| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 1996;62:737-743
© 1996 The Society of Thoracic Surgeons
Departments of Surgery and Anesthesiology, Columbia University College of Physicians & Surgeons, New York, New York
Accepted for publication April 19, 1996.
 |
Abstract |
|---|
 TopFootnotes Abstract IntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Methods. Edema was induced by coronary perfusion with Plegisol (750 mL, 289 mOsm/L) during a 1-minute aortic occlusion in 6 pigs. This was followed by whole blood reperfusion, inotropic support, and circulatory assistance until sinus rhythm and contractile function were restored. A control group (n = 6) was treated similarly, with 1 minute of electrically induced ventricular fibrillation and no coronary perfusion. Recorded data included electrocardiogram, left ventricular pressure and conductance, aortic flow, and two-dimensional echocardiography. Preload reduction by vena caval occlusion was used to define systolic and diastolic properties. Data were recorded at baseline and at 15-minute intervals for 90 minutes after reperfusion.
Results. In the edema group, average left ventricular mass (132 ± 7 [standard error of the mean] versus 106 ± 4 g) and ventricular stiffness constant (0.15 ± 0.02 versus 0.05 ± 0.01) increased after Plegisol versus baseline (p< 0.05), returning to normal after 45 minutes of reperfusion. In controls, mass (118 ± 6 versus 116 ± 4 g) and ventricular stiffness (0.06 ± 0.01 versus 0.05 ± 0.01) did not change significantly. There was no significant change in systolic function. Myocardial water content at the end of the study was not different for the two groups.
Conclusions. Crystalloid-induced edema and diastolic stiffness resolve after 45 minutes in pigs. This suggests that edema caused solely by cardioplegia during cardiac operations should not cause significant perioperative ventricular dysfunction.
|
Introduction |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Clinically, the formation and resolution of myocardial edema and effects on cardiac function are not well understood, in part because they are difficult to study in isolation. Myocardial edema caused solely by crystalloid coronary perfusion in rats resolves after 15 minutes of blood reperfusion [8], but the time course in larger animals is undefined. Accordingly, the purpose of this project was to develop a large animal model in which resolution of perfusion-induced myocardial edema could be studied in vivo, while left ventricular systolic and diastolic function were monitored. Because of the potential instability of conductance for measurement of ventricular volume in open-chest animals, two-dimensional echocardiography was employed as a reference standard during measurements of ventricular function.
|
Material and Methods |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
The protocol was carried out in 40- to 50-kg pigs, an edema group (n = 6) and a control group (n = 6). Pigs were anesthetized with acepromazine (0.5 mg/kg intramuscularly), ketamine (20 mg/kg intramuscularly), and atropine (1 to 2 mg intramuscularly). Animals were intubated and mechanically ventilated, maintaining arterial blood gas values within physiologic norms. Anesthesia was maintained with isoflurane (1.75% to 2.00%) and oxygen. A median sternotomy was performed while the electrocardiogram, peripheral arterial blood pressure, and body temperature were monitored. After a longitudinal pericardiotomy, the inferior and superior venae cavae, pulmonary artery, and aorta were isolated with snares. An ultrasonic flow probe (Transonic Systems Inc, Ithaca, NY) was placed on the ascending aorta for the assessment of cardiac output. A fluid-filled catheter was placed in the left atrium through a pursestring suture in the atrial appendage. A 5F ten-electrode, single-field conductance catheter (Webster Labs, Baldwin Park, CA) and a 5F micromanometer pressure catheter (Millar Instruments, Houston, TX) were placed in the left ventricle through a pursestring suture in the apex to measure left ventricular volume and pressure, respectively. The position of the conductance catheter was verified by echocardiography to ensure that contact of the electrodes with the endocardium was avoided. The number of conductance catheter electrode segments included in the ventricle was determined by two-dimensional echocardiography and by plotting the ventricular pressure-volume loop recorded by each of the electrode segments on a digital oscilloscope (MacLab; ADInstruments Inc, Milford, MA). Electrode segments producing counterclockwise rotation of the pressure-volume loops were included in the measurement of left ventricular conductance, whereas segments forming pressure-volume loops in a clockwise fashion were in the aorta and were excluded.
A pericardial well was created by sewing a plastic sheet to the pericardium and draping the free edges over the opened sternum. After instrumentation, this well was filled with water-soluble gel (Ultraphonic scanning gel; Pharmaceutical Innovations, Inc, Newark, NJ) for two-dimensional echocardiography. Echo gel has been shown to have minimal effects on parallel conductance in previous studies [9].
Before data collection, a 6-mL arterial blood sample was collected to measure blood resistivity with the rho cuvette using the Leycom Sigma-5 conductance module (Rijnsburg, Netherlands). After data collection, parallel conductance was measured using the hypertonic saline method described elsewhere [10].
Data digitized with an analog-to-digital converter (MacLab) and recorded on a digital computer (Macintosh Quadra 950; Apple Computer, Cupertino, CA) included electrocardiogram, left atrial and ventricular pressure, conductance, and aortic flow. Data were recorded with the lungs held at end-expiration during preload reduction by vena caval occlusion for 10 seconds.
Using a 5.0-MHz transducer, echocardiographic views were recorded (VingMed CFM 750; VingMed Sound, Inc, Salt Lake City, UT), including apical long-axis, apical two-chamber, and cross-sectional short-axis views. Techniques described elsewhere [11] allow calibrated left ventricular pressure and conductance to be displayed and temporally coordinated with two-dimensional echocardiographic views. These data allow end-diastolic echo area and conductance to be correlated. This relation has been shown in our laboratory to be linear, allowing presumptive validity of conductance calibration as long as the conductance-area relation is maintained. This technique was employed to confirm adequacy of conductance calibration during the experimental conditions of the present study.
Echo gel was added to the pericardial well as needed to maintain the level of the gel at 1 cm above the anterior surface of the heart. Gel pressure was measured with a fluid-filled catheter at midventricular level, and subsequently subtracted from the intracardiac pressures.
After completion of baseline data collection, a cardioplegia infusion catheter with a pressure monitoring port was inserted into the aortic root. After heparinization (300 IU/kg), a large cannula was placed in the right atrium to vent the right heart to a reservoir during aortic cross-clamping. Cannulas for left heart bypass (Biomedicus 520 D centrifugal pump, Minneapolis, MN) were placed in the left atrium and the left carotid artery. The bypass circuit, which did not include an oxygenator or reservoir, was primed with 150 mL of 5% albumin. The heart was arrested in the following manner: (1) cavae and subsequently the pulmonary artery were snared; (2) the aorta was cross-clamped; (3) the venous cannula in the right atrium was opened to vent the right heart; (4) cold (0° to 4°C) Plegisol (750 mL, 289 mOsm/L) was injected into the aortic root in the edema group, maintaining aortic root pressure at 55 to 65 mm Hg. In the control group, cardioplegia was omitted, and the heart was fibrillated using an alternating current transformer (Archer 273-1610; Radio Shack, Fort Worth, TX). In both groups, the average cross-clamp time was 1 hour 9 minutes.
After the arrest period, the snares and aortic clamp were released and the right atrial vent was clamped. Bypass was initiated, and the heart was massaged to avoid distention and maintain pulmonary flow. Each pig was resuscitated with defibrillation (50 J, mean number of shocks = 2 ± 1) and physiologic doses of dopamine (5 to 10 µg•kg-1•min-1 intravenously) and epinephrine (0.02 to 0.1 µg•kg-1•min-1 intravenously) until arterial pressure and heart rate stabilized. The duration of resuscitation was not significantly different between groups (10 ± 3 versus 9 ± 3 minutes; p = not significant by unpaired t test). Data collection for post-Plegisol left ventricular function studies was delayed until hemodynamic stability was confirmed 10 minutes after cessation of inotropic and mechanical support.
Blood resistivity, steady-state hemodynamics, and echocardiography were recorded every 15 minutes for 90 minutes. Data were also recorded during transient preload reduction by caval occlusion. At the completion of data collection, parallel conductance was measured. The heart was then arrested by infusing 60 to 100 mEq of cold potassium chloride solution into the aortic root. The heart was excised and trimmed, and the left ventricle was weighed. A sample of the left ventricle was weighed before and after drying to constant weight (48 hours) in an oven at 60°C. Myocardial water content (percent) was calculated as [(wet heart weight - dry heart weight)/(wet heart weight)]•100. Whole blood and albumin infusion were emphasized over crystalloid solutions for volume replacement. Hematocrit and hemoglobin were measured for each experiment before and after aortic cross-clamping. Blood resistivity, which decreases directly with progressive hemodilution, was also measured during the experiment before and after aortic cross-clamping. Changes in hematocrit, hemoglobin, and blood resistivity were compared before and after aortic cross-clamping using paired Student's t tests. Statistical difference was defined as p less than 0.05.
Data Analysis
Data digitized at 200-Hz intervals were analyzed using IGOR software (Wavemetrics, Inc, Lake Oswego, OR). The conductance data were corrected for parallel conductance [10] and calibrated against flow probe-derived stroke volume to determine left ventricular volume.
DIASTOLIC PROPERTIES.
End-diastole was defined as the point on the pressure trace preceding the systolic upstroke. End-diastolic pressure (EDP) and volume (EDV) were fitted to the equation EDP = 
eßEDV by the least squares method. ß is the left ventricular stiffness constant and is the base constant. Average mass, ß, and after blood reperfusion were compared with the baseline by repeated-measures analysis of variance. The qualitative effect of changes in and ß were then examined. Average base constants () and ventricular stiffness constants (ß) for each time period were grouped into three time intervals (before Plegisol, 16 to 45 minutes after Plegisol, and 46 to 90 minutes after Plegisol) and averaged. In the control and edema groups these average 's and ß's were then used to construct average end-diastolic pressure-volume relationships.
SYSTOLIC PROPERTIES.
End-systole was defined by the upper-left hand corner of the pressure-volume loop [12]. The slope of the linear regression line fitted to the end-systolic pressure-volume relation-defined systolic elastance. Stroke work was determined for each cardiac cycle by calculating the area within the pressure-volume loop. Preload recruitable stroke work was defined by the slope of the linear regression of stroke work versus end-diastolic volume. The volume intercepts were defined as the points where ventricular pressure or stroke work were zero. Contractility by end-systolic pressure-volume relationship (systolic elastance and volume intercept) and stroke work (preload recruitable stroke work and volume intercept) after blood reperfusion were compared with baseline for each animal in both groups by analysis of covariance.
ECHOCARDIOGRAPHY ANALYSIS.
The techniques employed for echocardiographic imaging and quantitative two-dimensional echocardiography have been previously developed and validated in this laboratory [7, 11, 13]. Left ventricular mass was calculated using Simpson's rule [7, 13] and three sections: two perpendicular long-axis views of the left ventricular excluding the papillary muscles [13] (apical long-axis [S1A], apical two-chamber [S2A]) and cross-sectional short axis just below the mitral valve leaflets at the largest diameter. Videotaped images were analyzed using a computerized light-pen (Varian, Salt Lake City, UT) to measure the long axis and epicardial and endocardial chamber areas [7, 13]. Wall volume was determined as the difference of the left ventricular epicardial and endocardial shells. Wall volume in cubic centimeters was multiplied by the specific gravity of the myocardium (1.05 g/cm3) for determination of ventricular mass [14, 15].
|
Results |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
. Left ventricular mass in this example increased from 90 to 122 g.
|
. One complete set of loops during a vena caval occlusion and fitted corresponding regression line are illustrated as well as regression lines at other time points during resolution of edema. The slopes of the regression lines were 2.4, 3.0, 2.7, and 2.5 at baseline, 26, 34, and 54 minutes respectively. Corresponding values for volume intercepts decreased from -12 to -20 mL over the same time period.
|
. Mass, , ß, and contractility did not change significantly after ventricular fibrillation. Average results for the edema group are presented in Table 2. Left ventricular mass (132 ± 7 versus 106 ± 4 g; p < 0.05) increased significantly after Plegisol perfusion; this resolved 30 minutes after blood reperfusion. The ventricular stiffness constant (ß) (0.15 ± 0.02 versus 0.05 ± 0.01; p < 0.05) also increased significantly after Plegisol perfusion; however, this resolved 45 minutes after reperfusion. Changes in systolic function and base constant () were not statistically significant.
|
|
. The effect of time on compliance is presented in Figure 3 for the control group and Figure 4 for the edema group. A statistically significant decrease in left ventricular compliance occurred after Plegisol infusion (see Fig 4); other changes, however, are not statistically different.
|
|
). Myocardial water content at the completion of the study was not significantly different for control versus edema groups (79.2% ± 0.9% versus 78.6% ± 1.4%) (see Table 3).
|
|
Comment |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
The present results indicate that myocardial edema induced by crystalloid coronary perfusion is associated with increased ventricular stiffness. These changes resolve after 45 minutes of reperfusion.
In our original design, we hoped that mechanical circulatory support would not be necessary. However, when adequate pressure and flow could not be maintained by massage alone, left atrial to aortic bypass was required. An oxygenator was avoided in the bypass circuit to prevent inflammatory responses that might contribute to edema.
Every effort was made to avoid systemic dilution. A low-volume bypass circuit was used, which was primed with 5% albumin. There was no significant change in hematocrit, hemoglobin, and blood resistivity before and after aortic cross-clamping. The fact that left ventricular mass did not change in the control group also suggests that fluid administration was properly managed.
Flow after Plegisol administration was two to three times normal cardioplegic flows (250 to 300 mL/min); however, aortic root pressures remained normal. This raises the possibility of aortic valve incompetence and subsequent left ventricular distention with possible endocardial ischemia. However, left ventricular distention was not observed during the experiment. Endocardial ischemia may have occurred as evident by the profound stunning requiring mechanical support, but ventricular distention was avoided.
Previous studies from this laboratory in the isolated, arrested, hypothermic pig heart [2, 3] demonstrated that myocardial edema is associated with predictable decreases in left ventricular volume at specified levels of left ventricular pressure. Changes in diastolic properties correlated more closely with changes in heart weight than with changes in myocardial water content of the right ventricle. The value for myocardial water content reported here in controls, 79.2% ± 0.9%, is similar to a previously reported value of 80% from our laboratory [3]. This is similar to the 78% to 80% values reported in normal dogs [16, 17] and humans [19], but is higher than the 75% to 76% reported in rats [8, 19].
In a previous study from our laboratory, perfusion-induced edema was only partially reversed by hypertonic crystalloid coronary perfusion in cold, arrested pig hearts [3]. The more complete resolution of edema in the present study could be related to effects of whole blood reperfusion (including oxygenation, buffering, free radical scavengers, and oncotic effects), mechanical compression of the myocardial cells and interstitium by systolic forces, or differences in the duration of ischemia employed. Cardiac massage may also have influenced the resolution of edema. The mechanisms involved are likely to include mechanical forces drawing fluids into the circulation, restoration of membrane function and endothelial integrity, and decreases in cellular osmolarity associated with synthesis of adenosine triphosphate and glycogen.
The relevance of the present results to myocardial edema encountered clinically is speculative. Plegisol, widely used for myocardial protection clinically in the past, is likely to have resulted in myocardial edema during cross-clamping. However, two-dimensional echocardiographic studies from this laboratory did not detect mass increases at the conclusion of cardiopulmonary bypass in such patients [18]. The time course of edema resolution in the present study is rapid enough to be complete by the conclusion of cardiopulmonary bypass in a routine operation. The effects of longer periods of aortic cross-clamping, myocardial injury, and hemodilution on this process need to be elucidated. Whole blood cardioplegia and hypertonic crystalloid cardioplegia are widely used for current surgery, making iatrogenic edema less likely. However, these issues may well be of importance in patients with significant intraoperative or perioperative ischemic myocardial injury, large volumes of crystalloid administration, or preservation problems in cardiac transplantation [20].
Although the importance of edema in myocardial protection has frequently been overlooked in the past, the present large animal model can be used to screen new technologies of myocardial protection for this potentially important side effect.
Edema is associated with reduced left ventricular compliance in ischemic myocardial injury [21, 22] and transplant rejection [23, 24], but it is not clear whether edema causes changes in diastolic properties in these conditions or whether both changes in compliance and edema are secondary manifestations of primary cellular damage (eg, rigor complexes [25]). This issue will be easier to resolve when improved techniques for analyzing ventricular compliance are applied to edematous states such as anasarca, sepsis, and generalized trauma in which levels of myocardial edema are similar to what can be achieved experimentally by coronary perfusion [19].
The present results do not demonstrate statistically significant changes in contractility during the reperfusion period. However, a period of depressed systolic function may have occurred. Assessment of contractility in this model is somewhat ambiguous because of inotropic drug administration and reflex effects during the resuscitation period. These influences may have altered systolic function but are unlikely to have influenced diastolic properties. Moreover, any confounding influence on diastolic properties should have equally affected both the edema and control groups, but statistically significant changes in diastolic properties were only observed in the edema group.
Only compliance abnormalities were observed in the present study, but the occurrence of such abnormalities clinically could be perceived differently. Thus, altered compliance would require elevated filling pressures to achieve any given level of cardiac performance. This might be perceived (incorrectly) as depression of ventricular contractility [26].
The use of conductance in the open chest is fraught with many potential problems, including the presence of metallic objects, changes in cardiac temperature, variable volume of the right ventricle, and changing lung volume, which can alter both parallel conductance and the slope constant for calibration [19, 27, 28]. In theory, this requires frequent recalibration when the conductance technique is employed. In the present experiment, parallel conductance was measured twice: once after the baseline hemodynamic recordings before aortic cross-clamping and once after completion of the 90-minute data collection. Repeated measurements of parallel conductance by hypertonic saline injection was avoided to prevent an intravascular hyperosmolar state in pigs causing premature resolution of myocardial edema. We used continuous monitoring of the echo-conductance relation to confirm validity of conductance during the period when measurement of parallel conductance was not practical. We have demonstrated that end-diastolic conductance and end-diastolic cross-sectional area of the left ventricle by two-dimensional echocardiography are linearly related during changes in volume associated with vena caval occlusion [11].
In summary, this article reports a model for the study of the time course of resolution of myocardial edema in the pig heart. In this large animal, perfusion-induced myocardial edema resolves after 30 minutes in the beating, blood-reperfused pig heart in vivo. The model may prove useful for studying the effects of longer periods of ischemia on resolution of edema and for examining functional effects of other techniques of cardioplegia.
|
Acknowledgments |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Supported in part by United States Public Health Service grant 1 RO1 HL-48109-03 and Columbia University Department of Surgery Research Committee. Doctor Dean is supported by National Institutes of Health National Research Service Award training grant HL09325-01. Doctor Spotnitz is the George H. Humphreys II Professor of Surgery.
|
Footnotes |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. Elahi, S. Asopa, A. Pflueger, N. Hakim, and B. Matata Acute kidney injury following cardiac surgery: impact of early versus late haemofiltration on morbidity and mortality Eur. J. Cardiothorac. Surg., May 1, 2009; 35(5): 854 - 863. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. M. Prasad, A. S. Al-Dadah, G. D. Byrd, T. P. Flagg, J. Gomes, R. J. Damiano Jr, C. G. Nichols, and J. S. Lawton Role of the Sarcolemmal Adenosine Triphosphate-Sensitive Potassium Channel in Hyperkalemic Cardioplegia-Induced Myocyte Swelling and Reduced Contractility Ann. Thorac. Surg., January 1, 2006; 81(1): 148 - 153. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Luckraz, M. B Gravenor, R. George, S. Taylor, A. Williams, S. Ashraf, V. Argano, and A. Youhana Long and short-term outcomes in patients requiring continuous renal replacement therapy post cardiopulmonary bypass Eur. J. Cardiothorac. Surg., May 1, 2005; 27(5): 906 - 909. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Korvald, O. P. Elvenes, T. Myrmel, and D. G. Sorlie Cardiac dysfunction and inefficiency after substrate-enriched warm blood cardioplegia Eur. J. Cardiothorac. Surg., September 1, 2001; 20(3): 555 - 564. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. E. Marik Pulmonary Artery Catheterization and Esophageal Doppler Monitoring in the ICU Chest, October 1, 1999; 116(4): 1085 - 1091. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. A. Borger, K. S. Wei, R. D. Weisel, J. S. Ikonomidis, V. Rao, G. Cohen, T. Shirai, A. S. Omran, S. C. Siu, and H. Rakowski Myocardial perfusion during warm antegrade and retrograde cardioplegia: a contrast echo study Ann. Thorac. Surg., September 1, 1999; 68(3): 955 - 961. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. A. Dean, M. M. Amirhamzeh, C.-X. Jia, S. E. Cabreriza, D. G. Rabkin, R. Sciacca, M. L. Dickstein, and H. M. Spotnitz REVERSAL OF IATROGENIC MYOCARDIAL EDEMA AND RELATED ABNORMALITIES OFDIASTOLIC PROPERTIES IN THE PIG LEFT VENTRICLE J. Thorac. Cardiovasc. Surg., May 1, 1998; 115(5): 1209 - 1214. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. A. Dean, C.-X. Jia, S. E. Cabreriza, P. F. Soto, D. G. Rabkin, M. J. Sardo, N. Chalik, and H. M. Spotnitz Retrograde Coronary Perfusion: Effects on Iatrogenic Edema and Diastolic Properties Ann. Thorac. Surg., February 1, 1998; 65(2): 449 - 453. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. M. R. Amirhamzeh, D. T. Hsu, S. E. Cabreriza, C.-X. Jia, and H. M. Spotnitz Myocardial Edema: Comparison of Effects on Filling Volume and Stiffness of the Left Ventricle in Rats and Pigs Ann. Thorac. Surg., May 1, 1997; 63(5): 1293 - 1297. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
