Coronary Microvascular Injury in Reperfused Acute Myocardial Infarction: A View From an Integrative Perspective
2018; Wiley; Volume: 7; Issue: 21 Linguagem: Inglês
10.1161/jaha.118.009949
ISSN2047-9980
AutoresMurat Sezer, Niels van Royen, Berrin Umman, Zehra Buğra, Heerajnarain Bulluck, Derek J. Hausenloy, Sabahattin Umman,
Tópico(s)Cardiac Ischemia and Reperfusion
ResumoHomeJournal of the American Heart AssociationVol. 7, No. 21Coronary Microvascular Injury in Reperfused Acute Myocardial Infarction: A View From an Integrative Perspective Open AccessReview ArticlePDF/EPUBAboutView PDFView EPUBSections ToolsAdd to favoritesDownload citationsTrack citations ShareShare onFacebookTwitterLinked InMendeleyRedditDiggEmail Jump toOpen AccessReview ArticlePDF/EPUBCoronary Microvascular Injury in Reperfused Acute Myocardial Infarction: A View From an Integrative Perspective Murat Sezer, MD, Niels van Royen, MD, PhD, Berrin Umman, MD, Zehra Bugra, MD, Heerajnarain Bulluck, MD, PhD, Derek J. Hausenloy, and MD, PhD, and Sabahattin UmmanMD Murat SezerMurat Sezer Istanbul Faculty of Medicine, Istanbul University, Istanbul, Turkey , Niels van RoyenNiels van Royen Radboud University Medical Center, Nijmegen, NL , Berrin UmmanBerrin Umman Istanbul Faculty of Medicine, Istanbul University, Istanbul, Turkey , Zehra BugraZehra Bugra Istanbul Faculty of Medicine, Istanbul University, Istanbul, Turkey , Heerajnarain BulluckHeerajnarain Bulluck The Hatter Cardiovascular Institute, Institute of Cardiovascular Science, University College London, London, United Kingdom Papworth Hospital NHS Trust, Cambridge, United Kingdom , Derek J. HausenloyDerek J. Hausenloy The Hatter Cardiovascular Institute, Institute of Cardiovascular Science, University College London, London, United Kingdom Papworth Hospital NHS Trust, Cambridge, United Kingdom National Heart Research Institute Singapore, National Heart Centre Singapore, Singapore Cardiovascular and Metabolic Disorders Program, Duke‐National University of Singapore, Singapore Yong Loo Lin School of Medicine, National University Singapore, Singapore The National Institute of Health Research University College London Hospitals Biomedical Research Centre, London, United Kingdom Barts Heart Centre, St Bartholomew's Hospital, London, United Kingdom , and Sabahattin UmmanSabahattin Umman Istanbul Faculty of Medicine, Istanbul University, Istanbul, Turkey Originally published24 Oct 2018https://doi.org/10.1161/JAHA.118.009949Journal of the American Heart Association. 2018;7:e009949Taking its incidence and prognostic importance into account, acute ST‐segment–elevation myocardial infarction (STEMI) can be regarded as one of the most important challenges faced in the field of clinical cardiology. Coronary artery disease and particularly acute myocardial infarction (AMI) are the leading causes of death and disability worldwide.1 Despite remarkable progress in the fight, particularly in the past 3 decades, there is still room for improvement. Indeed, the standardized 1‐year death rate for STEMI has nearly halved over a 25‐year period.2 This decrease in mortality is attributable to outstanding achievement accomplished in the limitation of final myocardial infarct (MI) size by introduction of efficacious reperfusion methods such as fibrinolysis and primary percutaneous coronary intervention (pPCI) during that period. Currently, reopening of the occluded epicardial coronary artery by timely pPCI is widely accepted as the most effective treatment for patients presenting with an acute STEMI in limiting final MI size and preserving left ventricular (LV) function.3 However, despite successful reperfusion by pPCI, mortality (15%)5 and particularly post‐MI morbidity still remain significant at 1 year.1 This disappointing course has been partly attributed to the potential detrimental effects of reperfusion itself. Indeed, reperfusion may lead to a further loss of cardiomyocytes that are succeeded to survive after initial ischemic insult in the subtended myocardial territory.Hemodynamic manifestations of this postreperfusion process include "no‐reflow phenomenon"—severe myocardial malperfusion despite restoration of epicardial coronary patency,8 which has been reported to occur in up to 50% of patients with STEMI following pPCI despite restoration of thrombolysis in myocardial flow 3 in the epicardial coronary artery.11 In general, myocardial no‐reflow phenomenon refers to severe microvascular injury that is known to be associated with impaired LV function12 and poor prognosis14 in patients undergoing successful pPCI. In addition, the magnitude of the preserved microvasculature at the acute phase is one of the major determinants of the long‐term functional and structural myocardial recovery.16 Although it occurs in every patient undergoing pPCI at varying intensity, identification of coronary microvascular injury depends on the diagnostic capability of the method used in its detection. Considering this high incidence and its important clinical consequences,14 a better understanding of the mechanisms underlying severe coronary microvascular injury resulting in myocardial malperfusion (myocardial tissue "no reflow") after pPCI is mandatory to be able to develop efficacious therapeutic interventions for preventing this complication of STEMI. Nevertheless, our current knowledge on the pathophysiology of microvascular damage after pPCI is poor, and hence recommendation of risk prediction and therapeutic interventions could be premature and limited.Notably, given the repeated failure of recent trials aiming to protect microcirculation during pPCI, it is obvious that identification of an appropriate therapeutic target linked to the underpinning mechanism has utmost clinical importance in the treatment of patients who develop extensive microvascular injury after pPCI. Although there is not yet a widely accepted and proven therapeutic intervention to limit postreperfusion microvascular injury, examination of microvascular integrity in the catheterization laboratory immediately upon completion of pPCI17 can provide a unique opportunity for timely identification of this patient subset with severe microvascular injury who may benefit from early adjunctive therapies within the therapeutic window.The aim of this review is to provide an integrative perspective on the pathophysiological mechanisms underlying post‐pPCI coronary microvascular injury and does not consider discussing the potential cellular and molecular mechanisms as cardiomyocyte targets for cardioprotection. In this review, we propose an integrative categorization scheme, which can provide comprehensive pathophysiological insights into post‐pPCI microvascular injury and potentially pave the way for therapeutic targeting of central pathologies behind the myocardial no‐reflow phenomenon in the future.Mechanisms of Microvascular Injury in Reperfused AMIObviously, reestablishing complete and sustained epicardial patency in a timely manner is the most critical step in salvaging ischemic myocardium from impending infarction. However, prompt restoration of coronary flow by reopening the occluded infarct‐related artery can itself paradoxically induce coronary microvascular injury and does not immediately terminate ongoing cardiomyocyte loss at the myocardial area at risk (AAR). During coronary occlusion and after reperfusion, dynamic pathological changes observed in both the microvascular and interstitial territories seem to contribute substantially to this progressive cardiomyocyte damage in the subtended myocardial area.18 Functional and structural consequences of the cortege of pathological changes temporally and spatially evolving at consecutive segments of myocardial circulation seem to determine the fate of the subtended myocardial territory (Table). In this regard, a new classification scheme using compartmental modeling (Figure 1) might help both to elucidate individual contributions of various factors to post‐pPCI microvascular injury and to reappraise any interplay between them.Table 1 Temporal and Spatial Changes Taking Place at the Consecutive Segments of Myocardial Circulation During Coronary Occlusion and After Mechanical ReperfusionPhaseSiteTotal Coronary ResistanceCoronary FlowEpicardial CoronariesArteriolesCapillariesCardiomyocytes and InterstitiumEarly ischemiaOccludedReactive dilatation (adaptive response)···Functional abnormality, cellular edemaImmeasurably high (predominantly epicardial)No‐flowProlonged ischemiaOccludedParalysis (both ischemic insult and adaptive response)Increased permeability, loss of integrity, in situ thrombosisIschemic necrosisImmeasurably high (predominantly epicardial)No‐flowInitial phase of reperfusionReopenedParalysisObstructed by plugs and destructedInterstitial edemaLow (predominantly microvasculara)OverflowLate phase of (established) reperfusionReopenedPartly recovered constrictor responseMore plugged, more leakyDeepened edema, even IMHHigh (predominantly microvasculara)Normal, slow or "no‐reflow"John Wiley & Sons, LtdTotal coronary resistance: epicardial resistance+microvascular resistance. IMH indicates intramyocardial hemorrhage.aUnless there is no additional epicardial lesion rather than the culprit one in the re‐opened infarct‐related artery.Download figureDownload PowerPointFigure 1 Compartmental schematization of the mechanisms behind post–primary percutaneous coronary intervention microvascular injury. Following reperfusion, factors contributing to microvascular injury at the subtended myocardial territory can be categorized under 2 major headings as intraluminal microvascular obstruction and extravascular compression of the microcirculation. Major pathologies contributing to luminal obstruction are distal thromboembolization, cellular plugging, in situ thrombosis, and vasospasm. External compression of microcirculation by cellular and/or interstitial edema, intramyocardial hemorrhage, and increased left ventricular filling pressures may also substantially contribute to microvascular impairment by generating an external compressive force on coronary microcirculation. LVEDP indicates left ventricular end‐diastolic pressure.Indeed, duration of ischemia is considered as the most important determinant of the magnitude of the microvascular damage and its recovery after STEMI.19 However, in a rat model of acute myocardial infarction, it has been recently demonstrated that ischemia alone induces only mild morphological changes in the coronary microcirculation with increased permeability. Nevertheless, ischemia followed by reperfusion has been shown to induce massive microvascular injury.21 Accordingly, on the temporal scale, beginning from the preocclusion period, interactions among the dynamic events evolved during both coronary occlusion and reperfusion seem to determine the eventual magnitude of microvascular damage and cardiomyocyte loss in the subtended microcirculatory territory over time (Figure 2). In addition, the functional status of the patient's coronary microcirculation before the STEMI (preexisting microvascular impairment),22 status of the infarct‐related artery beyond the culprit lesion, perfusion characteristics of the nonculprit vessels, presence of coexistent pathologies such as diabetes mellitus, and overall status and performance of the left ventricle and individual susceptibility23 contribute to the extent of microvascular damage at the subtended myocardial territory following reperfusion.Download figureDownload PowerPointFigure 2 Intraluminal and extravascular factors of microvascular injury operating on the temporal scale (before, during, and after reperfusion). On the temporal scale, in addition to the preexisting microvascular impairment, dynamic events evolving both during coronary occlusion and after reperfusion seem to determine the final magnitude of microvascular damage in the subtended microcirculatory territory. In the preocclusive period, patients' metabolic and inflammatory status and presence of diabetes mellitus, hypercholesterolemia, and hypertension may lead to a preexisting microvascular dysfunction. During epicardial occlusion, local procoagulant activity induced by local hypoxia in the downstream microcirculation may provide extremely suitable milieu for in situ microvascular thrombus formation. During this occlusive period, hypoxia‐induced endothelial disruption may also lead to loss of microvascular barrier function and microvascular leakage. After reperfusion, distal thromboembolization along with edema and intramyocardial hemorrhage developed in this destructed and leaky microcirculatory environment can markedly contribute to microvascular injury.Following reperfusion, mechanisms involved in the development of microvascular impairment can categorically be classified under 2 major headings as (1) intraluminal microvascular obstruction and (2) extravascular compression of the microcirculation. Indeed, intraluminal obstructive and extravascular compressive pathologies both seem to operate interconnectedly in the development of post‐pPCI microvascular injury. Primary contributors to intraluminal microvascular obstruction are distal thromboembolization, cellular plugging, in situ thrombosis, and vasospasm. Besides these intraluminal factors, cellular and/or interstitial edema, intramyocardial hemorrhage, and increased LV filling pressures are the main pathologies that may exacerbate microvascular impairment by generating an external compressive force on coronary microcirculation (Figure 1).Coronary Microvascular ObstructionCoronary Intraluminal PluggingIn the setting of pPCI, coronary microvascular (intraluminal) obstruction can mainly be caused by distal thromboembolism, circulating blood cells plugging and in situ microvascular thrombosis. Intraluminal microvascular obstruction (MVO) is regarded to be the leading pathology behind post‐pPCI myocardial malperfusion. Pathologically, MVO first appears in the infarct core and evolves spatially and temporally, which corresponds with progressive myocardial damage occurring in the AAR after reopening the infarct‐related artery. Consistently, the MVO zone has been shown to increase for up to 48 hours after reperfusion.24 Indeed, after a period of ischemia, blood flow cannot be restored in more than half of the capillaries in the myocardial AAR after reestablishing epicardial patency.25 In ischemic‐reperfused myocardium, myocardial blood flow in certain microvascular areas is hyperemic during the first minutes of reperfusion. Subsequently, regional blood flow in the AAR rapidly and progressively declines26 and reaches a plateau within 2 to 8 hours after reperfusion, resulting in a nearly 3‐fold increase in the anatomic MVO (no‐reflow) zone.27 This delayed progressive fall in blood flow in myocardial areas that initially received adequate perfusion seems to occur due to concurrently activated intraluminal obstructive and extravascular compressive pathologies triggered by reperfusion itself. Experimental19 and clinical studies28 consistently demonstrated a close correlation between zones of anatomic MVO (no‐reflow) and myocardial necrosis. Furthermore, in a rat model of coronary occlusion and reperfusion, it was shown that the no‐reflow phenomenon may persist for 1 month after reperfusion and could predict infarct expansion.29Factors that contribute to luminal MVO after pPCI will be reviewed below as classified in Figure 1.ThromboembolismIn the setting of pPCI, atherothrombotic embolization from the culprit lesion, reported to occur in 11% to 14.5% of the procedures,30 is believed to be one of the leading contributors of MVO occurring during pPCI. Distal embolization of atherothrombotic particles during pPCI may cause both mechanical obstructions by their mass effect and activation of pathways that trigger in situ coagulation and inflammatory responses in the downstream microcirculation.In both clinical and experimental trials, subjects with distal embolization early after reperfusion seem to have larger MI size and more extensive microvascular damage.32 Consistently, in a study where embolized particles during pPCI were quantified using intracoronary Doppler wire as high‐intensity transient signals, it was demonstrated that distal embolization might transiently reduce coronary flow, but this may not have a marked influence on MI size and LV function.34 In addition, it has been reported that angiographically visible distal embolization occurring during pPCI may contribute to myocardial damage within hours after symptom onset, but it does not seem to have a major impact as ischemic time increases.32In this context, it is expected for mechanical retrieval of epicardial thrombotic material using manual aspiration thrombectomy devices during pPCI to substantially limit distal embolization risk, resulting in significant reduction in the MVO zone and MI size. However, although significantly lower distal embolization rates seemed to be achieved with manual thromboaspiration compared with standard pPCI, this did not seem to translate into improved myocardial perfusion,35 contrary to what was initially observed in meta‐analyses.36 Even more importantly, most of the trials37 evaluating the efficacy of manual thrombus aspiration during pPCI failed to show any benefit of the mechanical removal of epicardial thrombus on the limitation of MI size. Furthermore, either routinely41 or selectively43 performed thromboaspiration did not show any clinical benefit in large‐scale clinical trials and in a recent meta‐analysis.44 All these negative results suggest that different approaches to tackling epicardial thrombus to initiate reperfusion, such as standard pPCI or thrombectomy followed by stenting, may not be associated with different pathological and clinical consequences. Moreover, reduction of distal embolization rates by manual thrombectomy does not necessarily translate into decreased infarct size and improved patient outcome. This makes the role of distal embolization from the proximal thrombotic occlusion in microvascular and myocardial damage during pPCI highly controversial and helps in understanding why the trials dealing with epicardial thrombus failed.In light of the current data, it can be concluded that distal thromboembolism may not be sufficient to produce extensive microvascular injury and may contribute to MVO only to a limited extent. Thus, it does not appear to be a major therapeutic target for preserving capillary integrity during pPCI.Cellular Plugging: Role of Circulating Blood CellsLeukocyte and platelet plugging45 and red cell aggregation46 can intensify post‐pPCI MVO. After successful epicardial recanalization, neutrophils can worsen microvascular reperfusion by adhering to the endothelium with platelets47 and by releasing cytokines or other factors48 that may reduce microvascular blood flow. Capillaries in the no‐reflow zone have been shown to contain extensive leukocyte plugging (capillary trapping).25 This leukocyte plugging can also lead to erythrocyte packing and rouleaux formation upstream from mechanical obstruction. In addition, platelets and neutrophils act synergistically in provoking microvascular injury after reperfusion.45 Accordingly, significant relationships have been shown between the presence of MVO and monocyte counts49 and platelet activity.50 In patients who underwent successful pPCI, higher admission neutrophil count,51 platelet volumes51 and neutrophil‐to‐lymphocyte ratio52 were shown to be associated with increased coronary microvascular resistance, suggesting the obstructive role of circulating blood cells supplied by reperfusion in post‐PCI microvascular impairment.In particular, maintenance of blood flow at the coronary microvascular segments, where capillary diameters are reduced under an erythrocyte diameter, largely depends on deformational capabilities of circulating blood cells and endothelial surface layer lubricity. However, following ischemia and reperfusion, the shedding of the glycocalyx layer, which covers the endothelial surface, makes the latter less permeable and more slippery and makes capillaries more prone to be obstructed by cellular plugging, leading to further increase in microvascular resistance.53 Despite these potential negative effects of circulatory cell plugging on microvascular perfusion, several trials that aimed at complement inhibition,55 leukocyte integrin receptor antagonism,56 and increased local platelet inhibition57 all yielded negative results in acute MI patients. Although it seems that circulating blood cells could potentially be involved in intraluminal plugging following pPCI, these might only play a limited role in acute post‐pPCI coronary microvascular impairment. Leukocyte infiltration may be more important in infarct healing and remodeling rather than the determination of the extent of the MVO zone and infarct size. Therefore, the role of leukocytes in postreperfusion myocardial damage is contentious, and they do not seem to be a viable therapeutic target for limiting reperfusion‐related microcirculatory damage.Humoral Factors (In Situ Thrombosis)During epicardial coronary occlusion, the local procoagulant milieu in the downstream microcirculation is extremely suitable for de novo (autochthonous) microvascular thrombus formation. Local hypoxia can immediately precipitate local coagulation by triggering homeostatic mechanisms at the injured endothelium, which may consequently induce microvascular thrombosis and in situ fibrin generation at the site of local damage.58 During occlusion of an epicardial coronary artery, tissue factor expressed from mainly hypoxic and injured endothelial cells, together with stasis, can strongly stimulate the coagulation cascade and de novo fibrin formation at the microvascular level. Additionally, dysfunctional endogen fibrinolysis after reperfusion following an ischemic period, as evidenced by significantly impaired tissue plasminogen activator release from the endothelium,59 can also lead to inadequate removal of fibrin deposits from the coronary microcirculation. Moreover, after its formation, it would not be easy to sweep away this in situ formed fibrin from the lumen via mere restoration of blood flow to the infarcted region because of the former's active adherence to the vessel wall via intercellular cadherin receptors.60 Indeed, all of the formed blood cells are already prone to being attached to the fibrin mesh formed in the microvasculature both passively and actively via the specific fibrin receptors on their surfaces.61 Thus, even a small amount of intraluminal fibrin can constitute a sticky trap for the formed blood cells supplied by reperfusion62 and may impede reperfusion once the occluded epicardial artery is reopened (Figure 3). Although in situ fibrin formation in cardiac microvasculature after ischemia and reperfusion has received almost no attention, autochthonous fibrin depositions in the downstream microcirculation have already been clearly shown in cerebral,63 intestinal,65 and renal66 ischemia‐reperfusion models.Download figureDownload PowerPointFigure 3 A vicious cycle of microvascular thrombosis: During epicardial occlusion, hypoxic endothelial damage, stasis, activated coagulation factors, and tissue factors trigger Virchow triad, which may lead to in situ microvascular fibrin generation. After reopening of the infarct‐related artery, formed blood cells supplied by reperfusion get entrapped in to the microvascular fibrin mesh, which may lead to further stasis and further fibrin generation.In particular, in a condition of slow flow (low shear stress), fibrinogen also contributes to impeding microvascular flow via facilitating red blood cell aggregation,67 mediating leukocyte‐endothelium bridging,68 and facilitating postischemic leukocyte‐thrombocyte interaction.69To this end, effective removal of fibrin(ogen) depositions from the microcirculation may result in better microvascular perfusion. In accordance with this perspective, adjunctive low‐dose intracoronary fibrinolytic drugs (streptokinase, urokinase) given immediately after successful pPCI were shown to be associated with significantly improved microvascular perfusion,70 decreased infarct size,71 preserved LV function71 and improved patient outcome.72 These encouraging results73 also served as a stimulus to further explore the effect of microvascular fibrin deposition removal using different fibrinolytic agent on patient outcome in a large‐scale ongoing clinical trial (T‐TIME [A Trial of Low‐Dose Adjunctive Alteplase During Primary PCI]).74In this context, autochthonous fibrin generation at the downstream coronary microcirculation leading to in situ thrombotic MVO may play a pivotal role in the pathogenesis of microvascular impairment induced by ischemia and reperfusion. However, it is evident that large‐scale clinical trials targeting in situ microvascular thrombosis, such as the T‐TIME trial, are required to reach more conclusive results.VasospasmCoronary microcirculation distal to the acute coronary occlusion is considered as maximally dilated due to exhausted autoregulatory function of arteriolar sphincters. However, even during severe myocardial ischemia, which is the most powerful stimulus known for vasodilation, a pharmacologically recruitable vasodilator reserve may persist.75 Furthermore, during myocardial ischemia/reperfusion (pPCI), the microcirculation remains highly responsive to α‐adrenergic coronary constrictor mediators. Impaired endothelial function by ischemia and reperfusion in conjunction with the release of soluble vasoconstrictor substances, such as serotonin and thromboxane A2 from ruptured plaque and platelet aggregates into the microcirculation may contribute to a vasospastic milieu during pPCI.76 Based on these rationales, the potential benefit of administration of adjunctive intracoronary vasodilators—in particular adenosine, which is a potent direct vasodilator of coronary microcirculation through stimulation of A2 receptors—in reducing microvascular injury in patients undergoing pPCI was examined in multiple clinical trials. Despite previous intriguing results,77 in a recent clinical trial, local adjunctive intracoronary adenosine and sodium nitroprusside administration targeting vasodilation of the subtended microvascular bed were shown to be ineffective in reducing MVO in patients undergoing primary PCI.79 Furthermore, in this latest and most powerful study,79 high‐dose adenosine appeared to be associated with increased infarct size and reduced ejection fraction compared with the control group. These findings suggested that vasodilator agents should not be used in the setting of pPCI to prevent reperfusion injury. In particular, vasomotor function (vasodilator reserve) regulating distal pressure in the reperfused territory can be crucial in obviating an uncontrolled and abrupt increase in capillary pressure during reperfusion that might otherwise be a protective mechanism against myocardial edema and intramyocardial hemorrhage. Therefore, when the negative results of the recent studies using potent vasodilators administered as an adjunct to pPCI are taken into account, vasomotor function at arteriolar level seems to be protected and not be suppressed.In light of the present data, at the current stage, the extent of contribution and potential role of vasospasm in coronary microvascular injury in the pPCI setting are highly controversial. Therefore, adjunctive pharmacological interventions targeting microvascular spasm to prevent microvascular injury seem not to be beneficial in the setting of pPCI.Extravascular Compression of the MicrocirculationExternal compression of the capillary bed by interstitial and cellular myocardial edema and intramyocardial hemorrhage (IMH) developed in the surrounding myocardium after reperfusion substantially contributes to post‐pPCI myocardial malperfusion mainly by increasing total microvascular resistance. Both interstitial and cellular myocardial edema and IMH emerge as a consequence of prolonged ischemia and reperfusion and subsequently become the major contributors to microvascular injury by generating an external compressive force on coronary microvasculature (Figure 4).Download figureDownload PowerPointFigure 4 Myocardial edema and intramyocardial hemorrhage initially emerge as a consequence of prolonged ischemia and reperfusion, and they subsequently become one of the main contributors of the microvascular impairment by generating an external compressive force on coronary microvasculature.The capillaries may not generate a significant resistance to myocardial blood flow in the healthy coronary circulation. However, the capillary network is also the most susceptible compartment to external compression generated by surrounding edema and IMH, as it is assumed to have the lowest radial force. The compressive force causes capillaries to shrink in diameter, with any such decrease resulting in an exponential increase in resistance, which may in turn lead to a substantial decrease in myocardial blood flow. Theoretically, extravascular pressure can simply be calculated as any extra volume added to the myocardial compartment divided by myocardial compliance. Situations like IMH and edema can lead to both an increase in the volume of interstitial space and a decrease in myocardial compliance80 and thus result in a marked increase in extravascular pressure that eventually causes substantial external microvascular compression.During total coronary occlusion, depending on the duration and severity of the ischemia, hypoxia‐induced endothelial disruption leading to loss of microvascular barrier function results in microvascular leakage (MVL),81 which is the central anatomic substrate underlying myocardial edema and hemorrhage occurring after establishment of reperfusion. Consistently, clinical cardiac magnetic resonance (CMR) studies have implicated a loss of microvascular barrier function in AMI manifested as edema and IMH.82 Therefore, edema and IMH, the main determinants of extravascular compressive force, can be regarded as a consequence of MVL. In a recent mouse I/R (Ischemia/Reperfusion) study, late gadolinium enhancement CMR, which is believed
Referência(s)