The best possible outcome in acute myocardial infarction (AMI) is the absence of any residual myocardial damage. This condition, usually obtained by early and effective reperfusion therapy, has been defined as aborted infarction.1 The classic diagnosis of aborted infarction is based on the detection of a minimal increase in cardiac enzymes together with a favourable evolution of the electrocardiographic modifications characteristic of AMI, in particular without Q-wave appearance.2 Although first used in the setting of thrombolytic therapy, the concept that AMI might abort is independent of therapy.
Rimar et al. reported the data of a registry including 2,382 AMI patients and identified a group of 98 patients with signs of spontaneous reperfusion: they had a better prognosis than the remaining patients and in 25% of cases fulfilled the enzymatic and electrocardiographic criteria of aborted infarction.3 Spontaneous reperfusion has also been postulated to interpret the apical ballooning syndrome (best known as Takotsubo cardiomyopathy) as a peculiar form of aborted infarction.4,5 In these patients, symptoms and signs highly suggestive of ST-elevation AMI (including severe but transient wall motion abnormalities usually located in the apical region of the left ventricle) are observed in the absence of detectable epicardial vessel obstructions on coronary angiography.6 However, the hypothesis that this syndrome may be caused by the complete spontaneous lysis of intracoronary thrombi remains in large part speculative and is not the most widely accepted.6 Another condition that must be considered in the differential diagnosis of aborted infarction is masquerading infarction.7 Under this definition are included the several diseases that may mimic ST-elevation AMI but are caused by other aetiologies and therefore do not entail myocardial necrosis and the related signs (such as enzyme rise or electrocardiographic evolutionary changes). Among these are acute pericarditis, ischaemia in prior AMI, left ventricular (LV) aneurysm, LV hypertrophy (LVH), early depolarisation, pre-excitation syndromes, Brugada syndrome, left bundle branch block and aortic dissection.8
Aborted Infarction and Thrombolysis
Most experience on the issue of MI comes from the setting of thrombolytic therapy, and particularly of early, pre-hospital treatment. Even before that, the notion of aborted infarction was established; various studies had demonstrated that anticipating thrombolytic treatment improved the rate of ST-segment resolution and reduced the incidence of Q-wave infarctions.9–11
The definition of aborted infarction relied first on the detection of symptoms and electrocardiographic signs suggestive of AMI (to exclude masquerading infarction and to avoid inappropriate thrombolysis), and then on the finding of a minor rise in cardiac enzymes (usually a creatine kinase [CK] rise less than two times the upper limit of normal or <250U/l, or a CK-MB peak <16U/l) together with typical evolutionary electrocardiographic changes (usually a prompt decrease in ST-segment elevation early after completion of thrombolysis, without subsequent Q-wave development).2,7,12–14 However, it must be considered that there is not a single, recognised definition of aborted infarction.15 Weaver et al., who first proposed the designation of ‘aborted infarction’, reported in the Microplasmin In Treatment of Ischemic stroke (MITI) trial the absence of scintigraphic signs of infarct scar in 40% of patients treated within three hours of symptom onset.1 The incidence of aborted infarction in the first study entirely devoted to this entity was 13% in the group treated with pre-hospital thrombolysis versus 4% in the in-hospital group.2 The same authors report in two subsequent studies their experience on other patient populations submitted to pre- versus in-hospital thrombolysis and demonstrated that the rate of aborted infarction ranged between 15.3 and 18.2% in the former group and was 4.5% in the latter group.13,14
In a site participating in the Global Utilization of Streptokinase and t-PA for Occluded Coronary Arteries (GUSTO-I) trial, Bahr et al. registered the presence of aborted infarction (defined as CK-MB <16U/l) in 10% of patients and minor myocardial damage (CK-MB <40U/l) in an additional 7% of subjects.12 The presence of prodromal angina was a major predictor of the later achievement of AMI abortion, while time to treatment was not found to be relevant in this cohort.12 The proportion of patients with aborted infarction and the importance of early treatment have since been confirmed by a large study of 5,470 patients enrolled in the ASsessment of the Safety and Efficacy of a New Thrombolytic regimen (ASSENT-3) trial, which registered an overall incidence of aborted infarction of 13.3%, with an increase to 25% in the subgroup treated less than one hour from symptom onset.7Figure 1 summarises the per cent incidence of aborted infarction over all AMI patients in these various studies.
The achievement of aborted infarction can be considered a favourable result in the setting of reperfusion therapy on the basis of its prognostic implications (see Figure 2). The one-year mortality of patients with aborted infarction in the study by Lamfers et al. was 2.2%, versus 11.6% in patients with established infarction.13 In the study by Bahr et al., the prognosis in the patients with aborted infarction was excellent, with a five-year death rate of only 5.9%, versus 24.5% in the patients with extensive necrosis.12 Taher et al. found less impressive differences: 3.9 versus 4.6% at 30 days and 7 versus 7.4% at one year for aborted and established infarctions, respectively.7 The baseline-adjusted mortality was nevertheless different between the two groups, with significantly lower odds ratios for the patients with aborted infarction: 0.76 for 30 days (p=0.005) and 0.70 for one year (p<0.05).7 In particular, the patients with aborted infarction and high-grade (>70%) ST-segment resolution represent a very-low-risk group (1% mortality at 30 days and 2.7% mortality at one year) compared with the remaining patients with aborted infarction, who had 5.9 and 9.3% mortality at 30 days and one year, respectively (p<0.002).7 On the basis of this evidence, the use of aborted infarction as a possible surrogate end-point in trials of ST-segment elevation AMI has recently been proposed.8
Aborted Infarction and Primary Percutaneous Coronary Intervention
Nowadays, the most advanced therapeutic regimen in AMI includes the execution of direct percutaneous coronary intervention (PCI).16 In this setting, the enzymatic measurements of infarct size are less reliable, particularly in the presence of reperfusion.17,18 Similarly, the electrocardiographic criteria used in the patients submitted to thrombolysis are probably inadequate in the setting of primary PCI.19 Indeed, there are no data about the incidence of aborted infarction, as defined by the usual enzymatic and electrocardiographic criteria, after primary PCI.8
Various studies have examined the extent of myocardial salvage achieved with PCI by comparing the scintigraphic area at risk before treatment and the post-reperfusion infarct size, showing that an almost complete perfusion recovery can be obtained.20–22 Moreover, the scintigraphic infarct size is effectively used as a surrogate end-point for assessing the efficacy of different therapeutic strategies in AMI patients.17,23
Finally, in their first definition of aborted infarction, Weaver et al. used the scintigraphic assessment of infarct size by thallium scan acquired approximately one month after AMI to classify the presence or absence of myocardial damage.1 Therefore, it could be reasonable to use the scintigraphic infarct size as a criterion to identify the occurrence of aborted infarction after primary PCI. Furthermore, the possibility of accurately evaluating both regional and global LV function through the electrocardiographic gating of perfusion single-photon emission computed tomography (gated SPECT) reinforces the reliability of a scintigraphic definition of aborted infarction, taking into account the most favourable prognostic meaning of a normal LV ejection fraction (LVEF) after MI.24–27
On the basis of these considerations, the author’s group defined as aborted infarction the finding of a completely normal myocardial perfusion with infarct size equal to 0, together with a normal LVEF and a normal regional wall motion and thickening in follow-up perfusion gated SPECT.28 According to these criteria, the incidence of aborted infarction in patients treated by primary PCI was 32 out of 208 patients (15%). In agreement with prior reports, patients with aborted infarction were more frequently female and had an older age than the other subjects.7 The delay of treatment in the authors’ series was longer than in the populations submitted to pre-hospital thrombolysis, but the overall incidence of aborted infarctions remained comparable (see Figure 1).28 This could be explained by the fact that primary PCI is likely more effective in obtaining reperfusion than thrombolysis, so the delay in starting treatment does not preclude the possibility of aborting AMI. A higher incidence of TIMI grade 3 coronary flow was observed before treatment in patients who subsequently fulfilled the criteria of aborted infarction.28 This observation is in agreement with the demonstrated value of spontaneous reperfusion in favouring aborted infarction and with the reports about the protective role of preserved coronary flow in the setting of primary PCI.3,29,30 Patients with aborted infarction also exhibited less potential necrosis before treatment, as suggested by the significantly lower ST-segment elevation in the leads exploring the infarct at admission.28
There are no studies about the prognosis of aborted infarction in patients treated with primary PCI. In principle, there are no grounds to suppose that their outcome should be different from that of the aborted infarctions after thrombolysis. Furthermore, with regard to this point it must be considered that a higher rate of elective PCI after index AMI was observed in patients with aborted infarction after thrombolysis.7
Because the effectiveness of primary PCI in restoring myocardial reperfusion is based on complete vessel reopening and taking into account the increasing use of coronary stenting in this setting, it would be reasonable to expect that this relatively less favourable outcome in patients with aborted infarction was no longer present after primary PCI.
Abortion of AMI may be regarded as the ideal result of any reperfusion strategy. To reach this goal, anticipating reperfusion is certainly important, as demonstrated by the experiences using pre-hospital thrombolysis. On the other hand, the effectiveness of reperfusion is essential and, therefore, primary PCI should nevertheless be preferred, so far as it does not result in a prolonged delay in therapy and it is performed in a centre with adequate expertise. A high incidence of aborted infarction in a population of AMI patients is certainly a marker of treatment success. However, there is not a universally accepted definition of aborted infarction, so that its use as a surrogate end-point for assessing the efficacy of different therapeutic strategies appears still problematic. Taken separately, several of the criteria used to define the presence of aborted infarction, such as limited release of myocardial enzymes, rapid resolution of ST-segment elevation, scintigraphic infarct size and preserved LV function, have already been used with good results as surrogate end-points for evaluating AMI treatment. Whether combining them in a single and established definition of aborted infarction could be helpful should be evaluated.