Statistics from Altmetric.com
- AMI, acute myocardial infarction
- AI, apoptotic index
- COX-2, cyclo-oxygenase-2
- HF, heart failure
- LV, left ventricular
- PCNA, proliferating cell nuclear antigen
- TUNEL, transferase mediated biotinylated UTP nick end labelling
Heart failure (HF) is less prevalent and has a better prognosis in women than in men. Also, the outlook after acute myocardial infarction (AMI) is more favourable in elderly women.1 Such sex related differences are probably related to sex specific cardiac remodelling processes.2 However, despite several hypothetical pathogenetic mechanisms, there are no established explanations.
Myocardial apoptosis is a pivotal determinant of left ventricular (LV) dysfunction and cardiac failure in experimental and clinical studies, including after AMI. Sex may indeed influence apoptosis, as shown in normal aging and in end stage HF.3 However, little is known about the role of sex in post-infarction apoptosis. We thus evaluated the influence of sex on expression of pro-apoptotic mediators and myocardiocyte apoptotic index (AI) in subjects dying after AMI.
Using established methods,4 we selected six females who had died after AMI with permanent infarct related artery occlusion at necropsy (< 30 hours after death); 15 males with similar characteristics were selected during the same period. Three men and one woman dying from non-cardiac causes were included as controls.
Re-infarction was excluded on clinicopathologic grounds in all cases. Infarct size was quantified at gross pathology on a four grade scale (small, moderate, moderate to extensive, and extensive).
Tissue specimens were obtained at peri-infarct and remote sites. In situ end labelling of DNA fragmentation by transferase-mediated biotinylated UTP nick end labelling (TUNEL) was performed (Apoptag, Oncor) and sections were stained with antibodies against muscle actin (DAKO-Carpintera; dilution 1:50), and activated caspase-3 (Cell Signaling Technology; dilution 1:50). Myocardiocytes were defined as apoptotic by co-localisation of TUNEL and caspase-3. Immunohistochemical staining for proliferating cell nuclear antigen (PCNA) (DAKO; dilution 1:100) and SC-35 (Sigma; dilution 1:200) was also performed. AI was defined as the ratio of myocardiocytes co-expressing TUNEL and caspase-3 positivity on nucleated cells per field (×250), calculated on 100 random fields. Muscle actin negative cells, TUNEL positive but PCNA or SC-35 positive cells were not included in cell counts. Immunohistochemistry for myocardial expression of the pro-apoptotic mitochondrial protein bax was performed in 14 cases (four females, 10 males; Santa Cruz; dilution 1:100). Immunoreactivity was quantified as percentage of positive myocardiocytes per field (×250) in an average of 50 fields. Immunohistochemical staining for the expression of cyclo-oxygenase-2 (COX-2), a potential marker of subacute myocardial ischaemia, was performed in four cases using a primary anti-COX-2 antibody (Santa Cruz; dilution 1:100). Results were described on a dicothomic (positive/negative) scale.
SPSS 10.1 was used as indicated for χ2 (Fisher corrected where appropriate), Mann-Whitney, Wilcoxon, Kruskal-Wallis, and Spearman tests, with two tailed significance at 0.05, were also used. Multivariable analysis was performed using a generalised linear model including univariate predictors of AI at the 0.10 level.
Demographic, clinical, and pathologic characteristics were not different between males and females (table 1).
Myocardial AI in peri-infarct regions correlated significantly with parameters of unfavourable LV remodelling (LV thickness, R = −0.48, p = 0.029, diameter to thickness ratio, R = +0.56, p = 0.008). Comparing specific patterns of remodelling (from compensatory hypertrophy to eccentric hypertrophy to overt dilatation), AI appeared lower in subjects with favourable LV remodelling, and significantly higher in those with eccentric hypertrophy or overt dilation (p = 0.046).
Myocardial AI in peri-infarct myocardium was 10-fold higher in men than in women, 25.9% (17.9–28.6) v 2.6% (0.5–14.1), p = 0.003 (fig 1A). A non-significant trend towards a twofold increase in AI in men (p = 0.066) was found at remote sites (fig 1B). Moreover, AI was significantly higher at peri-infarct versus remote regions in the overall population (p < 0.001), as well as in males (p = 0.001) and females only (p = 0.027). Apoptotic myocardiocytes were extremely rare in control hearts (0.01%, 0.01–0.02%), with no sex differences.
A higher number of myocardiocytes expressing cytoplasmic bax immunostaining was found in peri-infarct regions in males versus females, 55% (44–61) v 14% (2.5–42), p = 0.012) (fig 1C). A non-significant trend was also present in remote areas, 2.7% (0.8–4.1) v 0.9% (0.4–3.0), p = 0.23) (fig 1D). Moreover, bax expression at the peri-infarct site was significantly associated with macroscopic signs of unfavourable LV remodelling such as thickness (R = −0.68, p = 0.008), and diameter to thickness ratio (R = +0.82, p < 0.001) (fig 1E). The link between myocardial apoptosis, bax expression, and ischaemia was further strengthened by the finding of intense immunostaining for COX-2 myocardial expression in peri-infarct regions in subjects with higher than median myocardial apoptosis and bax staining (fig 2) (to see fig 2 visit the Heart website—http://www.heartjnl.com/supplement).
Multivariable analysis, assessing the independent role of the four borderline univariate predictors of increased peri-infarct AI (age, sex, HF, and trauma as cause of death), showed maleness to be a significant predictor of AI (p = 0.039).
To the best of our knowledge, the present study shows for the first time that in subjects dying after AMI peri-infarct apoptosis is greater in men than women. In this clinicopathologic model, males also show increased post-infarction myocardial expression of bax in comparison to females.
HF is more frequent and severe in men than in women.1 Sex related differences in cardiac remodelling are probably among the explanations of such sex differences in clinical outcomes.2 However, underlying mechanisms are still unclear, but may include the actions of oestrogens, telomerase, Akt, sodium/calcium exchanger, β adrenergic receptors, and the renin–angiotensin system.
Similarly to findings by Guerra and colleagues,3 we show sex specific differences in AI after AMI. We can thus hypothesise that males and females may have a different modulation of the apoptotic pathway in the peri-infarct region; females could be partially protected from ischaemia induced bax expression and pro-apoptotic activity in comparison to men, similar to findings in cardiac fibroblasts.5
A cause–effect link between sex, apoptosis, and LV remodelling cannot however be demonstrated by the present observational data, and great caution should be applied when extrapolating our results to clinical practice as the study was limited by the small sample size and the non-experimental design.
In conclusion, myocardial apoptosis in peri-infarct areas is higher in males dying late after AMI than in females, and this difference is associated with increased myocardial bax expression. These findings may explain the more aggressive course of post-infarction HF in men and the relatively more benign post-infarction remodelling in women, potentially suggesting an increased resistance to ischaemia in females. The recognition and thorough characterisation of the influences of sex on cardiac disease may indeed provide keys to cardiovascular pathophysiology, which may eventually benefit both sexes.
This study was supported by MIUR, Second University of Naples and FUTURA Inc. grants.
Dr G Biondi-Zoccai presented this study as an abstract at the 2002 Italian Society of Cardiology Congress winning a Young Investigator Award.
Figure 2 Myocardial specimens from peri-infarct region (original magnification x600): (A) Stained for in istue end labelling of DNA fragmentation (TUNEL). (B) Co-stained for TUNEL and activated caspase-3 expression. (C) Stained for cyclo-oxygenase-2 expression. (D) Stained for bax expression.
Files in this Data Supplement:
Correct Author list
Please note that the author list is incorrect. The corrected list is shown here:
G G L Biondi-Zoccai, A Abbate, R Bussani, D Camilot, F De Giorgio, M-P Di Marino, F Silvestri, F Baldi, L M Biasucci, and A Baldi
The error is much regretted
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.