| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
1Laboratory of Haemodynamics and Cardiovascular Technology, Ecole Polytechnique Fédérale de Lausanne, Lausanne; 2Department of Cardiovascular Surgery, Centre Hospitalier Universitaire Vaudois, Lausanne; 3Department of Cardiovascular Surgery, Inselspital, Bern; and 4Department of Cardiology, Centre Hospitalier Universitaire Vaudois, Lausanne, Switzerland
Submitted 4 September 2006 ; accepted in final form 4 December 2006
 |
ABSTRACT |
|---|
 TOP ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
PEP), ejection period (EP), and their ratio (
= EP/PEP) were calculated, together with the characteristic decay time during isovolumic relaxation (
) and the normalized elastance at end diastole (Eminn). MI provoked significant LV chamber dilatation, thus a loss in cardiac output (–33%), ejection fraction (–40%), and stroke volume (–30%) (P < 0.05). Also, it caused significant calcium increase (17-fold), fibrosis (2-fold), and LV hypertrophy. End-systolic elastance dropped from 0.66 ± 0.31 mmHg/µl (CTRL) to 0.34 ± 0.11 mmHg/µl (MI) (P < 0.05). Normalized elastance was significantly reduced in the MI group during the preejection, ejection, and diastolic periods (P < 0.05). The slope of En(tn) during the PEP and were significantly altered after MI (P < 0.05). Furthermore, and end-diastolic Eminn were both significantly augmented in the MI group. We conclude that the En(tn) differs quantitatively in all phases of the heart cycle, between normal and hearts post-MI. This should be considered when utilizing the single-beat concept. compliance; ischemia; contractility; ventricular function; hemodynamics; conductance volumetry
The aim of this study was therefore to validate whether the En(tn) is indeed independent of cardiac diseased state. For this purpose, we studied the infarcted rodent in terms of pressure-volume relation and derived the E(t) curves for control (sham operated) rats and rats where myocardial infarction (MI) was provoked by coronary artery ligation. To test the hypothesis that MI may alter the En(tn), the En(tn) were quantified and compared between the control and MI groups.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Male Sprague-Dawley rats were purchased from Charles River Breeding Laboratories. They were maintained in temperature- and humidity-controlled rooms with typical light-dark cycle and given standard chow and tap water ad libitum. All rats were handled in accordance with the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH publication no. 85-23, revised 1996), and our Institutional Animal Care and Use Committee approved the protocol.
Experimental MI
The myocardially infarcted group (MI) consisted of Sprague-Dawley rats (n = 9) with a baseline body weight (BW) of 234 ± 12 g. The rodents were anesthetized with 2% isofluorane (Forene, Abbott, Baar, Switzerland), intubated, and then ventilated with 100% oxygen at 60 cycles/min with a tidal volume of 2 ml (Harvard Apparatus model 683, Holliston, MA) before the surgical procedure was performed. The rats were placed on a heating pad to maintain body temperature, and disinfection was performed on the thorax. A left thoracotomy was performed at the third intercostal space to gain access to the heart. The pericardium was opened, and the left anterior descending coronary artery was located between the left atrium and the pulmonary artery and ligated with a 4.0 polypropylene (Ethicon, Somerville, NJ) to provoke the MI. This was confirmed by a change of color of the left ventricle (LV) from red to a purplish-gray distal to the ligation. The ribs were closed with two to three ligatures, a chest drain was inserted to avoid any pneumothorax, and the muscles surrounding the rib cage were closed before the skin was closed. Analgesia was given (Pro-Dafalgan) to attenuate any pain ensued by the rodents due to the MI (20 mg, Upsamedica, Baar, Switzerland). The anesthesia was gradually weaned, and when the rat began spontaneous breathing the intubation tube was removed. The rats were then returned to the animal house once complete recovery was observed. A similar sham-operated control (CTRL) (n = 10) with the rodents weighing 220 ± 10 g at baseline was performed without the MI.
Hemodynamic and Pressure-Volume Data
Eight weeks after the initial operation, the rats were again anesthetized and intubated. The right neck region was disinfected to provide access to the carotid artery. The skin was opened, and the right jugular vein and carotid artery were isolated. A 2-Fr conductance catheter (CC) (SPR 838 Aria, Millar Instruments) was inserted into the LV via the right carotid artery. Parallel conductance was measured after injection of 10% saline into the jugular vein using a 1-ml syringe (20 µl of Natrium chloratum, Sintetica, Mendrisio, Switzerland) in accordance with the method used by Baan et al. (2). An occlusion analysis was performed by temporarily occluding the inferior vena cava below the diaphragm via a mini laparotomy. Figure 1, left, shows pressure-volume loops recorded during such an inferior vena cava occlusion for a CTRL and a MI rat. On the same graph are shown the ESPVR and end-diastolic pressure-volume relationship, which in accordance with the classical varying elastance formulation are assumed to be linear and whose slopes define the maximum (Ees) and minimum elastance (Emin), respectively (28, 30). The linear ESPVR intercepts the volume axis at V0. Assuming that V0 does not change during the heart cycle, we can define the instantaneous elastance, E(t), as E(t) = P(t)/[V(t) – V0]. A typical waveform for the E(t) is shown in Fig. 1 for the two groups (Fig. 1, right).
|
Normalized Time-Varying Elastance Curve
The En(tn) was obtained by normalizing the elastance values to their peak value En = E/Ees and by normalizing times to the time of peak elastance tn = t/tmax. Superscript "n" denotes normalized values. A typical En(tn) is represented in Fig. 2. Comparison of the En(tn) curve between the CTRL and MI group is done at two levels. First, on a point-by-point basis, by comparing the values of the En at the same nondimensional time tn, tn varying over the entire cardiac cycle at 0.1 intervals. Second, by comparing four different parameters characterizing the En(tn), namely 1) the slope of the preejection period (PEP), 2) the slope of the ejection period (EP), 3) the characteristic time of the exponential decay during early diastole (t), and 4) its end-diastolic value, Eminn = Emin/Ees. These parameters are shown schematically in Fig. 2. The slopes were defined following Shisido et al. (29). In brief, PEP was defined as (En at end of isovolumic contraction phase – En at end diastole)/(tn at end of isovolumic contraction phase – tn at end diastole) and EP as (Eesn – En at end of isovolumic contraction phase)/(tn at Eesn – tn at end of isovolumic contraction phase). Constant was derived by fitting an exponential decay during early diastole (e–tn/, 1.1 
tn 
1.5). The ratio of the slopes = EP/PEP was also derived.
|
Cardiac infarct size and cardiac tissue preparation. After completion of hemodynamic measurements, hearts were excised and weighed. The infarct size area was determined as a percentage of the entire LV area, as reported previously (4). The entire upper segment of the heart was preserved in formalin for histological analysis. LV were then dissected and cut into two pieces: 1) LV apex was snap frozen, powdered, and kept at –80°C for calcium and collagen determinations, 2) middle LV segment was frozen in optimum cutting medium (OCT, Sakura) for immunohistochemistry analysis.
Cardiac tissue calcium content. The calcium content of the LV apex (µmol/g dry wt) was determined by atomic absorption spectrophotometry (AA10, Varian) after mineralization and acid digestion of the tissue (12).
Scleroprotein analysis. Collagen content was measured from a representative sample of the LV apex. After protein hydrolysis, the amount of collagen was estimated from the hydroxyproline content determined by colorimetric spectrophotometry (20).
Histology. Heart tissue samples were fixed in 4% formaldehyde and mounted in paraffin block, and slices were obtained with a microtom. After deparaffinization and hydratation, the samples were treated with either periodic acid Schiff staining (Sigma-Aldrich Chimie, Epalinges, Switzerland) protocol, to discriminate cell borders, or 0.1% picrosirius red for collagen (Sigma). The mean of the cardiomyocyte cross-sectional area (CSA) and diameter were calculated by photomicrographs of 100 cells/specimen with a computer-assisted image analysis system (Metamorph analysis) (6).
Immunohistochemistry. The detailed protocol has been described elsewhere (8). Briefly, cross-sectional slices of 5–10 µm were prepared with a cryostat (Cryo-star HM 560M). Sections were fixed for 10 min with 4% paraformaldehyde at room temperature and rinse in PBS. To block unspecific binding, slices were incubated for 30 min in 10% normal goat serum (NGS, Sigma) at room temperature. The following primary antibodies were incubated for 2 h at room temperature: mouse antibody against connexin 43 (BD Biosciences Pharmingen, Basel, Switzerland) and rabbit antibody against laminin (to discriminate the cell border; Sigma) were both diluted (1:100) in PBS containing 4% of NGS. After three washes with PBS, slices were incubated with secondary antibodies, goat Alexa 488 anti-mouse and goat Alexa 555 anti-rabbit (both 1 h at room temperature). The mean of cardiomyocyte length was calculated as described in the histology section.
Statistical analysis. Trans-thoracic echocardiography (TTE) was performed three times on each rat, and measurements were averaged. All parameters are reported as means ± SD. A Student's t-test was performed between the CTRL and MI groups, with P values of <0.05 considered significant. All analysis was done in SPSS (SPSS 11.5, SPSS, Chicago, IL).
|
RESULTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
The BW was 12% lower in the MI group (P < 0.05; Table 1). Heart weight in the MI group was significantly greater (21%) probably due to LV hypertrophy and chamber dilatation (P < 0.01). The lung weight over BW was reduced in the MI group, but not significantly. The relative infarcted area was 31 ± 6% in the MI group (P < 0.05; Table 1). HR, SBP, DBP, MBP, and pulse pressures did not differ between the two groups (Table 1).
|
Cardiac parameters derived from pressure-volume analysis for the CTRL group and the MI group are shown in Table 2. MI has caused the LVEDP to double, and EDV increased by 33%, indicative of chamber dilation (P < 0.05). SV decreased by 
30% (P < 0.05), and ESV doubled in the MI group, indicative of compromised contractile function (P < 0.01). EF and CO decreased significantly in the MI group (Table 2). Ees dropped by 48% in the MI group, but Emin was not different between the CTRL and the MI group (Table 2). Vo did not change significantly (Table 2). Despite a 10% decrease in the average value, dP/dtmax did not change significantly between the two groups. The slope of the dP/dtmax-EDV curve, however, did decrease significantly in the MI group (Table 2). The decrease in contractility was clearly seen in the ejection phase indexes (Table 2), with end-systolic elastance decreasing from 0.66 ± 0.31 mmHg/µl in CTRL to 0.34 ± 0.11 mmHg/µl in the MI group. As a consequence of chamber dilation and decrease in contractility, both CO and EF were significantly reduced in the MI group (Table 2). MI also affects active relaxation, with the IVR time being significantly augmented in the MI group (Table 2).
|
The mean En(tn) waveforms for the two groups were compared at 10 time intervals over the cardiac cycle, and statistical significance found at the intervals was marked with an asterisk or a cross (Fig. 3). Significant differences were found at the following points: start of PEP (P < 0.05), the main part of the EP (P < 0.05), the end of the IVR phase (P < 0.05), and the complete diastolic period (P < 0.01) (Fig. 3).
|
constant derived from the exponential decay time (e–t/) was found to be reduced in the MI group (P < 0.05; Table 3). This indicates a quicker relaxation time in the CTRL group. The normalized Emin was significantly elevated in the MI group (P < 0.05; Table 3). The ratio of slopes, , was doubled in the MI group (Table 3).
|
Significant increase in cardiac calcium content occurred in the MI hearts (Table 4). Collagen rose, which augmented the fibrosis in the MI group (Table 4, Fig. 4). The cardiomyocyte diameter as well as the CSA were 19% and 36%, respectively, higher (MI group) vs. CTRL group (P < 0.05; Table 4 and Fig. 4). No significant changes were observed in cardiomyocyte length (Table 4).
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
We have found that the En(tn), although similar in shape, differs quantitatively between normal and failing rat hearts, after induction of a MI. Differences in En(tn) were found during ejection, during IVR, and at end-diastolic phases (Fig. 3). Also, there was a difference in the slope of the En(tn) at the PEP, in the decay of En(tn) during the IVR phase (), and at the ratio of the EP/PEP.
The Use of Time-Varying Elastance for Assessing Cardiac Function in Health and Disease
Cardiac function has here been evaluated using pressure-volume analysis obtained invasively with a CC technique. The CC technique was first used by Baan in 1980 and has gained popularity as an accepted method to establish indexes of cardiac contraction (9, 14, 26). Pressure-volume loops combined with variations of preload permit the assessment of the ESPVR, end-diastolic pressure-volume relationship, dead volume, and, in general, the varying elastance curves. Suga and colleagues have shown that the E(t) is fairly independent of loading conditions, contractile state, and HR (31). However, subsequent studies revealed limitations to this simplified theory, such as ESPVR curvilinearity, afterload dependence under certain conditions, and that Vo may change during contraction (18, 26, 29). Furthermore, the ESPVR has been reported to be influenced by regional ischemia in acute and chronic animal models of heart failure (13, 21) and augmented in spontaneously hypertensive rats (5).
Uniqueness of the Normalized Time-Varying Elastance Curve
Senzaki et al. reported on human data and showed that the shape of the En(tn) is independent of different cardiac pathologies, afterload, preload, and contractility (28). In a later study by the same group, Georgakopoulos et al. showed that the shape of the En(tn) is identical in different animal species, the similarity in En(tn) being attributed to small differences in myofibril protein isoforms and protein kinetics among species (10).
Assuming a unique En(tn), Senzaki et al. advocated the use of such a global En(tn) to derive ESPVR values from noninvasive single-beat recordings of ventricular pressures and volumes (28). This could have potentially important clinical implications, because TTE could be used to estimate ventricular volume and peripheral arterial pressure for intraventricular systolic pressure, thereby providing a clinically feasible noninvasive estimate of cardiac contractility. Several other groups pursued this technique due to the fact that no vena cava occlusion is required, making it practical and feasible for clinical applications (11, 17, 19). A number of subsequent studies have shown, however, that the method proposed by Senzaki is inaccurate or not applicable. Kjorstad et al., for example, found the concept of a global En(tn) inaccurate and the single-beat-based method for deriving ESPVR imprecise and incapable of showing anticipated changes in contractility (16). When scrutinizing the results of Senzaki et al., one reveals large differences (standard deviations) on the En(tn), especially during the ejection and diastolic phases in the presence of aneurysms, dilated cardiomyopathy, and coronary artery disease with conserved LV function (28).
A more sensitive comparison of the En(tn) under different conditions and pathologies is obtained by looking at variations in the slope (time derivatives) of the En(tn) during different phases of the heart cycle (PEP and EP). Shisido et al. (29) used a bilinear approximation for the E(t), focusing primarily on the shape of the En(tn). The ratio of the two slopes of the EP to PEP, , supposedly constant in a global En(tn), was found to correlate to EF, Ees, and arterial elastance (29). Obviously, time derivatives of the En(tn) (dE*/dt) relate in a straight-forward manner to time derivatives of the pressure (dP/dt) during the isovolumic contraction and relaxation phases, but the relation becomes more complex during ejection. In our study, the slope of En(tn) during the PEP phase decreased significantly in MI rats. In consequence, rose in the MI group. Shisido et al. (29) found that this ratio was significantly attenuated during reduced contractility and vasoconstriction but augmented during vasodilatation. Kjorstad et al. (16) showed significant differences in values between controls and postischemic groups. In the same study, during the administration of methoxamine, to augment afterload, the En(tn) was altered in shape especially during the EP and diastolic phases, which we also found in the MI group (29).
Validation of the MI Model
We have based our study on the En(tn) on a comparison between a CTRL and a MI-induced failing heart model. A number of parameters relating to cardiac geometry and systolic and diastolic function were measured to validate our MI model. The infarcted hearts are dilated, as seen by the increased EDV and ESV values, a sign of heart failure. SV is compromised while HR is maintained, resulting in reduced CO and decreased SV and EF. In addition, there are changes in the diastolic portion of the pressure-volume relationship, namely an elevated LVEDP, indicating the presence of combined systolic and diastolic heart failure (32). Maximum elastance is compromised in MI, and this is also represented by a decrease in the slope of the dP/dtmax-EDV relation. During diastole, Eminn was greater in the MI group, and LVEDP rose as the ventricle transcends into failure. Characteristic relaxation times were also significantly prolonged, indicative of diastolic dysfunction.
All data presented here were derived from pressure-volume data obtained invasively using the CC. We have, however, performed in each animal detailed echo examination, in concordance with other research groups (22, 26, 27) and compared, on the same animal model, echo-derived and CC-derived data, and we showed, through a Bland-Altman analysis, that the data were similar and consistent (15).
Mechanisms of MI-Induced Alterations in the Normalized Varying Elastance Curve
The varying elastance curve reflects the time course of the apparent stiffness of the LV and thus is intimately related to the dynamics of contraction and active tension development in myocytes. Through a simple one-dimensional analysis, one can derive the typical force-velocity characteristics of the myocyte based on the E(t) curve. Therefore, the curve reflects the intrinsic contractile properties of the myocardium.
MI-induced remodeling involves myocyte hypertrophy and alteration in LV architecture to distribute wall stresses more evenly as the extracellular matrix forms a collagen scar to stabilize distending forces. In our study, this is validated by the rise in collagen (Table 4) and represented by the fibrosis in Fig. 4. Myocyte hypertrophy is visible microscopically with an up to 70% rise in cell volume and mural hypertrophy by in-series sarcomeric replication, without a change in sarcomere length (1). In our study, we have observed similar changes in myocyte diameter and CSA (Table 4), whereas length remained comparable (Table 4, Fig. 3).
Ischemia post-MI can cause ischemic contracture because, at the low ATP concentration and high-cytosolic calcium levels characteristic of the ischemic state, the cross bridges are no longer occupied by ATP and become permanently attached, thereby forming rigor bridges; thus the dominant form of the myosin head is in the strongly attached rigor state in flexion (23). The high force development may then cause hypercontracture, which exerts excess tension on the sarcolemma to cause microlesions that leak so that extracellular calcium can invade the cell (Table 4).
Study Limitations
The study was designed to prove or reject the hypothesis that En(tn) may not be preserved in the presence of heart disease and, in particular, after MI. There is a multitude of events provoked by MI at the molecular, cell, and tissue level, and we acknowledge that a complete study of how this cascade of events may differentially influence the elastance curve goes far beyond the objective of this work.
Our analysis assumed linear elastance relations of ESPVR that intersected at a common Vo. These are over simplifications, as previous studies have revealed that ESPVR can be nonlinear (3), can be influenced by afterload, and that Vo may change during contraction (18). It was also recently shown that the nature of curvilearity of the instantaneous pressure-volume relations (isochrones) changes over the heart cycle (7). The PEP and EP portions of the En(tn) are curvilinear, and thus their corresponding slopes vary with time. The derived slopes PEP and EP are only first-order approximations of the mean slope value of the PEP and EP portion, respectively. The choice of comparing "mean slopes" is thus limiting the analysis and was used for the sake of simplicity and for being consistent with previous literature.
In conclusion, when En(tn) is compared quantitatively between the CTRL and MI groups, statistical significance is found at the ejection phase and during diastole. Also, there is a difference in the slope of the En(tn) at the PEP, as well as a difference in characteristic decay time during the IVR phase. Therefore, despite the general similarity in the shape of the normalized varying elastance waveform, important quantitative differences may exist in the values and slopes of En(tn) curves between healthy and diseased hearts. These differences need to be taken into account when cardiac contractility is assessed based on a generalized E(t) curve in animal models or in the human in different physiological or pathological states.
|
GRANTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
