In this study, we applied a novel, dual-contrast MEMRI-DEMRI strategy to determine if viable myocardium was detectable within regions of positive, transmural DEMRI. MEMRI infarct volume was found to be significantly (39%) lower than DEMRI infarct volume in this 21-day swine IR model. This signal mismatch identified a border zone that was positive for both MEMRI signal (viable) and transmural DEMRI signal (non-viable), and which also displayed an intermediate SNR by MEMRI, compared to core infarct and remote zones. TEM analysis also revealed preservation of cell architecture and contractile elements within this border zone region. Overall, these results suggest that dual contrast MEMRI-DEMRI may allow accurate detection of viable myocardium within the infarct border zone that appears nonviable by DEMRI alone.
Although most patients who suffer an MI will survive the acute event31
, they often develop heart failure32
and a progressive decline in ventricular function that is believed to involve apoptosis and collagen deposition, particularly in border zone areas33,34
. Revascularization decisions for these patients are frequently made at intermediate and chronic timepoints post-MI and are often based upon viability assessment. In this study, a 21-day timepoint post-IR was chosen to explore this intermediate timepoint, and DEMRI infarct volume measurements were significantly higher than either MEMRI or TTC infarct volumes. These data suggest that DEMRI may overestimate infarct volume at intermediate post-IR timepoints, and that combination with MEMRI may accurately delineate these injured, but live populations of cardiomyocytes.
Intermediate MEMRI SNR measurements also reflected tissue heterogeneity within this mismatched border zone, which has not been previously reported using MEMRI in a myocardial IR model. While prior studies have reported heterogenous DEMRI signal in infarct border zones14
, Gd accumulation is non-specific and provides limited information on actual cell viability. Because of specific Mn2+
uptake into viable, functioning cells, the observed MEMRI SNR heterogeneity points to significant populations of cardiomyocytes that are alive with intact Ca2+
-channel function (MEMRI positive) in this region despite surrounding necrotic tissue (DEMRI positive). The intermediate degree of Mn2+
uptake within the border zone demonstrates that this region is not only visually and quantitatively distinct from the core infarct zone by TEM analysis, but biologically distinct as well. Prior MEMRI studies have imaged myocardial infarct areas16, 18, 24
; however, these studies focused either on the acute ischemia phase or have looked at chronic timepoints with ex vivo analysis. Intermediate MEMRI and DEMRI SNR, with intact cell structure, suggests that the DEMRI signal may still be evolving at 21 days post-IR and not yet reflective of true infarct size. Future studies may examine earlier and later time points to further characterize this mismatch region.
Viable myocardium that falls outside
the core infarct zone has been imaged previously using a combination of agents, including radio-labeled and fluorescent microparticles8, 25, 35–38
. However, in this study, MEMRI infarct volume was compared directly to DEMRI infarct volume and was found to be significantly smaller, pointing to viable cardiomyocytes within
areas of transmural DEMRI. A technical limitation of this study, and of DEMRI in general, is the potential for partial volume effects during infarct volume quantification. The use of 3D DEMRI images mitigated the partial volume effects by employing a thinner slice thickness (1.4mm) than standard 2D DEMRI images (10mm). As described above, a subset of DEMRI infarct volumes were analyzed using a semiautomatic FWHM method, and infarct sizes were not statistically different from visual tracing results. Our results are consistent with two recent studies, in which manual tracing of DEMRI volumes was equivalent to both FWHM- and Standard Deviation-based semiautomatic methods39, 40
. While some partial volume effect may contribute to border zone signal, the TEM analysis shows relative preservation of cytoarchitecture in this region, lending support to the overlap of the positive MEMRI and DEMRI signals in the border zone territories. These findings are consistent with previous publications that document DEMRI overestimation of infarct volume9–13
. Absolute and relative degrees of DEMRI infarct volume have also been observed to decrease over time following a myocardial infarction, which has been attributed to cell debris removal and resolution of myocardial edema11
. This diminishing zone of positive DEMRI underscores the need for improved imaging approaches to detect viable myocardial territories at early and intermediate time points post-MI with high accuracy to assist decisions on revascularization. Specifically, patients in the peri-infarct period may benefit from an imaging strategy that better delineates the amount of injured, yet viable myocardium that would be jeopardized if not revascularized.
The use of manganese as a MRI contrast agent has been limited by its potential for adverse cardiovascular effects, which is mainly attributed to its competition for the L-type Ca2+
-channel on the sarcolemmal membrane41
. To mitigate these effects, the EVP-1011 used for MEMRI contains calcium as well, and no toxicity was observed in swine dosed at 0.7cc/kg. Indeed, clinical manganese MRI imaging agents are already in use, namely mangafodopir trisodium (MnDPDP), which is FDA-approved for liver tumor imaging42
. Although no approval for cardiac imaging with manganese exists at present, EVP-1011 is currently being evaluated for FDA approval.
T2-weighted imaging (T2WI) has also been validated for characterizing peri-infarct zone biology, including ‘area-at-risk’ (AAR), in both the acute and subacute settings43, 44
. The presence of T2WI edema is particularly useful in the acute coronary syndrome (ACS) setting, as it was recently demonstrated to independently predict the presence of obstructive coronary disease and even 6-month survival in ACS patients45
. However, by 14–28 days post-infarct, myocardial edema may have dissipated completely46,47–49
, which lessens the utility of T2WI at intermediate and chronic timepoints. MEMRI-DEMRI dual contrast imaging affords unique information about infarct zone biology that may be particularly effective at intermediate and chronic timepoints post-infarct. Moreover, in contrast to T2WI, which assesses AAR lying primarily outside
, the MEMRI-DEMRI strategy presented herein detects live cardiomyocyte populations within
An important limitation of this study is that no revascularization data is available to confirm the functional viability of these overlapping DEMRI- and MEMRI-positive regions. Consequently, there is no evidence that these live cells are capable of contributing to overall ventricular function if they were to be salvaged with revascularization. Indeed, these populations of live cells may represent arrhythmogenic foci, as recent work has linked the border or ‘gray’ zone areas of DEMRI, which displays SNR heterogeneity, with increased propensity for inducible ventricular tachycardia51
and cardiovascular events27
. Future revascularization studies would be required to understand how these pockets of viable myocardium may affect long-term outcome, and whether aggressive restoration of coronary blood flow to these incompletely scarred regions is beneficial.
In summary, the infarct border zone represents a heterogeneous region comprised of complex post-infarction biology. In order to better characterize this dynamic process, contrast agents that enable both anatomical (DEMRI) and biological (MEMRI) information may be advantageous. This study demonstrates a non-invasive, dual-contrast MEMRI-DEMRI detection method for viable myocardium within the infarct border zone. Cells in the border zone display relatively intact cytoarchitecture, which suggests salvageable myocardium and/or arrhythmogenic foci. Future studies are needed to determine how these pockets of viable tissue may impact revascularization strategy in ischemic cardiomyopathy.