Association of dysglycaemia with persistent infarct core iron in patients with acute ST-segment elevation myocardial infarction

Background Dysglycaemia increases the risk of myocardial infarction and subsequent recurrent cardiovascular events. However, the role of dysglycaemia in ischemia/reperfusion injury with development of irreversible myocardial tissue alterations remains poorly understood. In this study we aimed to investigate the association of ongoing dysglycaemia with persistence of infarct core iron and their longitudinal changes over time in patients undergoing primary percutaneous coronary intervention (PCI) for acute ST-segment elevation myocardial infarction (STEMI). Methods We analyzed 348 STEMI patients treated with primary PCI between 2016 and 2021 that were included in the prospective MARINA-STEMI study (NCT04113356). Peripheral venous blood samples for glucose and glycated hemoglobin (HbA1c) measurements were drawn on admission and 4 months after STEMI. Cardiac magnetic resonance (CMR) imaging including T2 * mapping for infarct core iron assessment was performed at both time points. Associations of dysglycaemia with persistent infarct core iron and iron resolution at 4 months were calculated using multivariable regression analysis. Results Intramyocardial hemorrhage was observed in 147 (42%) patients at baseline. Of these, 89 (61%) had persistent infarct core iron 4 months after infarction with increasing rates across HbA1c levels (<5.7%: 33%, ≥5.7: 79%). Persistent infarct core iron was independently associated with ongoing dysglycaemia defined by HbA1c at 4 months (OR: 7.87 [95% CI: 2.60–23.78]; p < 0.001), after adjustment for patient characteristics and CMR parameters. The independent association was present even after exclusion of patients with diabetes (pre- and newly diagnosed, n = 16). Conclusions In STEMI patients treated with primary PCI, ongoing dysglycaemia defined by HbA1c is independently associated with persistent infarct core iron and a lower likelihood of iron resolution. These findings suggest a potential association between ongoing dysglycaemia and persistent infarct core iron, which warrants further investigation for therapeutic implications.


Background
Incomplete myocardial tissue reperfusion as a consequence of microvascular injury occurs in every second to third patient with acute STsegment elevation myocardial infarction (STEMI) [1,2].The presence and severity of microvascular injury is strongly associated with worse outcomes [3,4].Microvascular injury is a complex and heterogeneous phenomenon, mainly caused by severe ischemia/reperfusion injury together with distal embolization of thrombus and individual susceptibility [4].Cardiovascular magnetic resonance (CMR) imaging has become the reference standard for characterizing microvascular injury in myocardial infarction (MI) patients [4][5][6].Microvascular obstruction (MVO) and intramyocardial hemorrhage (IMH) are the two distinct entities of microvascular injury that can be visualized by CMR with high accuracy [6,7].IMH, the most severe form of microvascular injury, is characterized by extravasation of erythrocytes, resulting in iron deposition inside the infarcted myocardium [6].IMH has recently moved into clinical focus by demonstrating particularly strong prognostic implications in STEMI patients, incremental to other infarct severity parameters including infarct size and MVO [8][9][10].Furthermore, it was recently shown that IMH might not only be a consequence, but rather a determinant of final infarct size [11].
In contrast to MVO, which resolves within the first weeks after STEMI, persistence of infarct core hemorrhage can be detected in the chronic stage after MI in a significant proportion of patients [12] and has been shown to drive a persistent, proinflammatory burden within the infarct zone, leading to left ventricular dysfunction, adverse remodeling, and worse clinical outcomes [1,2,8,10].The underlying pathophysiology is still not completely understood, however, recent evidence demonstrated that iron-induced macrophage activation drives fatty infiltration of the myocardium contributing to unfavorable cardiac remodeling [13].
Dysglycaemia not only represents a major risk factor for the development of a first-time MI, but is also associated with higher rates of recurrent cardiovascular complications thereafter [14].Furthermore, it is well known that patients with STEMI and concomitant dysglycaemia have worse short-and long-term clinical outcomes [8,10].Considering the adverse effects of dysglycaemia on the microvasculature, and recently published data, demonstrating an association between dysglycaemia and more severe microvascular injury in the acute setting after MI [15], incomplete myocardial reperfusion due to IMH could be one explanation for the adverse outcome in STEMI patients with dysglycaemia [16,17].
We hypothesized that dysglycaemia in patients suffering acute STEMI might be associated with persistent infarct core iron as depicted by CMR imaging.Therefore, we investigated the association of dysglycaemia with persistence of infarct core iron and their longitudinal changes over time.

Study design
The current investigation is based on the Magnetic Resonance Imaging In Acute ST-Elevation Myocardial Infarction (MARINA-STEMI, NCT04113356) study, a prospective observational study that analyzed 348 STEMI patients at the coronary care unit of Innsbruck Medical University Hospital between 2016 and 2021.Briefly, MARINA STEMI was designed to evaluate the nature and clinical significance of myocardial tissue characteristics as determined by CMR imaging in STEMI patients.
Inclusion criteria were first STEMI, defined by clinical symptoms suggestive of ischemia and significant ST-segment elevation in at least 2 contiguous leads (> 0.1 mV in extremity leads; > 0.2 mV in precordial leads), who were treated by primary percutaneous coronary intervention (PCI) within 24 h following symptom onset.Exclusion criteria were defined as follows: age < 18 years, an estimated glomerular filtration rate < 30 ml/min/1.73m², Killip class ≥ 3 at the time of CMR imaging, any history of a previous MI or coronary intervention and any CMR contraindication (pacemaker, orbital foreign body, cerebral aneurysm clip, manifest claustrophobia, known or suggested contrast agent allergy to gadolinium).
The study was conducted in accordance with the Declaration of Helsinki and received approval by the research ethics committee of the Medical University of Innsbruck.Written informed consent was given by all patients before study inclusion.

Biomarker measurements
Blood samples for glucose and glycated hemoglobin A1c (HbA1c) analyses were obtained via peripheral venipuncture at hospital admission and at 4 months follow-up.
HbA1c at 4 months was available in 245 patients (70% of total cohort).
High-performance liquid chromatography was used for HbA1c measurements following the International Federation for Clinical Chemistry (IFCC) reference method [18].HbA1c was expressed as percentage according to the National Glycohemoglobin Standardization Program (NGSP) by applying the following IFCC/NGSP 'master equation' [19]: Non-fasting glucose levels were analyzed using whole-blood amperometry.Dysglycaemia was defined as having either pre-or manifest diabetes according to HbA1c values.Prediabetes was defined by a HbA1c value of ≥ 5.7%, manifest diabetes by a HbA1c value of ≥ 6.5% [20].

Cardiovascular magnetic resonance
CMR imaging was performed on a 1.5 Tesla Magnetom AVANTOscanner (Siemens®, Erlangen, Germany) within the first week after PCI and 4 months thereafter.Image acquisition and post-processing were performed following a previously published protocol [21].In brief, short-axis cine images acquired by electrocardiogram (ECG)-triggered balanced steady-state free precession bright-blood sequences were used to evaluate left ventricular volumes and function.Late gadolinium-enhanced (LGE) images were obtained 10-15 min after application of 0.2 mmol/kg contrast medium (Gadovist®, Bayer®, Leverkusen, Germany), applying an ECG-triggered phase-sensitive inversion recovery sequence.Infarct area was defined by hyper-enhancement at a threshold of + 5 standard deviations (SD) above the signal intensity of remote myocardial tissues of the opposite left ventricle.MVO was defined by hypo-enhancement within the infarct area on LGE images.
IMH and persistent infarct core iron were assessed by T2 * mapping using a breath-hold, cardiac gated gradient echo sequence with eight echoes obtained in three matching short-axis slices before administration of the contrast agent.A motion correction algorithm was applied to reduce movement artefacts.Typical imaging parameters were: echo time = 2.02-16.0ms (ΔTE = 2 ms), time to repetition = 18.2 ms, flip angle = 20, bandwidth = 815 Hz/pixel, In-plane pixel spacing 1.6 × 1.6 mm, and slice thickness = 8 mm.Motion corrected, colorcoded T2 * -maps were automatically inline generated by fitting signal intensities at each image pixel with an exponential model for the given echo times.
IMH was defined as a region of hypo-intense core within the infarcted area with a T2 * reduction below 20 ms at baseline.Persistent infarct core iron was defined as a region of hypo-intense core within the infarcted area with T2 * reduction below 20 ms at 4 months follow-up [8].A region of interest (ROI) size of at least 20 pixels was applied to determine T2 * values.
Inter-and intraobserver variability has been reported previously by our research group [12].

Statistical analyses
Statistical analysis was performed with SPSS Statistics 29.0.1 (IBM, Armonk, New York), MedCalc Version 22.0.22 (Ostend, Belgium) and R 4.2.0 (The R Foundation, Vienna, Austria).Distribution of data was tested using the Shapiro-Wilk test.Categorical variables are depicted as frequencies with corresponding percentages.Continuous variables are presented as mean ± SD or median with interquartile range (IQR), according to their distribution.Differences between groups were tested with Student's t-test, Mann-Whitney U-test, or Chi-square test as indicated.Univariable and multivariable logistic regression analyses were conducted to disclose significant and independent associations between persistent infarct core iron and iron resolution (dependent binary variable).All clinical, CMR and angiographic variables, as well as biomarkers of myocardial wall-stress, infarct size, and inflammation at baseline (Table 1) showing a p-value of < 0.10 in univariable testing were entered in the corresponding multivariable regression analysis, using the forced entry method.To allow for a better comparison of odds ratios (OR), all OR are presented for 1 standard deviation increase, unless otherwise stated.
The multivariable logistic regression model was also subjected to sensitivity analysis to estimate relative risks.Risk ratios were calculated using the generalized linear model with binomial family using R 4.2.0.For sensitivity analysis, HbA1c at 4 months was categorized according to clinically established cut-offs (HbA1c < 5.7%, HbA1c 5.7-6.4,HbA1c ≥6.5%).Other continuous variables were dichotomized according to their median.
A two-tailed p-value of < 0.05 was considered as statistically significant.

Dysglycaemia and infarct characteristics
CMR parameters at baseline and 4 months according to the presence and absence of IMH are shown in detail in Table 1.

Clinical outcome
Follow-up data were available in all patients.During the 4-months follow-up period, 13 (4%) patients experienced a MACE, including, 3 (1%) myocardial reinfarctions and 10 (3%) new congestive heart failures.No deaths were observed during follow-up.
A MACE was significantly more common in patients with persistent infarct core iron as compared to patients without persistent infarct core iron at 4 months follow-up (9 vs. 2%, p = 0.01).

Discussion
This study is the first to investigate a possible association of dysglycaemia with persistent infarct core iron, determined by T2 * -mapping, in a well-characterized cohort of STEMI patients treated according to current practice.The main finding was that higher HbA1c levels at 4 months follow-up were significantly and independently associated with persistent infarct core iron and a lower likelihood of iron resolution at 4 months after infarction.These findings suggest a pathophysiological link between dysglycaemia, persistence of infarct core iron and worse outcome after STEMI.Whether this pathophysiological link also represents an effective target for intervention should be investigated in further studies.

Intramyocardial hemorrhage in STEMI
In patients suffering STEMI, IMH and its persistence (most often referred to as "persistent infarct core iron") are increasingly studied tissue biomarkers that can be assessed by T2 * mapping.The presence of IMH drives delayed infarct healing and is responsible for fatty degeneration of the infarcted myocardium contributing to adverse remodeling and ultimately more frequent adverse clinical events [8,12,13].Previous studies using T2 * mapping reported IMH incidence rates of 23 to 50% in the acute setting after PCI for STEMI [7][8][9].Similarly, the present investigation found IMH in 42% of patients in the early phase after infarction.
In contrast to MVO, which resolves in the first weeks after STEMI, persistent infarct core iron, as a consequence of IMH, can be found in  the chronic stage after MI in a substantial subset of patients [12].In the present study, 61% of patients with IMH at baseline showed persistence of infarct core iron at 4 months after MI, which is in line with previous observations [23].Interestingly, it has recently been demonstrated that persistent infarct core iron can even remain up to a decade after STEMI and its presence is related to poor infarct healing [12].Furthermore, persistent infarct core residues were shown to be associated with a 4fold increase in all-cause death and heart failure [23].During the 4 months of follow-up in our study, we did not observe any deaths.However, a significant increase in the incidence of MACE was observed among patients with persistent infarct core iron.This increased incidence was primarily driven by a significant increase in new congestive heart failure.
In summary, IMH and persistent infarct core iron, assessed via T2 * mapping, are prevalent biomarkers associated with poor infarct healing and adverse outcome, persisting chronically in a significant subset of STEMI patients.

Dysglycaemia as a determinant of persistent infarct core iron in STEMI
Due to previous observations describing an association between dysglycaemia and microvascular damage in different organ systems [16], as well as impaired outcome in STEMI [14,24], a pathophysiological link between dysglycaemia and development of irreversible microvascular tissue injury can be assumed.
Although a number of previous studies evaluated the association between dysglycaemia and IMH in the acute setting after PCI for STEMI [9,17,25,26], data on the association between ongoing dysglycaemia and persistent infarct core iron in the chronic phase after STEMI are completely lacking.Furthermore, in most studies investigating the association between dysglycaemia and infarct characteristics, the definition of dysglycaemia relies on patient-reported diabetes diagnosis, hence the number of undiagnosed diabetes as well as prediabetes remains unclear.HbA1c, however, a measure of average blood glucose concentrations over weeks offers a more objective marker to describe dysglycaemia.
In the current investigation, dysglycaemia, defined by HbA1c levels at 4 months, was significantly and independently associated with persistent infarct core iron at 4 months after MI, even after adjustment for established biomarkers of infarct core iron, such as high-sensitivity cardiac troponin T [10].This study therefore expands previous data by demonstrating that dysglycaemia during STEMI might be an important and potentially modifiable risk factor for irreversible microvascular tissue injury that persists at least for up to 4 months post-MI.Interestingly, only a minority in our study had manifest diabetes, and therefore the primary finding was predominantly driven by patients with prediabetes.This is further corroborated since the exclusion of patients with manifest diabetes strengthened the association of HbA1c and persistent iron residues.One might therefore speculate that those patients with prediabetes could particularly benefit from interventions targeting dysglycaemia.In fact, IMH evolves progressively after STEMI (peak at day 3 post-infarction) [10] and decreases to some extent during the chronic phase after infarction [10].These findings could potentially indicate that IMH is susceptible to precise preventive measures or therapeutic strategies, should they be identified and established.Furthermore, it has recently been shown that IMH may not only be an expression of larger infarcts per se, but also a determinant of final infarct size itself.In fact, Liu and colleagues demonstrated that IMH was associated with a 3-fold loss of salvageable myocardium, resulting in an almost 2-fold increase in final infarct size in a canine model [11].Conclusively, dysglycaemia, primarily driven by prediabetes and identified through HbA1c levels at admission, was found to be a significant, potentially modifiable factor associated with persistent infarct core iron in STEMI patients, suggesting a need for precise preventive measures or therapeutic strategies.

Infarct core iron resolution
The underlying pathophysiological mechanisms by which dysglycaemia leads to persistence of infarct core iron are not completely understood and were not the scope of the present investigation.However, cell signaling dysfunction, formation of toxic metabolites (esp.advanced glycation end products) and altered redox potential due to preceding dysglycaemia might promote extensive and irreversible myocardial tissue damage, with less influence on reversible changes of the myocardial microvasculature [16,27].These hypotheses are strengthened by a large amount of preclinical evidence that dysglycaemia and diabetes can exacerbate myocardial ischemia/reperfusion injury [15,28].Interestingly, our analysis revealed that ongoing dysglycaemia (defined by HbA1c levels at 4 months) was significantly and independently associated with impaired iron resolution.It is therefore possible, that novel strategies targeting dysglycaemia beginning with HbA1c levels above 5.7 could mitigate the severity of persisting infarct core iron and improve outcome.It is important to emphasize that this is a hypothesis-generating observation and, as such, calls for rigorous investigation in future studies.

Clinical implications and potential therapeutic strategies for persistent infarct core iron
Prediabetes was observed in 38% of patients.In view of these findings, one could speculate that this so far neglected group of STEMI patients could benefit from early antidiabetic treatment to positively influence the acute and chronic pathophysiological processes leading to persistent core iron.
Data from previous preclinical and clinical studies indicate a possible benefit of insulin in preventing ischemia/reperfusion injury in the setting of acute MI [29].The effects of insulin that potentially attenuate ischemia/reperfusion damage were reported to go beyond glucoselowering, including anti-inflammatory, anti-thrombotic, anti-oxidative and anti-apoptotic cascades [30].Randomized studies testing the clinical benefit of insulin treatment in this setting are, however, very scarce.Furthermore, the existing literature on insulin treatment and hard clinical endpoints after infarction is conflicting [31,32].The possible effects of newer antidiabetic agents (e.g.sodium-glucose cotransporter 2 inhibitors) on infarct-related microvascular injury processes remain speculative and are currently being investigated (NCT04899479).
Our results thus suggest that screening for prediabetes and aiming for an optimal glycemic status post-STEMI may be advisable for these patients.Whether meaningful benefits accrue from specific interventions in STEMI patients with altered glucose metabolism should be further investigated.

Limitations
The observational design of our study limits the ability to definitively establish a causal relationship between HbA1c and persistent infarct core iron.However, the potential causal role of dysglycaemia in the persistence of infarct core iron is a compelling hypothesis and underscores the need for further investigation.Future experimental studies are needed to definitively establish a direct causal relationship and provide deeper insights into the dynamics of this relationship.
Our findings are not generalizable to unstable STEMI patients and those where a CMR examination was not possible.While our study provides important insights, it is important to emphasize that our reliance on T2 * maps derived from only three short-axis slices does not provide a complete representation of the left ventricle.Consequently, small regions of IMH or persistent infarct core iron may have been missed, and volumetric measurement of the infarct core over the entire ventricle was not possible.Another important limitation of this study is the lack of IMH and persistent infarct core iron quantification.Therefore, the relationship between the degree of dysglycaemia and the size of iron deposition cannot be determined.These limitations could potentially affect the precision of our findings and are an area for methodological improvement in future studies.Furthermore, follow-up HbA1c measurements were available in only 70% of the total cohort which represents another important limitation with respect to the secondary aim.Finally, the use of a single HbA1c measurement may not fully reflect the patient's glycemic status throughout the period from infarction to the 4-month follow-up.

Conclusion
In STEMI patients treated by contemporary primary PCI, ongoing dysglycaemia defined by HbA1c was significantly and independently associated with persistent infarct core iron and a lower likelihood of iron resolution.These data suggest a potential prognostically relevant pathophysiological relationship between dysglycaemia and irreversible microvascular damage after PCI for STEMI.Further research could explore whether dysglycaemia might serve as a therapeutic target to influence persistent infarct core iron and potential outcomes.

Table 1
Baseline characteristics of the study cohort (n = 348).

Table 3
Logistic regression analysis for the prediction of persistent infarct core iron.