Comparison of Flow Differences amoiig Venous Cannulas

Transcription

1 Comparison of Flow Differences amoiig Venous Cannulas Edward V. Bennett, Jr., MD., John G. Fewel, M.S., Jose Ybarra, B.S., Frederick L. Grover, M.D., and J. Kent Trinkle, M.D. ABSTRACT The efficiency of four methods of venous drainage during cardiopulmonary bypass was evaluated. Superior vena caval (SVC), inferior vena caval (IVC), right atrial (RA), and right ventricular (RV) pressures as well as RV blood flow were measured for tion with a single USCI 40F atrial, a single Sarns 51F cavoatrial, and double USCI 32F caval s with and without caval tourniquets. Measurements for all s were recorded in conditioned dogs during normothermic cardiopulmonary bypass with the heart beating, with the heart fibrillated, and with the apex of the heart elevated in the circumflex position. Venous drainage using double caval s with caval snares provided maximum decompression of the venous system (SVC pressure, - 3 f 1.52 mm Hg [standard deviation]; IVC pressure, t 2.67 mm Hg to mm Hg). However, decompression of the right side of the heart with this method was the poorest (RA pressure, mm Hg to 9.60 k 2.78 mm Hg; RV pressure, 5.67 t 0.82 mm Hg to mm Hg; RV vent flow, 96 t 4 ml/min to 370 t 50 ml/min). The caval s without tapes provided similar venous decompression but with better right-heart decompression (RA pressure, mm Hg to 0.60 t 2.51 mm Hg; RV pressure, 2.13 t 2.23 mm Hg to 11.6 t 0.89 mm Hg; RV vent flow, 29 k 5 ml/min to 75 t 4.5 ml/ min). The USCI 40F atrial provided less efficient decompression of the inferior vena cava (IVC pressure, 4.33 t 1.96 mm Hg to 7.17 t 3.46 mm Hg). The Sarns 51F cavoatrial had the most efficient right-heart decompression (RA pressure, t 2.19 mm Hg to -4.1 t 1.7 mm Hg; From the Division of Cardiothoracic Surgery of the University of Texas Health Science Center at San Antonio and the Audie Murphy Veterans Administration Hospital, San Antonio, TX. Accepted for publication Jan 5, Address reprint requests to Dr. Bennett, Division of Cardiothoracic Surgery, UTHSCSA, 7703 Floyd Curl Dr, San Antonio, TX RV pressure, -2.5 t 1.57 mm Hg to 4.6 t 1.2 mm Hg; RV vent flow, 0 rfr 7 ml/min to 26 rfr 9 ml/min). Placing the heart in the circumflex position adversely affected right heart decompression in all techniques except single tion with the 51F cavoatrial. Likewise, venous decompression was affected adversely in all techniques except double tion with caval snares. With the single 40F atrial, SVC and IVC pressures were both elevated. Only the SVC pressure was affected when the cavoatrial was used. A recent survey of 677 cardiovascular surgeons performing coronary artery bypass graft (CABG) procedures showed that 50.5% used one, 35.4% used two s, and 24.1% used two s with tapes [l]. Is one method really better than another, and do manipulations of the heart during CABG operation affect each method differently? A previous study from this laboratory showed that a large single venous catheter could drain the venous system effectively and decompress the right side of the heart when beating undisturbed in the pericardial cavity [2]. However, clinical experience with the cavoatrial has led us to believe that the catheter was positioned improperly during that study, invalidating the results. As described in that report, the cavoatrial was inserted to the 3 cm mark, and intermittent partial obstruction of the atrial opening was encountered [2]. The error was in positioning this according to the guide marker used in placing such s in dogs, which have a smaller right atrium; this positioned the atrial drainage ports too close to the inferior vena cava. The should be placed with the atrial drainage ports in the upper midatrial area for proper drainage. We undertook reevaluation of venous drainage using the USCI 40F, Sarns 51F cavoatrial, and double USCI 32F s with and without caval snares because we 59

2 60 The Annals of Thoracic Surgery Vol 36 No 1 July 1983 believed that the cavoatrial was positioned improperly in the first study, and because we wished to study the way that elevating the apex of the heart affected the efficiency of each method. The present study compared the efficiency of these venous drainage systems with the heart fibrillated to eliminate uncontrollable contractility variations in the right side of the heart at the time of measurement. This maintained coronary sinus flow and produced a situation of possible right ventricular (RV) distention. The portion of this investigation done with the heart beating served as a control against the previous study, except for the correct placement of the cavoatrial. Materials and Methods Three types of s were used to test the four techniques of venous drainage used most frequently in the clinical practice of coronary artery surgery [l]. These techniques and s were as follows: atrial tion through the atrial appendage into the midatrium using a USCI 40F venous ; cavoatrial tion through the atrial appendage using a Sarns 51F two-stage cavoatrial catheter positioned with the end in the inferior vena cava and the atrial ports in the upper midatrium, as determined by transatrial palpation; and caval tion using two USCI 32F catheters inserted through the atrial appendage and inferior atrial wall into the inferior and superior venae cavae, respectively, with and without caval tourniquets. Five conditioned adult mongrel dogs weighing from 22.7 to 30 kg each were anesthetized with chloralose, 100 mg per kilogram of body weight. Positive-pressure ventilation was maintained prior to cardiopulmonary bypass by a cuffed endotracheal tube using a Harvard pump respirator. A cutdown was made in the right side of the neck of each dog for placement of superior vena caval (SVC) and systemic arterial blood pressure catheters (Statham PE 150) through the internal jugular vein and carotid artery, respectively. The SVC catheter was positioned just below the confluence of the internal jugular and brachiocephalic veins in the superior vena cava. A cutdown was made in the right groin for placement of the inferior vena caval (IVC) pressure catheter (Statham PE 150), which was positioned in the midabdominal inferior vena cava by measurement and palpation from above the diaphragm. A median sternotomy was used to expose the heart. Following intravenous administration of sodium heparin, 3 mg per kilogram of body weight, an 18F arterial war; inserted into the right femoral artery and connected to an extracorporeal circuit consisting of a Bentley S-110 oxygenator and Sarns roller pump. Pressure catheters (14-gauge Jelco intracath) were inserted into the right atrium and right ventricle to measure right atrial (RA) and IW intracavitary pressures. Right ventricular and left ventricular (LV) vent catheters (Bardex 181F) were placed. The right vent was connected to a graduated cylinder positioned at the level of the right ventricle, approximately 40 inches above the oxygenator. The right side of the heart was isolated by snaring the main pulmonary artery with a tourniquet. Venous flow lo the right ventricle was measured in this manner. The LV vent was used for decompression of the ventricle only when the heart was fibrillated. The extracorporeal circuit was primed with Plasma-lyte solution, 20 ml/kg, and fresh heparinized whole blood was obtained from a donor animal just prior to each study. The ph was adjusted to 7.4 with sodium bicarbonate. The oxygenator was perfused with a mixture of 96% oxygen and 4% carbon dioxide at a rate of 4 liters per minute. The flow rate was kept between 70 and 80 ml/kg/min, with the mean arterial blood pressure at 60 mm Hg or higher. Normothermic cardiopulmonary bypass (37.5 C) was used in each procedure. A Beckman 411 recorder was used io monitor and record all pressures. All sets of catheters were used in each dog. The pressures and RV vent flow were recorded with the heart beating and with the heart electrically fibrillated, and measurements were repeated after the apex of the beating or fibrillated heart had been elevated out of the mediastinum toward the right hemithorax in the circumflex position. A 10-minute stabilization period was allowed following each manipulation and prior to obtaining measurements. The SVC, IVC, RA, and RV pressures and mean RV flow were compared for each -

3 61 Bennett et al: Flow Differences among Venous Cannulas tion technique. The data were analyzed using Student s independent t test. Results Right Ventricular Vent Flow With the beating heart of each study animal in the down position, double caval tion with caval snares provided the highest RV flow rate ( ml/min; Fig 1, Table). The 51F cavoatrial setup had the lowest RV vent flow at 0 t 7 ml/min (see Table). This value was significantly lower than those for the other techniques (p < 0.001). Results for the 40F atrial catheter and the double caval s without snares were not significantly different, but both techniques showed markedly better flows than the double caval s. With the heart of each dog electrically fibrillated in the down position, RV vent flow trends paralleled those found with the heart beating (see Table). Double caval tion provided the highest RV flow rate (370 t 50 ml/min), followed by single tion with the 40F atrial (75? 20 ml/ min), double caval tion (29 & 5 ml/min), and single tion using the 51F cavoatrial ( ml/min). The differences between RV vent flows in the 40F atrial and double caval s were significantly higher than in the 51F cavoatrial (p < 0.01) and the double s (p < 0.01). With the heart of each animal electrically fibrillated and elevated in the circumflex position, double caval tion with tapes showed the highest flow at 96? 4 ml/min (see Table). The double caval s without tapes had the next highest flow ( ml/min), followed by the 40F atrial (45? 8 ml/ min) and the 51F cavoatrial (26? 9 mi/ min). Results for both single- techniques were significantly lower than for double caval tion ( p < 0.01) and double caval tion (p < 0.01), but were not significantly different from each other ( p < 0.07). Superior Vena Caval Pressure The lowest SVC pressure with the empty, beating heart of each animal in the down position 200 I50._ c E \ -$ LL 50 0 ** P mno40fr USCI 0 51 Fr Cavoairial 0 32 Fr Caval 7 Snares 32 Fr.Caval F Snares ** p<.o1 -lo beating Fibrillated Fibrillated Down Down UP Fig 1. Right ventricular flow rates in 5 dogs during cardiopulmonary bypass using a single atrial (USCI 40F), a single cavoatrial (Sarns 51F), and double caval s (USCl32F) with and without caval tourniquets. (@ = pulmonary artery snare released.) was for the double caval s at mm Hg (Fig 2; see Table). The double caval s and the 51F cavoatrial had identical SVC pressures (1.0 t 1.26 mm Hg). The highest SVC pressure was noted in the single 40F atrial (1.75 & 0.99 mm Hg). With the empty, fibrillated heart of each animal in the down position, the lowest SVC pressure was also in the double caval s (- 3.0? 1.52 mm Hg; see Table). The 40F atrial and 51F cavoatrial had similar SVC pressures (2.17? 0.75 mm Hg and 2.17 t 0.98 mm Hg, respectively). Differences between SVC pressures with the heart beating or fibrillated and elevated in the circumflex position were not statistically significant. The lowest SVC pressure was again noted in the double caval s, while the 40F atrial had the highest SVC pressure (see Table). The SVC pressures for double caval tion were significantly lower (p < 0.05) than those for either single- technique. Although the SVC pressure was lower for the snared s than for the caval s, this variable only became statistically significant when the heart was elevated (p < 0.05). The 40F atrial catheter also

4 62 The Annals of Thoracic Surgery Vol 36 No 1 July 1983 Venous Drainage and Decompression of the Right Side of the Heart with Four Cannulation Techniques (N = 5)" Heart Empty, Heart Heart Heart Beating, Fibrillated, Beating, Fibrillated, Technique Down Down Elevated Elevated RV VENT FLOW^ Single 40F atrial 80 f f Single 51F cavoatrial O t 7 19f 8... can n u 1 a Double 32F caval s 250 t t Double 32F caval s 65 t 20 29f 5... svc PRESSURE' 45 f 8 26 f f 4.5 Single 40F atrial 1.75 t f t i 0.75 Single 51F cavoatrial 1.00? t ? f 0.82 Double 32F caval s f f f t 1.59 Double 32F caval s 1.00 f t f ? 1.25 IVC PRESSURE' Single 40F atrial 4.67 f t f f 2.70 Single 51F cavoatrial -0.50? f t Double 32F caval s f t f f 2.09 Double 32F caval s f t f f 2.56 RA PRESSURE' Single 40F atrial 0.96 f t ? f 1.30 Single 51F cavoatrial f t f f 2.19 Double 32F caval s 5.00 f f f f 2.78 Double 32F caval s 0.60 t ? f f 2.51 RV PRESSURE' Single 40F atrial 3.33 f f f t 1.47 Single 51F cavoatrial f t f f 1.20 Double 32F caval s 5.67 f ? t 1.51 Double 32F caval s 2.13 f f f t 0.89 "Values shown are mean 2 standard deviation. bdata expressed in milliliters per minute. 'Data expressed in millimeters of mercury. RV = right ventricular; SVC = superior vena caval; IVC = inferior vena caval; RA = right atrial.

5 63 Bennett et al: Flow Differences among Venous Cannulas 5 * pc.05 4] rrp< I"I 9) -I I T -4' I I I Beating Fibrillated Beating Fibrillated Dawn Down UP UP Fig 2. Superior vena caval pressures in 5 dogs during cardiopulmona y bypass using a single atrial, a single cavoatrial, and double caval s with and without caval snares. (See Fig 1 for data presentation.) 6 4 P2 E o " Beating Fibrillated Beating Fibrillated Dawn Down UP UP Fig 3. Inferior vena caval pressures in 5 dogs during cardiopulmona y bypass using a single atrial, a single cavoatrial, and double caval s with and without caval snares. (See Fig 1 for data presentation.) showed a significant change in SVC pressure with the heart elevated during fibrillation ( p < 0.05), but none of the other techniques demonstrated significant changes. Inferior Vena Caval Pressure The IVC pressures were consistently higher for the single 40F atrial (Fig 3; see Table). The double caval s had the lowest overall IVC pressures (see Table). However, these pressures were not statistically different from those for the 51F cavoatrial or the double caval s. In addition, IVC pressures with the 40F atrial were significantly higher than those for the other methods (p < 0.05). Right At rial Pressure The 51F cavoatrial consistently had the lowest RA pressures (Fig 4; see Table). The 40F atrial and double caval s without snares had similar values until the heart was fibrillated and elevated in the circumflex position; RA pressures with the 40F were then mm Hg compared with mm Hg for the unsnared double s (p < 0.05). The RA pressures for the double caval s were consistently the highest (see Table), and these measurements were obtained with the pulmonary artery snare released. The snared double caval s showed significantly higher RA pressures than the single 51F cavoatrial ( p < 0.01), the 40F atrial ( p < 0.05), or the double caval s ( p < 0.05). 6 4 m 2 I 0 E Beating Fibrillated Beating Down Down UP 1 Fibrillated UP Fig 4. Right atrial pressures in 5 dogs during cardiopulmonary bypass using a single atrial, a single cavoatrial, and double caval s with and without caval snares. (See Fig 1 for data presentation.) Right Ventricular Pressure The RV pressure was consistently lower for the 51F cavoatrial (Fig 5; see Table). The snared double caval s showed the highest RV pressures (see Table); again, these measurements were taken without the pulmonary artery snare. The double caval s without snares and the 40F atrial had RV pressures that did not differ significantly except when the heart was fibrillated and elevated (11.6 k 0.89 mm Hg versus mm Hg; p < 0.05). This difference is believed to result from less efficient drainage of the coronary sinus flow around the caval catheters when the heart is elevated. With the heart beating, the 51F cavoatrial showed a significantly lower RV pressure than the other techniques (p < 0.05). However,

6 64 The Annals of Thoracic Surgery Vol 36 No 1 July ' -6 Beating Fibrillated Beating Fibrillated Dawn Down UP UP Fig 5. Right ventricular pressures in 5 dogs during cardiopulmonary bypass using a single atrial, a single cavoatrial, and double caval s with and without caval snares. (PA = pulmonary artery; see Fig 1 for data presentation.) when the heart was fibrillated and in the down position, results for the cavoatrial, the 40F atrial, and the double caval s were not significantly different. With the heart fibrillated and elevated, the RV pressure for the double caval s rose almost as high as that for the double s (11.6? 0.89 mm Hg versus 12.5 * 1.51 mm Hg; see Table). Comment Controversy persists concerning which venous drainage system is most effective in decompressing the heart, and under what circumstances. Many manipulative techniques are used during coronary artery surgery for proper exposure of vessels to be grafted, especially the posterior marginal branches of the circumflex coronary artery and the posterior ventricular branches off the right system. In the present report, we have studied the effects of these manipulations on venous drainage and myocardial decompression with different methods of venous drainage and under conditions of high right-sided flow (perfused heart, beating and fibrillated) to test the capabilities of each drainage sytem. Arom and colleagues [2] showed that the amount of blood reaching the right ventricle during anoxic arrest was similarly low for each of several methods, with a range of mlimin to 19 t 5 ml/min. With the 51F cavoatrial in our study, the intraventricular pressure was negative at a ventricular flow of 0 i 7 mlimin. Dur- ing fibrillation the RV pressure was 2.2 t 0.9 mm Hg with a ventricular flow of 19 t 8 ml/ min. The 51F cavoatrial provided the best RA and RV decompression. The double caval s without tapes and the single 40F had equally good results except when the heart was in the circumflex position. With the heart elevated, the double caval s showed a rise in RV pressure that was probably due to restriction of flow around the s that help decompress the heart. The single 40F in the fibrillated circumflex position had an RV pressure not significantly different from that for the 51F cavoatrial but the RA pressure went from mm Hg to mm fig. Decompression of ihe systemic venous circulation, as measured by the SVC and IVC pressures, showed the SVC pressure to be lowest for the double caval s. All of the drainage techniques had SVC pressures less than mm Hg with the heart in physiological position, but with the heart in the circumflex position only the double caval s with and continued to have SVC pressures lower than 2.0 mm Hg (SVC pressures, 2.8? 0.84 mm Hg to 3.0 t 0.82 mm Hg for the 51F cavoatrial and 3.40 * 0.89 mm Hg to 4.3? mm Hg for the 40F atrial ). Inferior vena caval pressures showed uniformly good decompression of the lower venous system under all conditions tested for all techniques except single tion with the 40F atrial, which had IVC pressures significantly higher than those for the other methods. The IVC pressures for the 40F atrial ranged from 4.33? 1.96 mm Hg to mm Hg and showed much variability with the heart in the circumflex position. The IVC pressures for the other methods were not changed significantly by placing the heart in the circumflex position. In 1. summary: Double caval s provide maximum venous (decompression, especially of the superior venia cava. Decompression of

7 65 Bennett et al: Flow Differences among Venous Cannulas the right side of the heart is passive and in the heart being perfused requires venting of the right atrium by release of tapes or atrial suction. 2. Double caval s provide good and consistent venous and atrial decompression. The ventricle is decompressed well except when the perfused heart is in the circumflex position. 3. A large single (e.g., USCI 40F) can decompress both the venous system and right side of the heart, although venous and atrial drainage are much less efficient when the heart is in the circumflex position. 4. The Sarns 51F cavoatrial decompressed the venous system as efficiently as the double caval s. In fact, decompression of the atrium and ventricle were consistently much better with the cavoatrial than with any of the other methods. 5. Efficient venous and myocardial decompression using the 51F cavoatrial requires the atrial drainage ports to be positioned in the upper middle section of the atrium. Clinically, the position is correct when both the single and double marking bands on the are outside the atrium. References 1. Miller DW, Ivey TD, Bailey WW, et al: The practice of coronary artery bypass surgery in J Thorac Cardiovasc Surg 81:423, Arom KV, Ellestad C, Grover FL, Trinkle JK: Objective evaluation of the efficacy of various venous s. J Thorac Cardiovasc Surg 81:464, 1981