Cardiac surgery is associated with a high incidence of

Retrograde Autologous Priming of the Cardiopulmonary Bypass Circuit Reduces Blood Transfusion After Coronary Artery Surgery Subramaniam Balachandran, FRCA, Michael H. Cross, FRCA, Sivagnanam Karthikeyan, FRCA, Anilkumar Mulpur, FRCS, Stephen D. Hansbro, and Peter Hobson, BS The Yorkshire Heart Centre, Leeds General Infirmary, Leeds, United Kingdom Background. Hemodilution occurring with cardiopulmonary bypass imposes a risk for blood transfusion. Autologous priming of the cardiopulmonary bypass circuit at the initiation of bypass partially replaces the priming solution with autologous blood. We examined the efficacy of autologous priming of the circuit in reducing blood transfusion. Methods. One hundred and four patients were entered into a prospective, randomized, controlled study. Initiation of cardiopulmonary bypass was with or without autologous priming. Results. With autologous priming, a mean volume of 808.8 159.3 ml of priming solution was replaced with autologous blood. This allowed a higher hematocrit value on admission to the intensive care unit and at discharge from hospital. In all, 49% of the control group required a blood transfusion compared with 17% from the autologous priming group (p 0.0007). The mean volume of blood transfused was 277.6 363.8 ml in the control group compared with 70.1 173.5 ml in the autologous priming group (p 0.0005). Conclusions. Retrograde autologous priming of the bypass circuit reduces homologous blood transfusion owing to the reduction in bypass circuit priming volume. (Ann Thorac Surg 2002;73:1912 8) 2002 by The Society of Thoracic Surgeons Cardiac surgery is associated with a high incidence of homologous blood transfusion [1]. Concern about the side effects (immunologic reactions, viral transmission) of homologous blood transfusion have led to the development of many methods to conserve blood in cardiac surgery. These methods include preoperative autologous blood donation (PABD) [2], acute normovolemic hemodilution (ANH) [3], cell salvage [4], and the use of pharmacologic agents [5]. Hemodilution, which results from direct mixing of the patient s blood with the asanguineous cardiopulmonary bypass (CPB) circuit prime, may be an important risk factor for homologous red cell transfusion [6]. Displacing some of the circuit prime at the initiation of CPB with the patient s own circulating blood in both an antegrade direction through the venous cannula and a retrograde direction through the arterial cannula has become known as retrograde autologous priming (RAP) [7]. RAP of the CPB circuit reduces the priming volume of the CPB circuit and hence limits hemodilution at this time. This reduced hemodilution allows a higher HCT value to be maintained throughout CPB, which may be beneficial with regard to adverse outcomes [8]. After CPB the higher HCT may mean that the patient is less likely to reach a transfusion threshold and receive a homologous blood transfusion. Accepted for publication Feb 7, 2002. Address reprint requests to Dr Cross, Department of Anaesthesia, Leeds General Infirmary, Leeds LS1 3EX, UK; e-mail: michael.cross@ leedsth.nhs.uk. This study was designed to determine whether RAP of the CPB circuit performed primarily at the onset of CPB in an antegrade direction is useful in reducing homologous blood transfusion. Material and Methods Patient Population A prospective, randomized, controlled study was performed in patients presented for primary coronary artery bypass grafting (CABG). The sample size was 100 based on demonstrating a reduction in transfusion from 60% (local audit data 1999) to 40%, with a confidence level of 0.95 and a power of 0.8. After local research ethics committee approval (study reference no. 99/151; December 15, 1999) and informed consent, 104 patients were prospectively randomly assigned to control (no prime displacement, n 53) and RAP (prime displacement with the autologous blood at the initiation of bypass, n 51) groups. Exclusion criteria were patients aged less than 18 or greater than 80 years, a preoperative hematocrit (HCT) value less than 30%, left ventricular ejection fraction less than 30%, weight less than 50 kg or more than 115 kg, emergency CABG surgery, patients with neurologic deficits or a history of stroke, and patients who were receiving preoperative heparin or warfarin therapy. Both medical and nursing staff in the intensive care unit and the postoperative wards were blinded from the priming technique for purposes of the study. The primary end- 2002 by The Society of Thoracic Surgeons 0003-4975/02/$22.00 Published by Elsevier Science Inc PII S0003-4975(02)03513-0

Ann Thorac Surg BALACHANDRAN ET AL 2002;73:1912 8 RETROGRADE AUTOLOGOUS PRIMING OF THE CPB CIRCUIT 1913 point of the study was to compare RAP of the CPB circuit with the conventional priming in reducing the percentage of patients receiving homologous blood transfusion. Secondary endpoints included the volume of homologous blood transfused per patient, the trend in hematocrit during the perioperative period, and a number of safety issues. Anesthetic and Surgical Management Premedication and induction and maintenance of anesthesia were not restricted by the study. Blood samples were taken before induction of anesthesia for a baseline troponin T level and the initial HCT value. Acute normovolemic hemodilution (ANH) was performed before heparinisation aiming for a HCT during CPB of 20%. The calculated volume of blood to be removed during this process varied depending upon whether the patient was in the RAP or the control group. Mannitol (10%), at a dose of 0.5 g per kg body weight was given to patients in both groups and was completed before the commencement of bypass. Tranexamic acid 1 g was given prebypass, in the CPB prime and postbypass to all patients (3 g total). Patients were weaned from CPB after actively warming to 37 C and heparin was reversed with protamine. After termination of CPB and removal of cannulas, the remaining blood in the CPB circuit was collected and later transfused back to the patient. Intraoperative cell salvage was not used. Management of CPB The CPB circuit consisted of a hollow fiber membrane oxygenator (D905 Avant oxygenator; Sorin Biomedica, Mirandola, Italy) with an integral isolated cardiotomy reservoir. Roller pumps (Cobe), arterial line filter (Pall), and a set of polyvinylchloride tubing completed the circuit (Fig 1). The CPB circuit was initially primed with 1 L of Hartmann s solution, 1 L of gelofusine solution, and 5,000 units of porcine heparin. After exclusion of the prebypass filter, the final volume remaining in the CPB circuit was 1,700 ml. All patients were cooled to 32 C. The pump flow was maintained at between 2.0 and 2.4 L min 1 m 2. The MAP was maintained between 50 and 70 mm Hg by administration of phenylepherine or phentolamine as necessary. The HCT value was maintained above 18%, adding autologous blood initially or homologous blood as necessary. Technique of RAP To implement the RAP process, a quarter-inch recirculation line was diverted off the arterial line and connected to an empty 1-L prime bag. Positions A, B, and C (Fig 1A) denote the sites where clamps are applied or removed at various stages during the RAP process. Displacement of Prime From Arterial Line Prior to aortic cannulation clamps are applied at positions A, B, and C. When the aortic cannula is connected to the arterial line, the clamp at A is removed and the aortic pressure confirmed. The clamp at position C is slowly and partially removed to allow blood to flow Fig 1. (A) Diagrammatic representation of the cardiopulmonary bypass (CPB) circuit used. A, B, and C show positions where clamps are removed or applied at various stages during the retrograde autologous priming process. Displacement of priming solution from the arterial line: initially clamps are positioned at A, B, and C. When the aortic cannula is connected to the arterial line, the clamp at A is removed and the aortic pressure confirmed. The clamp at position C is slowly and partially removed to allow blood to flow retrograde from the patient s aorta into the arterial line displacing approximately 100 to 150 ml of priming solution into the prime bag. Once the arterial blood has reached the clamp at position B, the line clamp is reapplied at position A, and the clamp is removed from position B. (B) Displacement of priming solution from the venous side: once the venous line has been connected to the venous cannula, the variable occlusion clamp on the venous line is slowly released allowing venous blood to drain from the patient. At the same time, the arterial pump is slowly rotated at a sufficient flow (600 to 800 ml/min) to maintain a constant level in the venous reservoir. The venous blood slowly displaces the priming solution in the reservoir, oxygenator, and arterial line filter. Once the blood has reached the origin of the quarter-inch recirculation line, the clamp is removed from position A and reapplied at position C. The venous line is then fully opened and the pump flow is increased to establish full CPB.

1914 BALACHANDRAN ET AL Ann Thorac Surg RETROGRADE AUTOLOGOUS PRIMING OF THE CPB CIRCUIT 2002;73:1912 8 Table 1. Transfusion Protocol Packed red blood cells Platelets (4 units) Fresh frozen plasma (10 15 ml/kg) Cryoprecipitate (1 unit per 10 kg) Hematocrit 24% or clinically symptomatic anemia Platelets 90,000 and blood loss 150 ml/h Blood loss 300 ml during first postoperative hour Blood loss 100 ml/h for any 3 consecutive hours Blood loss 200 ml/h during any hour after first postoperative hour (except after turning patient) Blood loss 300 ml during first postoperative hour Blood loss 100 ml/h for any 3 consecutive hours Blood loss 200 ml/h during any hour after first postoperative hour (except after turning patient) Continued bleeding after fresh frozen plasma and platelets and recorded fibrinogen 1.5 g/l Indications for surgical reexploration: chest tube output 500 ml in the first hour; 400 ml/h in the first 2 hours; 300 ml/h in the first 3 hours; sudden massive bleeding. retrograde from the patient s aorta into the arterial line displacing approximately 150 ml of priming solution into the prime bag. Once the arterial blood has reached the clamp at position B, the line clamp is reapplied at position A and the clamp then removed from position B. Displacement of Prime From Venous Side Once the venous line has been connected to the venous cannula, the variable occlusion clamp on the venous line is slowly released allowing the venous blood to drain from the patient (Fig 1B). At the same time the arterial pump is slowly rotated at a sufficient flow (600 to 800 ml/min) to maintain a constant level in the venous reservoir. Once the venous blood has displaced the priming solution in the reservoir, oxygenator, and arterial line filter and reached the recirculation line, the line clamp is removed from position A and reapplied at position C. The venous clamp is then fully opened and the pump flow increased to establish full CPB. The time taken for displacement of the prime from the venous side is less than 2 minutes and hemodynamic stability can easily be maintained during this period using a small bolus of phenylepherine (50 g). All patients were admitted to the intensive care unit (ICU) and further management was similar in both groups. All data were collected prospectively and patients were followed up until they were discharged from hospital. A strict transfusion policy was followed for all patients during their entire hospital stay (Table 1). Statistical Analysis The demographic characteristics of the patients as well as the baseline and postoperative data were summarized with descriptive statistics. Spearman correlations were used to test for the strength of linear association between variables along with the Wilcoxon-Mann-Whitney tests when appropriate to test for group differences. Statistical significance was defined as p 0.05. Results Baseline Characteristics The groups were closely matched for age, body weight, body surface area, NYHA classification, comorbid risk factors, and preoperative hematologic data (preoperative HCT 43% in both groups). There were significantly more female patients in the RAP group (13 of 51, 25.49%) compared with the control group (4 of 49, 8.16%; p 0.02). Exclusions Four patients from the control group were excluded from the study for the following reasons: in 1 patient deep hypothermic circulatory arrest was used because of a difficulty in aortic cannulation; in 2 patients aprotinin was used intraoperatively (the study protocol did not allow for the use of aprotinin), and in 1 patient there was a violation of our study transfusion protocol. RAP and Intraoperative Events The intraoperative data are summarized in Table 2. In the RAP group more autologous blood was collected (ANH) in the prebypass period compared with the control group. This difference can be explained by the different expected dilution that would occur with CPB in the two groups when the target HCT during CPB was 20% in both groups. In 1 patient the RAP process could not be completed because of the development of acute atrial fibrillation. For the purposes of statistical analysis this patient was still included in the RAP group. Although the oxygenator reservoir volume in the RAP group was significantly lower (especially at the start of CPB) it was maintained above the safe level (300 ml). The net volume Table 2. Intraoperative Data Variable Control (n 49) RAP (n 51) p Value Autologous donation (ANH) 400 312 641 339 0.0006 (ml) RAP volume removed (ml) 809 159 RAP volume returned (ml) 588 343 ORV at 30 min (ml) 1126 578 723 484 0.0001 ORV at 60 min (ml) 940 519 673 330 0.0003 Total dose of phenylepherine 4.0 4.2 3.6 2.6 NS (mg) Net fluids added on bypass 323 279 29 523 0.0053 (ml) Homologous blood added on pump 1/49 1/51 NS Values are mean standard deviation where appropriate. p 0.05 is statistically significant. RAP retrograde autologous priming; ANH acute normovolemic hemodilution; ORV oxygenator reservoir volume; NS not significant.

Ann Thorac Surg BALACHANDRAN ET AL 2002;73:1912 8 RETROGRADE AUTOLOGOUS PRIMING OF THE CPB CIRCUIT 1915 Table 3. Postoperative Data and Clinical Outcome Variable Control (n 49) RAP (n 51) of clear fluid added to the CPB circuit (after return of the RAP volume if the patient had undergone RAP) was significantly less in the RAP group when compared with the control group (29.7 523.2 ml versus 322.7 279 ml; p 0.0053). Postoperative Events and Clinical Outcomes p Value Ventilation 24 hours 2 2 NS Tracheostomy 1 1 NS Hospital stay 10 days 12 13 NS Intra-aortic balloon pump 1 0 NS Q-wave infarction 8 12 NS Troponin T 1.1 g/l 4 6 NS Troponin T Q wave 3 4 NS Atrial fibrillation 7 8 NS Wound infection or delayed 10 7 NS healing New renal changes 1 1 NS New CNS changes 0 1 NS Values are number of patients in each group. NS not significant. There was no mortality (Table 3). There was no difference with respect to inotropic or vasopressor support, duration of ventilation, or hospital stay. In 2 patients (1 each in the control and RAP groups) new renal changes (creatinine 20% above the preoperative value) developed although neither patient required hemofiltration. In 1 patient in the RAP group a mild postoperative monoparesis developed and resolved spontaneously. Hematologic Data Preoperative hematocrit, platelet count, prothrombin time, and activated partial thromboplastin time were similar between the two groups. Postoperative platelet count, prothrombin time, and activated partial thromboplastin time were similar between control and RAP groups. Figure 2 shows the trend of HCT values from the preoperative period through bypass to hospital discharge. Blood Loss and Transfusion Requirements Twenty-four-hour chest tube output was not significantly different between the two groups (control 673 316 ml versus RAP 579 286 ml). However, patients in the RAP group required significantly less homologous blood transfusion. That may be explained by the higher hematocrit on admission to the ICU in the RAP group. Most of the blood and blood products were given in the first 24 hours postoperatively. Three patients (2 patients in the control group and 1 patient in the RAP group) received a blood transfusion after this. Table 4 summarizes data on blood loss and transfusion of blood and blood products. Comment Coronary artery bypass graft surgery is the most commonly performed cardiac operation and as many as 70% of patients require a blood transfusion despite improvements in cardiac surgical techniques. The transfusion rate varies between institution [9]. RAP of the CPB circuit is not a new technique. In 1960, Panico and Neptune [10] first described a method of autologously priming the CPB circuit to reduce the requirements for homologous blood, which was used to prime the CPB circuit at that time. This technique did not prove popular until the late 1990s when numerous authors published modifications of this technique of priming of the CPB circuit [6, 7, 11, 16, 17]. The technique of RAP varies depending upon the pump configuration and the CPB circuit design but all authors describe the technique of RAP in which the majority of prime displacement occurs in a retrograde direction through the aortic cannula. We have developed a technique in which the majority of prime displacement occurs in an antegrade direction. Acute Normovolemic Hemodilution and RAP A calculated amount of whole blood was removed before heparinization to yield a minimum HCT value on bypass of 20%. In our study, patients from the RAP group donated more autologous blood than did patients from the control group (641 339 ml versus 400 312 ml; p 0.0006), which compared favorably with the findings of Rosengart and colleagues [7]. The main reason for the RAP group donating more autologous blood was the expected reduction in crystalloid prime volume on bypass. We should not expect that this difference in autologous blood donation would have an impact on the homologous blood transfusion as evidenced by Kahraman S and associates [12]. In their study they found that the reduction in homologous blood transfusion was similar between the patients who had donated either 1 or 2 U of autologous blood intraoperatively. Despite the fact Fig 2. The HCT values at various time periods for control and retrograde autologous priming (RAP) groups. Patients from the RAP group had higher HCT values on CPB, on admission to ITU, and at the time of hospital discharge. (CPB cardiopulmonary bypass; HCT hematocrit value [%]; ITU intensive therapy unit; Pre op. preoperative.) p 0.05, statistically significant.

1916 BALACHANDRAN ET AL Ann Thorac Surg RETROGRADE AUTOLOGOUS PRIMING OF THE CPB CIRCUIT 2002;73:1912 8 Table 4. Blood Loss and Transfusion Requirements Variable Control (n 49) RAP (n 51) p Value 24-hour chest tube output (ml) 673 316 579 286 0.0626 RBC (units) 1 1.34 (0 5) 0.24 0.59 (0 3) 0.0003 RBC (ml) 278 364 70 174 0.0005 FFP (units) 0.24 0.97 (0 6) 0.04 0.28 (0 2) NS Platelets (units) 0.49 1.94 (0 12) 0.08 0.56 (0 4) NS Patients transfused (RBC) 24 (49%) 9 (17.6%) 0.0007 Patients transfused (FFP, platelets) 4 (8.2%) 1 (2%) NS Values are mean standard deviation where appropriate. p 0.05 is statistically significant. RAP retrograde autologous priming; RBC red blood cells; FFP fresh frozen plasma; NS not significant. therefore that there was a significant difference between the groups with respect to ANH volume removed, it would not appear that this is the reason for a significant reduction in homologous blood transfusion in the RAP group. Transfusion Triggers The HCT value below which a clinician considers that a patient requires a blood transfusion varies widely. We chose what would be considered as a moderately restrictive transfusion threshold. Hebert and colleagues [13] in their study of critical care patients found that the 30-day mortality was similar both in the restrictive strategy group (red blood cells were given when the hemoglobin was less than 7 g/dl and hemoglobin percentage maintained between 7 and 9 g/dl) and in the liberal strategy group (red blood cells were given when hemoglobin was less than 10 g/dl and hemoglobin percentage maintained between 10 and 12 g/dl). They also found that the mortality rates were significantly lower with the restrictive transfusion strategy among patients who were less acutely ill (8.7% versus 16.1%; p 0.03). The HCT on bypass frequently falls to a low level during many cardiac surgical procedures and minimum HCT level during CPB is an independent risk factor for mortality after coronary artery surgery. In a study of 2,500 patients Fang and colleagues [14] showed that there was an increased mortality for HCT levels less than 14% in the absence of significant risk factors but the HCT levels should be increased to 18% in the presence of risk factors if increased mortality was to be avoided. Jones and associates [15] randomly assigned 300 patients to a HCT transfusion trigger on bypass of either 21% or 15%. They found an actual decrease in morbidity and mortality in the low HCT group, although this difference was not statistically significant. In our study blood was given on bypass only when the HCT fell to 17% or less and in the postoperative period when the HCT fell to 23% or less. Technique of RAP The technique that we have developed to displace the priming solution from the CPB circuit uses predominantly antegrade flow of blood through the CPB circuit (venous cannula to arterial cannula) and with the exception of the blood that drains along the arterial line (approximately 150 ml) the process occurs at and not before the initiation of CPB. This is in contrast to the technique described by Rosengart and associates [7] and Shapira and colleagues [16] in which the displacement of blood was predominantly retrograde through the arterial cannula. The choice of technique will depend primarily upon the circuit configuration used. Efficacy of RAP We have demonstrated that RAP as a blood conservation method is an efficient, safe, and inexpensive technique of blood conservation in cardiac surgery. It appears to work by reducing hemodilution on bypass. The lowest HCT on CPB was higher in the RAP group when compared with patients from the control group (23.4% 3.4% versus 22.1% 2.6%; p 0.045). The RAP group consisted of more female patients (25.5% versus 8.2%; p 0.02). This has been recognized as a risk factor for transfusion and but despite that, only 3 out of 13 female patients in the RAP group received blood transfusion whereas in the control group 3 out 4 female patients received a blood transfusion. The overall transfusion rate was significantly less in the RAP group when compared with the control group (17.6% versus 49%; p 0.0009) which compares favorably with Rosengart and colleagues [7]. The precise technique employed for RAP will vary depending upon the CPB circuit configuration. The isolated cardiotomy reservoir, which is available with the Sorin reservoir, meant that the cardiotomy suction used after heparinization was kept completely separate from the clear priming solution. This is important because it allows easy identification of the point at which the clamps are applied at position C and removed at position A as clear fluid changes abruptly to blood in the tubing. Mixing of a variable amount of cardiotomy suction with the priming solution would make this much more difficult. As soon as CPB has been fully initiated the cardiotomy suction is added to the main reservoir. Although the technique of RAP appears simple and safe for all patients it would be useful to know whether there is a subgroup of patients who are particularly likely to benefit from RAP. To identify whether this is the case we have divided the patients into three subgroups and

Ann Thorac Surg BALACHANDRAN ET AL 2002;73:1912 8 RETROGRADE AUTOLOGOUS PRIMING OF THE CPB CIRCUIT 1917 analyzed data according to age ( 65 and 65 years), weight ( 80 and 80 kg) and preoperative HCT value ( 40 and 40%). In the RAP group there was no difference in the rate of homologous blood transfusion between patients aged less than 65 years and 65 years or older ( p 0.2154) or patients with preoperative HCT less than 40% and 40% or more ( p 0.2888) but patients who weighed less than 80 kg received more blood transfusions when compared with patients weighing more than 80 kg ( p 0.0128). In the control group there was a significant difference between patients aged less than 65 versus 65 years or more ( p 0.0155), patients who weighed less than 80 kg versus 80 kg or more ( p 0.0071), and patients with preoperative HCT less than 40% versus 40% or more ( p 0.0003) when comparing the rate of homologous blood transfusion. Comparing RAP with conventional priming of the CPB circuit it would appear that RAP adds little to the blood conservation strategy when all three factors are present ( 65 years, 80 kg, and 40% HCT). The transfusion rate was 0 of 21 in the RAP group and 1 of 12 in the control group. At the opposite end of the spectrum (age 65 years, weight 80 kg, and HCT 40%) 5 of 5 patients in the control group received a transfusion compared with 1 of 3 patients in the RAP group. If the technique is to be targeted toward a particular population, the presence of any of these three risk factors means that the risk of exposure to a blood transfusion can be significantly reduced using retrograde priming of the CPB circuit. There was no reduction in the 24-hour chest tube drainage in the RAP group (579 286 ml versus 673 316 ml; p 0.0626) and our finding was similar to that of Shapira and associates [16]. The reason why RAP group patients received less blood transfusion despite similar chest tube output was a higher HCT on admission to the ICU. One would expect that a reduction in the colloid oncotic pressure as a result of hemodilution might result in positive fluid balance (weight gain) in the postoperative period [17]. As RAP reduces hemodilution the consequent preservation of colloid osmotic pressure prevents the passage of fluid from the intravascular to extravascular compartment. That means there is a theoretic benefit of reduction in weight gain but to get a significant value, we would need a large number of patients and we were unable to demonstrate this. Safety of RAP Concerns about the safety of RAP of the CPB circuit mainly revolve around the fact that there is by definition a brief period of hypovolemia during the process, and before full CPB is instituted that may lead to hypotension. We attempted to overcome this problem by administering a small dose of phenylephrine prior to initiation of the process and in most cases a dose of 0.05 to 0.1 mg was used. This was given immediately before the venous line was unclamped and it allowed hemodynamic stability during the process. Except for 1 patient (in whom developed acute atrial fibrillation) the RAP process was completed without any untoward hemodynamic disturbances. Even though most patients required a small dose of phenylepherine for the RAP process, the total dose of phenylepherine used intraoperatively was similar between control and RAP groups. As our priming displacement occurs primarily in antegrade direction, it can be safely performed in relatively small patients (low BSA) in contradiction to priming displacement that occurs mainly through the aortic cannula [7]. In the patient undergoing CABG the highest risk period for adverse myocardial ischemia event is in the immediate postoperative period [18]. Various authors suggested either new Q wave on the ECG or increase in the troponin T level to monitor perioperative myocardial infarction (MI) but either of this on its own may of a doubtful value. Moderate increase in troponin T level after cardiac surgery can occur without MI. Inselmann and associates [19] in their study found that troponin T level increased to 0.90 0.17 g/l ( p 0.005 in comparison with the preoperative value of 0.08 0.02 g/l) and concluded that moderate elevations in troponin T were normal after CABG surgery in patients without postoperative MI. In a study of 302 consecutive patients undergoing coronary surgery, Svedjeholm and coworkers [20] found that more than 25% of the new Q waves were associated that plasma troponin T value below the reference level ( 0.2 g/l). In our study, we have used both troponin T ( 1.1 g/l) and new Q-wave changes to monitor the incidence of perioperative MI. In their multivariate analysis Spiess and colleagues [21] found that hematocrit value on admission to the ICU was the most significant independent predictor of Q-wave MI. One would expect because of its efficacy RAP may potentially increase HCT value on admission to the ICU and hence MI but we have found no difference in the incidence of postoperative MI (3 of 49 control; 4 of 51 RAP). On the other hand, by virtue of moderate increase in the HCT level it decreases the rate of homologous blood transfusion and its associated complications. Conclusions RAP of the CPB circuit, using a technique where prime displacement is mainly in an antegrade direction, significantly reduces homologous blood transfusion and this reduction is mainly due to a reduction in hemodilution with an associated increase in HCT value on bypass. It may be an additional arm to the existing blood conservation strategies in adult cardiac operations. We thank Dr Manzoor Nazir from the Yorkshire Heart Centre, United Kingdom, for his help in analyzing electrocardiographic data. References 1. Stover EP, Siegel LC, Parks R, et al. Variability in transfusion practice for coronary artery bypass surgery persists despite national consensus guidelines. Anesthesiology 1998;88: 327 33. 2. Owings DV, Kruskall MS, Thurer RL, Donovan LM. Autologous blood donations prior to elective cardiac surgery: safety and effect of subsequent blood use. JAMA 1989;262: 1963 8.

1918 BALACHANDRAN ET AL Ann Thorac Surg RETROGRADE AUTOLOGOUS PRIMING OF THE CPB CIRCUIT 2002;73:1912 8 3. Gillon J, Desmond M, Thomas MTG. Acute normovolemic hemodilution. Transfus Med 1999;9:259 64. 4. Hiratzka LF, Richardson JV, Brandt B III, et al. The effect of autologous blood salvage and the cost of routine coronary revascularization. Perfusion 1986;1:239 44. 5. Scott WJ, Kessler R, Wernly JA. Blood conservation in cardiac surgery. Ann Thorac Surg 1990;50:843 51. 6. Debois WJ, Sukhram Y, Mcrey J, et al. Reduction in homologous blood transfusion using a low prime circuit. J Extracorpor Technol 1996;28:58 62. 7. Rosengart TK, Debois W, O Hara M, et al. Retrograde autologous priming for cardiopulmonary bypass. A safe and effective means of decreasing hemodilution and transfusion requirements. J Thorac Cardiovasc Surg 1998;115:426 39. 8. DeFoe GR, Ross GS, Olmstead EM, et al. Lowest hematocrit on bypass and adverse outcomes associated with coronary artery bypass grafting. Ann Thorac Surg 2001;71:769 76. 9. Goodnough LT, Johnson MFM, Toy PTCY. The variability of transfusion practice in coronary artery bypass surgery. JAMA 1991;265:86 90. 10. Panico FG, Neptune WB. A mechanism to eliminate the donor blood prime from the pump oxygenator. Surg Forum 1960;10:605 9. 11. Cromer MJ, Wolk DR. A minimal priming technique that allows for a higher circulating haemoglobin on cardiopulmonary bypass. Perfusion 1998;13:311 13. 12. Kahraman S, Altunkaya H, Celebioglu B, et al. The effect of acute normovolemic hemodilution on homologous blood requirements and total estimated red blood cell volume lost. Acta Anesth Scand 1997;41:614 17. 13. Hebert PC, Wells G, Blajchman MA, et al. A multi-center, randomized, controlled clinical trial of transfusion requirements in critical care. N Engl J Med 1999;340:409 17. 14. Fang WC, Helm RE, Krieger KH, et al. Impact of minimum hematocrit during cardiopulmonary bypass on mortality in patients undergoing coronary artery surgery. Circulation 1997;96(suppl II):194 9. 15. Jones JW, Rawitscher RE, Mclean TR, Beall AC, Thornby JI. Benefit from combining blood conservation measures in cardiac surgery. Ann Thorac Surg 1991;41:541 6. 16. Shapira OM, Aldea GS, Treanor PR, et al. Reduction of allogenic blood transfusions after open heart operations by lowering cardiopulmonary bypass prime volume. Ann Thorac Surg 1998;65:724 30. 17. Jansen PG, te Velthuis H, Bulder ER, et al. Reduction in prime volume attenuates the hyperdynamic response after cardiopulmonary bypass. Ann Thorac Surg 1995;60:544 50. 18. Mangano DT. Perioperative cardiac morbidity. Anesthesiology 1990;72:153 84. 19. Inselmann G, Kohler K, Lange V, et al. Lipid peroxidation and cardiac troponin T release during routine cardiac surgery. Cardiology 1998;89:124 9. 20. Svedjeholm R, Dahlin LG, Lundberg C, et al. Are electrocardiographic Q-wave criteria reliable for diagnosis of perioperative myocardial infarction after coronary artery surgery? Eur J Cardiothorac Surg 1998;13:655 61. 21. Spiess BD, Ley C, Body SC, et al. Hematocrit value on intensive care unit entry influences the frequency of Q-wave myocardial infarction after coronary artery bypass grafting. J Thorac Cardiovasc Surg 1998;116:460 7.