardial haemorrhage contribute to the excess HF risk in patients with arterial hypertension. ³⁵ Furthermore, higher neurohormonal activation and more common LV remodelling was described in hypertensive patients after MI. ³⁶

Higher heart rate

A higher heart rate at admission was a risk factor for HF after acute MI in several studies. ⁵ , ¹⁸ , ²¹ , ²² , ³³ The risk rises by 7–23% for every 10 beats. ⁵ , ²¹ , ²² Tachycardia may reflect MI severity and imminent cardiac dysfunction.

Atrial fibrillation

New‐onset atrial fibrillation complicates 2–21% cases of MI and may reflect left atrial pressure increase and atrial fluid overload during MI. ³⁷ Atrial fibrillation increases the risk of HF after MI by 20–51%. ¹⁸ , ³³

Diabetes

After MI, the incidence of HF among diabetic patients is 60–70% higher than in patients without diabetes. ³⁸ After accounting for other co‐morbidities associated with diabetes, its presence still results in 30–42% higher risk of HF after MI. ⁵ , ⁶ , ³⁹ , ⁴⁰ Compared with non‐diabetic patients with similar infarct size, ⁴¹ similar systolic function, and infarct‐related coronary artery patency rate, ⁴² diabetic patients develop more often adverse LV remodelling and HF. ⁴³ This may be explained by a more common microvascular obstruction ⁴² and diastolic dysfunction in those with diabetes. ⁴⁴ The excess risk seems to be similar in patients with pre‐existing diabetes and diabetes diagnosed at the time of MI. ⁴¹ , ⁴⁵

Chronic kidney disease

Chronic kidney disease increases the risk of HF development after MI approximately two‐fold. ²⁰ Excess HF risk in CKD can be explained by accelerated atherosclerosis, more common MVD, atypical MI presentation, and lower odds of revascularization, which results in larger infarct size and more severe ventricular dysfunction. Moreover, CKD leads to fluid overload, secondary hypertension, anaemia, chronic inflammation, and alterations of the renin–angiotensin–aldosterone system. ⁴⁶ , ⁴⁷ Lower prescription of evidence‐based medications in CKD patients has also been reported. ⁴⁸

Ischaemic preconditioning and heart failure after myocardial infarction

Preconditioning is the process by which brief, repetitive episodes of ischaemia reduce the size of a subsequent MI. Compared with those without antecedent angina, patients with angina have a decreased risk of HF development during MI, lower risk of mortality or adverse LV remodelling after MI, and enhanced recovery of cardiac contractile function after MI. ⁴⁹ , ⁵⁰ The protective effect of angina has been described up to 3 months prior to MI. ⁵¹ The difference in outcomes in patients with and without angina preceding MI may be explained by ischaemic preconditioning or a larger extent of collateral circulation in patients with antecedent angina.

Whether the ischaemic preconditioning data could be applied to clinical care to improve outcomes remains to be seen. Remote ischaemic conditioning with transient ischaemia and reperfusion of the arm or leg has been shown in several small randomized controlled trials to reduce myocardial infarct size and increase myocardial salvage in patients with ST‐segment elevation MI (STEMI). ⁵² , ⁵³ However, in the recent large CONDI‐2 trail among 5401 STEMI patients, remote ischaemic conditioning by intermittent ischaemia and reperfusion applied to the arm did not decrease the risk of death or hospitalization for HF. ⁵⁴

Coronavirus disease 2019

Coronavirus disease 2019 (COVID‐19) caused by severe acute respiratory syndrome coronavirus 2 likely affects the risk of HF development after MI. However, hard clinical data are still lacking. There are several mechanisms by which COVID‐19 pandemic may influence the risk of HF development after MI.

First, the incidence of acute coronary syndrome (ACS) increases in the setting of viral infection, likely due to inflammation‐mediated plaque destabilization. The risk in the setting of COVID‐19 infection is unknown, but other viruses are associated with a 3‐fold to 10‐fold increased risk. ⁵⁵ However, a decrease in MI hospitalizations was observed in several countries, ⁵⁶ , ⁵⁷ , ⁵⁸ with a parallel increase in fatality and complication rates. Also, both patient‐related and system‐related delays were noted, with a 40% increase in symptom onset to coronary angiography time during the COVID‐19 pandemic. Thus, the absence or delay in coronary revascularization during MI may increase the proportion of patients with HF.

Second, there is growing evidence that COVID‐19 leads to direct myocardial injury. Within 24 h from admission for COVID‐19, troponin elevation is present in 36% of patients. ⁵⁹ Among recovering patients evaluated a mean of 71 days after confirmed COVID‐19 diagnosis, 78% of patients have demonstrable cardiac involvement via cardiac magnetic resonance imaging (MRI), 76% have detectable high‐sensitivity troponin, and 60% have evidence of active myocardial inflammation. ⁶⁰ The mechanisms responsible for myocardial injury during COVID‐19 infection may include inflammation, cytokine storm, hypercoagulable state with formation of microthrombi and macrothrombi, direct viral invasion of the myocardium, and myocardial supply/demand imbalance. ⁵⁹ MI in the setting of COVID‐19 infection likely increases not only the mortality risk ⁵⁶ but also the risk of subsequent HF development.

Biochemical markers

Biomarkers of infarct size

Cardiac troponin, a biomarker of choice in MI diagnostics, measured at plateau phase (48–72 h after MI symptom onset) is associated with MRI determined infarct size. ⁶¹ Similarly, peak levels of creatine kinase (CK) and CK‐MB are associated with infarct size on single‐photon emission computed tomography. ⁶² Several studies have shown the association of troponin or CK‐MB level with MACE, including HF. ¹⁹ , ⁶³ , ⁶⁴ Yet the association of peak troponin or CK‐MB with HF has not been seen in all investigations. ⁶⁵ , ⁶⁶

Natriuretic peptides

Alongside troponin, natriuretic peptides are associated with infarct size and cardiac dysfunction. ⁶⁷ , ⁶⁸ In addition to the magnitude of natriuretic peptides elevation, ⁶⁶ , ⁶⁹ its pattern is also associated with adverse events. While in some patients the brain natriuretic peptide increase after MI has a monophasic pattern with a peak at 16 h after admission, in others, the rise is biphasic with a second peak at 5 days. Patients demonstrating the biphasic pattern have a higher risk of LV remodelling and HF. ⁷⁰ , ⁷¹ Additionally, premorbid N‐terminal prohormone brain natriuretic peptide levels, as well as high‐sensitivity troponin T levels measured at a median time of 6 years before MI, have also been associated with adverse events, including HF. ⁷²

Inflammation markers

There is growing evidence that prolongation or expansion of the post‐infarction inflammatory response significantly contributes to LV remodelling and HF development. ⁷³ Numerous methods of inflammatory response quantification have shown promise in HF prediction. C‐reactive protein level predicted the risk of adverse events after MI, including HF, in several studies. ⁷⁴ , ⁷⁵ , ⁷⁶ , ⁷⁷ The neutrophil‐to‐lymphocyte ratio, an indicator of systemic inflammation, predicted MACE and HF in a meta‐analysis of 14 studies. ⁷⁸

Cytokines are strategic regulators of inflammation. In a study of 4939 patients with ACS, pro‐inflammatory cytokine interleukin 6 (IL‐6) was an independent predictor of MACE and HF. ⁷⁹ IL‐32 is a relatively novel pro‐inflammatory cytokine that induces the release of other inflammatory cytokines such as tumour necrosis factor‐α, IL‐1β, IL‐6, IL‐8, and IL‐18. Xuan et al. showed that IL‐32 is an independent predictor of cardiac death and HF among patients after MI. ⁸⁰

Renal biomarkers

The estimated glomerular filtration rate (eGFR) is independently associated with HF risk after MI. Fox et al. used a creatinine‐based MDRD equation for eGFR calculation to show that, after multivariate adjustments, the excess risk of HF attributable to renal dysfunction ranged from 30% to 90% depending on CKD stage. ⁸¹ Similar results were reported from the VALIANT study where the risk of HF rose by 10% for each 10 mL/min/1.73 m² decrease in eGFR. ¹⁸

Cystatin C is a sensitive marker of renal impairment that, unlike creatinine, is not affected by age, sex, and muscle mass. In the SOLID‐TIMI 52 study among patients with ACS, cystatin C provided incremental information for risk stratification, including HF hospitalization, independent of other biomarkers including eGFR. ⁸²

Biomarkers of fibrosis

Suppressor of tumourigenesis (ST2) is a member of the IL‐1 receptor family that is involved in the process of myocardial remodelling and fibrosis., ⁸³ , ⁸⁴ It has two isoforms: transmembrane ligand and soluble form. Binding of the soluble form (sST2) to IL‐33 prevents the beneficial effect of this IL on the reduction of cell death and fibrosis. While several studies demonstrated the prognostic utility of sST2 testing in HF, there is less evidence for the predictive value of sST2 after MI. Among consecutive MI patients from the Mayo Clinic, sST2 elevation was independently associated with excess risk of death and HF. Patients in the upper tertial of sST2 had three‐fold the risk of HF as compared with the lowest tertial. ⁸⁵

Galectin‐3, a β‐galactoside‐binding lectin mainly secreted by activated macrophages, is also reflective of fibrosis and cardiac remodelling in response to myocardial injury. ⁸⁶ The American College of Cardiology/American Heart Association guidelines recommend measurement of both sST2 and galectin‐3 for risk prediction in patients with HF. ⁸⁷ A recent study suggested an independent predictive value of galectin‐3 also among unselected MI patients. In a prospectively enrolled incident MI cohort, galectin‐3 was associated with increased risk of death and HF even after adjustment for MI severity, co‐morbidities, and sST2. ⁸⁸

Other biomarkers

Matrix metalloproteinases (MMPs) are proteolytic enzymes that degrade collagen and other proteins of the extracellular matrix. After MI, MMPs regulate the remodelling process by facilitating extracellular matrix turnover and inflammatory signalling. MMP‐8 and MMP‐9 were shown to predict LV remodelling ⁸⁹ , ⁹⁰ and adverse outcomes, including HF development. ⁹¹

Clusterin is a protein that regulates complement activity, apoptosis, and lipid transport. Proteomic analysis of plasma from patients after the first anterior MI identified increased plasma levels of clusterin to be associated with LV remodelling. ⁹² Whether clusterin is also associated with the risk HF after MI needs to be determined.

The prognostic utility of serial biomarker measurement and multi‐marker approach after MI was evaluated by Reinstadler et al., ⁹³ who measured several biomarkers as aspartate and alanine aminotransferases, high‐sensitivity troponin T, N‐terminal prohormone brain natriuretic peptide, lactate dehydrogenase, and high‐sensitivity C‐reactive protein daily for 4 days after admission for an MI. They reported that the peak level of the biomarkers studied was associated with LV remodelling, while the admission value was not. Furthermore, combined biomarker analysis was superior to any of the individual biomarkers. ⁹³ Thus, not only the selection of a biochemical biomarker but also timing of the measurement after MI may be of importance in HF risk prediction.

Genetic aspects

There is a paucity of data on genetic predictors of HF development after MI.

A recent study using weighted gene co‐expression network analysis identified genes BCL3, HCK, PPIF, S100A9, SERPINA1, and TBC1D9B to be involved in the inflammatory response, apoptosis, and HF development after MI. ⁹⁴

MicroRNAs are products of non‐coding DNA transcription consisting of approximately 22 nucleotides that act as significant regulators of mRNA translation. In the heart, they control various processes including cardiac cell death, cardiomyocyte regeneration, and cardiac fibroblast transformation into cardiomyocytes. ⁹⁵ A study by Niu et al. recognized miR‐142‐3p as a contributor to HF after MI. ⁹⁴ Shah et al. identified lower concentrations of miR‐17‐5p, miR‐20a‐5p, and miR‐106b‐5p to be associated with a higher incidence of HF after MI. ⁹⁶ In a study by Lakhani et al., miRNA‐24 and 29a levels were reduced in patients with acute MI and low ejection fraction, whereas miRNA‐34a, miRNA‐208b, and miRNA‐126 were increased in these patients. ⁹⁷

Imaging

Echocardiography

Echocardiography is a commonly used imaging method after MI. In regard to HF prediction, optimal timing of echocardiography is not well defined, as the early post‐MI examination may underestimate systolic function due to myocardial stunning. Therefore, repeated echocardiographic examinations after MI are recommended. ⁹⁸

Systolic function

Reduced LV ejection fraction (LVEF) is associated with the risk of HF development. ¹⁸ , ²¹ , ²⁴ A 5% decrease in LVEF determined by ventriculography performed during the MI hospitalization increases the risk of HF development after the hospital discharge by 12–18%. ²² , ²⁴ Similarly, a 5% decrease in LVEF evaluated by echocardiography 5–20 months after MI increases the risk of HF by 20%. ¹⁸

The wall motion score index (WMSI) reflects wall motion abnormalities better than LVEF because compensatory hyperkinesia of non‐affected regions may compensate for the impaired systolic function. ⁹⁸ , ⁹⁹ In a study of 144 patients with MI, WMSI ≥ 1.5 identified people at increased risk of cardiac death, unstable angina, and HF, independent of LVEF. ¹⁰⁰ In a study by Møller et al., each 0.2 increase in WMSI was associated with hazard ratio of 1.21 (95% confidence interval 1.07–1.37, P = 0.002) for HF development and hazard ratio of 1.15 (95% confidence interval 1.10–1.21, P < 0.0001) for mortality. ⁹⁹ A study by Jurado‐Román et al. deemed WMS a more powerful predictor of mortality and HF than LVEF. ¹⁰¹

Right ventricular (RV) dysfunction significantly contributes to HF development after MI. The tricuspid annular plane systolic excursion (TAPSE) is the most commonly used parameter to evaluate RV systolic function. In patients after MI, RV dysfunction defined by TAPSE ≤ 14 mm was able to predict early cardiac events, including cardiogenic shock. ¹⁰² However, TAPSE assesses only longitudinal contraction of RV and as such provides only partial information on RV function. Fractional area change reflects RV function better than TAPSE. In several studies among patients after MI, decreased fractional area change was associated with an increased risk of adverse events, including HF. ¹⁰³ , ¹⁰⁴

Diastolic function

Standard Doppler examination of transmitral flow provides valuable information in patients after MI. A meta‐analysis of 12 studies by Møller et al. found that restrictive filling pattern is associated with an increased risk of all‐cause mortality and HF. ¹⁰⁵ Tissue Doppler‐derived parameter of E/e′ over 15 is also a strong predictor of mortality and HF development after MI. ¹⁰⁶

Left ventricular remodelling

Left ventricular remodelling is commonly defined as a 20% increase in LV end‐diastolic volume. ⁹ Post‐MI remodelling is exacerbated by a larger infarct size, transmural MI, microvascular obstruction, myocardial haemorrhage, and advanced age. In the contemporary era, almost half of patients after MI demonstrate LV remodelling on echocardiography within 1 year from MI. Among patients with LV remodelling, the risk of hospitalization for HF is 2.7 times higher than in those without LV remodelling. ¹⁰⁷

Speckle tracking echocardiography

Speckle tracking echocardiography (STE) measures regional and global myocardial deformation. Global longitudinal (apico‐basal) strain is the most used and is superior to LVEF measurement, especially in the early phases of systolic dysfunction. In a study by Ersbøll et al. among MI patients with LVEF > 40%, global longitudinal (apico‐basal) strain higher than −14 was associated with a five times higher risk of HF and 12 times higher risk of cardiac death. ¹⁰⁸ 3D STE is a novel method offering more realistic and accurate models of LV than 2D STE. ¹⁰⁹ Global area strain, one of the 3D STE parameters, combines both longitudinal and circumferential strains. Among patients after MI, global area strain is an independent predictor of MACE and HF hospitalization, superior to conventional 2D echocardiography parameters. ¹¹⁰ , ¹¹¹

Myocardial contrast echocardiography

Myocardial contrast echocardiography (MCE) visualizes myocardial perfusion by intravenous or intracoronary infusion of microbubbles. MCE can distinguish reversible and irreversible ischaemia, thus detecting myocardial viability. ⁹⁸ Lack of perfusion signals caused by microvascular obstruction on MCE is consistent with the results of cardiac magnetic resonance. ⁹⁸ To detect no‐reflow, MCE should be ideally performed 24–48 h after coronary intervention for MI. In several studies, no‐reflow after MI was an independent predictor of LV recovery and adverse outcomes including HF. ¹¹² , ¹¹³

Stress echocardiography

Dobutamine stress echo‐derived parameters such as infarction zone non‐viability and ischaemia/infarction at a distance were identified as independent predictors of adverse outcomes, including HF, in patients 2 to 7 days after MI. ¹¹⁴

Cardiac magnetic resonance

Infarct size

Magnetic resonance imaging is currently the gold standard imaging modality for quantifying infarct size using late gadolinium enhancement. According to a meta‐analysis of studies measuring infarct size by MRI or single‐photon emission computed tomography in patients after STEMI, for every 5% increase in MI size, the risk of hospitalization for HF increases by 20%. ¹¹⁵

On the other hand, a recent prospective study showed no additional long‐term prognostic value of infarct size measured by MRI over LVEF in patients after non‐STEMI. ¹¹⁶ The difference in prognostic value in STEMI and non‐STEMI may be explained by a different magnitude of myocardial damage.

Microvascular obstruction

Microvascular obstruction refers to the lack of perfusion in the coronary microcirculation, despite revascularization of the epicardial vessels. Microvascular obstruction can be identified as a hypointense core within the area of hyperenhancement on either early (referred to as early microvascular obstruction) or late gadolinium enhancement (late microvascular obstruction). Microvascular obstruction is associated with larger MI size and adverse remodelling. In a recent individual patient data meta‐analysis, microvascular obstruction increase by 10% elevated the risk of hospitalization for HF by 80% and all‐cause mortality by 114%. ¹¹⁷

Intramyocardial haemorrhage

If the microvascular injury after MI is severe and the integrity of microcirculation is compromised, extravasation of red blood cells into the myocardium can occur. Red blood cell extravasation is referred to as intramyocardial haemorrhage and can be detected by MRI as a hypointense zone within the MI core on T2* imaging or mapping. Several studies suggest that iron deposits from red blood cells trigger pro‐inflammatory response and lead to adverse LV remodelling. ¹¹⁸ , ¹¹⁹ Smaller studies in patients after STEMI suggested that myocardial haemorrhage was more closely associated with adverse outcomes, including HF, than microvascular obstruction. ¹²⁰ . ¹²¹

Multiple scars

A MRI substudy of the third Danish Study of Optimal Acute Treatment of Patients with ST‐segment Elevation Myocardial Infarction (DANAMI 3) proved that multiple scars (characterized by late gadolinium enhancement in more than one myocardial areas remote from acute infarction area) were associated with almost three‐fold risk of all‐cause mortality and HF hospitalization after adjustment for clinical risk factors and MI size. ¹²²

Molecular imaging

Molecular imaging is an emerging method studying different phases of the post‐MI period at a molecular level. The principle is based on the existence of specific tracers that bind to molecules of interest. Various methods of nuclear medicine have shown potential to predict adverse LV remodelling [e.g. tracers binding to MMP‐2 or MMP‐9 within the infarct zone, angiotensin‐converting enzyme inhibitor (ACEi) and angiotensin receptor antagonist‐based tracers, growth factor receptors, and α_vβ₃ integrin tracers]. ¹²³ Currently, data showing predictive value of these methods for HF prediction after MI are lacking.

Remote monitoring

Among patients with established HF, various remote monitoring strategies have been tested to detect worsening of HF and reduce the risk of HF readmission. Several implantable ¹²⁴ (using data from cardioverter defibrillators and cardiac resynchronization therapy) and wearable ¹²⁵ devices have shown the capability to detect HF exacerbation.

However, most of the studies with implantable devices could not detect the mortality benefit of telemonitoring. ¹²⁶ , ¹²⁷ , ¹²⁸ The exception was IN‐TIME trial (Biotronik Home Monitoring technology), which documented improvement of a composite clinical score and particularly all‐cause mortality in remote monitoring group of patients. ¹²⁹ Pooled analysis of the three trials with the same monitoring system with daily transmissions (TRUST, ECOST, and IN‐TIME) confirmed 38% and 36% reduction of all‐cause mortality and the composite endpoint of all‐cause mortality or hospitalization for HF worsening, respectively. The benefit of this form of monitoring appears to be driven by the prevention of HF exacerbation, mainly due to early detection of arrhythmias and/or loss of biventricular pacing. ¹³⁰ Another technology that was capable to reduce HF hospitalizations was the CardioMEMS device—an implantable pulmonary artery pressure monitor. ¹³¹ Yet no remote monitoring study has so far targeted at‐risk population after MI.

Guideline‐recommended therapies

Pharmacotherapy

Beta‐blockers

Beta‐blockers interfere with the harmful effects of sustained activation of the sympathetic nervous system, particularly by blocking the β₁‐adrenergic receptors. Together with ACEi, angiotensin receptor blockers, and statins, beta‐blockers indirectly inhibit MMPs. ¹³²

The evidence for the favourable effect of beta‐blockers on post‐MI outcomes comes mainly from the pre‐thrombolytic era. In a meta‐analysis of 31 randomized studies, long‐term beta‐blocker use reduced all‐cause mortality after MI by 23%. ¹³³ However, none of the studies specifically involved patients with systolic dysfunction and HF during MI hospitalization. ¹³⁴

In the SAVE and AIRE studies, which originally analysed the effect of ACEi in MI patients with LV systolic dysfunction, beta‐blocker use reduced the risk of progression to severe HF by 21% and 42%, respectively. ¹³⁵ , ¹³⁶ In the CAPRICORN study, carvedilol administered in MI patients with systolic dysfunction (LVEF < 40%) reduced all‐cause mortality by 23% and HF hospitalization by 14% compared with placebo. ¹³⁴ This effect was additional to ACEi. ¹³⁴

Angiotensin‐converting enzyme inhibitors and angiotensin receptor blockers

Activation of the renin–angiotensin–aldosterone system actively participates in the process of LV remodelling, myocardial fibrosis, and HF development after MI. ACEi by blocking conversion of angiotensin I into angiotensin II supresses vasoconstriction and aldosterone secretion mediated by angiotensin II. Angiotensin receptor blockers block this action of angiotensin II by interfering with the binding of angiotensin II to its receptor.

Early initiation of ACEi within 0–36 h from MI symptom onset reduces 30 day mortality and HF by 7% and 4%, respectively. ¹³⁷ The absolute benefit is greater in high‐risk groups (such as Killip Class II/III, heart rate >100 b.p.m. at entry) and anterior MI. Importantly, 40% of the survival benefit occurred on the first day of treatment, underscoring the value of initiating ACEIs early, as long as patients have adequate blood pressure. ¹³⁷ The positive effect of ACEi post‐MI was proved also in long‐term studies. ¹³⁸ , ¹³⁹ , ¹⁴⁰ ACEIs started between 3 and 16 days post‐MI reduce the relative risk of mortality by 26% and readmission for HF by 27%. ¹⁴¹

Angiotensin receptor blockers are used in patients with intolerance of ACEi. Various studies have shown favourable effect of losartan or valsartan on mortality and HF hospitalization. ³⁹ , ¹⁴²

Mineralocorticoid receptor antagonist

Aldosterone by its action on distal nephron increases sodium and water reabsorption leading to an expansion of the extracellular fluid. In the heart, mineralocorticoid receptor activation triggers inflammation, hypertrophy, and fibrosis. Mineralocorticoid receptor antagonist eplerenone was tested in the EPHESUS study, in which patients after MI with LVEF < 40% and HF or diabetes were enrolled. As compared with placebo, eplerenone reduced all‐cause mortality and HF hospitalization by 15% and death from cardiovascular causes by 17%. ¹⁴³ This effect was present only if eplerenone was administered in the first 7 days after MI. ¹⁴⁴

Statins

Statins are lipid‐lowering drugs with pleiotropic effects. Besides inhibition of 3‐hydroxy‐3‐methylglutarylcoenzyme A reductase, the key enzyme in cholesterol synthesis, statins exert endothelium‐stabilizing, anti‐inflammatory, and antiproliferative effects on cells involved in atherosclerosis. ¹⁴⁵ In the IDEAL and PROVE IT‐TIMI 22 studies, high doses of statin (atorvastatin 80 mg daily) decreased HF risk by 26% and 45% as compared with low to moderate statin dose, respectively. ¹⁴⁶ , ¹⁴⁷ Early administration of statins (within 24 h of hospitalization) is associated with a 2.5‐fold risk reduction of HF hospitalization and a three‐fold reduction in in‐hospital mortality. ¹⁴⁸

Percutaneous coronary intervention

A decrease in HF after MI since the adoption of PCI has been well documented. ¹⁰ , ¹⁴⁹ , ¹⁵⁰ , ¹⁵¹ In population‐based studies from Sweden, Western Australia, Denmark, and Olmsted County in the USA, increase in PCI rates led to 20–41% reduction in HF after MI. ¹⁰ , ¹⁴⁹ , ¹⁵⁰ , ¹⁵¹

Conclusions

Development of HF after MI is associated with adverse events, impaired quality of life, and lower survival. As reviewed in this paper, a wide range of clinical, laboratory, and diagnostic findings are associated with HF development after MI. However, precise, cost‐effective, and accurate scoring system integrating clinical risk factors, genetics, biomarkers, and imaging methods for HF prediction and prognostication after MI is lacking. Better identification of patients at risk of HF development after MI is needed because timely initiation of guideline‐directed HF therapy can reduce the risk of further LV remodelling, morbidity, and mortality.

Conflict of interest

None declared.

Funding

This study was supported by the Ministry of Health of the Czech Republic (Ministerstvo Zdravotnictví Ceské Republiky), Grant No. NV 19‐09‐00125. All rights reserved.

Author contributions

D.J. performed the literature search and wrote the original draft. P.W. performed the literature search and critically revised the work. J.S., J. Kautzner, V.S., V.A., V.M., and J. Kettner critically revised the work.

Jenča, D. , Melenovský, V. , Stehlik, J. , Staněk, V. , Kettner, J. , Kautzner, J. , Adámková, V. , and Wohlfahrt, P. (2021) Heart failure after myocardial infarction: incidence and predictors. ESC Heart Failure, 8: 222–237. 10.1002/ehf2.13144.