Clinical Evaluation of Priming Solutions for Pump Oxygenator Perfusion

.. Clinical Evaluation of Priming Solutions for Pump Oxygenator Perfusion William H. Lee, Jr., M.D., Joseph W. Rubin, M.D., and Mary P. Huggins, B.S. ABSTRACT Various hemodilution agents are now used routinely to prime heart-lung machines for cardiac operations. Hemodilution has resulted in considerable conservation of blood as well as diminution of plasma and corpuscle damage by decreasing the concentration of these elements in blood during extracorporeal circulation. Controversy has existed regarding the relative efficacy of various hemodilution solutions. This study covers 68 patients, divided into three groups, for whom hemodilution was done as follows: (1) the pump was primed with a 5% dextrose solution containing no colloid; (2) Ringer s lactate solution containing approximately 1 % low-molecular-weight dextran was used; and (3) a new plasma expander, hydroxyethyl starch, was used as the colloid component of an electrolyte solution. Evaluations and comparisons were carried out for flow rates, blood pressure, urine volume, hematocrit, BUN, blood loss, clotting factors, and the patient s clinical course with regard to pulmonary and neurological complications. We conclude that a colloid is beneficial, especially with longer perfusions. D ilution of the blood prime for heart-lung machines used in extracorporeal circulation was reported in 1961 by Zuhdi and colleagues [ 131, who employed 5% dextrose in water (D5W) and also in 1961 by Long and co-workers [l 11, who advocated a solution of dextran 4. Zuhdi s group was concerned with the use of hemodilution for conservation of blood and Long s with improving the flow characteristics of blood by enhancement of fluidity. Since 1961 the concept of diluting blood for priming of heartlung machines has gained widespread popularity. The advantages of hemodilution in cardiac surgical operations include conservation of freshly drawn blood, thus markedly reducing the logistical difficulties of blood banking; improved patient safety by eliminating the need for mixing multiple units of cross-matched donor blood in the oxygenator circuit; and diminution ofboth plasma protein and corpuscle trauma by the relative decrease in the concentration of these elements present in the perfusion fluid during extracorporeal circulation. The potential hazards of interfacial protein denaturation in oxygenator devices and the reduc- From the Division of Thoracic and Cardiovascular Surgery, Medical University of South Carolina, Charleston, S.C. The authors would like to express their grateful appreciation to Dr. C. Boyd Loadholt for his helpful and meticulous statistical analysis of the data in this report. Presented at the International College of Surgeons, Fifth Asian Federation Congress, Bombay, India, Nov. 9, 1973. Accepted for publication Nov. 4, 1974. Address reprint requests to Dr. Lee, Division of Thoracic and Cardiovascular Surgery, Medical University of South Carolina, 8 Barre St., Charleston, S.C. 2941. VOL. 19, NO. 5, MAY, 1975 529

LEE, RUBIN, AND HUGGINS tion of these hazards by lowering protein concentration in the bulk phase during perfusion have recently been reviewed in some detail [8]. Controversy has existed regarding the relative merits and efficacy of various hemodiluting agents. The main categories of solutions used for hemodilution may be grouped into noncolloid priming solutions, such as D5W, and colloid priming solutions consisting of a balanced electrolyte solution rendered isooncotic by the addition of appropriate amounts of albumin, dextran, or most recently a new plasma expander, hydroxyethyl starch. DsW offers the advantages of ready availability, practically total lack of side-effects or toxicity, and economy. The colloidal perfusion fluids, however, have the added theoretical advantage of providing a more physiological oncotic pressure, thus avoiding the potential for leakage of water and crystalloid into the interstitial tissues, especially the pulmonary parenchyma during or immediately following perfusion. The present study was designed to compare the relative efficacy and safety of three hemodilution per fu sion flu ids. Three groups of patients underwent cardiac operations utilizing extracorporeal circulation. In the first group the pump was primed with DsW (23 patients). In the second group Ringer s lactate solution containing approximately 1% dextran 4 in normal saline was used to prime the oxygenator circuit (22 patients). In the third group a similar electrolyte solution containing approximately 1% hydroxyethyl starch was used for hemodilution (23 patients). Materials and Methods The patient popuiation for these investigations consisted of 68 persons consecutively admitted to the Medical University Hospital for electively scheduled cardiac surgical procedures. Most were adults. Ages ranged from 1 to 67 years and were similarly distributed in the three groups. Approximately 8% of the operations were valve replacements or myocardial revascularization procedures. Around 2% of the operative procedures were for the correction of congenital heart disease. The types of operations were similarly distributed throughout the three groups. All perfusions were carried out using the Bentley disposable oxygenator and heat exchanger and a Pemco or an Olsen roller pump with Pall or Swank blood filters on either the suction or arterial lines of the circuit. Clinical and laboratory observations and comparisons were carried out for the following factors: perfusion flow rates, arterial and central venous blood pressure, urine volume, hematocrit, BUN, blood loss and replacement, fibrinogen level, thrombin time, clotting time, prothrombin time, plasma hemoglobin following perfusion, length of perfusion, platelet count, blood gases, and the patient s clinical course, especially with regard to pulmonary and central nervous system complications. The extent of hemodilution varied somewhat in accordance with the body weight of the patient. However, the ratio of volume of prime to the patient s calculated blood volume generally ranged from 5 to 1% with a mean of 64%. (Thus, if this ratio was 176, the resulting mixture in the patient and oxygenator circuit represented a 1 : 1 dilution of blood by priming solution, or an actual 5% hemodilution.) 53 THE ANNALS OF THORACIC SURGERY

Prime fw Pump Oxygenators 5 4 T &X GLUCOSE Group HES omvp E3 DEXTRAN Group 1 3 t V 2 FIG. 1, Pumpflow rates during bypass. (HES = hydroxyethyl starch.) 1 - At Start of Bypass 2-4 Min Later 1 Hour Later Results Figure 1 displays the perfusion flow rates during extracorporeal circulation for the three groups of patients at the start of bypass, approximately 3 minutes later, and approximately one hour later. The patients perfused with DSW exhibited moderate but significantly higher flow rates for the first 3 minutes of the perfusion. However, after one hour the differences in flow rates between the three groups were not of statistical significance. There were no significant differences in flow rates between the hydroxyethyl starch group and the dextran group. All flow rates were the maximum that could be obtained during each perfusion and generally ranged between 75 and 1 ml per kilogram of body weight. Despite the higher flow rates in the noncolloid perfusion group, arterial blood pressure during bypass was significantly lower in the dextrose group than in the colloid groups for most of the perfusion (Fig. 2). The patients perfused with hydroxyethyl starch priming fluid maintained higher mean arterial blood pressures than the other two groups, and this difference was statistically significant during the early and late phases of the perfusion. In order to examine and FIG. 2. Mean arterial blood pressure during bypass. (HES = hydroxyethyl starch.) 1 9 8 7 6 E 5 4 3 2 1 T Y Start a Bypas 2-4 1 nuta OnrHour Lator VOL. 19. NO. 5, MAY, 1975 531

LEE, RUBIN, AND HUGGINS.3 T mgwcose Gnup oms Group rn DEXTRAN Qrmp.25 @.*.OIS.OlO.oos FIG. 3. starch.).ooo At start of Bypau 2-4 Minuhr Lahr OnaHour Lotv Mean arterial blood pressure per unit flow rate during bypass. (HES = hydroxyethyl compare the relationship between mean arterial blood pressure and blood flow rate, calculations were made for mean arterial blood pressure per unit of flow rate increment during the bypass. These results are displayed in Figure 3 and indicate that both of the colloid perfusion fluids resulted in higher blood pressures per unit of flow rate than did the dextrose prime. In an attempt to assess quantitatively a potential for pulmonary interstitial edema following perfusion, the two lowest values for arterial blood POP (Fig. 4A) during the first 18 hours postoperatively were averaged for each group and compared. During this period, patients having an arterial Po2 less than 8 mm Hg were ventilated on a volume respirator with 5% oxygen. The higher arterial Po2 16 14c 12C I" 8 E 6 4 2 c A 6 FIG. 4. (A) Lowest Po, during the first 18 hours Postoperatively. (B) Urine volume during thefirst 24 hours postoperatively. (HES = hydroxyethyl starch.) 532 THE ANNALS OF THORACIC SURGERY

Prime for Pump Oxygenators I 5 25 FIG. 5. starch.) Plasm hemoglobin immediately postoperatively and pump time. (HES = hydroxyethyl values noted in the hydroxyethyl starch and dextran prime groups represent a statistically significant difference when compared with the group primed with D,W. There was no renal impairment in any of the groups as assessed by urine volume (Fig. 4B) and BUN. Clotting times, thrombin times, fibrinogen assay, and prothrombin activity did not vary significantly between the three groups of patients. There was a modest reduction in fibrinogen in all three groups and a mild diminution in prothrombin activity. The incidence of postoperative bleeding requiring transfusion or reoperation was 4% in the dextran-primed group, 13% in the hydroxyethyl starch group, and 17% in the dextrose group. All bleeding problems were controlled by either drug therapy or reoperation or both. The differences in the incidence of postoperative bleeding in the three groups did not reach statistical significance. Similarly, the occurrence of neurological abnormality or behavioral aberration during the early postoperative period was 9% in the dextran-primed group, 13% in the hydroxyethyl starch group, and 22% in the dextrose group. These differences also failed to reach statistical significance. Most of the central nervous system complications manifested as transient delirium, e.g., psychotic behavior, confusion, hallucination, disorientation, or hostility. There was a close correlation between the lerigth of perfusion and the degree of hemolysis regardless of the priming agent used (Fig. 5). Although both perfusion time and plasma hemoglobin levels were higher in the D5W prime group, the differences were not of statistical significance. All blood transfused during operation and for the first three days after operation was type-specific, homologous blood anticoagulated with citratephosphate-dextrose solution and used within five days of procurement. No significant difference in the total blood transfused for the first four days (Fig. 6) could be detected between the three groups of patients. Similarly, there was no significant variance in the hematocrit. Somewhat greater volumes of blood were VOL. 19, NO. 5, MAY, 1975 533

LEE, RUBIN, AND HUGGINS 25 3 15 8 rc 2 lo 1 so FIG. 6. (A) Hematocrit on the day of operation and for two days postoperatively. (B) Total blood given ad operation and for the next three days. (HES = hydroxyethyl starch.) used at operation in the dextrose-prime group (Fig. 7), but these higher volumes of transfused blood were matched in the hydroxyethyl starch and dextran groups during the next three days. The objective assessment of pulmonary complications among the three groups of patients was quite difficult, as almost all of the patients manifested some degree of atelectasis, pulmonary interstitial edema, pneumonitis, or a combination of these. However, careful review of the clinical course and roentgenograms, which were made at least daily, did not reveal any apparently significant differences among the three groups. Approximately one-fourth of the patients experienced pulmonary complications judged to be moderately severe. An additional 6 1 % of the patients experienced mild to moderate pulmonary complications. All were managed successfully by appropriate drug or respirator therapy or both. Comment The three priming solutions used in these investigations seem to represent safe and effective hemodiluting agents to conserve blood, minimize blood mis- 2 FIG. 7. Blood given at operation and for the next three postoperative days. (HES = hydroxyethyl starch.) 15 : 1 5 During Nrrt 3 Dg 534 THE ANNALS OF THORACIC SURGERY

Prime for Pump Oxygenators match reactions, and decrease the concentration of plasma proteins and blood corpuscular elements during perfusion. The main theoretical advantage of the colloid over the noncolloid prime would be the maintenance of a more nearly physiological oncotic pressure in the priming fluid. According to Guyton and Lindsey [4], the critical pulmonary venous and left atrial pressure for transudation of plasma into the pulmonary parenchyma is directly related to the preservation of normal plasma protein concentrations. In their experiments, fluid transudation into the lungs occurred at a pressure of about 24 mm Hg when the plasma protein concentration was normal. When the concentration was reduced to about 5% of control, fluid transudation occurred when the left atrial pressure rose above a critical value of about 11 mm Hg. Extrapolation from these experiments suggests that in the patient with left atrial or pulmonary venous hypertension from any cause, maintaining normal colligative pressure in the pulmonary circulation is of critical value in preventing pulmonary interstitial edema. One might then expect that the dextrose-prime group would exhibit a higher incidence of pulmonary complications and respiratory dysfunction than the colloid-prime groups because of fluid leakage into the lungs in patients having postoperative cardiac decompensation and elevated pulmonary capillary pressure secondary to increased left atrial or left ventricular end-diastolic pressures. Such was not observed in comparing these three groups of patients. Several considerations are relevant to the similarity in incidence of pulmonary problems in the dextrose- and colloid-prime groups. In the first place, the degree of hemodilution would rarely represent more than a 5% dilution of the patient s blood volume, as previously discussed, and thus the colloid concentration in the resultant mixture would still be above the critical level at which fluid transudation occurs unless the patient experienced overt cardiac failure. In addition, the rapid mobilization of protein reserves following, for example, clinical plasmaphoresis would tend to quickly restore the plasma protein concentration toward control values. In addition, these patients were subjected to rather vigorous diuresis and were treated with intravenous albumin whenever hypoproteinemia (as determined by serum refractive index) or pulmonary interstitial edema began to emerge in the postoperative period. The data regarding perfusion flow rates and mean arterial pressure during bypass suggested that the dextrose prime reduced blood viscosity (or resistance to flow), thus providing higher flow rates at lower pressures than in the colloidprime group. The benefit or hazard of this phenomenon is as yet undetermined. In previous reports [5,91, we showed that mean blood pressures below 5 mm Hg during perfusion correlate significantly with an increased incidence of both pulmonary and cerebral dysfunction following perfusion. These investigations were conducted with a colloid-prime hemodilution agent, however, and the optimal relationship between mean blood pressure and various organ perfusion rates during bypass has not been well defined for noncolloid perfusion solutions. The data in this small group of patients suggest that individuals primed with dextrose were not adversely affected by somewhat lower blood pressures as long as perfusion flow rates were well maintained. One disadvantage of the dextrose VOL. 19, NO. 5, MAY, 1975 535

LEE, RUBIN, AND HUGGINS prime, however, may be inferred from the observation that the loss of noncolloid priming fluid over the course of a longer perfusion necessitates the addition of more blood to the perfusion circuit during the operation, thus obviating some of the advantages of hemodilution. The similar though slightly higher hematocrit level in the dextrose-prime group immediately postoperatively reflects the larger volumes of blood transfused in the perioperative period. Dextran has long been recognized as an antithrombotic agent that interferes to some degree with the coagulation process [2,6,121, especially clot propagation. Dextran was initially chosen as the colloid for hemodilution because of this property in an effort to reduce the incidence of thromboembolism following implantation of prosthetic heart valves 131. Hydroxyethyl starch is a safe and effective plasma expander which does not exhibit the antithrombotic properties of dextran [ 1,7, 1, 123. In these clinical patients no difference could be detected in the antithrombotic or anticoagulant properties of the three agents, nor was there any significant difference in the incidence of hemorrhagic complications. References 1. Arrants, J. E., Cooper, N., and Lee, W. H., Jr. The effects of a new plasma expander (hydroxyethyl starch) on intravascular clot formation. Am Surg 35:465, 1969. 2. Bloom, W. L., Harmer, D., Bryant, M., and Brewer, S. S. Coating of blood vessel surfaces and blood cells: A new concept in the prevention of intravascular thrombosis. Circulation 26:69, 1962. 3. Dalton, D. H., Jr., Arrants, J. E., Hairston, P., and Lee, W. H., Jr. The problem of thromboembolism following heart valve replacement. J SC Med Assoc 65:29, 1969. 4. Guyton, A. C., and Lindsey, A. W. Effect of elevated left atrial pressure and decreased plasma protein concentration on the development of pulmonary edema. Circ Res 7:649, 1959. 5. Lee, W. H., Jr., Brady, M. P., Rowe, J. M., and Miller, W. C., Jr. Effects of extracorporeal circulation upon behavior, personality, and brain function: 11. Hemodynamic, metabolic, and psychometric correlations. Ann Surg 173: 113, 1971. 6. Lee, W. H., Jr., Cooper, N., Russell, H. E., Ballenger, J., and Murner, E. S. Observations on the clinical use of dextran for the treatment of vascular disease. Actu Chir Scud [Suppll 387:61, 1968. 7. Lee, W. H., Jr., Cooper, N., Weidner, M. G., Jr., and Murner, E. S. Clinical evaluation of a new plasma expander, hydroxyethyl starch.j Trauma 8:381, 1968. 8. Lee, W. H., Jr., and Hairston, P. Structural effects upon blood proteins at the gas-blood interface. Fed Proc 3:1615, 1971. 9. Lee, W. H., Jr., Miller, W., Jr., Rowe, J., Hairston, P., and Brady, M. P. Effects of extracorporeal circulation on personality and cerebration. Ann Thorac Surg 7:562, 1969. 1. Lee, W. H., Jr., and Weidner, M. G. Preliminary Reports on Limited Studies of the Use of Hydroxyethyl Starch Solution in Man. In Proceedings of the Committee on Plasma Expanders, National Research Council and National Academy of Science, Cleveland, Ohio, October, 1965. 11. Long, D. M., Jr., Sanchez, L., Varco, R. L., and Lillehei, C. W. The use of low molecular weight dextran and serum albumin as plasma expanders in extracorporeal circulation. Surgery 5:12, 1961. 12. Russell, H. E., Jr., Bradham, R: R., and Lee, W. H., Jr. An evaluation of infusion therapy (including dextran) for venous thrombosis. Circulation 33:839, 1966. 13. Zuhdi, N., McCollough, B., Carey, J., Krieger, C., and Greer, A. Hypothermic perfusion for open-heart surgical procedures: Report on the use of a heart-lung machine primed with 5% dextrose in water inducing hemodilution. J Znt Coll Surg 35:319, 1961. 536 THE ANNALS OF THORACIC SURGERY