Article Text

Download PDFPDF
Original article
Early measurements of plasma matrix metalloproteinase-2 predict infarct size and ventricular dysfunction in ST-elevation myocardial infarction
  1. Lennart Nilsson1,2,
  2. Jonas Hallén3,4,
  3. Dan Atar3,4,
  4. Lena Jonasson1,2,
  5. Eva Swahn1,2
  1. 1Division of Cardiology, Department of Medical and Health Sciences, Linköping University, Linköping, Sweden
  2. 2Department of Cardiology, University Hospital, Linköping, Sweden
  3. 3Department of Cardiology, Oslo University Hospital Ullevål, Oslo, Norway
  4. 4Institute for Clinical medicine, University of Oslo, Oslo, Norway
  1. Correspondence to Dr Lennart Nilsson, Department of Cardiology, University Hospital, Linköping S-58185, Sweden; lennart.nilsson{at}


Background Immediate reopening of the acutely occluded infarct-related artery via primary PCI is the preferred treatment in ST-elevation myocardial infarction (STEMI). However, the sudden reinitiation of blood flow can lead to a local acute inflammatory response with further endothelial and myocardial damage, so-called reperfusion injury. The activation of matrix metalloproteinases (MMPs) is suggested to be a key event in this process.

Objectives To investigate circulating MMPs, tissue inhibitors of metalloproteinases (TIMPs) and myeloperoxidase (MPO) in relation to infarct size, left ventricular dysfunction and remodelling in a STEMI population undergoing PCI.

Methods 58 Patients with STEMI undergoing primary PCI were included. Blood samples were collected at baseline before PCI and at 12, 24 and 48 h for later analysis of MMPs, TIMPs and MPO by ELISA. Infarct size, left ventricular (LV) dysfunction and remodelling were assessed by cardiac MRI at 5 days and 4 month after STEMI.

Results Plasma MMP-2 at 0 and 12 h showed a consistent and significant correlation with infarct size and LV dysfunction measured both at 5 days and at 4 months and correlated well with troponin I measurements. For TIMP-1 and TIMP-2 some support was found for associations with infarct size and LV dysfunction, but these were not as consistent as for MMP-2. MMP-8, MMP-9 and MPO did not overall correlate with measures of infarct size, LV dysfunction or remodelling.

Conclusions In patients with STEMI, circulating levels of MMP-2, measured early and even before reperfusion therapy, are strongly associated with infarct size and LV dysfunction. This provides further evidence for the role of MMP-2 in ischaemia-reperfusion injury.

  • Matrix metalloproteinase
  • ST-elevation myocardial infarction
  • ischaemia-reperfusion injury
  • coronary artery disease
  • myocardial ischaemia and infarction (IHD)
  • acute coronary syndrome
  • risk stratification
  • risk factors
  • st-t changes

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Coronary artery disease is the leading cause of death in the Western world. In many patients the first clinical presentation of coronary artery disease is an acute myocardial infarction. Immediate reopening of the acutely occluded infarct-related artery via primary percutaneous coronary intervention (PCI) is the preferred treatment to salvage ischaemic myocardium in the setting of ST-elevation myocardial infarction (STEMI).1 2 However, the sudden reinitiation of blood flow achieved with primary PCI can give rise to a local inflammatory assault with further endothelial and myocardial damage, so-called reperfusion injury.3 4 The underlying mechanisms of reperfusion injury have not been fully elucidated, but evidence suggests membrane damage induced by oxygen radicals, intracellular calcium overload and tissue injury related to infiltration and activation of leucocytes.5–8

This study is a prespecified substudy of the F.I.R.E. trial, which studied the effects of the fibrin–derived peptide Bβ15-42 (FX06) on reperfusion damage in patients undergoing primary PCI for STEMI.9 10 FX06 is a naturally occurring fragment of fibrin. In vitro and in vivo studies have shown that FX06 binds to the vascular endothelial (VE)-cadherin receptor on endothelial cells, thereby inhibiting leucocyte transmigration through gap junctions and tissue inflammation injury.11 12 The positive effects of FX06 on the inflammatory response are demonstrated by a decrease in cytokines and interleukins following coronary occlusion and reperfusion in pigs.13 In occlusion-reperfusion studies in rats, administration of FX06 caused a significant 40% reduction in infarct size compared with placebo.12

Matrix metalloproteinases (MMPs) are known to play a crucial role in the destabilisation of atherosclerotic plaques, resulting in acute coronary syndromes.14 Accumulating data also suggest a key role for MMPs in myocardial ischaemia-reperfusion injury. Data from animal models show that such injury is mediated by an increased MMP activity thus creating a proteolytic environment.15 Experimental ischaemia-reperfusion injury consistently induces an increased expression of MMP-2 and MMP-9 in the myocardial tissue, where these MMPs act on various intracellular and extracellular targets to mediate tissue damage.16–18 Tissue inhibitors of metalloproteinases (TIMPs) are endogenous inhibitors and important regulators of MMP activity. As shown in experimental studies, decreased expression or increased consumption of TIMPs might create an imbalance between MMPs and TIMPs in the myocardium contributing to the ischaemia-reperfusion injury.19 Impaired microvascular function, as part of ischaemia-reperfusion injury, is associated with recruitment and activation of neutrophils. Neutrophils are the main source of MMP-8, MMP-9 and myeloperoxidase (MPO). Decreased bioavailability of nitric oxide secondary to neutrophilic release of MPO is a proposed mechanism of microvascular dysfunction.20 The role of MMP-8 in ischaemia-reperfusion injury is still unknown.

Plasma levels of MMPs in patients with acute coronary syndrome or with post-acute coronary syndrome have consistently been found to be markers of left ventricular (LV) dysfunction, remodelling, future cardiovascular events and poor prognosis.21–26 However, to our knowledge, there is no study available on the temporal changes of MMPs and TIMPs in plasma in patients with STEMI undergoing primary PCI.

Cardiac MRI (CMR) has consistently been found to accurately determine the size of myocardial infarct and LV volumes used to calculate measures of LV dysfunction and remodelling.27 28 They all reflect the extent of myocardial ischaemia-reperfusion injury and are strongly associated with poor prognosis.29

The aim of this study was to analyse circulating levels of MMPs, TIMPs and MPO during the first 48 h in patients with STEMI undergoing primary PCI. Furthermore, we wanted to investigate potential correlations between these biomarkers and measures of infarct size, LV dysfunction and remodelling as assessed by CMR. Lastly, we aimed to study the effect of FX06 on plasma levels of MMPs, TIMPs and MPO.

Materials and methods

Study population

This study is a prespecified substudy of the F.I.R.E. trial. The study design and results of the F.I.R.E. trial (, NCT00326976) have been published elsewhere.9 10 In brief, the F.I.R.E. trial was a double-blind, randomised, placebo-controlled multicentre trial performed in 2006–8, including 234 patients, to study the effects of FX06 on reperfusion damage in patients undergoing primary PCI for STEMI presenting within 6 h from onset of symptoms. Apart from the study medication, patients were treated according to current guidelines. No treatment differences attributable to FX06 were observed in primary or secondary outcome measures.10 This substudy was designed by three of the authors (LN, LJ, ES). Complete sample sets were obtained from 58 subjects at 10 of the participating sites. The study protocol was approved by all local ethic committees and written informed consent was obtained from all patients.

Cardiac MRI

The CMR protocol has previously been described in detail.10 All CMR studies were analysed at the central MR Evaluation Centre, University Hospital Basel (Basel, Switzerland) by a single experienced CMR reader who was blinded to study groups, followed by a blinded review by a level III CMR expert. Intraobserver variability was assessed for the primary reader in a subset of 40 randomly chosen studies and the intraclass correlation was 0.85 for 26 studies from day 5.

Biochemical analysis

Blood samples for MMPs, TIMPs and MPO were collected in Vacutainer tubes (using sodium heparin as anticoagulant) at baseline immediately before primary PCI and at 12, 24 and 48 h. Samples were centrifuged within 30 min to separate plasma, which was then stored immediately at −20°C and within a week at −70°C until analysed. MMP-2, MMP-8, MMP-9, TIMP-1, TIMP-2 and MPO were analysed using a commercially available ELISA kit (R&D Systems Europe, Abingdon, UK). The assays for MMP-2, MMP-8 and MMP-9 recognise both pro-MMP and MMP. The TIMP-1 and TIMP-2 assays measure the amount of free TIMP and to some extent also TIMP complexed with MMPs. The lower limits of detection were 0.16 ng/ml (MMP-2), 0.02 ng/ml (MMP-8), 0.156 ng/ml (MMP-9), 0.08 ng/ml (TIMP-1), 0.011 ng/ml (TIMP-2) and 0.100 ng/ml (MPO), respectively. The interassay coefficients of variation ranged from 2.4% to 4.7% for MMP-2, 1.7% to 4.5% for MMP-8, 1.1% to 3.8% for MMP9, 0.8% to 7.6% for TIMP-1, 1.3% to 2.4% for TIMP-2 and 2.5% to 2.8% for MPO. Blood samples for measurement of serum cardiac troponin I were obtained at 24 and 48 h after admission to hospital. All samples were analysed in a blinded core laboratory (Spranger Laboratories, Ingolstadt, Germany). Cardiac troponin I was measured on the Abbott AxSYM System (Abbott Diagnostics, Abbott Park, Illinois, USA) using the second-generation AxSYM troponin-I ADV assay. The lower limit of detection of the assay is 0.020 ng/ml with a 10% coefficient of variation of 0.16 ng/ml.

Statistical methods

Based on previous data on plasma MMPs and TIMPs,30 we calculated that a sample size of 21 subjects in each treatment group (42 subjects in total) would be sufficient to give 80% power to detect an estimated 20% change in plasma levels of the biomarkers. In this study we included 58 patients of whom 27 received placebo and 31 FX06 treatment.

Baseline and procedural variables are presented as mean with SD or median with IQR for non-Gaussian distributed data. Biochemical parameters are all presented as median with IQR. Differences between groups were tested by Kruskal–Wallis and Mann–Whitney U tests. Spearman correlations were used to assess associations between variables. Repeated-measures analysis of variance by rank (Friedman test) was used to test time-dependent changes in biochemical parameters. Two-tailed p values <0.05 were considered to be statistically significant. No formal statistical tests were used to deal with the problem of multiple comparisons. However, given the large number of tests performed, thoughtful consideration of the specific context, the pretest probability and the biological plausibility of each statistically significant association was an important issue. All statistical analyses were performed using SPSS V.17 (SPSS Inc).


There were no differences between the FX06-treated group and the placebo group in the levels of any of the biomarkers at any time point. Thus, for further analysis both treatment groups were analysed as one group. Baseline characteristics of the 58 study participants are summarised in table 1. Of note, mean age of the study group was 61 years and about one-third were women. Time from symptom to PCI was just over 3 h, and before PCI 93% had TIMI 0 flow, whereas after primary PCI 98% had TIMI 2–3 flow. The median infarct size was 21% and 17% of LV volume at 5 days and 4 months, respectively.

Table 1

Baseline, procedural and outcome measures of the study population (n=58)

Levels of circulating MMPs, TIMPs and MPO during first 48 h in patients with STEMI undergoing primary PCI

Plasma levels of MMP-2, MMP-8, MMP-9, TIMP-1, TIMP-2 and MPO at 0–48 h are summarised in table 2. Of note, plasma MMP-2 decreased significantly from baseline to 12–24 h and then returned to baseline levels. Plasma TIMP-2 showed a pattern similar to MMP-2. Plasma MMP-9 showed high levels already at 0 h and then declined significantly over time, whereas TIMP-1 increased significantly from baseline to 12 h and remained at a high level up to 48 h. Also, plasma MPO had its highest level at baseline, decreased markedly at 12 h, increased slightly at 24 h and finally reached its lowest level at 48 h. Thus, MPO tended to show a biphasic pattern over time, similar to the pattern seen for MMP-8.

Table 2

Plasma levels of MMPs, TIMPs and MPO during the first 48 h after STEMI and reperfusion treatment (n=58)

Correlations between MMPs, TIMPs, MPO and measures of infarct size

The infarct size, defined as the late gadolinium enhancement zone (LGE zone) on CMR examination, was assessed at 5 days and 4 months after STEMI and primary PCI. All study participants were included in the day 5 examination, whereas 12 of the 58 patients did not complete the MRI at 4 months. Correlations between MMPs, TIMPs, MPO and LGE zone are summarised in table 3. We found a strong positive correlation between MMP-2 at 0–12 h and LGE zone both at 5 days and at 4 months. Figure 1 shows infarct size (LGE zone in % of left ventricular mass) at 4 months according to quartiles of MMP-2 at baseline. There were significant differences in infarct size between the first and fourth quartiles (p<0.05) as well as the third and fourth quartiles (p<0.05). TIMP-2 at 0–12 h correlated significantly with LGE zone at 5 days, but not significantly at 4 months. Furthermore, peak values of TIMP-1 (at 24–48 h) showed a positive correlation with infarct size. For MMP-8, MMP-9 and MPO we found no significant correlations with LGE zone.

Table 3

Correlations between MMPs, TIMPs, MPO and measures of infarct size

Figure 1

Infarct size (late gadolinium enhancement zone as percentage of left ventricular (LV) mass) at 4 months according to quartiles of matrix metalloproteinase-2 (MMP-2) at baseline. Median and IQR are shown.

Correlations between MMPs, TIMPs, MPO and measures of LV dysfunction and remodelling

Findings are summarised in table 4. Of interest, plasma MMP-2 at 0–12 h showed a significant negative correlation with LV ejection fraction (LVEF)% at 5 days and 4 months. A similar but somewhat weaker negative correlation was observed for MMP-9 at the same time points. Peak values of TIMP-1 at 24–48 h also correlated negatively with LVEF% at 5 days and tended to correlate at 4 months. TIMP-2 at 0–12 h tended to have a negative correlation with LVEF% at 5 days, with a weaker non-significant association at 4 months. Plasma MPO at several time points was found to have significant and borderline significant negative correlations with LVEF%. For change in end-diastolic volume index from 5 days to 4 months (dEDVI) and change in end-systolic volume index from 5 days to 4 months (dESVI), there were tendencies towards a negative correlation with TIMP-1 at 24–48 h, but otherwise we found no consistent correlations with the other biomarkers.

Table 4

Correlations between MMPs, TIMPs, MPO and measures of left ventricular dysfunction and remodelling

Correlation between MMPs, TIMPs, MPO and troponin I

Correlations with troponin I measured at 24 and 48 h are summarised in table 5. Again, we found consistent significant correlations between MMP-2 at 0–12 h and troponin I at both time points. In parallel, TIMP-2 at 0–12 h also had significant positive correlations with troponin I. Furthermore, plasma TIMP-1 at 24–48 h showed positive correlations with troponin I at 24 and 48 h. MMP-8, MMP-9 and MPO at 12–48 h were found to have significant or borderline significant correlations with troponin I at 24 h, with weaker associations to the 48 h troponin I measurement.

Table 5

Correlations between MMPs, TIMPs, MPO and troponin I (n=58)


The main finding of the study was the consistent and highly significant correlation between baseline and 12 h levels of plasma MMP-2 and outcome measures of infarct size and LV dysfunction. Individuals with the highest levels of MMP-2 had the largest infarcts and developed a more pronounced LV dysfunction, which are both markers of a poor prognosis. Similar but weaker correlations were found for baseline and 12 h levels of plasma TIMP-2 and peak levels of TIMP-1 at 24 and 48 h. Notably, MMP-9 did not consistently correlate with any of the outcome measures.

Numerous experimental studies and animal models have demonstrated the role of MMP-2 in activating different pathways of reversible and irreversible myocardial ischaemia-reperfusion damage.31 To our knowledge, this is the first study in a pure STEMI population to show a positive association between early measurements of MMP-2 in plasma and the extent of the final infarct size.

LV remodelling was assessed by calculating changes in end-systolic and end-diastolic LV volume indices (dESVI and dEDVI) from day 5 to 4 months after STEMI. None of the biomarkers showed significant correlation with these measures of remodelling. The lack of association between MMP-2 and remodelling, despite strong correlation with infarct size and LV dysfunction (LVEF), has several possible explanations. The process of remodelling, occurring over a longer period of time, might involve pathways not dependent on MMP activation or might not be determined by the early activation of MMP in STEMI. Furthermore, in this homogeneous STEMI population of patients, arriving at the PCI laboratory within 6 h of symptom onset and receiving state-of-the-art treatment both in the acute phase and during follow-up, infarct sizes were overall rather small and the process of remodelling might have been inhibited to a large degree.

This study was designed and sized to analyse the effect of FX06 on MMPs, MPO and TIMPs. No significant differences or trends to differences were found between the treatment groups. Also, the F.I.R.E study had an overall negative result with no effect of FX06 on primary or secondary outcomes.10 Thus, we think that analysis of all study participants together as one group in relation to outcome measures is well justified.

Plasma levels of MMPs, TIMPs and MPO were analysed at baseline (ie, within 6 h of symptom onset) and at 12, 24 and 48 h after primary PCI. Thus, baseline measurements reflect levels seen in circulation during myocardial ischaemia, whereas the later measurements are a result of both ischaemia and reperfusion phases. This study was not designed to assess changes in biomarkers in STEMI as compared with healthy individuals or stable coronary artery disease. These changes have been studied and described previously by several groups.32 33 However, analysis of circulating levels of MMPs, TIMPs and MPO may give other important insights into which pathways are activated in the different phases of acute ischaemia and reperfusion. The enhanced levels of MMP-9 and MPO in plasma at baseline probably reflect activation of neutrophils and release of these markers at the site of thrombosis and occlusion of the infarct-related artery.34 However, the lack of association between plasma MMP-9, MPO and measures of infarct size and LV dysfunction does not necessarily exclude the possibility that these enzymes are involved in the process of myocardial damage during ischaemia-reperfusion but rather that plasma levels are mostly determined by release from circulating neutrophils and not by inflammatory cells in the myocardium.

This study has some limitations. First, we included a few more patients then needed according to power calculations, but nevertheless the size of the study population might have been too small, especially since the biomarkers had a non-normal distribution with values within a rather wide range. Second, study participants were randomly assigned to receive FX06 or placebo as adjunctive treatment. Since FX06 did not have an impact on primary or secondary outcomes in the F.I.R.E trial and did not affect levels of biomarkers in our substudy, we decided to consider all patients as one group for the correlation studies. However, we cannot totally exclude the possibility that some minor effects of FX06 influence the associations between biomarkers and measures of infarct size, LV dysfunction and remodelling. Third, we chose to analyse plasma antigen concentrations of MMPs and MPO rather than enzymatic activity. In our experience and according to previous data from our group and several others, determination of antigen concentration is a more robust and stable way to assess circulating MMPs and still gives valuable information about pathological conditions and future prognosis. Lastly, TIMP-4 is highly expressed in the myocardium and has been shown to be an important player in ischaemia-reperfusion injury.35 However, TIMP-4 was not analysed in this study.

To summarise, in a homogeneous STEMI population we found that baseline and 12 h levels of MMP-2 in plasma were associated with myocardial infarct size and with LV dysfunction at follow-up. Whether or not MMP-2 has a causal relationship with these measures cannot be established from this study, but needs to be further investigated.


The authors thank Professor Peter Buser and Dr Jens Bremerich (University Hospital Basel, Basel, Switzerland) for reading cardiac magnetic resonance images. We are grateful to Ylva Lindegårdh for excellent technical assistance.



  • See Editorial, p 1

  • Funding This project was supported by grants from the Swedish Heart-Lung foundation and the Foundation for Old Servants.

  • Competing interests None to declare.

  • Patient consent Obtained.

  • Ethics approval The study protocol was approved by all local ethic committees.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Linked Articles