Distal coronary embolization following acute myocardial infarction increases early infarct size and late left ventricular wall thinning in a porcine model

Our study demonstrates that in a well-controlled model of moderate to severe AMI, there is an incremental effect of DCE associated with increased inflammatory markers and larger infarct size at early time points and increased ventricular wall thinning at later time points. Molecular studies indicated increased MMP-2 activity and decreased Akt phosphorylation within the infarct zone of DCE treated pigs at the early post infarct period, suggesting potential mechanisms for the adverse changes in late ventricular remodelling.

These studies were performed in a porcine model of 60 min of ischemic time prior to the embolization intervention. This ischemic time period was carefully selected to ensure the presence of a clinically relevant, patchy (ie nontransmural) infarct, which would be at least moderate in size. Previous work from our group has shown that extending the ischemic time in this model to 90 min causes transmural, large infarcts with extensive MVO that would be difficult to detect incremental damage by distal embolization. Reducing the ischemic time to 45 min produces a small or even negligible infarct without MVO that would be less representative of the infarct sizes typically encountered in clinical practice [13]. Although 60 min may seem a very short ischemic period for patients presenting with acute myocardial infarction, Hedstrom et al have previously shown large differences in the ischemic time required for 50 % infarction of area at risk between pigs and humans (37 min versus 288 min, respectively) [18]. The 14 % infarct size in our control group seems consistent with the 12.7 % infarct size documented in human studies in which the ischemic time ranges from 2.5 to 6 h, which represents the typical presentation time of most acute myocardial infarctions treated by reperfusion [19]. In contrast, the infarct size of patients presenting beyond the 6 h period was 17.9 %.

NR has been attributed to microvascular injury, due to several factors, including endothelial cell damage, myocyte swelling, inflammatory cell accumulation, reactive oxygen species and platelet-fibrin thrombi [5]. Previous experimental studies in animal models have suggested that treatment of NR may not actually affect infarct size since the microvascular damage was confined histologically to the irreversibly injured core of myocardial necrosis [5, 20]. However, these studies were not performed in an AMI model with distal embolization, which commonly occurs during PCI reperfusion therapy. In the current porcine AMI model, 80 % of pigs had normal (TIMI-3) flow grade after 60 min of ischemia alone (ie. without DCE), which still resulted in a moderate-severe sized infarct with MVO in the majority of pigs. The NR initially present in all of the DCE pigs immediately post injection did improve spontaneously over the 10-min period in 30 % of the pigs. By 3d, there were no longer differences in NR between the DCE and control groups (approximately 30 % in both groups). The presence of no-reflow in control animals at 10 min/3 days and the resolution of no-reflow in DCE animals at 10 min/3 days could not be attributed to differences in perfusion territory size within each group and time point. The spontaneous improvement in TIMI flow grades following DCE in our porcine study parallels the human setting, where NR following primary angioplasty is frequently transient and improves either spontaneously or in response to vasoactive agents during the revascularization procedure [4]. TIMI myocardial perfusion grade in patients with myocardial infarction treated with primary angioplasty has also been shown to improve over the initial few days [21]. However, even transient NR during PCI appears to adversely affect prognosis in patients, including in-hospital, 30 days and 6 months mortality [22, 23].

The DCE group had a modest but non-significant increase in persistent MVO at the early time point. In previous studies, the presence and size of MVO predicts larger infarct size, worse scar thinning and infarction expansion [24]. In the current study, there was an approximately 30 % increase in infarct size at 3d in the DCE group, indicating an initially more extensive infarct in the presence of DCE causing NR. Using CMR to determine the myocardial perfusion territory at risk, our study demonstrated that the increased infarct size at the early time point (3d) in DCE treated animals was due to a more extensive infarction within the perfusion territory at risk (59 % compared to 40 % in controls).

However, it was very interesting to note that at 6 weeks, there were no longer differences in overall infarct size between the two groups. There was a marked reduction (approximately 50 %) in overall infarct size in both groups. An AMI patient study from our group previously reported a 41 % reduction in infarct size in serial CMRs done initially at 1–48 h post AMI and again at 2–3 weeks [25]. This infarct shrinkage has been shown to be due to scar formation and infarct contraction [14]. Of note, these differences in infarct size in the current porcine study did not alter the overall ejection fractions or ventricular volumes. The accentuated infarct thinning in the DCE group would partially explain the lack of differences in infarct size between the two groups at the late time point.

The extent of infarct transmurality has been related to LV remodeling and dysfunction during healing following anterior infarction [26]. Furthermore, increased infarct transmurality measured by LGE CMR has also been identified as a risk factor for life threatening arrhythmias and cardiac death in patients with chronic myocardial infarction [27].

We have previously shown that additional effects of distal embolization include activation of inflammatory (TNF-alpha, IL-6) and thrombotic markers as a consequence of injecting inert polystyrene microspheres into porcine coronary arteries [28]. This activation occurred before the onset of NR, with progressive increase up to the point of NR. In the current study, two potential mechanisms may contribute to the remodeling changes as a result of DCE following ischemia-reperfusion injury. First, there was an increase in pro- and active MMP-2 levels within DCE samples compared to controls. Several studies have indicated that increased MMP levels occur post AMI and may adversely affect LV remodeling. In preclinical AMI models, MMP expression and activity increased following reperfusion [29]. It has also been shown that MMP-2 knockout mice have lower incidence of cardiac rupture and adverse LV remodeling [30]. Clinical studies have demonstrated an association between higher serum MMP-2 and adverse LV remodeling [31]. In our study, the increased myocardial tissue MMP-2 levels at 3d post-AMI in the DCE group may have contributed to the proteolytic milieu that resulted in enhanced LV infarct wall thinning at 6 weeks.

Secondly, our study showed reduced Akt phosphorylation in myocardial samples from the infarct core in the DCE group compared to controls. Akt is constitutively expressed within the myocardium and controls a myriad of cellular functions including cell death, differentiation and metabolism [32]. The activation (phosphorylation) of Akt has been shown to be associated with myocyte survival and reduction in infarct size in a number of pre-clinical studies [33‐35]. The higher level of phosphorylated Akt in the infarct of the control group suggests that a better pro-survival tissue environment may have contributed to the observed smaller early infarct size and less late ventricular thinning seen in controls.

Many studies have described the effects of DCE, without prior AMI, on the myocardium [28, 36‐38]. These studies relied on microspheres to simulate thrombotic material and demonstrated altered coronary flow, contractile dysfunction, and an inflammatory response [38]. Very few studies, however, have investigated the role of microembolization following AMI. Skyschally et al. recently demonstrated that injection of microspheres following 90 min of ischemia resulted in a significant increase in infarct size measured using triphenyl tetrazolium chloride staining [12]. Our results also illustrate an early (3d) increase in infarct size measured using CMR. Skyschally et al., however, did not report the late effects of embolization. The results from our study suggest that the differences in infarct size between groups do not persist into the chronic phase following AMI. Moreover, the use of microspheres by the previous studies may not accurately reflect the endogenous embolization material. We have utilized biologically active microthrombi derived from porcine blood which is more representative of the embolization material present clinically. Previous studies have also utilized microthrombi derived from blood [39], however, these studies have not demonstrated its effects in the context of AMI.