The importance of no-reflow/microvascular obstruction in the STEMI patient

This editorial refers to ‘Relationship between microvascular obstruction and adverse events following primary percutaneous coronary intervention for ST-segement elevation myocardial infarction: an individual patient data pooled analysis from seven randomized trials’†, by S. de Waha et al., on page 3502. Experimental studies showed that when a fluorescent dye, thioflavin S, was injected into the vasculature after removal of a proximal coronary artery occlusion in the dog, portions of the subendocardial microvasculature failed to perfuse.1 Under ultraviolet light, the heart slices revealed well-delineated non-fluorescent zones, devoid of the dye. These defects appeared within secomds of releasing a coronary artery occlusion of 90 min in duration, despite the fact that the epicardial coronary artery was widely patent. Electron microscopy of these non-fluorescent, no-reflow zones showed clear-cut evidence of microvascular damage, including focal swelling of the endothelium with formation of blisters or membrane-bound blebs that protruded into the lumen of capillaries and small blood vessels, which appeared to block the flow of red blood cells. Occasionally, swollen cardiomyocytes appeared to be compressing adjacent vessels. Endothelial pinocytotic vesicles were often absent; breaks in the endothelial continuum were observed, as were occasional fibrin tactoids and microvascular haemorrhage.1 Clearly microvascular damage was contributing to the no-reflow phenomenon. The size of the no-reflow zone was observed to expand over the first few hours of reperfusion, suggesting that this phenomenon was indeed a manifestation of reperfusion injury.2 , 3 In general, zones of no-reflow were confined to the ischaemic risk zone and tended to occur well within the zone of necrosis, as defined by triphenyltetrazolium chloride staining for viable cardiomyocytes. In addition, electron microscopy studies showed that myocyte cell death appeared to precede microvascular damage.4 Whereas it is unlikely that no-reflow contributes directly to myocardial cell death, it can impede the healing process by limiting blood flow into and out of the necrotic zone, reducing those cellular elements important to the removal of necrotic debris and limiting cellular elements and cytokines important to the building of a strong scar. In fact, experimental studies showed that the worse the no-reflow phenomenon, the more severe was infarct thinning and adverse left ventricular remodelling.5 In experimental studies, carried out in animals with otherwise clean coronary arteries, we have observed that neither thrombolytic agents nor anticoagulants have reduced no-reflow size.6 , 7 Therapies such as early reperfusion, pre-conditioning, and hypothermia that consistently reduce myocardial infarct size have reduced no-reflow in our experimental models. No-reflow has been described in numerous clinical reports.8 In the clinical setting, microvascular obstruction may be more complicated than in the experimental studies described above. Atherosclerotic debris and micro-emboli certainly could contribute in the setting of primary percutaneous coronary intervention and exacerbate plugging of the microvasculature on top of the localized endothelial swelling and blebs that are so prominent in the experimental studies. No-reflow in humans is typically diagnosed by imaging studies, including the use of echo contrast agents, nuclear studies, and, most recently, using contrast agents such as gadolinium with cardiac magnetic resonance imaging. Studies by Wu et al.9 showed that the perfusion defects observed using gadolinium and magnetic resonance imaging and referred to as microvascular obstruction correlated closely to the non-fluorescent zones observed using the dye thioflavin S, that we had referred to as ‘no-reflow’. Several studies showed that the extent of microvascular obstruction or no-reflow demonstrated by various imaging techniques correlated with worse clinical outcomes including increases in mortality and adverse left ventricular remodelling.10–14 Some analyses also showed that the presence of no-reflow predicted worse clinical outcome independent of myocardial infarct size.12 , 14 The authors of the paper in the current issue15 are to be congratulated for carrying out the largest analyses to date of the relationship between microvascular obstruction in ST-segment elevation myocardial infarction (STEMI) patients (assessed by cardiac magnetic resonance imaging using late gadolinium enhancement) and clinical outcome. They analysed seven reports from clinical trials that included 1688 patients with STEMI. Median duration of their clinical follow-up was 1 year. Some degree of microvascular obstruction was observed in 57% of the patients, a number that is higher than in many other reports, and should raise awareness of this clinical problem. Their study verifies that the degree of microvascular obstruction correlates with mortality as well as hospitalization for heart failure. In addition, they showed that no-reflow was also more common in patients with left anterior descending coronary artery occlusion, baseline TIMI flow of 0/1, and longer symptom to device times. The presence of microvascular obstruction did not differ by sex, by age, by the presence of hypertension or hyperlipidaemia, or by current smoking. However it was slightly more common in diabetics. No-reflow did not correlate with re-infarction, but this was not really a surprise as re-infarction would most probably involve pathology in the proximal coronary arteries, not the distal microvasculature. The increases in mortality and heart failure are probably related to worse healing with no-reflow that encompasses an increase in myocardial infarct expansion (thinning and stretching of the necrotic wall), regional and global left ventricular dilatation, and eccentric hypertrophy of the non-infarcted walls (all elements of adverse left ventricular remodelling). Of note, microvascular obstruction contributed to mortality independently of myocardial infarct size, a finding that was also observed in previous smaller analyses. Despite the prevalence of no-reflow following STEMI and its independent association with increasing mortality, there have been few large clinical trials of therapies, specifically aimed at reducing no-reflow. The study by de Waha et al.15 details a carefully performed analysis that should make clinicians more aware of the phenomenon. It is a wake-up call for the need to develop new therapies that target this specific aspect of reperfusion injury. What are some of the options? Various vasodilators such as nitroprusside, adenosine, nitroglycerin, and calcium channel blockers including verapamil and diltiazem, and mechanical removal of the thrombus have been used with variable success.16 However, there is a need for larger, systematic clinical trials that focus on therapies for no-reflow. Over the last 20 years, the focus has been on reducing myocardial infarct size. Early reperfusion using primary percutaneous coronary interventions (stenting and/or angioplasty) and thrombolytic agents have reduced myocardial infarct size and improved clinical outcomes. However, as pointed out by Meenes et al.,17 further reductions in door to balloon times have not further improved mortality. Attempts to reduce myocardial infarct size above and beyond that achieved with early reperfusion with adjunctive pharmacological agents have mostly been disappointing in the clinical realm, although a few have been effective (such as high dose i.v. adenosine infusions and hyperoxygenation). Recently, studies of remote ischaemic conditioning have shown a reduction in biomarkers of necrotic myocardium and improvement in the salvage index in patients with STEMI.18 Attempts to reduce adverse left ventricular remodelling have been successful, but reductions in adverse clinical outcomes with agents such as angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, and beta-blockers have in general been modest (with reductions in adverse events of ∼20–25%). The time has come for fresh approaches to improve further outcomes following STEMI. One-month mortality rates are still >10% in some studies, and nearly a quarter of older patients develop heart failure after their STEMI. Microvascular obstruction/no-reflow is a common problem in STEMI and represents a fresh and deserving target for therapy. One approach might be to consider developing therapies that target microvascular damage that do not necessarily affect infarct size. In studies from our laboratory, agents that reduce infarct size must be on board during ischaemia in order to reduce infarct size. In some cases, that could delay treatment in the catheterization laboratory and delay door to device time. However, we have recently observed that a therapy may be given after reperfusion, even as late as 30 min after reperfusion has commenced, and still have a benefit on no-reflow. Using rabbit and rodent models of experimental myocardial infarction, we have observed that initiation of therapeutic hypothermia minutes after re-establishing patency of a proximal coronary artery occlusion markedly reduced no-reflow, without affecting myocardial infarct size.19 , 20 Previous studies from our research group showed that inducing hypothermia during the period of ischaemia reduced myocardial infarct size;21 however, this could potentially slow door to device time. Initiating therapy after re-establishing patency of the large epicardial infarct-related coronary artery would not. Reducing no-reflow, by instituting a single period of hypothermia after reperfusion has occurred, has the potential to prevent myocardial infarct expansion and adverse left ventricular remodelling (including dilatation of the ventricle), and possibly reduce heart failure and mortality. There is an unmet need to study further the no-reflow phenomenon, develop therapies that specifically target no-reflow, and determine the long-term consequences of such therapies. The study by de Waha et al.15 is an important contribution to our understanding of the prevalence of no-reflow in STEMI and its long-term adverse effects on clinical outcome. Hopefully more cardiologists will become aware of its importance after reading this important study. Conflict of interest: R.A.K. receives research support from Faraday Pharmaceuticals and Servier.