Adverse Hemodvnamic Effects of Pericardial Closure Soon After Open Heart Operation J
Steven Hunter, FRCS, Geoffrey H. Smith, FRCS, and Gianni D. Angelini, MD, MCh, FRCS Department of Cardiac Surgery, The University of Sheffield, Northern General Hospital, Sheffield, United Kingdom
The short-term hemodynamic effects of pericardial closure on cardiac function were studied during steady-state anesthesia and ventilation in 10 patients (6 men) (mean age, 59 f 9 years) who underwent an open-heart valve operation. Observations were made after the heart was decannulated, both while the pericardium was open and after it had been closed, and then after closure of the chest after the pericardium had been reopened by removing the pericardial suture through the chest wall. The effect of closing the pericardium before closing the chest was an immediate reduction in cardiac output (thermodilution) of 1.39 f 0.24 L/min from 5.09 2 0.40 Wmin (p < 0.001). The heart rate remained stable, but there was a decrease in stroke volume of 29% and an increase in systemic vascular resistance of 34% (both, p < 0.01). The mean arterial pressure increased slightly by 2% (not significant). Opening the pericardium (1.5 to 2 hours
after the end of the operation) while the chest remained closed was followed by an increase in cardiac output of 1.33 f 0.15 L/min from 4.12 f 0.62 L/min (p < 0.001). As the heart rate and the mean blood pressure changed insignificantly, there was an increase in stroke volume of 15 2 3 mL from 53 2 5 mL and a reduction in systemic vascular resistance of 473 2 83 dyne s cm-5 from 1,721 2 181 dyne . s cm-5 (both, p < 0.01). The data suggest that pericardial closure early after an open heart operation consistently lowers stroke volume and cardiac output; mean blood pressure is maintained by an increase in systemic vascular resistance. Thus, monitoring blood pressure alone is not a reliable guide to decisions on whether to close the pericardium or leave it open in individual patients.
T
he physiological function of the intact pericardium with its fluid is to lubricate and isolate the heart from other structures in the thorax, thus preventing adhesions and the spread of infection [l].Congenital absence of the pericardium or surgical removal, however, has no serious effects [Z]. Traditionally surgeons were against closure of the pericardium after an open heart operation in the belief that this would increase the risk of tamponade [3]. More recent studies have suggested, however, that closure of the pericardium may reduce postoperative bleeding [4,51 and protect the heart at repeat sternotomy [ 6 ] . Controversy remains regarding the effects of pericardial closure on cardiac function, particularly on blood pressure and cardiac output. Reduction in cardiac output [7-91 and blood pressure [101 has been reported after pericardial closure. All these studies [7-101 included patients who had undergone coronary artery bypass grafting, and consequently there was the possibility of a constrictive effect of pericardial closure on the grafts as well as on the heart. In the work of Damen and Bolton [lo], measurements were carried out only intraoperatively while the sternotomy remained opened, a situation that may not be directly relevant to the clinical status after chest closure.
In a recent study in pigs, our group [ll]showed that after sternotomy and pericardiotomy, pericardial closure consistently lowered cardiac output and stroke volume, whereas systemic blood pressure was maintained by an increase in systemic vascular resistance. These changes were reversed by opening the pericardium. However, no cardiac operation was performed, and extrapolation of animal data to humans cannot be absolute. The aim of the present study was, therefore, to assess the hemodynamic effects of closing and opening the pericardium while the chest remained open and after it had been closed in 10 patients who underwent a valve operation.
Accepted for publication Aug 20, 1991.
Hernodynamic Measurements
Address reprint requests to h4r Angelini, Department of Cardiac Surgery, The University of Sheffield, Clinical Sciences Centre, Northern General Hospital, Sheffield S5 7AU, United Kingdom.
0 1992 by The Society of
Thoracic Surgeons
(Ann Thorac Surg 1992;53:425-9)
Material a n d Methods With informed consent, studies were performed on 10 patients who required a valve procedure. There were 6 male patients, and the mean age was 59 + 9 years (+ standard deviation). Six patients underwent mitral valve replacement, 3 had aortic valve replacement, and 1 had mitral valve repair and annuloplasty. Anesthetic and operative techniques were standardized. Throughout the study period, the patients did not receive vasopressor, vasodilator, or inotropic pharmacological agents. A peripheral electrocardiographic lead was monitored in the usual manner, and the heart rate was recorded. A 0003-4975/92/$5.00
426
HUNTER ET AL HEMODYNAMIC EFFECTS OF PERICARDIAL CLOSURE
Fig I . The study protocol. The cross-hatched areas represent the experimental periods during which the chest, the pericardium, or both were closed. During period 3 , the open heart procedure was performed. Between periods 7 and 8, the patient was transferred from the operating theater to the cardiac intensive care unit. The lines at the foot of the figure denote the periods used for the statistical comparisons. (A = pericardial closure with the chest open (periods 4 and 5 versus period 3); B = chest closure with the pericardium closed (period 7 versus' period 5); C = comparison between period 8 and period 7 (time delay between these periods was 2.5 to 2 hours [see text]); D = pericardial opening with the chest closed (periods 9 and 10 versus period 8).)
Ann Thorac Surg 1992;534259
Experimental Period
3
2
1
4
5
6
7
8
9
10
1'
lo-
Chest Closed
Pericardium Closed
I lime Delay (mins)
l
l
-1'
lU
1'
lo-
A
U
Statistical Comparisons
20-gauge Jelco cannula (Critikon Ltd, Tampa, FL) was placed percutaneously into a radial artery for continuous measurement of blood pressure (Viggo/Spectramedpressure transducers; BOC Health Care Co, Swindon, Wiltshire, UK) and for sampling of blood for arterial gases. The tip of a 7F Swan-Ganz flow-directed thermodilution catheter (Baxter International Inc, Santa Ana, CA) was positioned in the pulmonary artery through a percutaneous internal jugular venotomy. Right atrial, pulmonary artery, and pulmonary capillary wedge pressures were recorded, and cardiac output was measured in duplicate (or in triplicate if the difference between the first two measurements was > 10%) using an SAT 1 oximeter/ cardiac output computer (American Edwards Laboratories, Santa Ana, CA). Stroke volume and systemic and pulmonary vascular resistances were calculated using standard formulas [12].
Experimental Protocol A maximum of ten sets of data were recorded for each patient, as summarized in Figure 1. Baseline hemodynamic measurements were recorded with the chest closed (period l), and then after the chest was opened through a median sternotomy, the measurements were repeated (period 2). The pericardium was opened through a vertical incision, after which the measurements were again repeated (period 3). The operative procedure was performed during period 3. The mean cardiopulmonary bypass time was 81 f 5 minutes, with a mean aortic cross-clamp time of 44 & 5 minutes. After decannulation, 20 minutes were allowed to elapse so that a stable state was reached before the hemodynamic measurements were repeated. The pericardium was then closed using a continuous 1-0 Ethilon suture (Ethicon Ltd, Edinburgh, UK), which was exteriorized at each end of the skin wound. The ends of the suture were secured to the surgical drapes to keep the pericardium tightly closed. Sets of data were recorded 1
B
4 %
minute and 10 minutes later (periods 4 and 5, respectively). The sternotomy incision was then closed in layers, and further sets of data were recorded at 1 minute and 10 minutes (periods 6 and 7). With the ends of the pericardial suture taped securely to the skin, the patient was transferred from the operating theater to the cardiac intensive care unit. With the patient in a steady state of anesthesia and ventilation, the hemodynamic measurements were repeated (period 8). The time between periods 7 and 8 was a mean of 105 minutes (range, 90 to 125 minutes). Then, while the chest remained closed, the pericardium was opened by pulling out the pericardial suture, and sets of data were recorded at 1 minute and 10 minutes (periods 9 and 10, respectively). In the first 5 patients, radiopaque metal clips were applied to the edges of the pericardium. A chest roentgenogram was obtained to confirm that the pericardium remained closed until the suture was removed and again to confirm that the edges became widely separated when it was withdrawn. The most important variables after four key interventions are summarized in Figure 2. These interventions were pericardial closure with the chest open (data set A, corresponding to the changes between periods 4 and 5 compared with period 3 in Fig l), chest closure with the pericardium closed (data set B, periods 6 and 7 compared with period 5), transfer of the patient from the operating theater to the cardiac intensive care unit with both chest and pericardium closed (data set C, period 8 compared with period 7), and pericardial opening while the chest remained closed (data set D, periods 9 and 10 compared with period 8).
Statistical Analysis Data are expressed as the mean f the standard error of the mean unless stated otherwise. In Figure 2, data are given as mean changes compared with the value preceding the intervention. The statistical significance of the
HUNTERETAL HEMODYNAMIC EFFECTS OF PERlCARDlAL CLOSURE
Ann Thorac Surg
1992;534259
1.00
1
0.00
-1.00
-2.00
A
1
I
B
C
U
B
C
D
30 20 10 0
-10 -20 -30
A
!XSl'EMIC VASCUUR RESISTANCE (we.aec.crn-3
900,
0
-T -900
I
p (0.01
: : I I
1
I
A
B
C
I
D
I
Fig 2 . Summay of mean changes (? standard error of the mean) in cardiac output, stroke volume, and systemic vascular resistance during four key periods of study. Open bars are change at 1 minute and hatched bars, change at 10 minutes. The statistical comparison is with the value preceding the intervention. (A = pericardium closed, chest open; B = pericardium closed, chest closed; C = transfer from operating theater to cardiac intensive care unit; D = pericardium open, chest closed.)
changes in the measured variables was determined by using paired Student's t tests. The overall level of significance was set at a probability equal to 0.05 for two-tailed tests.
Results When the chest was opened by median sternotomy, there were no significant hemodynamic changes: heart rate,
427
right atrial pressure, mean pulmonary artery pressure, and stroke volume remained constant; cardiac output fell insignificantly by 2%; and mean arterial pressure and calculated systemic vascular resistance rose insignificantly by 2%and 3%, respectively. Pericardiotomy caused a rise in heart rate of 3 2 1 beats per minute from 62 f 4 beats per minute and in cardiac output of 0.04 f 0.05 L/min from 3.43 2 0.10 L/min. Mean arterial pressure fell by 2 f 3 mm Hg from 76 2 4 mm Hg; systemic vascular resistance, by 59 f 65 dyne * s * cm-5 from 1,651 f 98 dyne s . cmP5;and stroke volume, by 2 2 1 mL from 56 f 2 mL. None of these changes were significant. There were no changes in the right heart pressures. After discontinuation of cardiopulmonary bypass, closing the pericardium while the chest remained open caused an immediate reduction in cardiac output of 1.39 +0.24 L/min from 5.09 f 0.40 L/min (p < 0.001) (Table 1). There was no further change after 10 minutes (see Fig 2, data set A). As the heart rate remained stable (75 f 5 to 74 f 4 beats per minute), there was a decrease in stroke volume of 21 2 5 mL from 72 f 8 mL (p < 0.01). There was no initial change in mean arterial pressure 1 minute after closure of the pericardium, but there was an insigruficant increase of 5 f 3 mm Hg from 76 f 5 mm Hg after 10 minutes. The central venous pressure (right atrial) remained unchanged at 6 f 1 mm Hg. Systemic vascular resistance increased by 400 f 95 dyne . s . cm-5 from 1,193 f 147 dyne . s . cm-' (p < 0.01). There were no changes in mean pulmonary artery pressure or in calculated pulmonary vascular resistance after pericardial closure. Closing the chest caused no major change in any of the hemodynamic variables (see Table 1; Fig 2, data set B). Immediately after the chest was closed, the heart rate fell slightly (by ll%),and the mean arterial pressure increased slightly (by 5%), but neither was significant; however, after 10 minutes, both approximated the values recorded after pericardial closure (see Table 1). After the patients were transferred from the operating theater to the cardiac intensive care unit, there was an increase in cardiac output of 0.64 f 0.47 Wmin from 3.48 2 0.19 L/min (see Fig 2, data set C). There was also a slight increase in heart rate (by 3%), mean arterial pressure (by 8%), stroke volume (by lo%), and systemic vascular resistance (by 2%),but none were significant (see Table 1). There was no change in the pressures of the right side of the heart. Opening of the pericardium while the chest remained closed was followed by an immediate and persistent rise in cardiac output of 1.19 & 0.14 L/min from 4.12 2 0.60 L/min (p < 0.001) (see Fig 2, data set D). Because there was an insignificant increase in heart rate from 77 f 5 beats per minute to 81 f 5 beats per minute, there was a corresponding increase in stroke volume of 14 2 3 mL from 53 f 5 mL (p < 0.01). Mean arterial pressure did not change, and calculated systemic vascular resistance fell by 464 f 99 dyne . s . cm-5 from 1,721 f 181 dyne . s . cm-5 (p < 0.01). These changes all persisted during the 10 minutes after the pericardium had been opened (see Table 1). There was no change in the right heart pressures or the pulmonary vascular resistance.
428
HUNTER ET AL HEMODYNAMIC EFFECTS OF PERICARDIAL CLOSURE
Ann Thorac Surg 1992;53:4259
Table 1 . Hemodynamic Measurements During Each Study Periodaph Period Variable
3
4
5
6
7
Heart rate 75 f 5 74 f 4 76 2 5 6824 75 f 5 77 f 5 (peats/min) Mean arterial 76 2 5 77 f 4 81 f 4 85f3 79 f 4 8522 pressure (mm Hg) 5.09 f 0.4 3.7 5 0.2F 3.6 f 0.23' 3.43 0.24' 3.48 f 0.19' 4.12 f 0.6 Cardiac output (Llmin) Stroke 51 4' 49 f 5' 52 f 5' 48 f 4' 53 f 5 72 2 8 volume (mu Systemic 1,193 f 147 1,593 5 134' 1,692 f 126' 1,802 f 118' 1,684 f 118' 1,721 f 181 vascular resistance (dyne * s * ~ m - ~ )
*
*
Periods 3 to 10 are as described in Figure 1. 3. Significance: p < 0.001 versus period 8. a
9
8
10
81 f 5
83
84f3
84f3
5.31
f
0.6d
67
f
6'
1,257 -+ 96'
5.45
68
f
f
6
0.62d
f 7'
1,248 f 115'
Data are shown as the standard error of the mean. Significance: p < 0.001 versus period Significance: p < 0.01 versus period 3. Significance: p < 0.01 versus period 8.
Comment This study showed that pericardial closure lowered cardiac output irrespective of whether the chest was open or closed. Heart rate and mean arterial pressure remained constant because of an immediate elevation of the systemic vascular resistance, which was maintained after the chest was closed. The reversal of these effects was observed when the pericardium was reopened. This is in agreement with data from our animal model [ll], in which the use of a small epicardial echocardiographic probe enabled us to demonstrate an increase in the size of the left ventricle when the pericardium was opened, a finding compatible with constriction of the heart chamber as the underlying cause of the hemodynamic changes induced by pericardial closure. Such a restrictive effect may be greater after the period of cardiopulmonary bypass and cardioplegic arrest when some degree of edema to the heart is to be expected. This may be further enhanced by the presence of chest drains and residual blood in the pericardial cavity. Between the last measurement in the operating theater and the first in the cardiac intensive care unit, a mean time of 105 minutes lapsed during which there was no major compensation of the depressed cardiac output or stroke volume. These results compare favorably with those in previous studies [7,81 in which a continuous suture was withdrawn through the chest wall 11to 15 hours after the sternum had been closed. They suggest that the depression in cardiac output and stroke volume is not temporary but persists as long as the pericardial closure exerts a restrictive effect on the heart. A previous report of hypotension after pericardial closure [lo] was not confirmed in this study. This may be explained by differences in study design. Propranolol hydrochloride was used in the cardioplegic solution, and
'
this may have had an effect on the systemic blood pressure. Also, the effect of pericardial closure was assessed only intraoperatively while the sternotomy remained opened. Furthermore, the study included patients who underwent coronary artery bypass grafting, in whom pericardial closure may have had a constrictive effect on the grafts as well as on the heart, although there was no report of electrocardiographic evidence of ischemia. The design of our study, in which only patients who had had a valve procedure were entered, allowed us to avoid the confounding influences caused by the presence of bypass grafts. The fact that the rapid fall in cardiac output and stroke volume resulted in an increase in systemic vascular resistance, which maintained arterial blood pressure, may be an argument for closing the pericardium in patients who remain in stable condition and in a good circulatory state as they come off bypass. However, even in these patients, a reduction in cardiac output and an increase in peripheral systemic vascular resistance may reduce peripheral perfusion, cause acidosis, and slow the rate of rewarming after operation. In conclusion, this study suggests that measuring systemic arterial blood pressure and other simple hemodynamic indices is insufficient to demonstrate the circulatory effect of pericardial closure in individual patients. Any decision as to whether or not to close the pericardium can be made in the operating theater while the chest is still open by measuring the effect of closure on cardiac output and systemic vascular resistance. In the critically ill patient with a low cardiac output, poor peripheral perfusion, or other problems when coming off bypass or who already requires inotropic support, any further acute embarrassment to the circulation should be avoided by leaving the pericardium open.
429
A n n Thorac Surg IYY2;33:42.%Y
HUNTER ET AL HEMODYNAMIC EFFECTS OF I’ERICARDIAL CLOSURE
References
postoperative LV systolic function [Abstract]. Circulation 1985;72(Suppl 3):1301 Daughters GT, Mcad CW, Alderman EL, Derby GC, Schwarzkopf A, Miller DC. Effects of the pericardium on LV diastolic filling in man early after cardiac surgery [Abstract]. Circulation 1985;72(Suppl 3):411. Jarvinen A, Peltola K, Rasanen J, Heikkila J. Immediate haemodynamic effects of pcricardial closure after open heart surgery. Scand J Thorac Cardiovasc Surg 1987;21:1314. Damen J, Bolton DT. Acute haemodynamic effects of pericardial closure in man. Acta Anaesthesiol S c a d 1989;33:207-9. Aiigelini GD, Fraser AG, Koning MMG, Smyllie MB, Sutherland GR, Verdouw PD. Adverse hemodynamic effects and echocardiographic consequences of pericardial closure soon after sternotomy and pericardiotomy. Circulation 1990; 82(Suppl 4):397406. Bertrand ME, Widimsky J. Vascular resistance. Eur Heart J 1985;6(Suppl C):19.
1. Holt JP. The normal pericardium. Am J Cardiol 1970;26: 455-65. 2. Shabetai R, Mangiardi L, Bhargava V, Ross J, Higgins CB. The pericardium and cardiac function. Prog Cardiovasc Dis 1979;22:107-34. 3 . Gibbon JH. Surgery of the chest. Philadelphia: W.B. Saunders, 1969143. 4. Nandi P, Leung JSM, Cheung KL. Closure of pericardium after open heart surgery: a way to prevent post-operative cardiac tamponade. Br Heart J 1976;39:1319-23. 5. Asanza L, Rao G, Voleti C, Hartstein ML, Wisoff BG. Should the pericardium be closed after a n open-heart operation? Ann Thorac Surg 1976;22:5324. 6. Bahn CH, Annest LS, Miyamoto M. Pericardial closure. Am J Surg 1986;151:612-5. 7. Frist WH, Daughters GT, Mead CW, Derby GC, Schwarzkopf A, Miller DC. Pericardial closure adversely affects early
8.
9.
10.
11.
12.
Steven Hunter, FRCS, Geoffrey H. Smith, FRCS, and Gianni D. Angelini, MD, MCh, FRCS Department of Cardiac Surgery, The University of Sheffield, Northern General Hospital, Sheffield, United Kingdom
The short-term hemodynamic effects of pericardial closure on cardiac function were studied during steady-state anesthesia and ventilation in 10 patients (6 men) (mean age, 59 f 9 years) who underwent an open-heart valve operation. Observations were made after the heart was decannulated, both while the pericardium was open and after it had been closed, and then after closure of the chest after the pericardium had been reopened by removing the pericardial suture through the chest wall. The effect of closing the pericardium before closing the chest was an immediate reduction in cardiac output (thermodilution) of 1.39 f 0.24 L/min from 5.09 2 0.40 Wmin (p < 0.001). The heart rate remained stable, but there was a decrease in stroke volume of 29% and an increase in systemic vascular resistance of 34% (both, p < 0.01). The mean arterial pressure increased slightly by 2% (not significant). Opening the pericardium (1.5 to 2 hours
after the end of the operation) while the chest remained closed was followed by an increase in cardiac output of 1.33 f 0.15 L/min from 4.12 f 0.62 L/min (p < 0.001). As the heart rate and the mean blood pressure changed insignificantly, there was an increase in stroke volume of 15 2 3 mL from 53 2 5 mL and a reduction in systemic vascular resistance of 473 2 83 dyne s cm-5 from 1,721 2 181 dyne . s cm-5 (both, p < 0.01). The data suggest that pericardial closure early after an open heart operation consistently lowers stroke volume and cardiac output; mean blood pressure is maintained by an increase in systemic vascular resistance. Thus, monitoring blood pressure alone is not a reliable guide to decisions on whether to close the pericardium or leave it open in individual patients.
T
he physiological function of the intact pericardium with its fluid is to lubricate and isolate the heart from other structures in the thorax, thus preventing adhesions and the spread of infection [l].Congenital absence of the pericardium or surgical removal, however, has no serious effects [Z]. Traditionally surgeons were against closure of the pericardium after an open heart operation in the belief that this would increase the risk of tamponade [3]. More recent studies have suggested, however, that closure of the pericardium may reduce postoperative bleeding [4,51 and protect the heart at repeat sternotomy [ 6 ] . Controversy remains regarding the effects of pericardial closure on cardiac function, particularly on blood pressure and cardiac output. Reduction in cardiac output [7-91 and blood pressure [101 has been reported after pericardial closure. All these studies [7-101 included patients who had undergone coronary artery bypass grafting, and consequently there was the possibility of a constrictive effect of pericardial closure on the grafts as well as on the heart. In the work of Damen and Bolton [lo], measurements were carried out only intraoperatively while the sternotomy remained opened, a situation that may not be directly relevant to the clinical status after chest closure.
In a recent study in pigs, our group [ll]showed that after sternotomy and pericardiotomy, pericardial closure consistently lowered cardiac output and stroke volume, whereas systemic blood pressure was maintained by an increase in systemic vascular resistance. These changes were reversed by opening the pericardium. However, no cardiac operation was performed, and extrapolation of animal data to humans cannot be absolute. The aim of the present study was, therefore, to assess the hemodynamic effects of closing and opening the pericardium while the chest remained open and after it had been closed in 10 patients who underwent a valve operation.
Accepted for publication Aug 20, 1991.
Hernodynamic Measurements
Address reprint requests to h4r Angelini, Department of Cardiac Surgery, The University of Sheffield, Clinical Sciences Centre, Northern General Hospital, Sheffield S5 7AU, United Kingdom.
0 1992 by The Society of
Thoracic Surgeons
(Ann Thorac Surg 1992;53:425-9)
Material a n d Methods With informed consent, studies were performed on 10 patients who required a valve procedure. There were 6 male patients, and the mean age was 59 + 9 years (+ standard deviation). Six patients underwent mitral valve replacement, 3 had aortic valve replacement, and 1 had mitral valve repair and annuloplasty. Anesthetic and operative techniques were standardized. Throughout the study period, the patients did not receive vasopressor, vasodilator, or inotropic pharmacological agents. A peripheral electrocardiographic lead was monitored in the usual manner, and the heart rate was recorded. A 0003-4975/92/$5.00
426
HUNTER ET AL HEMODYNAMIC EFFECTS OF PERICARDIAL CLOSURE
Fig I . The study protocol. The cross-hatched areas represent the experimental periods during which the chest, the pericardium, or both were closed. During period 3 , the open heart procedure was performed. Between periods 7 and 8, the patient was transferred from the operating theater to the cardiac intensive care unit. The lines at the foot of the figure denote the periods used for the statistical comparisons. (A = pericardial closure with the chest open (periods 4 and 5 versus period 3); B = chest closure with the pericardium closed (period 7 versus' period 5); C = comparison between period 8 and period 7 (time delay between these periods was 2.5 to 2 hours [see text]); D = pericardial opening with the chest closed (periods 9 and 10 versus period 8).)
Ann Thorac Surg 1992;534259
Experimental Period
3
2
1
4
5
6
7
8
9
10
1'
lo-
Chest Closed
Pericardium Closed
I lime Delay (mins)
l
l
-1'
lU
1'
lo-
A
U
Statistical Comparisons
20-gauge Jelco cannula (Critikon Ltd, Tampa, FL) was placed percutaneously into a radial artery for continuous measurement of blood pressure (Viggo/Spectramedpressure transducers; BOC Health Care Co, Swindon, Wiltshire, UK) and for sampling of blood for arterial gases. The tip of a 7F Swan-Ganz flow-directed thermodilution catheter (Baxter International Inc, Santa Ana, CA) was positioned in the pulmonary artery through a percutaneous internal jugular venotomy. Right atrial, pulmonary artery, and pulmonary capillary wedge pressures were recorded, and cardiac output was measured in duplicate (or in triplicate if the difference between the first two measurements was > 10%) using an SAT 1 oximeter/ cardiac output computer (American Edwards Laboratories, Santa Ana, CA). Stroke volume and systemic and pulmonary vascular resistances were calculated using standard formulas [12].
Experimental Protocol A maximum of ten sets of data were recorded for each patient, as summarized in Figure 1. Baseline hemodynamic measurements were recorded with the chest closed (period l), and then after the chest was opened through a median sternotomy, the measurements were repeated (period 2). The pericardium was opened through a vertical incision, after which the measurements were again repeated (period 3). The operative procedure was performed during period 3. The mean cardiopulmonary bypass time was 81 f 5 minutes, with a mean aortic cross-clamp time of 44 & 5 minutes. After decannulation, 20 minutes were allowed to elapse so that a stable state was reached before the hemodynamic measurements were repeated. The pericardium was then closed using a continuous 1-0 Ethilon suture (Ethicon Ltd, Edinburgh, UK), which was exteriorized at each end of the skin wound. The ends of the suture were secured to the surgical drapes to keep the pericardium tightly closed. Sets of data were recorded 1
B
4 %
minute and 10 minutes later (periods 4 and 5, respectively). The sternotomy incision was then closed in layers, and further sets of data were recorded at 1 minute and 10 minutes (periods 6 and 7). With the ends of the pericardial suture taped securely to the skin, the patient was transferred from the operating theater to the cardiac intensive care unit. With the patient in a steady state of anesthesia and ventilation, the hemodynamic measurements were repeated (period 8). The time between periods 7 and 8 was a mean of 105 minutes (range, 90 to 125 minutes). Then, while the chest remained closed, the pericardium was opened by pulling out the pericardial suture, and sets of data were recorded at 1 minute and 10 minutes (periods 9 and 10, respectively). In the first 5 patients, radiopaque metal clips were applied to the edges of the pericardium. A chest roentgenogram was obtained to confirm that the pericardium remained closed until the suture was removed and again to confirm that the edges became widely separated when it was withdrawn. The most important variables after four key interventions are summarized in Figure 2. These interventions were pericardial closure with the chest open (data set A, corresponding to the changes between periods 4 and 5 compared with period 3 in Fig l), chest closure with the pericardium closed (data set B, periods 6 and 7 compared with period 5), transfer of the patient from the operating theater to the cardiac intensive care unit with both chest and pericardium closed (data set C, period 8 compared with period 7), and pericardial opening while the chest remained closed (data set D, periods 9 and 10 compared with period 8).
Statistical Analysis Data are expressed as the mean f the standard error of the mean unless stated otherwise. In Figure 2, data are given as mean changes compared with the value preceding the intervention. The statistical significance of the
HUNTERETAL HEMODYNAMIC EFFECTS OF PERlCARDlAL CLOSURE
Ann Thorac Surg
1992;534259
1.00
1
0.00
-1.00
-2.00
A
1
I
B
C
U
B
C
D
30 20 10 0
-10 -20 -30
A
!XSl'EMIC VASCUUR RESISTANCE (we.aec.crn-3
900,
0
-T -900
I
p (0.01
: : I I
1
I
A
B
C
I
D
I
Fig 2 . Summay of mean changes (? standard error of the mean) in cardiac output, stroke volume, and systemic vascular resistance during four key periods of study. Open bars are change at 1 minute and hatched bars, change at 10 minutes. The statistical comparison is with the value preceding the intervention. (A = pericardium closed, chest open; B = pericardium closed, chest closed; C = transfer from operating theater to cardiac intensive care unit; D = pericardium open, chest closed.)
changes in the measured variables was determined by using paired Student's t tests. The overall level of significance was set at a probability equal to 0.05 for two-tailed tests.
Results When the chest was opened by median sternotomy, there were no significant hemodynamic changes: heart rate,
427
right atrial pressure, mean pulmonary artery pressure, and stroke volume remained constant; cardiac output fell insignificantly by 2%; and mean arterial pressure and calculated systemic vascular resistance rose insignificantly by 2%and 3%, respectively. Pericardiotomy caused a rise in heart rate of 3 2 1 beats per minute from 62 f 4 beats per minute and in cardiac output of 0.04 f 0.05 L/min from 3.43 2 0.10 L/min. Mean arterial pressure fell by 2 f 3 mm Hg from 76 2 4 mm Hg; systemic vascular resistance, by 59 f 65 dyne * s * cm-5 from 1,651 f 98 dyne s . cmP5;and stroke volume, by 2 2 1 mL from 56 f 2 mL. None of these changes were significant. There were no changes in the right heart pressures. After discontinuation of cardiopulmonary bypass, closing the pericardium while the chest remained open caused an immediate reduction in cardiac output of 1.39 +0.24 L/min from 5.09 f 0.40 L/min (p < 0.001) (Table 1). There was no further change after 10 minutes (see Fig 2, data set A). As the heart rate remained stable (75 f 5 to 74 f 4 beats per minute), there was a decrease in stroke volume of 21 2 5 mL from 72 f 8 mL (p < 0.01). There was no initial change in mean arterial pressure 1 minute after closure of the pericardium, but there was an insigruficant increase of 5 f 3 mm Hg from 76 f 5 mm Hg after 10 minutes. The central venous pressure (right atrial) remained unchanged at 6 f 1 mm Hg. Systemic vascular resistance increased by 400 f 95 dyne . s . cm-5 from 1,193 f 147 dyne . s . cm-' (p < 0.01). There were no changes in mean pulmonary artery pressure or in calculated pulmonary vascular resistance after pericardial closure. Closing the chest caused no major change in any of the hemodynamic variables (see Table 1; Fig 2, data set B). Immediately after the chest was closed, the heart rate fell slightly (by ll%),and the mean arterial pressure increased slightly (by 5%), but neither was significant; however, after 10 minutes, both approximated the values recorded after pericardial closure (see Table 1). After the patients were transferred from the operating theater to the cardiac intensive care unit, there was an increase in cardiac output of 0.64 f 0.47 Wmin from 3.48 2 0.19 L/min (see Fig 2, data set C). There was also a slight increase in heart rate (by 3%), mean arterial pressure (by 8%), stroke volume (by lo%), and systemic vascular resistance (by 2%),but none were significant (see Table 1). There was no change in the pressures of the right side of the heart. Opening of the pericardium while the chest remained closed was followed by an immediate and persistent rise in cardiac output of 1.19 & 0.14 L/min from 4.12 2 0.60 L/min (p < 0.001) (see Fig 2, data set D). Because there was an insignificant increase in heart rate from 77 f 5 beats per minute to 81 f 5 beats per minute, there was a corresponding increase in stroke volume of 14 2 3 mL from 53 f 5 mL (p < 0.01). Mean arterial pressure did not change, and calculated systemic vascular resistance fell by 464 f 99 dyne . s . cm-5 from 1,721 f 181 dyne . s . cm-5 (p < 0.01). These changes all persisted during the 10 minutes after the pericardium had been opened (see Table 1). There was no change in the right heart pressures or the pulmonary vascular resistance.
428
HUNTER ET AL HEMODYNAMIC EFFECTS OF PERICARDIAL CLOSURE
Ann Thorac Surg 1992;53:4259
Table 1 . Hemodynamic Measurements During Each Study Periodaph Period Variable
3
4
5
6
7
Heart rate 75 f 5 74 f 4 76 2 5 6824 75 f 5 77 f 5 (peats/min) Mean arterial 76 2 5 77 f 4 81 f 4 85f3 79 f 4 8522 pressure (mm Hg) 5.09 f 0.4 3.7 5 0.2F 3.6 f 0.23' 3.43 0.24' 3.48 f 0.19' 4.12 f 0.6 Cardiac output (Llmin) Stroke 51 4' 49 f 5' 52 f 5' 48 f 4' 53 f 5 72 2 8 volume (mu Systemic 1,193 f 147 1,593 5 134' 1,692 f 126' 1,802 f 118' 1,684 f 118' 1,721 f 181 vascular resistance (dyne * s * ~ m - ~ )
*
*
Periods 3 to 10 are as described in Figure 1. 3. Significance: p < 0.001 versus period 8. a
9
8
10
81 f 5
83
84f3
84f3
5.31
f
0.6d
67
f
6'
1,257 -+ 96'
5.45
68
f
f
6
0.62d
f 7'
1,248 f 115'
Data are shown as the standard error of the mean. Significance: p < 0.001 versus period Significance: p < 0.01 versus period 3. Significance: p < 0.01 versus period 8.
Comment This study showed that pericardial closure lowered cardiac output irrespective of whether the chest was open or closed. Heart rate and mean arterial pressure remained constant because of an immediate elevation of the systemic vascular resistance, which was maintained after the chest was closed. The reversal of these effects was observed when the pericardium was reopened. This is in agreement with data from our animal model [ll], in which the use of a small epicardial echocardiographic probe enabled us to demonstrate an increase in the size of the left ventricle when the pericardium was opened, a finding compatible with constriction of the heart chamber as the underlying cause of the hemodynamic changes induced by pericardial closure. Such a restrictive effect may be greater after the period of cardiopulmonary bypass and cardioplegic arrest when some degree of edema to the heart is to be expected. This may be further enhanced by the presence of chest drains and residual blood in the pericardial cavity. Between the last measurement in the operating theater and the first in the cardiac intensive care unit, a mean time of 105 minutes lapsed during which there was no major compensation of the depressed cardiac output or stroke volume. These results compare favorably with those in previous studies [7,81 in which a continuous suture was withdrawn through the chest wall 11to 15 hours after the sternum had been closed. They suggest that the depression in cardiac output and stroke volume is not temporary but persists as long as the pericardial closure exerts a restrictive effect on the heart. A previous report of hypotension after pericardial closure [lo] was not confirmed in this study. This may be explained by differences in study design. Propranolol hydrochloride was used in the cardioplegic solution, and
'
this may have had an effect on the systemic blood pressure. Also, the effect of pericardial closure was assessed only intraoperatively while the sternotomy remained opened. Furthermore, the study included patients who underwent coronary artery bypass grafting, in whom pericardial closure may have had a constrictive effect on the grafts as well as on the heart, although there was no report of electrocardiographic evidence of ischemia. The design of our study, in which only patients who had had a valve procedure were entered, allowed us to avoid the confounding influences caused by the presence of bypass grafts. The fact that the rapid fall in cardiac output and stroke volume resulted in an increase in systemic vascular resistance, which maintained arterial blood pressure, may be an argument for closing the pericardium in patients who remain in stable condition and in a good circulatory state as they come off bypass. However, even in these patients, a reduction in cardiac output and an increase in peripheral systemic vascular resistance may reduce peripheral perfusion, cause acidosis, and slow the rate of rewarming after operation. In conclusion, this study suggests that measuring systemic arterial blood pressure and other simple hemodynamic indices is insufficient to demonstrate the circulatory effect of pericardial closure in individual patients. Any decision as to whether or not to close the pericardium can be made in the operating theater while the chest is still open by measuring the effect of closure on cardiac output and systemic vascular resistance. In the critically ill patient with a low cardiac output, poor peripheral perfusion, or other problems when coming off bypass or who already requires inotropic support, any further acute embarrassment to the circulation should be avoided by leaving the pericardium open.
429
A n n Thorac Surg IYY2;33:42.%Y
HUNTER ET AL HEMODYNAMIC EFFECTS OF I’ERICARDIAL CLOSURE
References
postoperative LV systolic function [Abstract]. Circulation 1985;72(Suppl 3):1301 Daughters GT, Mcad CW, Alderman EL, Derby GC, Schwarzkopf A, Miller DC. Effects of the pericardium on LV diastolic filling in man early after cardiac surgery [Abstract]. Circulation 1985;72(Suppl 3):411. Jarvinen A, Peltola K, Rasanen J, Heikkila J. Immediate haemodynamic effects of pcricardial closure after open heart surgery. Scand J Thorac Cardiovasc Surg 1987;21:1314. Damen J, Bolton DT. Acute haemodynamic effects of pericardial closure in man. Acta Anaesthesiol S c a d 1989;33:207-9. Aiigelini GD, Fraser AG, Koning MMG, Smyllie MB, Sutherland GR, Verdouw PD. Adverse hemodynamic effects and echocardiographic consequences of pericardial closure soon after sternotomy and pericardiotomy. Circulation 1990; 82(Suppl 4):397406. Bertrand ME, Widimsky J. Vascular resistance. Eur Heart J 1985;6(Suppl C):19.
1. Holt JP. The normal pericardium. Am J Cardiol 1970;26: 455-65. 2. Shabetai R, Mangiardi L, Bhargava V, Ross J, Higgins CB. The pericardium and cardiac function. Prog Cardiovasc Dis 1979;22:107-34. 3 . Gibbon JH. Surgery of the chest. Philadelphia: W.B. Saunders, 1969143. 4. Nandi P, Leung JSM, Cheung KL. Closure of pericardium after open heart surgery: a way to prevent post-operative cardiac tamponade. Br Heart J 1976;39:1319-23. 5. Asanza L, Rao G, Voleti C, Hartstein ML, Wisoff BG. Should the pericardium be closed after a n open-heart operation? Ann Thorac Surg 1976;22:5324. 6. Bahn CH, Annest LS, Miyamoto M. Pericardial closure. Am J Surg 1986;151:612-5. 7. Frist WH, Daughters GT, Mead CW, Derby GC, Schwarzkopf A, Miller DC. Pericardial closure adversely affects early
8.
9.
10.
11.
12.
