E associated with many pathophysiological changes

Transcription

1 Quantitation of Particulate Microemboli During Cardiopulmonary Bypass: Experimental and Clinical Studies JinFeng Liu, MD, ZhaoKang Su, MD, and WenXiang Ding, MD Department of Pediatric Cardiothoracic Surgery, Xin Hua Hospital, Shanghai Second Medical University, Shanghai, People's Republic of China An electronic particlesize analyzer (Coulter Counter ZM) was used to quantitate particulate microemboli 15 to 8 pm in sue during cardiopulmonary bypass. Through both laboratory studies and clinical research, we confirmed three main causes of microemboli: (1) infusion of banked blood stored for more than 3 days; (2) use of cardiotomy reservoirs; and (3) use of bubble oxygenators. The regression equation between number of particles and blood storage time was Y = X (T =.886; p <.1). The number of microemboli from cardiotomy reservoirs was 2.8 to 5.1 times that from other sources (p <.1). The number of solid particles from bubble oxygenators was 1.8 to 3.2 times that from membrane oxygenators (p <.1). Electron microscopy showed that a large number of solid particles more than 2 pm in size were formed during heartlung bypass. They obstructed the microcirculation and damaged pulmonary capillary endothelial and alveolar epithelial cells. The degree of histological damage was related to the number and size of microemboli and the duration of pulmonary microcirculatory obstruction. (Ann Thorac Surg 1992;54:119622) xtracorporeal circulation in open heart surgery is E associated with many pathophysiological changes resulting in impairment of the brain, heart, lungs, and kidneys [l]. One major complication is microembolism, which causes microcirculatory obstruction [2, 31. There are two kinds of microemboli, gaseous particles and solid particles. The occurrence of air embolism has been markedly decreased by improvements in operative technique and in filter and oxygenator design, whereas solid particles have become more important as causes of embolism [41. Methods for the quantitation of microparticles have included microscopic examination of the filter membrane after filtration, measurement of screen filtration pressure, and ultrasound. The first two can be influenced by environmental factors and cannot differentiate cell or platelet adhesion from aggregation. Ultrasound cannot detect particles smaller than 5 pm and is more sensitive to air bubbles than solid particles. The electronic particle counter (Coulter Counter ZM, Coulter Electronics, Luton, Bedfordshire, England), introduced in the 197s, is very sensitive and accurate in quantitating solid particles [591. The Coulter Counter ZM is a precision dualthreshold particlecounting and sizing instrument for matter in the overall size range of.4 to 8 pm. Using the electronic particlesize analyzer and the Cell DYN 9 blood analyzer (SequoiaTurner Corporation, Mountain View, CA), we carried out experiments to Accepted for publication July 2, Address reprint requests to Dr Liu, Department of Pediatric Cardiothoracic Surgery, Xin Hua Hospital, 1665 Kong Jiang Rd, Shanghai, People's Republic of China 292. investigate the causes of destruction of blood cells and the formation of solid microparticles. We examined the correlation between the number and the size of the particles, their effect on the pulmonary ultrastructure, and methods to prevent and reduce solid microparticles in blood during extracorporeal circulation. Material and Methods In Vitro Study The in vitro study involved a perfusion circuit and acidcitratedextrose bovine blood stored for 1 day to 3 days. It comprised four experimental groups: group a, perfusion circuit with a roller pump (Sarns 74); group b, perfusion circuit with a roller pump and a bubble oxygenator (Shanghai Medical Diagnostic Instrument Factory); group c, perfusion circuit with a roller pump and a membrane oxygenator (Scientific Instrument, FuDan University, Shanghai); and group d, perfusion circuit with a roller pump, a membrane oxygenator, and a cardiotomy suction Table 1. Clinical Data on the Two Patient Groups Bubble Oxygenator Membrane Variable (n = 1) (n = 1) Age (Y) 6.3 * Weight (kg) f 2.5 Hematocrit f.31 Bypass flow (ml. kg' * min') 1G5 1OC15 Gasiblood ratio 3:l.51:l Bypass time (min) 48 t ? by The Society of Thoracic Surgeons 34975/92/$5.

2 Ann Thorac Surg 1992;M: LIUETAL : apump bpumpbo cpumpmo dpumpmosuction 25 4 : apump bpumpbo cpumpmo dpumpmosuction ' 3 29 a M 24 U : 5 19 A I I u B 25 3 I a pump bpumpbo cpumpmo dpumpmosuction a pump bpumpbo cpumpm dpumpmosuction " Fig 2. Number of microparticles 15 to 8 /.m in diameter detected in the in vitro circuits used in groups a, b, c, and d. (BO = bubble oxygenator; MO = membrane oxygenator.) count, with particular attention paid to the number and distribution of particles 15 to 8 pm in diameter. The experiment was repeated five times for each group. Eightyfour bags of acidcitratedextrose banked blood were randomly selected for study. The blood was divided into seven groups according to storage time, with 12 bags per group. Red blood cell, white blood cell, platelet, and particle counts were done before and after blood passed the filter. Animal Study Twenty dogs weighing 1 to 16 kg (mean weight, 12.6 * 2.1 kg) were divided into three groups. In group 1 (n = 4), cardiopulmonary bypass (CPB) was undertaken with a membrane oxygenator; this was the control group. In group 2 (n = 8), standard polyvinyl chloride particles 13, 2, 45, and 75 pm in size were injected into the right atrium before bypass. During bypass, rectal temperature was maintained at 28" to 3 C and hematocrit, at.2 to.25. The perfusion rate was 1 to 15 ml kg' min'. Cardiopulmonary bypass lasted 3 hours. The gas to blood flow ratio for the oxygenator was 1:1, and the heparin dose was 2 mg/kg. No filter was used in the extracorporeal C Fig 1. Destruction of (A) red blood cells (RBC), (B) white blood cells (WBC), and ( platelets (PLT) relative to duration of perfusion on the circuit in the four in vitro experimental groups (ad). (BO = bubble oxygenator; MO = membrane oxygenator.) system. The blood prime was heparinized (2 mg/1 ml), and the hematocrit was.25 to.3. The perfusion rate was 2 Llmin for 3 hours. The gas to blood flow ratio was 3:l for the bubble oxygenator,.51:l for the membrane oxygenator, and 13:l for the cardiotomy suction system. Blood samples were taken before the perfusion circuit was started and every hour during perfusion for determination of red blood cell, white blood cell, and platelet counts, plasmafree hemoglobin levels, and microparticle apump 71. bpumpbo cpumpmo 61 dpumpmosuctlon * p Fig 3. Increase in plasma free hemoglobin (HB) relative to duration of perfusion on the circuit in the in vitro experimental groups (ad). (BO = bubble oxygenator; MO = membrane oxygenator.)

3 1198 LIUETAL Ann Thorac Surg 1992;54: Table 2. Comparison of Four Variables Between Fresh and Banked Blood Fresh Days of Storage of Banked Blood Heparinized Variable Blood RBC 3.9 f f f f f f r_.5 (X L) WBC 7.2 r_ f f.8 3.5? & f f.4 (X 17L) Plt 2 f 3 14 f 3 7 f 2. 3 f 1. <1 (1 (1 <1 (X loy/l) Particles 3.9 f.7 7.3? f f f f f (X 1VL) Plt = platelets; RBC = red blood cells; WBC = white blood cells. circuit. Blood samples for the measurement of particles 15 to 8 pm in size were collected from the femoral artery and the venous line before bypass and at 3, 6, 9, 12, and 18 minutes after bypass. Biopsy samples from the middle of the right lung were taken before and after bypass for electron microscopic examination. In group 3 (n = 8), standard 13, 2, 45, and 75pm particles were injected into the right atrium after the chest was opened. Arterial and venous blood samples were taken before and at 3 and 6 minutes after injection for particle measurement and electron microscopic examination. Clinical Study Twenty patients with noncyanotic congenital heart diseases who were undergoing an open heart operation were selected for the study. They were divided into two equal groups: a bubble oxygenator group and a membrane oxygenator group (Table 1). In the extracorporeal circuit system, blood collected from the cardiotomy suction reservoir passed through a mesh with 55pm pores (Shanghai Xin Hua Hospital Pediatric Cardiothoracic Surgery Laboratories). Heparin, 2 mg/kg, was given by intracardiac injection, and 1 to 2 mg/1 ml was added to the prime. The activated clotting time was prolonged to greater than 6 seconds during CPB. Before bypass, blood samples for particle measurement were taken from the patient's arterial blood and the prime. Blood samples also were obtained from the extracorporeal venous line, the outlet of the oxygenator, the arterial line distal to the pump, and the cardiotomy reservoir 5 minutes and 3 minutes after the beginning of bypass and immediately after the termination of bypass and were examined for the presence of 15 to 8pm particles and red blood cell, white blood cell, and platelet counts, hematocrit, and plasma free hemoglobin level. Several kinds of filters were compared: (1) blood transfusion filters with a pore size of 4 pm (Shanghai Xin Hua Hospital Pediatric Cardiothoracic Surgery Laboratories), 55 pm (Pall Biomedical Products), and 17 pm (Shanghai Medical Diagnostic Instrument Factory); (2) prime filter, pore size of 5 pm (Shiley Inc); (3) arterial line filter, pore size of 15 pm (Shanghai Xin Hua Hospital Pediatric Cardiothoracic Surgery Laboratories); and (4) cardiotomy reservoir filter, pore size of 55 pm (Shanghai Xin Hua Hospital Pediatric Cardiothoracic Surgery Laboratories). Blood particles were counted before and after filtration. Filter meshes and filter membrane surfaces were examined by light microscopy. Results In Vitro Study The in vitro experiments demonstrated destruction of blood elements by the perfusion circuit. The order of the severity of destruction by different parts of the circuit was as follows: cardiotomy reservoir, bubble oxygenator, membrane oxygenator, and roller pump (Fig 1). There was no significant difference between the last two components of the circuit (p >.5). The degree of destruction, which was most significant in the first hour, was linearly correlated with the time of perfusion on the circuit. The number of microparticles in the blood was significantly increased (p <.5) in the first hour on the perfusion circuit in all groups, but there was no significant change thereafter (p >.5) (Fig 2). The increasing level of plasma free hemoglobin was most closely related to the time on the perfusion circuit, especially in the circuit containing a cardiotomy reservoir (group d) where the degree of increase was three to eight times higher than in the other components (p <.1) (Fig 3). The studies showed that stored acidcitratedextrose blood contains many 15 to 8pm particles. The number 2 6 L4 U r=.886 P < Y= Xt blood storage time (days) Fig 4. Relationship between blood storage time and number of microparticles 15 to 8 pm in diameter.

4 Ann Thorac Surg 1992;54: LIU ETAL 1199 Table 3. Changes in Pulmonary Ultrastructure in Animal Experiments" Swelling, Rupture, Discontinuity of Lysosomal Granules and Denaturation of Edema and Rupture Collagen and of Plts and WBCs in Types I, I1 Epithelial Groupb of Endothelial Cells Elastin Fibrils Interstitial Tissue Cells 1 (n = 4) 2' (n = 8) ' (n = 8) f t f 75 a The degrees of danger were as follows: = no damage; f = doubtful damage; = definite damage; = obvious damage; = severe damage. Group 1 had a membrane oxygenator and served as the control group; group 2, a membrane oxygenator and injection of standardized particles; group 3, injection of standardsized particles. ' Minuses indicate the size of the particles injected in micrometers. Flts = platelets; WBC = white blood cells. Fig 5. Changes in pulmonay ultrastructure in animal experiments: (A) edema and rupture of endothelial cells (arrow); (B) discontinuity of collagen and elastin fibrils (arrow); (C) lysosomal granules of platelets and white blood cells in interstitial tissue (arrow); and (D) swelling, rupture, and denaturation of types I and II epithelial cells (arrow). (A, X36; B and C, ~5; D, ~6.) t A C B D

5 12 LIUETAL Ann Thorac Surg 1992;54: Fig 6. Number of microparticles from different components of the cardiopulmonay bypass circuit \ a Venous line Roller pump rl Oxygenator x 2 u1.#i c, 15 & Cardiotomy reservoir U Ism of particles in blood stored for 5 days was almost 11 times higher than the number in fresh blood (Table 2). The regression equation between particulate number and blood storage time was Y = X (r =.886; p <.1) (Fig 4). Animal Study The animal experiments showed that pathological changes in pulmonary ultrastructure after extracorporeal circulation were linearly correlated with the number of microparticles produced during bypass. The size of the microparticles was also important; those greater than 2 pm in size caused damage to the pulmonary ultrastructure. The bigger the particles, the more severe the damage (Table 3; Fig 5). Clinical Study These studies demonstrated that the tubing system of the extracorporeal circuit contained many exogenous microparticles. After simple perfusion of the extracorporeal circuit primed with normal saline solution for 1 minutes, the 15 to 8pm particle count reached 199? 69/mL, 14 times higher than before perfusion ( p <.1). The main source of microparticles during extracorporeal circulation was the cardiotomy reservoir, which produced 2.8 to 5.1 times the number from other sources (p <.1) (Fig 6). Microparticles produced by the bubble oxygenator were 1.8 to 3.2 times more numerous than the number from the membrane oxygenator ( p <.1) (Fig 7). With the prolongation of extracorporeal circulation, the difference became even more significant. We found that the prime filter (pore size, 5 pm) could remove 87% of the particles in the prime; the cardiotomy reservoir filter (pore size, 55 pm) removed 57% of the particles resulting from suction; and the blood transfusion filter with a pore size of 55 pm removed 59% of the particles in banked blood. Filters with a pore size larger than 15 pm were ineffective in removing microparticles (Fig 8). Comment The sources of solid particles in blood during extracorporeal circulation can be classified as intrinsic and extrinsic. After simple bypass of the extracorporeal circuit primed with normal saline solution for 1 minutes, the number of particles 15 to 8 pm in size reached 199 C 69/mL, 14 times higher than the count before bypass. The application of a Dacron filter (pore size, 5 pm) can remove 87.4% of these particles. Under a light microscope, many foreign microparticles can be observed on the surface.of the filter mesh. Another important source of extrinsic particles is silicone oil. Our animal study demonstrated silicone oil emboli in the heart, kidneys, and lungs after CPB with a bubble oxygenator for 2 hours (Fig 9). The roller pump can cause a large number of particles to break off from the compressed silicone rubber tube, an event that occurs most frequently when a new tube is used [lo]. The particles 1 to 25 pm in diameter reached their highest number, 1,5O/mL, 2 minutes after bypass and decreased markedly after 3 to 4 hours [lo]. Intrinsic particles are mainly produced by damage to f 1 U 3 8 i6 u 4 d = 2 bypass time Fig 7. Comparison of number of microparticles produced with a membrane oxygenator (MO) versus a bubble oxygenator (8) in the cardiopulmonary bypass circuit. lh) En

6 Ann Thorac Surg 1992; 54: LIUETAL 121 before filter after filter type of filter (pm) Fig 8. Effectiveness of filters of different pore sizes in reducing microparticle count. blood elements, denaturation of protein, and aggregation of platelets. We found the major sources of bloodderived particles during extracorporeal perfusion to be (in descending order) banked blood, cardiotomy reservoir, and bubble oxygenator. In pediatric practice, to prevent excessive hemodilution, blood must often be added to the prime for extracorporeal circulation. We have shown that stored acidcitrate dextrose blood in clinical use contains large quantities of 15 to 8pm particles, which is correlated with storage time (Y =.886; p <.1). Of these particles, 94.6% are less than 5 pm in size. Red blood cells, white blood cells, and platelets, especially the last, decrease in number with prolongation of storage time [ll]. The main source of microparticles during extracorporeal circulation is the cardiotomy reservoir [12]. Our study indicates that the number of blood microparticles produced by this source is related to the platelet count. The blood microparticles increased up to 5.1 times after passage through the cardiotomy reservoir in the early phase of bypass because of the relatively high platelet count. Our study also showed that the number of particles 15 to 8 pm in size produced by the bubble oxygenator is 1.8 to 3.2 times greater than the number produced with the membrane oxygenator [ Microparticles produced during extracorporeal circulation can cause embolism in tissues and organs, especially the lungs. The diameter of pulmonary capillaries and small arteries in children is much smaller than in adults [16]. The diameter of a smaller pulmonary artery is 1 to 12 pm and that of a branch, 2 to 25 pm. The capillaries are about 1 to 12 pm in diameter, which allows passage of red blood cells (7 pm). After CPB, pulmonary microcirculation might be obstructed by microparticles and degraded red blood cells, white blood cells, and platelets. We also found obvious changes in the pulmonary ultrastructure with obstruction by polyvinyl chloride particles. The basic pathological process is widespread degeneration of cellular components in the interstitial spaces characterized by edema, rupture of capillary endothelial cells, and denaturation of types I and I1 epithelial cells. Lysosoma1 rupture and granules of white blood cells and platelets can be observed in the interstitial spaces. Also, 4 cell and protein exudation is present in alveoli. The degree and the range of histological damage are related not only to the quantity and size of blood microparticles but also to the duration of obstruction of the microcirculation. There is still controversy regarding the harmful effects of particles smaller than 35 to 4 pm [7]. Microemboli of this size can cause detectable organ damage only when they obstruct a considerable number of small vessels. To prevent and to reduce tissue damage caused by blood microparticles during CPB, we recommend the following: (1) Avoid or reduce the use of stored blood during CPB. This not only relieves demands on the blood supply but, more importantly, can reduce the complications from transfusion. Currently, we usually do not use banked blood in open heart procedures with CPB in patients weighing more than 1 kg. For patients with a low body weight and hematocrit, fresh blood is used whenever possible. (2) Use a membrane oxygenator to reduce both damage to blood elements and embolism, as A B Fig 9. Silicone oil emboli in lungs and heart after cardiopulmonay bypass with a bubble oxygenator for 2 hours: (A) Silicone oil emboli alone (arrow) and (B) Silicone oil emboli (arrow) in the heart. (Both, X36 before 11% reduction.)

7 122 LIUETAL EMBOLI DURING CI'B Ann Thorac Surg 1992;54:119&22 this type of oxygenator avoids direct contact of air with blood. In a previous study, we [17] showed that the time of mechanical ventilation was 1 hours shorter in patients with a membrane oxygenator than in those who had a bubble oxygenator. Respiratory secretions and pulmonary complications after operation were also markedly reduced in patients who had a membrane oxygenator ( p <.1). (3) Reduce damage from the cardiotomy system by controlling negative suction pressure. The use of deep hypothermic circulatory arrest in infants may also be of help. (4) Use filters with different pore sizes to remove as many microparticles as possible, as the production of microparticles during CPB cannot be completely eliminated at present. The arterial line filter should provide a large surface area of 65 to 8 cm * Ll min', and its pore size should be 4 to 8 pm [7]. Filters with a pore size greater than 15 pm are ineffective in removing microparticles. We gratefully acknowledge the helpful suggestions and support of Dr Richard A. Jonas, The Children's Hospital, Boston, MA. References 1. Ratliff NB, Young WG Jr, Hackel BB, et al. Pulmonary injury secondary to extracorporeal circulation. J Thorac Cardiovasc Surg 1973;65: Allardyce DB, Yoshida SH, Ashmore PG. The importance of microembolism in the pathogenesis of organ dysfunction caused by prolonged use of the pump oxygenator. J Thorac Cardiovasc Surg 1966;52: Ashmore PG, Cvitek V, Ambrase P. The incidence and effects of particulate aggregation and microembolism in pumpoxygenator systems. J Thorac Cardiovasc Surg 1968;55: Solis RT, Noon GP, Beall AC Jr, DeBakey ME. Particulate microembolism during cardiac operation. Ann Thorac Surg 1974;17: Loop FD, Szabo J, Rowlinson RD, Urbanek K. Events related to microembolism during extracorporeal perfusion in man: effectiveness of inline filtration recorded by ultrasound. Ann Thorac Surg 1976;21: Clark RE, Deitz DR, Miller JG. Continuous detection of microemboli during cardiopulmonary bypass in animals and man. Circulation 1976;54(Suppl 3): Edmunds LH Jr, Williams W. Microemboli and the use of filters during cardiopulmonary bypass. In: Utley JR, ed. Pathophysiology and techniques of cardiopulmonary bypass; vol 2. Baltimore: Williams & Wilkins, 1983:lOl Solis RT, Wright CB, Gibbs MB, et al. Quantitative studies of microaggregate formation in vitro and in vivo. Chest 1974; 65(Suppl 4): Solis RT, Kennedy PS, Beall AC Jr, et al. Cardiopulmonary bypass microembolization and platelet aggregation. Circulation 1975;52: Stoney WS, Alford WC Jr, Burrus GR, Glassford DM Jr, Thomas CS Jr. Air embolism and other accidents using pump oxygenators. Ann Thorac Surg 198;29: Liu JF, Su ZK, Ding WX. Particle in banked blood and filtration by micropore filters. Chin J Hematol 199;11: de long JCF, ten Duis HJ, Smit Sibinga CT, Wildevuur CRH. Hematologic aspects of cardiotomy suction in cardiac operations. J Thorac Cardiovasc Surg 198;79: Clark RE, Beauchamp RA, Magrath RA, et al. Comparison of bubble and membrane oxygenators in short and long perfusions. J Thorac Cardiovasc Surg 1979;78: Connell RS, Page US, Bartley TD, Bigelow JC, Webb ML. The effect on pulmonary ultrastructure of Dacronwool filtration during cardiopulmonary bypass. Ann Thorac Surg 1973;15: Bisio JM, Connell RS, Harrison MM. The formation and effect of stored platelet concentrate microemboli on pulmonary ultrastructure. Surg Gynecol Obstet 1982; Tian Niu. Microcirculation. Beijing: Chinese Science Company Publishing, 198: Cao DF, Ding WX, Su ZK. Using FuDan AL2 hollowfibre membrane oxygenator in undertaking infant cardiac surgery with deep hypothermia and circulatory arrest: a report of 6 cases. J Clin Pediatr 1988;6:3623.