Prognostication in Different Heart Failure Phenotypes: The Role of Circulating Biomarkers

1. Introduction

Heart failure (HF) remains an important clinical entity that has increased in prevalence worldwide due to improved survival after a HF diagnosis [1, 2]. Recent studies have shown sufficient differences in the aetiology, pathophysiology, clinical presentation and outcomes, as well as the prognosis, between HF with a preserved ejection fraction (HFpEF) and HF with a reduced ejection fraction (HFrEF) [3–5].

HFpEF is a phenotypic and heterogeneous clinical syndrome characterized by cardiovascular (CV) disease and dysmetabolic and inflammatory states associated with both advanced age and various non-CV co-morbidities, which finally lead to the impairment of myocardial structure and function, unless under the condition of declining global EF <45% [6, 7]. Although global left ventricular EF >50% is currently used to differentiate between reduced and preserved cardiac pump function, this cut-off point is widely discussed as a likely inadequate criterion [8, 9] However, old age, being female, suffering from diabetes mellitus, hypertension, atrial fibrillation and chronic kidney disease are strong predictors of HFpEF's development [10–12]. Based on evidence from endomyocardial biopsies, some of the specific cardiac structural phenotypes to be targeted in HFpEF may be represented by myocyte hypertrophy and interstitial fibrosis [13, 14]. HFrEF has been described as a disease of aged elderly subjects, with a male predominance that is frequently associated with dilation cardiomyopathy, ischaemia, inflammatory and diabetic aetiology, and rarely with arterial and pulmonary hypertension [15, 16]. Cell loss due to ischaemia, apoptosis and necrosis, myocardial inflammation associated with oxidative stress, expanded interstitial fibrosis leading to the disintegrity of the cardiac wall, increased passive myocardial stiffness, the worsening of cardiac configuration and contractile function are common in HFrEF's development [17].

Many questions remain unanswered regarding differences in the molecular signals that initiate the development of HFpEF and HFrEF [18]. In this context, it might be possible to appropriately stratify at risk HFpEF and HFrEF patients by using biomarkers. Recently, brain natriuretic peptides, cardiac specific troponins, metabolomic intermediates, Galectin-3, ST2, cardiotrophin-1, soluble endoglin, growth differentiation factor 15 and other new biological markers associated with HF's development have been widely investigated [5, 6, 12, 13, 17, 19]. However, the current data on the interrelationship of these biomarkers and phenotypes of HF are limited. The aim of the review is devoted to the accumulation of knowledge regarding the utility of cardiac biomarkers, aiming to reclassify patients with different phenotypes of HF.

2. Biomarkers in HF Risk Stratification

The routine use of biomarkers might help to stratify the patients with HFrEF and HFpEF at higher risk of death and clinical outcomes. The current guidelines — the 2012 European Society of Cardiology (ESC) Guidelines for the Diagnosis and Treatment of Acute and Chronic Heart Failure and the 2013 American College of Cardiology Foundation/American Heart Association (ACCF/AHA) Guideline for the Management of Heart Failure — are well accepted by many clinicians regarding HFrEF's prognostication. Indeed, HFpEF is the one that really requires the improvement of biomarkers for diagnosis and prognosis. In this context, many biological markers, which reflect several faces of the pathogenesis of HF, have been investigated in detail, but only natriuretic peptides, soluble ST2, Galectin-3 and highly sensitive cardiac-specific troponins have been validated thus far. Table 1 offers summarized evidence regarding the predictive role of biomarkers in patients with different HF phenotypes.

3. Brain Natriuretic Peptides

Within the last two decades, cardiac natriuretic peptides (BNP and NT-proBNP) have been defined as biomarkers that we may use to screen for LV systolic dysfunction in patients with symptoms suggestive of HF. BNP and NT-proBNP are now included in the current guidelines for HF diagnosis, management and risk assessment because of their high specificity and sensitivity [19]. Despite BNP and NT-proBNP improving discrimination modestly for HF above and beyond conventional risk factors, and substantially improving the risk classification for HF, peak concentrations of BNP and NT-proBNP and serial measurements of NT-proBNP levels in longitude are not able to allow the differentiation of HF phenotypes [20, 21]. However, there were important differences in the prognostic value of NT-proBNP in HFpEF versus HFrEF in the NT- proBNP-guided arm of the TIME-CHF study [22]. NT-proBNP has demonstrated less prognostic value in HFpEF as compared to HFrEF, and has not predicted a development of HFpEF or HFrEF. NT-proBNP lost significance as a risk stratifier in ambulatory patients with stable HF and probably also in those who have HFpEF. There are attampts to use of sing sample measurement of mid-regional atrial natriuretic peptide (mr-ANP) and NT-proANP in order to screen HFpEF and HFrEF in individuals, when the diagnosis of HF is not obvious. In this setting, the diagnostic value and prognostic ability for HF-related mortality and CV hospitalization for both mr-ANP and NT-proANP were not superior to those of NT-proBNP [23].

4. Cardiac Troponins

Recent studies have shown that elevated levels of highly sensitive cardiac troponin I (hs-cTnI) and T (hs-cTnT) as biomarkers of subclinical myocardial injury may provide to be clinically useful prognostic information, concerning both the future risk of HF's manifestation in asymptomatic subjects and the risk of fatal events and primary/re-admissions in the hospital in those with already established symptomatic acute, acutely decompensated and chronic stable HF related to ischaemic and non-ischaemic causes [24–27]. Moreover, cardiac troponin mutations are considered a cause of impaired relaxation in the mutant cardiac myocytes due to myofibril hypersensitivity to Ca²⁺ [28].

OPEN IN VIEWER

Table 1.

Predictive role of biomarkers in patients with different HF phenotypes

Biomarkers	Patient population	The most important findings	References
Natriuretic peptides	Exerted dyspnoea	Predictor of HF risk manifestation, risk of admission in the hospital and HF-related deaths	[19]
	Known HFpEF and HFrEF	Biomarkers independently predicted HF-related outcomes, CV mortality, all-cause death, admission in the hospital, but they did not predict a development of HFpEF or HFrEF	[19, 20, 21]
Cardiac troponins	Ischaemia-induced HF	Predictors of HF manifestation risk in asymptomatic subjects Predictors of death in HFpEF and HFrEF	[24, 25]
	Ischaemia-induced and non-ischaemia-related HF	Predictors of death, primary/re-admissions in the hospital	[25–27]
Galectin-3	General population	Prognosticator of HF risk, risk of death from any cause	[33]
	Known HF patients	Predictor of CV death, HF-related deaths, primary and re-admission in the hospital	[35–37]
Soluble ST2	General population	Predictor of higher risk of all-cause mortality, HF manifestation	[42]
	Known HF patients	Independent predictor of CV deaths, HF-related deaths, admission in the hospital	[44]
Cardiotrophin-1	Known ischaemia-induced HF patients	Predictor of CV clinical outcomes	[53, 54]
Endoglin	Patients at higher risk of CV disease	Predictor of CV events/ outcomes, HF manifestation	[58–60]
Growth differentiation factor 15	Patients with known CV disease	Predictor of HF manifestation	[64, 65]
	Known HF patients	Predictor of HF-related outcomes	[68]

Abbreviation: CV, cardiovascular; HF, heart failure; HFpEF, HF with preserved ejection fraction; HFrEF, HF with reduced ejection fraction.

Cardiac-specific troponins exhibited the strongest associations with hospitalization, survival and outcomes in cases of HF; there are expectations regarding the ability of troponins to emerge as an aetiology-dependent relation to phenotypes of HF. Seliger et al. [29] hypothesized that hs-cTnT would identify HF risk among older adults with left ventricular hypertrophy (LVH). In the Cardiovascular Health Study, its authors found that the adjusted risk of HFrEF was 7.8 times higher among those with the highest tertile of hs-cTnT and LVH (HR=7.83; 95% CI: 4.43–13.83). Patients with LVH and longitudinal increases in hs-cTnT or NT-proBNP were approximately three times more likely to develop HF (primarily HFrEF), compared with those without LVH and with stable biomarkers. Thus, in this study, the authors were not able to find sufficient advantages regarding hs-cTnT compared NT-proBNP in order to characterize sub-phenotypes of HF. In another study, Neeland et al. [30] reported that identifying a malignant sub-phenotype of LVH was the better predictive surrogate marker than a limited elevated level of hs-cTnT, and even increased NT-proBNP among asymptomatic individuals with a high risk of progression to HF and CV death in the general population. Therefore, there was evidence that the higher levels of cTnT and NT-proBNP correlated well with the risk of HF in older adults, but were not associated with phenotypes of HF [31]. Overall, the circulating level of the cell injury biomarker is not a powerful tool for HF-phenotype detection.

5. Systematic Metabolomic Biomarkers

Zordoky et al. [32] suggested that a systematic metabolomic analysis would reveal a novel metabolomic fingerprint of HFpEF that will help us to understand its pathophysiology and assist us in establishing new biomarkers for its diagnosis. Compared to non-HF control, HFpEF patients demonstrated higher serum concentrations of acylcarnitines, carnitine, creatinine, betaine and amino acids, and lower levels of phosphatidylcholines, lysophosphatidylcholines and sphingomyelins. Medium- and long-chain acylcarnitines and ketone bodies were higher in HFpEF than in HFrEF patients. The authors suggested that this abovementioned metabolomic fingerprint has been utilized to identify two novel panels of metabolites that can separate HFpEF patients from both non-HF controls and HFrEF patients. However, this assumption requires further investigation.

6. Conclusion

Several reports have shown that biomarkers reflecting the differentiation of fibroblasts into myofibroblasts, subsequently altering collagen turnover, cardiac fibrosis and inflammation, might have diagnostic and predictive value in HFpEF and HFrEF. The biomarkers with most predictive value in determining the early stage of HF's development were sST2, Galectin-3, CT-1 and GDF-15. However, increased plasma concentrations of these biomarkers were not specific for a distinct disease group of HFpEF and HFrEF. Finally, more investigations are required to determine the role of novel biomarkers in predicting HF and the determination of the early stage of HFpEF and HFrEF's development.