ORIGINAL ARTICLE

Subendocardial fibrosis in remote myocardium results from reduction of coronary driving pressure during acute infarction in rats

Clovis de Carvalho Frimm; Marcia Kiyomi Koike; Mariana Cúri

University of São Paulo Medical School and Institute of Mathematics and Statistics – University of São Paulo - São Paulo - Brazil

Correspondence

ABSTRACT

OBJECTIVE: To investigate the role of hemodynamic changes occurring during acute MI in subsequent fibrosis deposition within non-MI.

METHODS: By using the rat model of MI, 3 groups of 7 rats each [sham, SMI (MI <30%), and LMI (MI >30%)] were compared. Systemic and left ventricular (LV) hemodynamics were recorded 10 minutes before and after coronary artery ligature. Collagen volume fraction (CVF) was calculated in picrosirius red-stained heart tissue sections 4 weeks later.

RESULTS: Before surgery, all hemodynamic variables were comparable among groups. After surgery, LV end-diastolic pressure increased and coronary driving pressure decreased significantly in the LMI compared with the sham group. LV dP/dt_max and dP/dt_min of both the SMI and LMI groups were statistically different from those of the sham group. CVF within non-MI interventricular septum and right ventricle did not differ between each MI group and the sham group. Otherwise, subendocardial (SE) CVF was statistically greater in the LMI group. SE CVF correlated negatively with post-MI systemic blood pressure and coronary driving pressure, and positively with post-MI LV dP/dt_min. Stepwise regression analysis identified post-MI coronary driving pressure as an independent predictor of SE CVF.

CONCLUSION: LV remodeling in rats with MI is characterized by predominant SE collagen deposition in non-MI and results from a reduction in myocardial perfusion pressure occurring early on in the setting of MI.

Key words: coronary driving pressure, subendocardial fibrosis, hemodynamic, myocardial infarction, ventricular remodeling

Cardiac remodeling following myocardial infarction (MI) is characterized by scar formation with wall thinning at the site of myocyte fiber loss and progressive left ventricular (LV) dilatation with collagen deposition within noninfarcted myocardium (non- MI) ¹.

Ventricular dilatation is a primary consequence of MI location and size ², but other factors influence fibrosis accumulation, such as the cardiac renin-angiotensin system, endothelins, catecholamines, and inflammatory mediators ^3,4. However, the role of hemodynamic changes occurring during ongoing infarction have not been investigated ⁵. It is still also unclear where fibrosis deposition actually takes place, because subendocardial (SE) regions of non-MI have rarely been contemplated in post-MI morphometric studies ^6,7.

The aim of this study was to investigate, using the experimental MI model in the rat, the role of hemodynamic changes taking place acutely after left coronary artery ligature in subsequent fibrosis accumulation within non-MI.

Methods

Twenty-one male Wistar rats weighing 275±5g were used for the experiments. All procedures were carried out in accordance with the norms of the Brazilian College of Animal Experiments and conformed to the "Guide for the Care and Use of Laboratory Animals." Our Institutional Ethical Committee approved the protocol.

Animals were anesthetized with Ketamine chloride, 50 mg.kg^-1 i.p. and Pentobarbital sodium, 25 mg.kg^-1 i.p. and put under mechanical ventilation with a rodent ventilator (Model 683, Harvard Apparatus Inc., MA USA). Systemic and LV blood pressures were obtained from femoral and carotid arteries, respectively. The catheters were connected to pressure transducers and coupled to a calibrated preamplifier (General Purpose Amplifier 4 - model 2, Stemtech Inc. WI, USA). The pressure tracings were recorded by using a computerized system processor (AT/Codas, Dataq Instruments Inc., OH, USA).

After hemodynamic stabilization, pressure tracings were recorded for 10 minutes before opening the chest wall and continued throughout the entire surgical procedure. The thoracic cage was closed and residual air drained from the pleural space. When new steady hemodynamic signals were observed, an additional 10-minute period of hemodynamic recording was performed. The average of beat to beat hemodynamic measures recorded during the 2 aforementioned 10-minute periods was analyzed.

Heart rate (beats.min^-1), LV systolic pressure (LVSP, mm Hg), LV end-diastolic pressure (LVEDP, mm Hg), LV dP/dt_max, LV dP/dt_min (mm Hg.s^-1) and systemic systolic blood pressure (SBP, mmHg), and diastolic blood pressure (DBP, mmHg) were recorded. To obtain an estimate of coronary blood flow, coronary-driving pressure (CDP) was calculated as the difference between DBP and LVEDP ⁸.

MI was produced by ligature of the left coronary artery, by using a modification of a previously described technique ⁹. Sham-operated rats were operated on similarly except for not tying the ligature around the coronary artery.

During recovery, after observation of spontaneous respiration, mechanical ventilation was concluded and arterial catheters were withdrawn. The animals were returned to their individual cages.

Four weeks after the experiment, animals were anesthetized with Pentobarbital sodium, 30 mg.kg^-1 i.p. The heart, lungs, and liver were removed, cleaned, and weighed. Atria and large vessels were removed before weighing the heart. MI was demonstrated by grossly visible scarring of the LV free wall.

A coronal slice of the heart including both ventricles was obtained at the equatorial plane where the largest surface of infarction was detected. Tissue fixation was performed in 10 percent buffered formalin.

Six-micron paraffin embedded sections were cut and stained with Sirius Red 3BA in saturated picric acid solution ¹⁰. By using an image analysis system (Leica Q500 iW, Leica Imaging Systems ltd., Cambridge, UK), these sections were analyzed morphometrically. Fibrillar collagen was identified in the picrosirius-stained sections by its red colored appearance.

A videocamera equipped with a macro lens permitting the visualization of the entire coronal section of each heart was used to identify MI and non-MI regions and to obtain infarct size. The ratios between endocardial infarct surface length and endocardial total LV circumference and the ratios between epicardial infarct surface length and epicardial total LV circumference were calculated and averaged to obtain infarct size ².

Using a microscopic x10 objective, fibrillar collagen within MI and non-MI was estimated as a collagen volume fraction (CVF, %). CVF was determined as the percentage of red-stained connective tissue areas per total myocardial area, excluding perivascular collagen. Non-MI CVF was addressed separately in 3 distinct regions of each tissue section examined: the inner third of the non-MI corresponding to the LV subendocardium (SE), the medium third of the non-MI interventricular septal wall (IVS), and the right ventricular myocardium (RV).

MI rats were classified into groups SMI and LMI, according to the presence of small (<30%) or large MI (>30%), respectively.

ANOVA, complemented by the Bonferroni t test, was used for comparing quantitative structural variables between MI groups and the sham group. Repeated-measures analysis of variance, complemented by the Wald test, was used to evaluate the effects of MI on hemodynamic variables. Normality and equal variance were verified in all analyses. Data were expressed as mean ± S.E.M. Statistical significance was established at a p <0.05.

The potential relationships between postexperiment hemodynamic variables and subsequent fibrosis deposition within non-MI was assessed by using Pearson's correlation coefficient. Multiple linear regression analysis was performed for detecting among the hemodynamic variables which were the best predictors for subsequent fibrosis deposition. The stepwise selection method was used, with p-values of 0.10 and 0.05 considered significant for entering a variable into or removing it from the model, respectively. The statistical analysis was performed with SAS (Statistical Analysis System) software ¹¹.

Results

The 3 study groups comprised 7 rats each. Infarct size was 20.4±1.5% (range: 15.1 to 25.3%) in the SMI group and 50.5 ± 2.4% (range: 40.8 to 58%) in the LMI group.

Heart, lungs, and liver to body weight ratios were significantly greater in the LMI than in the sham group (table I).

Figure 1 illustrates the hemodynamic variables computed before and after the experiment. Before surgery, all 3 groups were statistically comparable. Surgery did not affect any of the hemodynamic variables in the sham group.

LVEDP increased fivefold in the LMI group, resulting in significantly higher values than in the sham group. The twofold increase in LVEDP observed in the SMI group was not statistically different from that in the sham group.

CDP significantly decreased only in the LMI group (-30%). Postexperiment CDP was comparable among all 3 groups.

A statistically significant decrease in LV dP/dt_max (-36%) was demonstrated for the LMI group alone, and LV dP/dt_min did not change significantly. Despite these nonuniform and not always significant changes, after surgery LV dP/dt_max and LV dP/dt_min turned out to be statistically different between each the MI groups and the sham group.

SBP and DBP had a tendency to decrease, particularly in the LMI group (-5% and -7%, -9% and -7%, -23% and -19% for sham, SMI, and LMI groups, respectively) but remained comparable among groups after surgery.

Surgery did not change heart rate in any group (sham, 345±17 to 330±18; SMI 323±16 to 317±14; LMI 353±14 to 328±16 beats.min^-1).

Table II depicts CVF determined within MI and non-MI regions. MI CVF was comparable between the SMI and LMI groups. Non-MI CVF in IVS and in RV was comparable among the 3 groups. Otherwise, LMI showed a greater non-MI SE CVF than that in the sham group.

In figure 2, four heart tissue sections are depicted under progressive magnification representing an animal included in the LMI group. An evident increase can be observed in collagen fibers revealed as a thick layer of fibrosis scattered within the non-MI SE LV.

As CVF was predominantly found within the SE layer of non-MI LV, the amount of collagen fibers found at this location was used for further statistical analyses. Of the hemodynamic variables investigated, non-MI SE CVF correlated inversely with SBP (r = -0.51, p = 0.02), DBP (r = -0.44, p = 0.04), and CDP (r = -0.52, p = 0.02), and directly with LV dP/dt_min (r = 0.44, p =0.04).

CDP was the only hemodynamic variable identified as independently related to non-MI SE CVF (parameter estimate = -0.061, standard error = 0.02, Rsqr = 0.27, p = 0.02). Figure 3 graphically represents the linear correlation found between non-MI SE, CVF, and CDP.

Discussion

The present study confirmed previous findings showing that experimental left coronary artery ligature in rats produces immediate systemic and LV hemodynamic changes ¹². These changes occurred predominantly in animals with the largest MI and corresponded to decreases in both LV dP/dt_max and CDP and to increases in both LVEDP and LV dP/dt_min.

In particular, our findings allude to a relationship between acute post-MI CDP and fibrosis deposition within non-MI examined 4 weeks later. The increase in non-MI collagen content was found rather typically at the SE region. Recently, experiments using the isolated rat heart preparation have shown that baseline flow may be reduced by as much as 28% immediately after coronary artery ligature ¹³.

The in vivo nature of the present study permitted us to establish a potential connection between acute hemodynamic changes jeopardizing blood flow to non-MI and late development of SE fibrosis.

As hemodynamics was more severely impaired in the LMI group, we speculate that the acute impairment of LV systolic performance ² chiefly contributed to the reduction in coronary perfusion pressure and blood flow ⁸. Ventricular diastolic dysfunction would further limit coronary perfusion and aggravate ischemia.

We recognize that although CDP has been demonstrated to be an independent predictor of subsequent SE fibrosis, the associated r-square value was of not so great a magnitude. In part, this may be due to the relatively early hemodynamic measurements undertaken in the present study. Although studies performed in humans ¹⁴ and in dogs ¹⁵ suggest that an initial transitory improvement may occur in post-MI hemodynamics, in rats these parameters usually deteriorated during the first 24 hours ¹² and beyond ^16-19. Further ischemia at the SE region may occur during the progression of remodeling as a consequence of reduced coronary blood flow reserve ^20-22.

Controversy regarding the pathogenesis of collagen accumulation within non-MI, involving different local and systemic mediators ^3,4,23-26, still remains. The present findings indicate that underperfusion to non-MI should also be taken into account as a potential mechanism, particularly to explain SE collagen fiber accumulation.

The predominance of SE over interstitial fibrosis has been reported before ^27,28. This pattern of collagen deposition is rather in support of being representative of a reparative scarring process in response to impeding myocardial perfusion. It may also explain reports of remote wall motion abnormalities occurring early on and persisting for as long as 2 months after MI ^29,30.

Ventricular remodeling has been characterized by thinning and expansion of the infarcted wall, on the one hand, and by enlargement of the ventricular cavity, on the other, occurring as early as 2 days ³¹ after MI, or even earlier ²⁹. It has been attributed to the occurrence of side to side slippage of remnant myocyte fibers ³². In disagreement with this assumption is the fact that LV epicardial circumference length remains unchanged up to 21 days after MI ³¹. Alternatively, it is quite conceivable to suppose that early enlargement of the ventricular cavity may be, in part, the result of loss of myocyte cells jeopardized by poor coronary perfusion taking place under unfavorable acute hemodynamic conditions, such as those reported in the present study.

Based on the present findings, we speculate that impeding ischemia of non-MI occurs acutely after MI, particularly affecting the SE region, and may result from hemodynamic changes interfering with CDP. This appears to play a chief role in subsequent fibrosis found in this region. To be confirmed, this hypothesis deserves further investigation.

Acknowledgments

We are grateful to Dr Maria de Lourdes Higuchi for her technical assistance and for providing the facilities for processing and analyzing the tissue sections. We also thank Dr Irineu Tadeu Velasco for providing the facilities of the experimental laboratory of the University of São Paulo Medical School (LIM-51) where this study was performed.

Correspondence to

Clovis de Carvalho Frimm

Faculdade de Medicina da USP, LIM 51

Av. Dr. Arnaldo, 455

01246-903 - São Paulo, SP, Brasil

E-mail:

frimm@emercli.fm.usp.br