Persistent Infarct–Related Artery Occlusion Is Associated With an Increased Myocardial Apoptosis at Postmortem Examination in Humans Late After an Acute Myocardial Infarction
Background— Myocardial apoptosis persists beyond the acute phases of acute myocardial infarction (AMI) and is associated with left ventricular (LV) remodeling. Infarct-related artery (IRA) patency is considered a favorable prognostic factor after AMI and may be associated with more favorable LV remodeling because of reduced apoptosis at the site of AMI. The aim of this study was to assess the influence of IRA status on apoptotic rate (AR) in the hearts of subjects dying late after AMI.
Methods and Results— We used colocalization for in situ end-labeling of DNA fragmentation and immunohistochemistry for caspase-3 to calculate the AR at time of death (12 to 62 days after AMI) in 16 hearts with persistently occluded IRAs and in 8 hearts with patent IRAs. No significant differences were found when comparing the clinical characteristics of the 2 groups. Occluded IRA was associated with significantly higher AR at site of infarction (25.8% [interquartile range 20.9% to 28.5%] versus 2.3% [interquartile range 0.6% to 5.0%], P<0.001). This strong correlation between IRA occlusion and AR remained statistically significant even after correction for clinical characteristics such as sex, age, history of previous additional AMI or heart failure, transmural AMI, anterior AMI, fibrinolytic treatment, time from AMI to death, trauma as cause of death, and multivessel coronary disease (P=0.003).
Conclusions— A significantly higher AR was associated with persistent IRA occlusion late post-AMI. These data may suggest that the post-AMI benefits observed with a patent IRA (the “open-artery hypothesis”) may in part be due to reduced myocardial apoptosis.
Received May 28, 2002; revision received July 10, 2002; accepted July 11, 2002.
The “open-artery hypothesis” was clearly formulated in 1993 by Kim and Braunwald.1 Experimental, retrospective, and prospective studies1,2⇓ have described the potential benefits of infarct-related artery (IRA) patency late after acute myocardial infarction (AMI) for reducing left ventricular (LV) remodeling and decreasing mortality (irrespective of the amount of myocardial necrosis). Experimental studies have challenged the idea of necrosis as the prevalent modality of myocardiocyte death in AMI, with a significantly greater number of myocardiocytes undergoing apoptosis than necrosis.3 Recently, apoptotic rate (AR) was shown to be persistently elevated at the site of infarction even late post-AMI in humans4 and animal models,5 and anecdotic data on low grade apoptosis (versus high grade) in 2 hearts of subjects with an open IRA at autopsy suggested a correlation between IRA status and myocardial apoptosis.4 The goal of the present study was to evaluate the influence of IRA patency on AR in the hearts of subjects who died 12 to 62 days after AMI.
Selection of the Samples
We compared 16 hearts collected at autopsy from subjects who died after an AMI and who had persistent IRA occlusion at time of death (as defined by complete absence of residual lumen at pathological examination due to atheroma and/or thrombosis) with 8 hearts from subjects with similar characteristics but patent IRAs at time of death. The cause of death was trauma in 7 cases, whereas congestive heart failure and co-morbidities were present in the others. None of these subjects suffered from a re-infarction, as indicated by clinical data and serial creatine kinase-MB determination in the days preceding death. Characteristics of the patients are shown in Table 1.
Gross examination of the hearts was performed to measure LV parameters and to define the infarcted area and the IRA. Arteries were defined as occluded if there was a failure to demonstrate residual lumen. Tissue specimens were obtained at sites of infarction and in regions of the left ventricle remote from it. In situ end-labeling of DNA fragmentation (TUNEL) was performed using the peroxidase-based Apoptag kit (Oncor). TUNEL-positive cells were detected according to the supplier’s instructions. Several series of TUNEL-stained sections, as well as of other consecutive sections were subsequently stained for different markers. Sections were processed with antibodies against muscle actin (mouse monoclonal anti-human actin HHF35, DAKO, Carpintera, Calif) and caspase-3 (rabbit polyclonal anti-human caspase-3, Upstate Biotechnology, Lake Placid, NY) and visualized by the streptavidin-biotin system (DAKO), using either 3-amino-9-ethylcarbazide or diaminobenzidine as the final chromogen. The optimal working dilutions and the suitable negative and positive controls for both TUNEL and caspase-3 have been defined elsewhere.4 Myocardiocytes were defined as apoptotic if colocalization of markers of DNA fragmentation (TUNEL) and caspase-3 was evident (Figure, A), because high immunohistochemical expression of caspase-3 as detected with the antibody used is apparent in myocardiocytes undergoing apoptosis, colocalizes to TUNEL-positive myocardiocytes, and corresponds mostly to increased expression of its activated form.4,6,7⇓⇓ The AR, expressed as the ratio of number of myocardiocytes co-expressing TUNEL and caspase-3 positivity on nucleated cells per field (250×) at light microscopy, was calculated and compared in different specimens by 2 observers blinded to the protocol. We examined 100 random fields per case, with an average of 90 cells per field. Myocardiocytes co-expressing TUNEL-positivity and specific staining for markers of DNA synthesis (using mouse monoclonal anti-human PCNA PC10 antibody [DAKO]; 0.7% of cells) or markers of transcription activity (RNA splicing factor SC-35; using mouse monoclonal anti SC-35 [Sigma, Milan, Italy]; 10% of cells) were not included in the cell count, following protocol defined elsewhere.4 The correction for possible confounding factors (DNA synthesis and/or RNA splicing) as well as the absence of ongoing necrosis in all specimens further enhanced reliability of the double assay used to define apoptotic cells, as previous reports have discussed potential limitations of the TUNEL technique alone.8
SPSS 10.0 for Windows (SPSS) was used. A χ2 test was used to compare discrete variables. Quantitative results are expressed as median and interquartile range. Non-parametric tests were used to compare ARs among different regions of each subject (Wilcoxon test for paired data) and among different subjects (U Mann Whitney for non-paired data). ANOVA by multiple regression was performed. Clinical factors used in multivariable analyses were sex, age, history of heart failure and/or previous additional AMI (more than 6 months earlier than the most recent AMI), transmural necrosis, anterior AMI, recent use of fibrinolysis for AMI, time to death post-AMI, a traumatic cause of death, and multivessel coronary disease.
Clinical and Pathological Characteristics
No differences in demographic, clinical (including peak creatine kinase levels at time of AMI), and gross pathological characteristics were found between the 2 groups (Table 1). Areas of scarring consistent with previous necrotic cell death were demonstrable in the infarcted area without signs suggestive of ongoing necrosis. All but 4 cases were transmural infarcts. Twelve patients (50%) had previously experience an AMI (more than 6 months before the most recent MI).
The AR at the site of infarction was significantly higher in subjects with IRA occlusions (25.8% [20.9% to 28.5%]) versus patients with patent IRAs (2.3% [0.6% to 4.9%]; P<0.001) (Figure, B). In the overall population, the AR was significantly higher at site of infarction versus remote regions (19.5% [3.0% to 27.8%] versus 0.5% [0.3% to 0.7%]; P<0.001). This was more evident in hearts with IRA occlusions (37-fold increase of apoptosis at site of AMI versus remote site; P<0.001) (Figure, C) and less evident in cases of patent IRA (2.3% [0.6% to 4.9%] versus 0.3% [0.3 to 0.4%]; P=0.027) (Figure, D). No significant differences were demonstrable in relation to clinical variables such as cause of death (trauma versus other causes, P=0.39), time to death (P=0.99), use of fibrinolytics (P=0.77), and history of heart failure (P=0.197) (Table 2). Moreover, IRA occlusion remained significantly associated with elevated rates of myocardial apoptosis even after correction for several clinical and pathological factors (P=0.003) (Table 2).
To the best of our knowledge, the present study shows for the first time that a significantly greater myocardiocyte loss due to apoptosis late post-AMI is present in subjects with persistent IRA occlusion than in those with an open artery. This fact is reliably demonstrated by the presence of a higher rate of myocardiocytes positive for TUNEL and caspase-3 immunocytochemistry in subjects with occluded IRAs. The occurrence of an unfavorable LV remodeling because of progressive LV dilation and dysfunction could be correlated with the presence of apoptosis at the site of infarction late after AMI.4–5⇓
Our data may help to explain the “open-artery hypothesis,”1 which suggests the possibility of long-term post-AMI improved LV function deriving from a patent IRA. The mechanisms underlying this hypothesis are not completely understood. In our study, the association between the finding of a patent IRA at time of death and lower AR at the site of infarction suggests that the clinical benefits observed with an open IRA may be in part due to reduced myocardiocyte loss and, possibly, prevention of LV remodeling.
A cause-effect link, however, cannot be assumed by the data presented in this analysis, and an open artery late after AMI may or may not be associated with the direct cause of reduced apoptosis. Interestingly, a link between coronary artery occlusion and apoptosis even late post-AMI has been suggested by an experimental model of AMI in rats.5 Ligation of the descending anterior artery was associated with persistently elevated ARs at the border zone of infarction up to 12 weeks later, with progressive LV dilatation.5 Moreover, although an earlier article by Gottlieb et al9 suggested an increased apoptotic rate in an ischemia-reperfusion model, a more recent experimental study in rats has shown that although reperfusion after myocardial ischemia accelerates apoptosis in the non-salvageable myocardiocytes, reperfusion is also associated with significantly lower total number of cells undergoing apoptosis.10
Further studies will be necessary to investigate the significance of apoptosis late post-AMI, the potential protective role of an open artery, and the eventual role of other features such as regenerating myocardiocytes.
Our study demonstrates that postmortem samples of human hearts with persistent IRA occlusion late after acute AMI have a 10-fold higher AR than hearts with a patent artery at time of death. These data suggest that the beneficial effects described for open IRA post -AMI may be, at least in part, due to reduced apoptosis at site of infarction.
- ↵Kim CB, Braunwald E. Potential benefits of late reperfusion of infarcted myocardium: the open artery hypothesis. Circulation. 1993; 88: 2426–2436.
- ↵Palojoki E, Saraste A, Eriksson A, et al. Cardiomyocyte apoptosis and ventricular remodeling after myocardial infarction in rats. Am J Physiol. 2001; 280: H2726–H2730.
- ↵Narula J, Pandey P, Arbustini E, et al. Apoptosis in heart failure: release of cytochrome c from mitochondria and activation of caspase-3 in human cardiomyopathy. Proc Natl Acad Sci U S A. 1999; 96: 8144–8149.
- ↵Fliss H, Gattinger D. Apoptosis in ischemic and reperfused rat myocardium. Circ Res. 1996; 79: 949–956.