Circulating biomarkers as predictors of left ventricular remodeling after myocardial infarction.

INTRODUCTION: The main impact of myocardial infarction is shifting from acute mortality to adverse remodeling and chronic left ventricle dysfunction. Several circulating biomarkers are explored for better risk stratification of these patients. Biomarker testing is a very attractive idea, since it is non-invasive, not operator-dependent and widely available. AIM: In the present paper we analyze data from the years 2005-2020 about circulating biomarkers of remodeling after myocardial infarction.
MATERIAL AND METHODS: We assessed 53 articles, which examined 160 relations between biomarkers and remodeling. We analyze inclusion criteria for individual studies, time points of serum collection and remodeling assessment as well as imaging methods.
RESULTS: The main groups of assessed biomarkers included B-type natriuretic peptides, markers of cardiomyocyte injury and necrosis, markers of inflammatory response, markers of extracellular matrix turnover, microRNAs and hormones. The most common method of remodeling assessment was echocardiography and the most frequent time point for remodeling evaluation was 6 months.
CONCLUSIONS: The present analysis shows that although a relatively large number biomarkers were tested, selecting one ideal marker is still a challenge. A combination of biomarkers from different groups might be appropriate for predicting remodeling. Data presented in this analysis might be helpful for designing future studies, evaluating clinical use of an individual biomarker or a combination of different biomarkers. Copyright:

Introduction

Mortality during the acute phase of myocardial infarction (MI) has steadily decreased over the past 3 decades [1, 2]. The main impact of MI is shifting from acute mortality to adverse remodeling, chronic left ventricle (LV) dysfunction and eventually clinically apparent heart failure [1, 3]. Occurrence of adverse remodeling increases long-term mortality after MI [4]. Several biomarkers are screened in order to identify patients who are at risk of LV remodeling development. Biomarker testing is a very attractive idea, since it is non-invasive, not operator-dependent and widely available. However, because of the complex pathophysiology of remodeling, selecting one ideal marker is challenging.

The aim of this narrative review was to assess and discuss data about circulating biomarkers of remodeling in patients after MI.

Data assessment

We performed a Medline search of articles published in the years 2005–2020 using the keywords: “myocardial infarction AND ventricular remodeling AND biomarkers”. We examined original studies of patients, admitted with acute MI, reporting measurement of ≥ 1 circulating biomarker. Articles with a follow-up of LV imaging and presenting LV volumes as an indicator of remodeling were analyzed. Studies with sample size of less than 30 patients and with follow-up of < 1 month were excluded. Finally, we selected and assessed 53 studies, which examined 160 relations between biomarkers and remodeling. In Table I we present details about examined publications. Main groups of assessed biomarkers included: B-type natriuretic peptides (BNPs); markers of cardiomyocyte injury and necrosis (troponin, creatinine kinase); markers of inflammatory response including C-reactive protein (CRP), white blood count (WBC), soluble ST2 and galetctin-3; markers of extracellular matrix turnover including matrix metalloproteinases (MMPs), tissue inhibitors of matrix metalloproteinases (TIMPs) and collagen propeptides; microRNAs and hormones (aldosterone, cortisol, norepinephrine, copeptin) (Figure 1).

Table I

Studies of circulating biomarkers associated with left ventricle adverse remodeling after myocardial infarction in chronological order of publication date

Article details	Biomarkers	Patient no. main incl. criteria	LVAR assessment method	LVAR definition	Time of serum collection	LVAR evaluation time	Correlationwith LVAR
Jirmar et al. Int Heart J 2005 [22]	PIIINPPICP	35STEMIPCI	Echocardiography	LVEDV	Admission, day 2, 4, 7,1 month	Day 1, 4,1, 6 months	PositivePositive
Matsunaga et al. Int J Cardiol 2005 [23]	MMP-2 + MMP-9	52STEMIPCI	Echocardiography	LVEDVILVESVI	Week 2	Admission, week 2,6 months	Positive
Wagner et al. J Card Fail 2006 [24]	MMP-9	109STEMIPCI	EchocardiographyVentriculography	LVEDVLVESV	Admission	Admission,6 months	Positive
Hirayama et al. Am J Cardiol 2006 [25]	BNP	106Firstanterior MIPCI	Ventriculography	LVEDV	1, 6 months	1, 6 months	Positive
Webb et al. Circulation 2006 [26]	MMP-9Other biomarkers:MMP-2MMP-7MMP-8TIMP-1TIMP-2	32STEMINSTEMI	Echocardiography	LVEDV	Day 1, 2–5,1, 3, 6 months	Day 1, 5,1, 3, 6 months	PositiveNot associatedNot associatedNot associatedNot associatedNot associated
Orn et al. J Card Fail 2007 [27]	MMP-2MMP-9NT-proBNP	52STEMINSTEMIPCIFibrinolysis	CMR	LVEDVI	Admission,1 month,1, 4 years	4 years	Not associatedPositivePositive
Kelly et al. Eur Heart J 2008 [28]	TIMP-1MMP-9NT-proBNP	404STEMINSTEMIFibrinolyisConservative	Echocardiography	LVEDVLVESV	Day 1, discharge	Discharge,6 months	PositivePositivePositive
Kelly et al. J Card Fail 2008 [29]	Copeptin	274STEMINSTEMIPCIFibrinolysis	Echocardiography	LVEDVLVESV	Discharge	Discharge, mean of 155 days	Positive
Kuribara et al. J Cardiol 2009 [30]	DNaseI	45STEMINSTEMIPCI	Echocardiography	LVEDVLVESV	Admission, day 2, 3, 7, 14, 6 months	Admission,6 months	Positive
Garcia-Alvarez et al. Am J Cardiol 2009 [31]	BNP	82STEMIPCIFibrinolysis	Echocardiography CMR	> 20% increase in LVEDV	Day 4,1, 6 months	6 months	Positive
Weir et al. Eur J Heart Fail 2009 [32]	ApelinOther biomarkers:NT-proBNPNorepinephrine	100LVEF < 40%STEMINSTEMIPCIFibrinolysis	CMR	LVEDVILVESVI	Day 2,6 months	Discharge,6 months	Not associatedPositivePositive
Fertin et al. Am J Cardiol 2010 [33]	BNPTnICRP	246First anteriorQ-wave MIPCIFibrinolysis	Echocardiography	> 20% increase in LVEDV	Discharge, 1, 3, 12 months	Discharge,3, 12 months	PositivePositiveNot associated
Weir et al. Cytokine 2010 [34]	MCP-1	100LVEF < 40%STEMINSTEMIPCIFibrinolysis	CMR	LVESVI	Day 2,3, 6 months	Day 2,3, 6 months	Negative
Weir et al. J Am Coll Cardiol 2010 [35]	ST2 proteinOther biomarkers:NT-proBNPAldosteroneNorepinephrine	100LVEF < 40%STEMINSTEMIPCIFibrinolysis	CMR	LVEDVILVESVI	Admission,3, 6 months	Admission,3, 6 months	PositivePositivePositivePositive
Weir et al. J Thromb Thrombolysis 2010 [36]	t-PAvWFMMP-2MMP-3MMP-9BNP	100LVEF < 40%STEMINSTEMIPCIFibrinolysis	CMR	LVESVI	Day 2,3, 6 months	Day 2,3, 6 months	PositivePositiveNot associatedPositiveNot associatedPositive
Kelly et al. Biomarkers 2010 [37]	Procalcitonin	273STEMINSTEMIFibrinolysisConservative	Echocardiography	LVEDVLVESV	Discharge	Discharge,4 months	Positive
Hallén et al. Heart 2010 [38]	TnI	132STEMIPCI	CMR	LVEDVILVESVI	Day 1, 2	Day 5,4 months	Positive
Lamblin et al. Eur J Heart Fail 2011 [39]	Hepatocyte growth factor	246First anterior Q-wave MIPCI Fibrinolysis	Echocardiography	LVEDVLVESV	Discharge, 1, 3, 12 months	Discharge,3, 12 months	Positive
Weir et al. Eur J Heart Fail 2011 [40]	AldosteroneCortisol metabolites	50LVEF < 40%STEMINSTEMIPCIFibrinolysis	CMR	LVESVI	Admission	Admission,6 months	PositivePositive
Dominguez-Rodriguez et al. Am J Cardiol 2011 [41]	GDF15Other biomarkers:TnIBNP	97STEMIPCI	Echocardiography	> 20% increase in LVEDV	Day 1	First 4 days, 12 months	PositiveNot associatedNot associated
Aoki et al. J Cardiol 2011 [42]	Peak PBMCFPGPeak WBCPeak monocyte	131STEMIPCI	Ventriculography	> 10% increase in LVEDVI	Day 1–5	Admission,6 months	PositivePositivePositivePositive
Erkol et al. Atherosclerosis 2012 [43]	OsteoprotegerinOther biomarkers:Peak TnI	92STEMIPCI	Echocardiography	> 20% increase in LVEDV	Admission	Day 1,6 months	PositivePositive
Wyderka et al. Mediators Inflamm 2012 [44]	CD34+/CXCR4+	50STEMIPCI	Echocardiography	LVEF	Admission,12 months	Admission,12 months	Negative
Devaux et al. J Card Fail 2012 [45]	VEGFB	290STEMIPCI	Echocardiography	LVEDV	Day 4	Discharge,6 months	Negative
Fertin et al. J Cardiol 2012 [46]	sFas ligandOther biomarkers:BNP	246First anteriorQ-wave MIPCI Fibrinolysis	Echocardiography	LVEDVLVESV	1 month	Discharge,3, 12 months	Not associatedPositive
Urbano-Moral et al. Heart 2012 [47]	NT-proBNPTnThsCRPMMP-9PINP	112STEMIPCI	Echocardiography	> 20% increase in LVEDV	Discharge	Discharge,6 months	PositivePositivePositivePositiveNot associated
Weir et al. Cytokine 2012 [48]	IL-21Other biomarkers:MMP-2MMP-3MMP-9TIMP-1TIMP-2TIMP-4MCP-1BNP	100LVEF < 40%STEMINSTEMIPCIFibrinolysis	CMR	LVESVILVEDVI	Admission,6 months	Admission,6 months	PositiveNot associatedPositiveNegativeNegativePositivePositivePositivePositive
Devaux et al. Cir Cardiovasc Genet 2013 [49]	miR-150	90First STEMIFibrinolysisConservative	Echocardiography	LVEDV	Day 3–4	Discharge,6 months	Negative
Bauters et al. Int J Cardiol 2013 [50]	miR-133amiR-423-5p	246AnteriorQ-wave MIPCIFibrinolysis	Echocardiography	LVEDV	Admission, 1, 3, 12 months	Discharge,3, 12 months	Not associatedNot associated
Mather et al. Int J Cardiol 2013 [51]	hsCRPTnINT-proBNPH-FABP	48First STEMIPCI	CMR	LVEDVILVESVI	Day 2,1 week,1, 3 months	Day 2,1 week,1, 3 months	PositivePositivePositiveNot associated
Meng et al. Postgrad Med J 2013 [52]	CatestatinOther biomarkers:BNP	31STEMIPCI	Echocardiography	> 20% increase in LVEDV	Admission,day 3, 7,3 months	Week 1,3 months	PositivePositive
Weir et al. Circ Heart Fail 2013 [53]	Galectin 3	100LVEF < 40%STEMINSTEMIPCIFibrinolysis	CMR	LVESVI	Admission,6 months	Admission,6 months	Not associated
Eschalier et al. Circ Heart Fail 2013 [54]	PINPPIIINPPICPOther biomarkers:BNPTnICRP	246First anteriorQ-wave MIPCIFibrinolysis	Echocardiography	> 20% increase in LVEDV	1 month	Discharge,12 months	Not associatedNot associatedPositivePositivePositiveNot associated
Reinstadler et al. Heart 2013 [55]	Copeptin	54STEMIPCI	CMR	LVEDVLVESV	Day 2	Admission,4 months	Positive
Kleczynski et al. Dis Markers 2013 [56]	NT-proBNP	45STEMIPCI	CMR	LVEDVLVESV	Admission,6 months	6 months	Positive
Fertin et al. PLoS One 2013 [57]	MMP-1MMP-2MMP-3MMP-8MMP-9MMP-13TIMP1TIMP-2TIMP-3TIMP-4	246First anterior MIPCIFibrinolysis	Echocardiography	> 20% increase in LVEDV	Admission,3 months,1 year	Discharge,1, 3, months, 1 year	Not associatedNot associatedNot associatedNot associatedPositivePositiveNot associatedNot associatedNot associatedNot associated
Lv et al. Int J Mol Sci 2014 [58]	miR-208bmiR-34aOther biomarkers:TnTPeak CKBNP	359PCIFibrinolysis	Echocardiography	> 10% increase in LVEDV	Admission	Baseline,6 months	PositivePositivePositiveNot associatedPositive
Kumarswamy et al. Circ Res 2014 [59]	Mitochondrial long noncoding RNA uc022bqs.1	246First anteriorQ-wave MIPCIFibrinolysis	Echocardiography	> 20% increase in LVEDV	Day 3–7, 1, 3, 12 months	Day 3–7,3, 12 months	Positive
Manhenke et al. Eur Heart J 2014 [60]	PINPMMP-2MMP-3Other biomarkers:TnThsCRPNT-proBNP	42First STEMIPCI	CMR	LVEDVILVSVI	Admission, day 2, 7,2, 12 months	Day 2, 7,2, 12 months	NegativeNegativePositivePositivePositivePositive
Liu et al. Cardiology 2015 [61]	miR-146amiR-21Other biomarkers:NT-proBNPCRPTnICK-MB	198STEMIPCI	Echocardiography	> 20% increase in LVEDV	Admission	Day 5,1 year	PositivePositivePositivePositiveNot associatedPositive
Abdel Hamid et al. J Interv Cardiol 2016 [62]	Circulating endothelial cells	78PCIFibrinolysis	Echocardiography	> 20% increase in LVEDV	Day 1	Day 2,1 month	Positive
Türkoğlu et al. Coron Artery Dis 2016 [63]	M30 antigenM60 antigenOther biomarkers:BNP	255STEMIPCI	Echocardiography	> 20% increase in LVEDV	Day 1	Day 1,6 months	PositivePositivePositive
Reindl et al. Heart 2017 [64]	FGF 23Other biomarkers:cTnThsCRPNTproBNP	88STEMIPCI	CMR	> 20% increase in LVEDV	Day 2	Day 2,4 months	PositivePositivePositivePositive
Grabmaier et al. Int J Cardiol 2017 [65]	miR-1miR-29bmiR-21	44STEMIPCI	CMR	LVEDV	Day 4, 9,6 months	Day 4,6 months	Not associatedNegativeNot associated
Hendriks et al. Int J Cardiovasc Imaging 2017 [66]	Peak CKPeak CK-MBPeak TnTNT-proBNp	271First STEMIPCI	CMR	LVEDVILVESVI	Admission, week 2, 6	4 months	PositivePositivePositivePositive
Hsu et al. Int J Med Sci 2017 [67]	BNP decrease ratioPeak CK-MBPeak TnICRP	97STEMINSTEMIPCI	Echocardiography	> 20% increase in LVEDV	Day 2, 7,3 months	Day 2, 7,3 months	NegativePositiveNot associatedNot associated
Di Tano et al. Heart 2017 [68]	Galectin 3Other biomarkers:NT-proBNP	103First STEMILAD culpritPCI	Echocardiography	> 15% increase in LVESV	Day 2,1, 6 months	Day 2,1, 6 months	PositiveNot associated
Miñana et al. Int J Cardiol 2018 [69]	ST2 proteinOther biomarkers:TnTNT-proBNP	109First STEMIPCI	CMR	LVEDVILVESVI	Day 1	1 week,6 months	PositiveNot associatedNot associated
de Gonzalo-Calvo et al. Sci Rep 2018 [70]	miR-1254	70First STEMIPCI	CMR	LVESVI	Admission	Week 1,6 months	Negative
Orrem et al. Int J Cardiol 2018 [71]	IL-1RasIL-1RAcPsIL-1R2sIL1-R1Other biomarkers:Peak TnTPeak CRPNTproBNP	320STEMIPCI	CMR	LVEDVILVESVI	Admission, day 1,4, 12 months	Day 2,4 months	Not associated Not associated PositiveNot associatedPositivePositiveNot associated
Padoan et al. Int J Cardiol 2019 [72]	Vitamin DOther biomarkers:CRPPeak TnI	253STEMINSTEMIPCICABG	Echocardiography	> 15% increase in LVESV	During hospitalization	During hospitalization, 4 months	NegativePositivePositive
Garcia et al. Int J Mol Sci 2019 [73]	Peak CKTnINT-proBNPCRPWBCNeutrophil countCreatinine	64STEMIPCIFibrinolysis	CMR	> 10% increase in LVESV	Day 2	Admission,3, 12 months	PositiveNot associatedNot associatedPositivePositivePositiveNot associated
Reindl et al. Eur Heart J Acute Cardiovasc Care 2019 [74]	TSHOther biomarkers:Peak TnTPeak CRP	102STEMIPCI	CMR	> 20% increase in LVEDV	Day 1,4 months	Week 1,4 months	NegativePositivePositive

PIIINP – type III procollagen propeptide, PICP – carboxy terminal propeptide of type I collagen, MMP – matrix metalloproteinases, BNP – B-type natriuretic peptide, TIMP – tissue inhibitor of MMP, Tn – troponin, CRP – C reactive protein, MCP – monocyte chemoattractant protein, tPA – tissue plasminogen activator, vWF – von Willebrand Factor, GDF – growth differentiating factor, PBMC – peripheral blood mononuclear count, FPG – fasting plasma glucose, WBC – white blood count, VEGFB – vascular endothelial growth factor B, PINP – procollagen type I amino terminal propeptide, Il – interleukin, miR – micro RNA, HFABP – heart type fatty acid binding protein, CK – creatinine kinase, FGF – fibroblast growth factor, TSH – thyroid stimulating hormone, Incl – inclusion, MI – myocardial infarction, STEMI – ST elevation MI, NSTEMI – non-ST elevation MI, PCI – percutaneous coronary intervention, LAD – left anterior descending, LVEF – left ventricle ejection fraction, LVAR – left ventricular adverse remodeling, CMR – cardiac magnetic resonance, LVEDV(i) – left ventricle end diastolic volume (index), LVESV(i) – left ventricle end systolic volume (index).

Figure 1

Groups of most commonly assessed biomarkers. Data are shown as number of studies evaluating groups of biomarkers

BNP – B type natriuretic peptide, ECM – extracellular matrix.

A positive correlation between examined biomarkers and remodeling was found in 101 (63%), a negative correlation was found in 13 (8%) and no significant association was found in 46 (29%) cases. Figure 2 presents the relationships between the most common individual biomarkers and remodeling. BNPs, troponin, CRP and creatinine kinase were the most frequent biomarkers and they were positively correlated with remodeling. MMP-9 was the most commonly analyzed member of metalloproteinases. It occurred in 9 studies and in 7 a positive correlation with remodeling was reported. MMP-2 was assessed in 7 studies, but in 5 reports no significant association with remodeling was found. MMP-3 was analyzed in 4 studies and in 3 it was positively correlated with remodeling. Less frequent biomarkers included soluble ST2, TIMPs and procollagen type I amino terminal propeptide (PINP).

Figure 2

Relationships between individual biomarkers and remodeling. A – Data are shown as number of studies evaluating specific biomarkers. B – Data are shown as number of patients enrolled in studies evaluating biomarkers

BNP – B type natriuretic peptide, Tn – troponin, CRP – C reactive protein, MMP – matrix metalloproteinase, CK – creatinine kinase, TIMP – tissue inhibitor of MMP, PINP – procollagen type I amino terminal propeptide.

The majority of presented studies (68%) included ST-elevation MI (STEMI) patients exclusively. In most studies (57%) patients were treated with primary percutaneous coronary intervention (PCI). In 38% of studies patients underwent PCI and fibrinolysis and in 5% of studies patients underwent fibrinolysis or conservative treatment only. In Figure 3 we show the most commonly assessed biomarkers in patients treated exclusively with primary PCI. We observed that TIMPs were less frequently and microRNA-21 was relatively more frequently assessed in studies which included patients treated exclusively with primary PCI.

Figure 3

Relationships between individual biomarkers and remodeling in patients treated exclusively with primary percutaneous coronary intervention. Data are shown as number of studies evaluating specific biomarkers

BNP – B type natriuretic peptide, Tn – troponin, CRP – C reactive protein, MMP – matrix metalloproteinase, CK – creatinine kinase, miR – microRNA, PINP – procollagen type I amino terminal propeptide.

In the presented articles remodeling was defined as an increase in LV end diastolic volume (LVEDV) or less often LV end systolic volume (LVESV) during follow-up. Twenty (38%) studies utilized specific cut-off values for LV volume increase. Most commonly it was a 20% increase in LVEDV. Echocardiography and cardiac magnetic resonance (CMR) were the most common methods of remodeling assessment. Echocardiography was used in 57% and CMR was used in 41% of studies. In more recent studies, from the years 2015–2019, CMR was used in 57% of cases and echocardiography in 43%. Time points of LVAR assessment differed vastly among analyzed papers. The shortest period of LVAR evaluation after MI was 1 month (1 study), the longest was 4 years (also in 1 study). The most frequent time point for LVAR assessment was 6 months (73% of studies).

Description of biomarkers

The present analysis shows that a relatively large number of circulating biomarkers were tested, which reflects the complex pathophysiology of remodeling. Main groups of assessed biomarkers included BNPs, markers of cardiomyocyte injury and necrosis, markers of inflammatory response, markers of extracellular matrix turnover and microRNAs.

B-type natriuretic peptides

BNP is secreted predominantly from heart ventricles. It is a marker of volume overload and high filling pressure. In response to myocardial wall stretch, pre-proBNP is synthesized and processed to proBNP, which is further processed to the biologically inactive N-terminal prohormone fragment (NT-proBNP) and biologically active BNP [5]. Biological effects of BNP include diuresis, natriuresis, vasodilatation and inhibition of the renin-angiotensin system. BNP is an established biomarker of LV systolic dysfunction and heart failure progression [6]. Higher BNP concentrations in patients after MI were reported to predict long-term mortality [6]. According to ESC guidelines BNP and NT-proBNP provide prognostic information regarding the risk of death and acute heart failure in MI patients [7]. Although the cut-off values are different for BNP and NT-proBNP, the guidelines give no indication which marker presents better accuracy for heart failure [7]. In the present analysis NT-proBNP was analyzed in 13 studies and BNP was assessed in 14 reports. Both markers were positively correlated with remodeling.

Cardiac troponins

The cardiac troponin complex consists of 3 subunits: troponin C, troponin T and troponin I. Troponin I and T form an actin-myosin complex and are released into peripheral blood after myocyte injury. Elevated concentration of troponin I and T is a diagnostic marker of acute coronary syndromes. Peak levels of both troponin I and T are predictive for mortality, recurrent MI and newly developed post-MI heart failure. Early troponin measurement provides an estimate of infarct size [5]. Although both troponins present comparable diagnostic accuracy for MI, troponin T provides greater prognostic value [7]. Currently, high sensitivity (hs) troponin assays are recommended for diagnosis and prognosis of MI instead of conventional assays. In the present analysis troponin I was examined in 10 studies and troponin T was assessed in 8 studies. Both troponins were positively correlated with remodeling.

Markers of inflammatory response

C-reactive protein is an acute phase protein of hepatic origin. Myocardial ischemia is associated with the systemic inflammatory response with increased production of acute phase proteins including CRP, partly as a response to stimulation by interleukin-6, which is released from the infarct zone. Levels of CRP increase in the first hours of MI and peak approximately at day 2. Elevated CRP concentrations are associated with adverse clinical outcome after MI, larger infarct size, microvascular obstruction and higher mortality in patients with heart failure [8]. In the present analysis CRP was assessed in 12 publications. In 9 studies, it was positively correlated with remodeling. Several studies assessed high-sensitivity (hs) CRP, which was also positively associated with remodeling.

Soluble suppression of tumorigenicity-2 (sST2) is the soluble form of interleukin-1 receptor-like 1 and is a protein biomarker of cardiac stress. Serum levels of sST2 were reported to be higher in patients with heart failure. In patients with MI, higher concentrations of sST2 predicted mortality and occurrence of post-MI heart failure [5]. In the present analysis sST2 was assessed in 3 studies and in 2 it was positively correlated with remodeling.

Extracellular matrix turnover

Extracellular matrix (ECM) surrounds cardiomyocytes and forms a scaffold which maintains the LV shape and geometry. ECM rearrangement corresponds to a balance between degradation and synthesis of extracellular components, regulated by MMPs and TIMPs [9]. MMPs are members of zinc-dependent endopeptidases, which degrade several ECM proteins and thus modulate physiological and pathological processes including MI and congestive heart failure. MMPs consist of 25 enzymes which are endogenously inhibited by TIMPs, a family comprising 4 members (TIMP-1, -2, - 3 and -4) [10]. The ECM turnover during remodeling is regulated through the balance of MMPs and TIMPs, levels of both of which rise after MI. In the present analysis MMP-9 was the most frequent analyzed member of MMPs. It was assessed in 9 studies and in 7 a positive correlation with remodeling was reported. The second most commonly assessed biomarker from this group was MMP-3, which appeared in 4 studies and in 3 a positive correlation with remodeling was observed. The relationship between levels of TIMPs and remodeling was inconclusive in the present analysis.

Collagen synthesis begins in fibroblasts which produce procollagen. In the ECM, the amino-terminal and carboxy-terminal propeptides are separated by endopeptidases and released into the circulation. They can be used as markers of collagen synthesis. Collagen type I and III are principal structural proteins found in the myocardium. PINP is a marker of type I collagen synthesis. It was reported to be associated with reverse remodeling and inversely correlated with LV volumes in patients undergoing resynchronization therapy [11]. In the present analysis PINP was assessed in 3 studies and in 1 it was negatively correlated with remodeling. In 2 studies, no significant association with remodeling was reported.

MicroRNAs

MicroRNAs are small noncoding RNA molecules with regulatory functions. They participate in various cardiovascular processes through post-transcriptional regulation of gene expression. MicroRNAs are related to the regulation of cardiomyocyte apoptosis and fibrosis [12]. In the present analysis microRNAs were tested in 6 studies; however, the most frequently assessed microRNA-21 appeared only in 2 studies and in 1 analysis a positive correlation with remodeling was reported; thus selecting a biomarker of remodeling from the microRNA family is limited.

Methods and time points of remodeling assessment

Remodeling is defined as molecular, cellular and interstitial changes resulting from myocardial ischemia [13]. Clinical assessment of LV remodeling is based on detection of increase of LV volumes. In the present analysis the most common cut-off value was a 20% increase in LVEDV. Cardiac magnetic resonance is considered to be a gold standard for remodeling assessment due to accurate and reproducible measurements of LV volumes [14]. CMR is a more precise method with reduced operator variability compared to echocardiography. In addition, CMR with late gadolinium enhancement has the ability to distinguish between reversible and irreversible myocardial injury. CMR may also provide more precise information about scar formation, transmural necrosis and microvascular obstruction [15-18]. In the present analysis the rate of studies utilizing CMR was 41% and increased in more recent publications. Despite this, echocardiography remains the fastest and most accessible method which is used not only in clinical practice but also in clinical trials. Transthoracic echocardiography is also recommended in all patients with acute MI to evaluate global and regional function of LV [7].

Remodeling is a time-dependent process, which can continue up to 6–12 months after MI with infarct extension occurring in weeks to months after reperfusion [19]. Earlier assessment might not reflect the full remodeling process. A frequently selected time point for remodeling evaluation is 6 months after MI. Time points of blood collection are also vital. In several analyzed studies, serial blood sampling during index hospitalization and follow-up was utilized, which is helpful in determining the strongest association with remodeling. However, we think that the most clinically useful is the relationship between remodeling and levels of biomarkers measured in the acute phase of MI. Nowadays, biomarker guided therapy in patients after MI is not a standard approach. On the other hand, identification of high risk individuals could allow implementation of follow-up with more frequent LV assessment after hospital discharge.

Future directions

Association of classic biomarkers including BNPs, cardiac troponin and CRP with post-MI remodeling is widely documented. These biomarkers are readily available, routinely assessed in MI patients and their measurement is relatively inexpensive. In the present analysis MMP-9 was frequently examined and positively correlated with remodeling. However, measurement of MMP-9 activity is challenging due to its complex in vivo regulation. MMPs are synthesized as inactive zymogens, and must be enzymatically activated by hydrolyzation of a propeptide domain. Their activity is further regulated by TIMPs. Typical methods such as western blot, ELISA or immunohistochemistry are reported to be not sufficient to accurately describe MMPs’ in vivo activity [20]. The ideal biomarker should not only allow improvement of clinical decisions but also be easily detectable from blood. The main idea of biomarker testing is their wide availability and no inter/intra-operator variability. The present analysis shows that a relatively large number of different biomarkers were assessed. Due to the complex pathophysiology of remodeling, selecting one marker is challenging. What is more, several biomarkers including MMPs, TIMPs and microRNAs occur in many types; thus despite being tested in a relatively large amount of studies, individual biomarkers appeared in a limited number of reports. Perhaps at a recent stage of studies, single biomarker testing might be not sufficient for remodeling prediction. A combination of biomarkers from different groups, reflecting different pathways of remodeling, might be appropriate. Reinstadler et al. showed that combined biomarker testing including NT-proBNP, troponin T, CRP, lactate dehydrogenase and liver transaminases improved the predictive value for remodeling compared to single biomarker assessment [21].

Conflict of interest

The authors declare no conflict of interest.