Serial changes in the left ventricular (LV) volume and shape described as LV remodeling affect LV function and have been demonstrated to be important predictors of mortality and morbidity after acute myocardial infarction (AMI) (1). Previous studies reported that LV remodeling is regulated by multiple components, such as mechanical, neurohormonal and therapeutic factors (2- 3).

Aldosterone (ALD) displays both myocardial and renal effects that may have profound implications for LV remodeling (4). Increased plasma ALD levels were demonstrated to be associated with increased LV mass in patients with hypertension as well as in a population-based sample (5- 7). In the Randomized Aldactone Evaluation Study (RALES), the ALD receptor antagonist spironolactone was shown to reduce the mortality of patients with congestive heart failure (CHF) (8). Moreover, the substudy of RALES demonstrated that the beneficial outcome in RALES might combine with the effect of spironolactone to suppress cardiac collagen synthesis (9). Furthermore, ALD levels have been found to be increased in some patients with myocardial infarction (MI) with profound implications for cardiac remodeling and long-term prognosis (10- 11). Moreover, we recently demonstrated that the suppression of plasma ALD during the acute phase of AMI prevented post-infarct LV remodeling (12). Modena et al. (13) reported that combination therapy with an angiotensin-converting enzyme (ACE) inhibitor and an ALD receptor antagonist nine days after onset of AMI could prevent LV remodeling compared with an ACE inhibitor alone. These results suggested that endogenous ALD and a mineralocorticoid receptor may play important roles in the progression of post-infarct LV remodeling.

We reported that plasma ALD was extracted through the heart in patients with chronic CHF (14) and that the transcardiac gradient of plasma ALD was correlated with LV end-diastolic volume index (LVEDVI) and plasma levels of procollagen type III aminoterminal peptide (PIIINP), a marker of myocardial fibrosis (15), including patients with MI after more than three months. However, the cause and effect of the relation between the transcardiac gradient of plasma ALD and LVEDVI remain unknown because we could not repeatedly measure these parameters.

Recently, it has been reported that the mineralocorticoid receptor, which mediates the action of ALD, is expressed in cardiomyocytes, endothelial cells and fibroblasts in the human heart (16- 19). In addition, the existence and the increase of ALD synthase in rats with infarcted hearts has been reported (20), suggesting the possibility of cardiac production of ALD in patients with AMI. However, whether plasma ALD is extracted or produced through the heart in patients with AMI has not yet been elucidated.

Therefore, in this study we evaluated: 1) whether plasma ALD is extracted or produced through the heart during the acute phase, and 2) if plasma ALD is extracted, whether the transcardiac extraction of plasma ALD during the acute phase is related to LV remodeling in patients with AMI.

We prospectively studied 64 patients with first MI. Admission criteria was the same as previously reported (12). Patients who had prior MI, significant stenosis of an infarct-nonrelated coronary artery and residual stenosis (>70%) of an infarct-related coronary artery were excluded from this study.

Patients were classified into the high-extraction group or the low-extraction group, based on the transcardiac extraction of ALD. The cutoff level was the median value for transcardiac extraction of ALD at the acute phase.

All patients gave informed consent, and the study was approved by the Committee on Human Investigation at our institution.

All patients underwent cardiac catheterization by the femoral approach. The infarct-related artery was visualized in five views with contrast injections. Patients with persistent occlusion of the infarct-related vessel underwent percutaneous transluminal coronary angioplasty after standard techniques. The patients who could not obtain less than 70% patency and/or Thrombolysis In Myocardial Infarction (TIMI) flow grade 3 were excluded from this study.

After obtaining revascularization (patency ≥70% and TIMI flow grade 3) and assuring hemodynamic stability, right-sided cardiac catheterization using a 7F Swan-Ganz catheter and measurement of LV end-diastolic pressure using a 5F pig tail catheter were performed.

Blood samples for measuring plasma ALD were collected simultaneously from the aortic root (Ao) and coronary sinus (CS) as previous reported (14). Blood samples for measuring the plasma levels of PIIINP were drawn from the CS. After angioplasty and blood sampling were done, contrast left ventriculography was performed.

After admission to the coronary care unit, all patients received oral aspirin and/or ticlopidine. Angiotensin-converting enzyme inhibitors (enalapril) were administrated to all patients. If required, oral nitrates, calcium antagonists, beta-adrenergic blocking agents and/or diuretics were added and continued. However, ALD receptor antagonists such as spironolactone were prohibited in this study.

Repeat cardiac catheterization and contrast left ventriculography were performed one month after the initial catheterization to determine culprit artery patency and LV function. Left ventriculography performed by contrast medium were analyzed for LV ejection fraction (LVEF) and LV volume by cardiologists who were unaware of the patients’ data at the acute phase and after one month. Left ventricular ejection fraction was calculated by the area-length method. Hemodynamic measurement by Swan-Ganz catheter and blood sampling from the Ao and CS were also performed. Patients with significant restenosis (>70%) of the culprit lesion were excluded from the study.

Plasma concentrations of atrial natriuretic peptide and brain natriuretic peptide were measured with a specific immunoradiometric assay using a commercial kit (Shionogi, Osaka, Japan) as previously reported (21). Plasma ALD levels were measured using a commercial radioimmunoassay kit, and plasma levels of PIIINP were measured with a specific immunoradiometric assay using a commercial kit (CIS Bio International, Nagoya, Japan) as previously reported (14).

All results are expressed as the mean ± SEM. Categoric data were compared against a chi-squared distribution. Student t test was used for continuous variables between groups, and paired t test was used for within-group comparison. Linear regression analysis was used to determine the relation between continuous variables. To evaluate the contribution of ALD extraction at the acute phase to LVEDVI one month after onset, univariate and stepwise multivariate analysis were used among the 12 variables. A p value of <0.05 was considered significant.

Sixty-four consecutive patients who met the entry criteria were enrolled (Table le1). One patient died of lethal arrhythmia; two died of CHF. Four patients were excluded from this study due to restenosis of the culprit coronary artery. Therefore, 57 of 64 patients enrolled in the trial completed the entire protocol. There were no differences in baseline characteristics including infarct lesion, in-hospital therapy and the dose of ACE inhibitor (enalapril). However, maximum creatine phosphate (CK) was significantly higher in the high-extraction group than it was in the low-extraction group.

In 57 consecutive patients with AMI, plasma ALD at the acute phase was significantly lower in the CS than it was in the Ao (84.7 ± 6.3 pg/ml vs. 105.5 ± 8.0 pg/ml, p < 0.0001) (Figure 1). One month after onset, plasma ALD was also significantly lower in the CS compared with that in the Ao (46.2 ± 2.9 pg/ml vs. 59.8 ± 3.5 pg/ml, p < 0.0001).

At the acute phase, mean pulmonary arterial pressure, mean right atrial pressure and pulmonary capillary wedge pressure were significantly higher in the high-extraction group compared with that in the low-extraction group (Table le2). One month after onset, there were no significant differences between the two groups in all hemodynamic parameters.

Regarding LV function and volumes at the acute phase, there were no significant differences in LVEF, LVEDVI and LV end-systolic volume index (LVESVI). However, after one month, the LVEF was significantly lower in the high-extraction group compared with that in the low-extraction group, and LVEDVI and LVESVI were significantly higher in the high-extraction group than they were in the low-extraction group. The absolute change in LVEDVI was significantly higher in the high-extraction group than it was in low-extraction group.

The plasma ALD level in Ao at the acute phase and after one month were significantly higher in the high-extraction group and in the low-extraction group (Table le3). The transcardiac ALD extraction at the acute phase and after one month were also significantly higher in the high-extraction group compared with those in the low-extraction group. The plasma PIIINP level in the CS after one month was significantly higher in the high-extraction group compared with that in the low-extraction group, while there was no significant difference observed at the acute phase.

There was a significant positive correlation between the transcardiac extraction of ALD at the acute phase and the LVEDVI after one month (Figure 2). Moreover, there was also a significant positive correlation between the transcardiac extraction of ALD at the acute phase and the absolute change in LVEDVI (Figure 2). There was a significant positive correlation between the transcardiac extraction of ALD at the acute phase and the plasma level of PIIINP in CS after one month (Figure 3), and there was also a significant positive correlation between the plasma level of PIIINP in CS and LVEDVI after one month (Figure 3).

(Figure 4) shows the significant positive correlation between the plasma level of ALD in the Ao at the acute phase, the transcardiac extraction of ALD at the acute phase (r = 0.735, p < 0.0001) and the transcardiac extraction of ALD after one month (r = 0.539, p < 0.0001).

(Table le4)shows the results of univariate and multivariate analysis among 12 variables related to the acute phase to assess factors regulating LVEDVI one month after onset. According to stepwise multivariate analysis, only a high level of transcardiac extraction of ALD at the acute phase (p < 0.0001), a high level of maximum CK (p = 0.0010) and poor performance of LVEF at the acute phase (p = 0.0168) were significant independent predictors of a large LVEDVI after one month.

The results of this study provide the first evidence that: 1) plasma ALD is extracted through the heart at the acute phase of AMI, 2) the transcardiac gradient of plasma ALD at the acute phase correlates with the LVEDVI and plasma levels of PIIINP one month after onset. These results indicate that ALD is extracted through the heart and ALD extraction stimulates LV remodeling in patients with AMI.

Recently, it has been shown that the mineralocorticoid receptor, which mediates the action of ALD, is expressed in the human heart (16- 19). Although ALD synthase was not detected in the autopsy sample of normal subjects (19), the existence and upregulation of ALD synthase in the heart was reported in rats with MI, suggesting cardiac production of ALD in patients with AMI. However, whether plasma ALD is extracted or produced through the heart in patients with AMI has not been yet elucidated. In this study, we demonstrated for the first time that plasma ALD is extracted through the heart at the acute phase of AMI. We previously reported (14) that about 20% of the plasma ALD level was extracted between the Ao and the CS in patients with CHF, including previous MI more than three months after the onset. This study showed that the ALD extraction ratio through the heart was about 17% and 20% at the acute phase and at one month, respectively. Taken together with our previous study, ALD extraction through the heart was sustained from the acute stage to the chronic stage of MI. Although our findings did not deny the possibility of local ALD production in the heart in patients with AMI, the predominant extraction of ALD was shown because the transcardiac gradient of ALD reflects the total net of ALD extraction and production through the heart. Silvestre et al. (20) reported that both ALD synthase messenger RNA and ALD level increased by 2- to 3.7-fold in the ventricle of rats with MI, while plasma ALD level was not increased compared with that in sham-operated rat. Therefore, local production of ALD in the heart may not be sufficient to increase the plasma ALD level, which is consistent with our findings.

In this study, we demonstrated a significant positive correlation of the transcardiac ALD gradient at the acute phase with LVEDVI after one month and the absolute change in LVEDVI in patients with AMI, indicating that circulating ALD is extracted through the heart and promotes cardiac fibrosis and collagen synthesis via the mineralocorticoid receptor in these patients.

Left ventricular remodeling was shown to be regulated by multiple factors, including mechanical, neurohumoral and therapeutic factors (2- 3). According to stepwise multivariate analysis, only a high transcardiac gradient of plasma ALD at the acute phase, high levels of maximum CK and poor LVEF at the acute phase among the 12 acute phase variables were significant independent predictors of a large LVEDVI at one month, suggesting that extraction of ALD, at least at the acute phase, as well as infarct size play a significant role in modulating LV remodeling after AMI. With regard to therapy during the acute to subacute period, there was no difference between the two groups in therapeutic strategy, including mechanical support or use of drugs including the dose of ACE inhibitor. In this study, all patients received revascularization therapy and oral ACE inhibitor administration, which have been shown to prevent LV remodeling after AMI (22- 23).

Changes in PIIINP that have been shown to be induced by AMI in humans may reflect both synthesis and degradation of collagen (15,24). Aldosterone inhibition has been shown to reduce post-infarction collagen synthesis and progressive LV dilation in patients with subacute MI (13). We found a positive correlation in patients with AMI between transcardiac ALD extraction at the acute phase and the plasma PIIINP level in CS one month after onset. Previous findings that ALD stimulated collagen synthesis in isolated fibroblasts (25) and that PIIINP is a biochemical marker of myocardial fibrosis (15) support our results. In addition, plasma PIIINP after one month was significantly higher in the high-extraction group than it was in the low-extraction group, suggesting that ALD extraction promotes ventricular fibrosis. Furthermore, high levels of plasma PIIINP, in relation to ventricular fibrosis, were reported to be associated with poor LV function, remodeling and prognosis (15,26), also supporting our data.

Therefore, our hypothesis, that the sustained extraction of ALD is an important modulator of the progression of post-infarct LV remodeling, is supported not only by the hemodynamic parameters but also biochemical markers.

We demonstrated that transcardiac extraction of plasma ALD at the acute phase correlates with LVEDVI after one month of onset and that plasma ALD extraction at the acute phase also correlates with plasma levels of ALD in the Ao at the acute phase. Therefore, it would be possible to predict the magnitude of LV remodeling by measuring the plasma ALD level. The effect of ALD on cardiac fibrosis was shown to be a direct effect of ALD interacting with mineralocorticoid receptors located on cardiomyocytes (18) or an indirect effect mediated by upregulation of angiotensin II receptors (27). In this study, all the patients were administered ACE inhibitors; therefore, most of the effect of ALD on LV remodeling in this study is considered to be an effect conveyed via mineralocorticoid receptors. Moreover, the acute expansion of the infarct region of the LV wall is reported to occur within one week. Taken together with this study, our findings suggest that administration of the mineralocorticoid receptor antagonist immediately after AMI is useful in combination with ACE inhibitors to prevent post-infarct LV remodeling.

Patients with multivessel disease and prior MI were excluded from this study. This study also excluded patients with restenosis of the culprit lesion because significant stenosis (>70%) might reduce coronary flow and perfusion. To verify the effect of ALD extraction under such complex conditions, further studies are needed. Being unable to measure the total amount of ALD extraction because the CS flow was not measured is also a limitation of this study. Further studies are needed to evaluate how the amount of ALD extraction affects LV performance.

In patients with AMI, plasma ALD at the acute phase was significantly lower in the CS than it was in the Ao, suggesting ALD extraction across the heart. The transcardiac gradient of plasma ALD at the acute phase was correlated with the plasma levels of PIIINP and LVEDVI one month after onset. Moreover, transcardiac ALD extraction at the acute phase affected LVEDVI after one month, independent of infarct size. These findings suggest that elevated transcardiac ALD extraction through the heart plays an important role in the regulation of post-infarct LV remodeling in patients with AMI.

The authors thank Ms. Ikuko Sakaguchi for excellent technical assistance and Mr. Daniel Mrozek for assistance in preparing the manuscript.