Effects of Elevated Coronary Sinus Pressure on Coronary Blood Flow and Left Ventricular Function
Implications After the Fontan Operation
Background After the Fontan operation there is elevated systemic venous pressure, and the coronary sinus pressure (CSP) may also be elevated depending on the operative technique. Elevated CSP can potentially alter coronary perfusion and thereby be a cause for postoperative left ventricular (LV) dysfunction.
Methods and Results The effects of elevated CSP on coronary blood flow (CBF) and LV function were evaluated in 14 isolated blood-perfused juvenile lamb hearts. After baseline measurements were made, CSP was elevated by a 10-minute inflation of a balloon catheter inserted into the coronary sinus via the hemiazygos vein in 7 hearts (CSHT group) to cause moderate (phase I, ≈15 mm Hg) and severe (phase II, ≈30 mm Hg) elevations of mean CSP at a constant coronary perfusion pressure (80 mm Hg). The results were compared with results from 7 hearts continuously perfused without elevation of CSP (C group). Mean CSP in the CSHT group was elevated from 0.4±1.9 to 16.4±2.4 mm Hg during phase I and to 32.6±3.6 mm Hg during phase II. CBF in the CSHT group decreased to 89.7±5.2% in phase I and to 79.0±13.2% in phase II, and these values were significantly lower than those in the C group (98.5±6.7% in phase I and 106.8±16.0% in phase II; P<.05 each). There were no significant differences in maximum developed pressure (DP), max+dP/dt, max−dP/dt, or LV end-diastolic pressure (LVEDP) at a fixed volume between the CSHT group and the C group either in phase I or phase II. The time constant of pressure decline during LV isovolumic relaxation (tau) showed no significant difference in phase I, but in phase II tau was significantly higher in the CSHT group (116.2±7.8%) than that in the C group (106.3±8.5%; P<.05).
Conclusions Elevated CSP on a short-term basis did not affect the LV systolic function indexes (max DP, max+dP/dt), max−dP/dt, or LVEDP at a fixed volume, but tau did appear to worsen and CBF decreased during CSP elevation. These actions might have deleterious effects on the LV over a longer time period.
Systemic venous pressure after the Fontan procedure is generally higher than normal to maintain pulmonary perfusion without a ventricular pumping chamber in the pulmonary circuit.
Systemic venous pressure can also be increased by the elevation of left atrial (LA) pressure that results from ventricular dysfunction, an elevation of pulmonary vascular resistance, or an increase in cardiac output during exercise. If the coronary sinus drains into the systemic venous chamber, then systemic venous hypertension leads to the elevation of coronary venous pressure. Elevated CSP may affect coronary perfusion and subsequently reduce ventricular function.
Previous studies examining this question in isolated blood-perfused heart or in situ heart preparations have yielded various results.1 2 3 The present study used an isolated blood-perfused heart model in juvenile lambs to elucidate the effects of elevated CSP on CBF and LV function.
An isolated blood-perfused heart model, which has been described in detail in previous studies from our laboratory,4 was used for this study. A total of 14 juvenile lambs (body wt, 6.0 to 9.8 kg; age, 9 to 36 days) were studied. After ketamine injection (40 mg/kg IM), they were intubated and mechanically ventilated with inhalation of a 1:1 mixture of O2 and N2O and 0.5% halothane. Through a median sternotomy, a 22F arterial cannula was inserted retrograde into the brachiocephalic artery after systemic heparinization (2000 U). Coronary perfusion was maintained with heparinized homologous blood using a roller pump (Coronary Perfusion Pump, Olson Medical Products Inc) and a bubble oxygenator (Bentley Bio-2 Infant Blood Oxygenator, Baxter Healthcare Corp). The perfusate was oxygenated with a mixture of 20% O2, 5% CO2, and 75% N2, and coronary perfusion pressure was maintained constant at 80 mm Hg throughout the experiment. The heart was isolated and placed on a temperature-controlled water bath. The temperature of the perfusate and the water bath was maintained at 37°C. Both superior and inferior cavae were ligated, and all of the coronary venous return was collected through a 32F venous cannula that was inserted into the RV retrograde via the pulmonary artery. An 8F vascular sheath was placed into the hemiazygos vein, which connects to the great coronary vein in this species, and a 6F side-hole balloon catheter (Berman angiographic catheter, Arrow International Inc) was placed into the RA through the coronary sinus to elevate the CSP by partial obstruction of coronary venous return. The side hole in this catheter is proximal to the balloon so that by inflating the balloon the pressure upstream of the balloon is measured.
A latex balloon containing a pressure transducer (model SPC-350, Millar Instruments Inc) was placed into the LV through an apex to measure the LV function. A Foley balloon catheter (14F) was inserted into the left atrium to prevent the LV balloon from herniating into the left atrium and to vent blood and air from the LV (Fig 1⇓). A pair of the sonomicrometry crystals (SL5-2, Triton Technology Inc) was implanted just beneath the epicardial surface of the LV free wall along the equatorial circumference to measure epicardial segment length using a sonomicrometer (model 120, Triton Technology Inc).
All animals in this study received humane care as described by the “Principles of Laboratory Animal Care” formulated by the National Society for Medical Research and the “Guide for the Care and Use of Laboratory Animals” prepared by the National Academy of Sciences and published by the National Institutes of Health (NIH Publication No. 86-23, revised 1985).
Perfusion pressure was monitored by a small catheter placed in the arterial cannula, and CSP by the side-hole balloon catheter placed in the coronary sinus.
CBF was measured by a 6-mm in-line–type electromagnetic flow probe and flowmeter (models FF-060T and MFV-3100, Nihon Kohden), which was connected to the cannula inserted into the pulmonary artery. This flow represents the total CBF from both the coronary sinus and thebesian vein.
LV function was measured during isovolumic contraction by inflating the intraventricular balloon with 1-mL increments of saline until an LV end-diastolic pressure (EDP) of 20 mm Hg was reached. LV pressure and its first derivative (dP/dt) were recorded at each volume. Systolic function was evaluated by measuring the maximum developed pressure (DP) and maximum LV positive dP/dt (max+dP/dt). Maximum LV negative dP/dt (max−dP/dt), isovolumic LVEDP, and tau (time constant of pressure decline during LV isovolumic relaxation) at a balloon volume of V10 were measured to assess diastolic function. The volume V10 was defined as the balloon volume needed to produce an EDP of 10 mm Hg during baseline measurement. Calculation of tau was according to zero-symptote method described by Weiss et al.5
Epicardial segment length on the equational circumference of LV free wall was measured by a pair of microsonometry crystals at a balloon volume of V10. Change in epicardial circumferential segment length was directly related to wall thickness change when the isovolumic balloon was inflated, and the wall thickness change was considered to represent the change of the ventricular wall volume.6
All analog physiological data were digitized in real time at a speed of 120 Hz and recorded on a desktop computer (model 4DX2-66V, Gateway 2000) using a commercially available software package (dataflow, Crystal Biotech).
Baseline measurements were made after a 20-minute equilibrium period. In seven hearts (CSHT group), CSP was elevated by partial obstruction of the ostium of the coronary sinus by the balloon catheter to cause moderate (phase I, ≈15 mm Hg) and then severe (phase II, ≈30 mm Hg) elevations of mean CSP. Partial obstruction of the ostium of the coronary sinus was achieved by pulling the balloon that was inflated within the RA toward the coronary sinus. This procedure enabled obstruction of all of the cardiac veins from the LV. A 10-minute equilibrium period was allowed before each hemodynamic measurement in each phase. Coronary perfusion pressure was maintained constant at 80 mm Hg throughout the experiment. In another seven hearts (C group), measurements were made at the same time points without elevation of CSP. All data are given as percent of baseline values, and comparison was made between CSHT group and C group.
All values were expressed as mean±SD and analyzed by a statistical analysis software package (fisher, Nakayama Shoten and spss, SPSS Inc). An unpaired Wilcoxon-Mann-Whitney test was used to compare the differences between the two groups, and ANOVA was used to compare the differences between the phases. A value of P<.05 was considered significant.
There were no significant differences between the groups in baseline values (Table 1⇓).
Mean CSP during baseline measurement was 0.4±1.9 mm Hg in the CSHT group and −0.3±3.2 mm Hg in the C group. CSP was 16.4±2.4 mm Hg during phase I and 32.6±3.6 mm Hg during phase II in the CSHT group. In the C group, it was −0.8±2.3 mm Hg during phase I and −0.2±2.5 mm Hg during phase II.
CBF in the CSHT group significantly decreased, to 89.7±5.2% in phase I (P<.05 versus baseline) and further to 79.0±13.2% in phase II (P<.05 versus baseline and phase I). CBF in the CSHT group was significantly lower than that in the C group in both phase I and phase II (Table 2⇓, Fig 2⇓).
Although max+dP/dt and max−dP/dt in both groups decreased during phases I and II, there were no significant differences in max DP, max+dP/dt, or max−dP/dt between the CSHT group and the C group either in phase I or phase II. There were no significant differences in isovolumic LVEDP between the two groups in phase I or phase II. tau showed no significant difference in phase I, but tau in the CSHT group decreased significantly during phase II and was significantly higher than that in the C group in phase II (Table 2⇑, Fig 3⇓).
Epicardial Segment Length
Isovolumic end-diastolic segment length of the LV free wall did not change either in the CSHT group or the C group throughout the experiment (Table 2⇑).
Effect of Elevated Coronary Venous Pressure on CBF
Short-term elevation of CSP with a constant coronary arterial pressure results theoretically in a decrease of the effective perfusion gradient across the coronary vascular bed and thereby may cause a decrease in CBF. Many years ago, Gregg and Dewald7 reported that the short-term complete occlusion of the coronary vein decreased CBF. Recent studies8 9 have also demonstrated a decrease in CBF concomitant with the elevation of the coronary venous pressure during prolonged diastole of the ventricle. However, conditions such as complete occlusion of the coronary vein or prolonged diastole do not precisely simulate the postoperative condition after a Fontan operation.
Studies by Ilbawi1 and Ward et al,2 whose purposes were to examine CBF during the elevation of CSP using open-chest animals, simulated more closely the postoperative hemodynamics after the Fontan operation. Ilbawi et al inserted a catheter into the coronary sinus externally and closed the opening of coronary sinus to the RA to force all of the venous return from the coronary sinus to drain into an external graduated reservoir. Although they demonstrated a decrease in CBF during the elevation of CSP, the mean flow at baseline (13 mL/min) and the mean flow during the elevation of CSP to 25 mm Hg (2 or 3 mL/min) in their preparation seemed quite low, given the weight of the dogs (7 to 10 kg) and the mean cardiac output (3.5 L · min−1 · m−2). Canine myocardial blood flow in the other experiments in dogs has been reported to be 0.62 to 0.85 mL · min−1 · g−1 of myocardium.10 11 Ilbawi et al also assumed that the venous return from the coronary sinus represented all CBF, but they did not measure or account for collateral flow through the thebesian veins into the RA and RV, which might increase during elevation of CSP. In contrast, Ward and colleagues used radioactive microspheres to determine CBF, and they were unable to demonstrate a decrease in CBF in adult sheep using a balloon catheter to elevate CSP. They inflated the balloon within the coronary sinus, but we have found that the coronary veins from the posterior part of the LV have their openings in very close proximity to the ostium of the coronary sinus. Thus, inflation of the balloon within the coronary sinus might be inadequate to elevate the pressure of all the cardiac veins from the LV. In addition, it has recently been emphasized that there is a rich network of venous collaterals (thebesian system) that can allow coronary venous drainage even in the presence of complete coronary sinus occlusion.12 The extent of this collateral development has been reported to be different among animal species and increases with age.13 Therefore, the differences that might exist between species and ages of the animals used in experiments might result in a different conclusion.
In our model, we used the electromagnetic flowmeter to measure the total venous return from both the coronary sinus and the RA and RV by draining all cardiac venous blood through a cannula placed in the RV. This technique enabled us to account for the venous collaterals. We used juvenile lambs instead of adult sheep because most of the patients who undergo the Fontan operation are children. Under these conditions, we did demonstrate a decrease in total CBF during the elevation of the CSP at a constant coronary arterial pressure.
Effect on LV Systolic Function
The independent effects on LV function of CBF and coronary perfusion pressure in the absence of ischemia remain in dispute despite several studies. Gregg14 reported that increasing coronary arterial pressure and flow increased myocardial oxygen consumption in isolated and in situ dog hearts, and Opie15 found that coronary arterial pressure and flow rate had a significant impact on LV function. Abel and Reis16 and Templeton et al17 concluded that CBF was an independent determinant of the contractile state of LV rather than coronary perfusion pressure, but Arnold et al18 concluded that perfusion pressure was the major determinant of LV function. Most of these studies speculated that the increase in perfusion pressure or flow caused an increased blood volume within the coronary vasculature and that this erectile effect improved the contractile function of the ventricle. On the other hand, Downey19 has reported that the LV function was maintained unless the CBF decreased to the level that was set by autoregulatory mechanisms, and supernormal CBF did not increase the contraction of LV.
Many of the previous studies to elevate CSP have focused on the possibility of the coronary vein as an alternative source of blood supply to an ischemic area such as those supplied by obstructed coronary arteries, and less attention has been paid to coronary venous pressure as a determinant of LV function. Scharf and Bromberger-Barnea20 used isolated blood-perfused dog hearts and reported that elevated CSP produced by complete occlusion increased contractility of the LV. However, they kept total CBF constant and coronary perfusion pressure did increase during occlusion. As a result, coronary vascular volume increased as did the contractility of LV, probably by increasing wall thickness and thereby decreasing wall stress. In the experiment of Ilbawi et al,1 elevated CSP reduced LV systolic function and, subsequently, cardiac output. They speculated that the reduction of LV function was due to the decrease of CBF.
In our experiment, systolic indexes of LV function in the CSHT group were slightly higher than in the C group, but the differences were not statistically significant. Although the elevation of CSP may have an erectile effect on the coronary venous system, ventricular wall volume did not increase in our experiment and these erectile effects might be canceled by the decrease in CBF.
Effect on LV Diastolic Function
Previous reports have addressed the relation between the stiffness or distensibility of LV and coronary circulation. An increase in coronary perfusion pressure or flow rate has been reported to increase the volume of blood in the coronary vasculature and thereby to increase the stiffness (reduce the distensibility) of the LV.21 22 Olsen et al22 reported that coronary perfusion pressure, and not CBF, was a more direct determinant of cardiac diastolic properties, although this study was carried out in a potassium-arrested heart rather than a normally contracting heart. In contrast, Vogel et al6 concluded that CBF appeared to be a more important determinant of wall thickness than perfusion pressure. Watanabe et al3 reported that increased coronary venous pressure decreased LV diastolic distensibility with an increase in LV volume. In their study, CBF decreased slightly during the elevation of RA and RV pressures while segment length on the surface of LV increased, which indicated an increase in LV volume.
In our study, elevation of coronary venous pressure tended to increase LVEDP at a fixed balloon volume, although this elevation was not statistically significant. In addition, elevated CSP prolonged tau, in spite of a reduction in CBF and unchanged segment length on the LV. These results indicate that the disturbance of LV relaxation occurred with elevation of CSP without an increase in LV volume.
A number of reports have shown that diastolic functional abnormalities preceded systolic dysfunction in patients with coronary artery disease23 and valvular disease.24 If the process of the deterioration of ventricular function after Fontan operation was similar to the process in patients with these heart diseases, the abnormalities of diastolic ventricular function might be found in the early period of ventricular deterioration.
Among the diastolic ventricular functional indexes, tau is one of the indexes during relaxation. Although ventricular EDP and stiffness or compliance are also diastolic indexes, some studies have reported that a prolongation of tau occurs without a measurable abnormality in ventricular stiffness or compliance.25 26 These reports suggest that tau may be a more sensitive index for detecting diastolic ventricular abnormalities.
Implications for the Fontan Operation
Importance of Diastolic Abnormalities in Fontan Operation
Recent studies emphasized the importance of diastolic ventricular function before and after Fontan operation.27 28 29 30 After the Fontan operation, diastolic dysfunction can be expected to result in elevated pulmonary venous (LA) pressure, which will elevate pulmonary artery and systemic venous pressures. Patients with ventricular hypertrophy especially after pulmonary artery banding were reported to have a significant operative risk,27 and patients with increased ventricular mass or mass/volume ratio had poor clinical outcomes because the diastolic properties of their ventricles may be disturbed.28 29 Therefore, preoperative assessment of diastolic function has been recommended as an important criterion when patients are considered for this operation.30
In addition, in the early postoperative period, LA pressure is likely to be higher because of the effects of ischemia during the operation. Pulmonary vascular resistance is also likely to be higher after cardiopulmonary bypass, and the positive airway pressure delivered by the mechanical ventilator may make pulmonary vascular resistance higher.31 32 Therefore, elevation of systemic venous pressure can easily develop during the early postoperative period after the Fontan operation. Moreover, short-term volume unloading of ventricle observed during the transition period to Fontan circulation can result in relative hypertrophy of the ventricle, and this hypertrophy may reduce the diastolic function of the ventricle.33 34 The adverse effects of elevated coronary venous pressure on diastolic function as shown in this study might exaggerate the diastolic dysfunction during the early period after the Fontan operation.
In this study, abnormality of the diastolic function was found only in phase II, in which mean CSP reached 30 mm Hg. CSP after Fontan operation at rest generally does not reach this high venous pressure. However, Barber et al35 have reported that systemic venous pressure after the Fontan operation reaches levels up to 27 mm Hg during exercise, and high coronary venous pressure during exercise might be predicted. CBF during exercise must be higher than at rest in response to the increase in oxygen consumption of the heart. During exercise in these patients, the diastolic properties of LV and CBF might be impaired by the higher systemic venous pressure as shown in this study. Impairment of CBF might result in myocardial ischemia and in dysfunction of the ventricle.
Selection of the Operative Method
Many variations of Fontan-type procedures have been described. The coronary sinus drains into the systemic venous atrium (RA) in patients with classic tricuspid atresia after direct atriopulmonary anastomosis with closure of atrial septal defect, but the coronary sinus drains into the pulmonary venous atrium when the lateral tunnel technique with cavopulmonary anastomosis is used. One of the potential advantages of this type of operation has been thought to be a more favorable flow pattern with less energy loss in the venous pathway on the right side,36 but another may be the lower coronary venous pressure after the operation. Although the results of our study show relatively subtle changes in diastolic function short term, we suggest that coronary veins are better connected to the low pressure atrium (ie, pulmonary venous atrium) in the Fontan operation to preserve ventricular function.
If the coronary sinus is connected to the systemic venous atrium and CSP is elevated, an atrial right-to-left shunt may develop through the thebesian venous system, after the development of a collateral pathway from the coronary sinus to the pulmonary venous atrium, in which pressure will be lower than in the systemic venous atrium. We have seen some patients who developed such an atrial right-to-left shunt with resulting arterial desaturation after Fontan operation. Use of the lateral tunnel method, which allows coronary venous drainage to the lower pressure atrium, may help to avoid this atrial shunt.
Limitations of the Study
We used an isolated blood-perfused model to keep constant perfusion pressure, but this model is a non-working heart except when the intraventricular balloon is inflated and therefore it is not truly physiological. Also, this study examined only short-term effects of the elevation of CSP and not long-term effects. Evaluation of the long-term effects may be most important to improve the long-term results after the Fontan operation, and more refined study may be necessary to clarify the long-term effects of coronary sinus hypertension on ventricular function.
Selected Abbreviations and Acronyms
|CBF||=||coronary blood flow|
|CSP||=||coronary sinus pressure|
|LV||=||left ventricle/left ventricular|
|tau||=||time constant of pressure decline during LV isovolumic relaxation|
This work was supported by the Research Fund of the Cardiac Surgery Department, Children’s Hospital. We sincerely thank Mark A. Cioffi for his technical assistance.
- Copyright © 1995 by American Heart Association
Watanabe J, Levine MJ, Bellett F, Johnson RG, Grossman W. Effects of coronary venous pressure on left ventricular diastolic distensibility. Circ Res. 1990;67:923-932.
Weiss JL, Frederiksen JW, Weisfeldt ML. Hemodynamic determinants of the time-course of fall in canine left ventricular pressure. J Clin Invest. 1976;58:751-760.
Vogel WM, Apstein CS, Briggs LL, Gaasch WH, Ahn J. Acute alternations in left ventricular diastolic chamber stiffness: role of the ‘erectile’ effect of coronary arterial pressure and flow in normal and damaged hearts. Circ Res. 1982;51:465-478.
Gregg DE, Dewald D. The immediate effects of the occlusion of the coronary veins on the dynamics of the coronary circulation. Am J Physiol. 1938;124:444-456.
Bellamy RF, Lowensohn HS, Ehlich W, Baer RW. Effect of coronary sinus occlusion on coronary pressure-flow relations. Am J Physiol. 1980;239:H57-H64.
Uhrig P, Baer R, Vlahakes G, Hoffman J. Effect of coronary sinus pressure elevation on coronary flow. Circulation. 1981;64(suppl IV):IV-38. Abstract.
Simpson PJ, Schelm JA, Smith GF. Therapeutic defibrination with ancrod does not protect canine myocardium from reperfusion injury. J Pharmacol Exp Ther. 1991;256:780-786.
Ratajczyk-Pakalska E. The coronary venous anatomy. In: Meerbaum S, ed. Myocardial Perfusion, Reperfusion, Coronary Venous Retroperfusion. New York, NY: Springer; 1990:51-91.
Gregg DE. Effect of coronary perfusion pressure or coronary flow on oxygen usage of the myocardium. Circ Res. 1963;13:497-500.
Abel RM, Reis RL. Effects of coronary blood flow and perfusion pressure on left ventricular contractility in dogs. Circ Res. 1970;27:961-971.
Templeton GH, Wildenthal K, Michell JH. Influence of coronary blood flow on left ventricular contractility and stiffness. Am J Physiol. 1972;223:1216-1220.
Arnold G, Kosche F, Miessner E, Neitzert A, Lochner W. The importance of the perfusion pressure in the coronary arteries for the contractility and the oxygen consumption of the heart. Pflugers Arch. 1968;229:339-356.
Downey JM. Myocardial contractile force as a function of coronary blood flow. Am J Physiol. 1976;230:1-6.
Scharf SM, Bromberger-Barnea B. Influence of coronary flow and pressure on cardiac function and coronary vascular volume. Am J Physiol. 1973;224:918-925.
Gaasch WH, Bing OHL, Franklin A, Rhodes D, Bernard SH, Weintraub RM. The influence of acute alternations in coronary blood flow on left ventricular diastolic compliance and wall thickness. Eur J Cardiol. 1978;7(suppl):147-161.
Olsen CO, Attarian DE, Jones RN, Hill RC, Sink JD, Lee KL, Wechsler AS. The coronary pressure-flow determinants of left ventricular compliance in dogs. Circ Res. 1981;49:856-865.
Aroesty JM, McKay RG, Heller GV, Royal HD, Als AV, Grossman W. Simultaneous assessment of left ventricular systolic and diastolic dysfunction during pacing-induced ischemia. Circulation. 1985;71:889-900.
Shintani H, Glantz SA. Effect of disrupting the mitral apparatus on left ventricular function in dogs. Circulation. 1993;87:2001-2015.
Whittenberger JL, McGregor M, Berglund E, Borst HG. Influence of state of inflation of the lung on pulmonary vascular resistance. J Appl Physiol. 1960;15:878-882.
Meliones JN, Bove EL, Dekeon MK, Custer JR, Moler FW, Callow LR, Wilton NC, Rosen DB. High-frequency jet ventilation improves cardiac function after the Fontan procedure. Circulation. 1991;84(suppl III):III-364-III-368.
Gewillig MH, Lundström UR, Deanfield JE, Bull C, Franklin RC, Graham TP Jr, Wyse RK. Impact of Fontan operation on left ventricular size and contractility in tricuspid atresia. Circulation. 1990;81:118-127.
Penny DJ, Lincoln C, Shore DF, Xiao HB, Rigby ML, Redington AN. The early response of the systemic ventricle during transition to the Fontan circulation: an acute hypertrophic cardiomyopathy? Cardiol Young. 1992;2:78-84.