D clinical hemostatic defect produced by cardiopulmonary

Transcription

1 Safetv and TheraDeutic Effectiveness of Reinfused Shed Blood After Open Heart Surgery Trevor C. Axford, MD, Joseph A. Dearani, MD, Gina Ragno, Hollace MacGregor, BS, Manisha A. Patel, BA, C. Robert Valeri, MD, and Shukri F. Khuri, MD Departments of Surgery, BrocktoniWest Roxbury Veterans Administration Medical Center, Brigham and Women's Hospital, and Harvard Medical School; and Naval Blood Research Laboratory, Boston University School of Medicine, Boston, Massachusetts This prospective study was designed to determine whether use of nonwashed shed mediastinal blood exacerbated platelet and related hematologic dysfunctions after cardiopulmonary bypass, compared with the alternative use of autologous and homologous standard liquid preserved blood for volume support. Thirty-two patients undergoing cardiopulmonary bypass for open heart operations were randomized to receive either nonwashed shed mediastinal blood (group ; n = 6) or liquid preserved packed red blood cells (group 2; n = 6) for transfusion therapy in the management of postoperative bleeding. Patient blood samples and bleeding times were obtained preoperatively, after cardiopulmonary bypass but before transfusions, 2 and 24 hours after transfusion, and on postoperative days 2,3, and 7. Group patients received an average of 70 f 90 ml (range, 300 to,700 ml) of nonwashed shed mediastinal blood containing significantly greater (p < 0.000) amounts of fibrin degradation products and D-dimer protein. Of the hematologic, microaggregate, and plasma protein measurements performed, only the protein C level was significantly greater in group (p < 0.05) after transfusion. Patient bleeding times were not significantly different between the groups at any of the time points, and the total postoperative blood loss was not different between the groups. There was a trend toward less need for homologous transfusion in group (p < 0.). This study documents the safety and ease of using nonwashed shed mediastinal blood as a primary blood volume support after an open heart operafion. ( 2994;57:625-22) espite recent advances in our understanding of the D clinical hemostatic defect produced by cardiopulmonary bypass, manifested principally as a platelet dysfunction [l-3, postoperative bleeding after cardiac operations remains an important source of morbidity and is responsible for increasing the frequency of homologous blood transfusions. In addition, evidence from patients suffering acute myocardial infarctions who are treated with tissue-type plasminogen activator indicates that there may be a plasmin-mediated derangement of platelet aggregation in vivo, manifested by a prolonged template bleeding time, that correlates with spontaneous bleeding complications that arise after thrombolytic therapy [4]. Previous studies of the transfusion of shed mediastinal blood after cardiac operations have shown there is a reduction in the homologous blood requirement but no increase in the postoperative blood loss [5, 6. This prospective, randomized study was designed to test the hypothesis that the use of nonwashed shed mediastinal blood may exacerbate platelet dysfunction after cardiopulmonary bypass, compared with the alternative use of autologous or homologous standard liquid Accepted for publication May 3, 993. Address reprint requests to Dr Khuri, VA Medical Center, 400 VFW l'kwy, West Roxbury, MA The opinions or assertions contained herein are those of the authors and are not to be construed as official or reflecting the views of the Navy Department or Naval Service at large. preserved blood for volume replacement. Specifically, we sought to determine whether the presence of substances that inhibit platelet function, including fibrin degradation products and D-dimer proteins, in nonwashed shed mediastinal blood caused the functional thrombocytopathy present after extracorporeal circulation to be worse. In addition, we sought to identify any effect of shed mediastinal blood transfusion on the activation of fibrinolytic pathways in vivo. We also hoped to determine what the safe volume of nonwashed shed mediastinal blood was that could be transfused without precipitating hemorrhagic complications. Material and Methods Pa tien t s The experimental protocol for this study was approved by our institutional human research committee. Between June 988 and August 989, 03 patients (mean age, 60 5 years; range, 4 to 72 years) who gave informed consent to participation in this study underwent cardiopulmonary bypass for open heart operations at the West Roxbury Veterans Administration Medical Center. All operations were performed by one of two surgeons using identical operative and myocardial protection techniques. Randomization Criteria Randomization was done preoperatively, irrespective of the planned procedure, by arbitrarily assigning the pa by The Society of Thoracic Surgeons /94/$7.00

2 66 AXFORD ET AL 994; tient to receive either shed mediastinal blood or packed red blood cells if transfusion proved necessary postoperatively. To ensure a homogeneous patient population for comparison, strict criteria for inclusion in the study were established before initiation of the study. The goals of these criteria were to sample patients who were bleeding as the result of nonsurgical causes (ie, blood loss unrelated to an error in surgical technique), and who demonstrated a need for volume expansion using red blood cell products during the immediate (less than 2 hours) postoperative period. Patients who bled less than 400 ml into the chest tube-pleur-evac collecting system (Deknatel, Fall River, MA) within the first 4 hours after complete reversal of the heparin effects with protamine (as documented by return of the activated clotting time [ACT] to, or below, preoperative values) were excluded from the study because, in our experience, such patients rarely need transfusion with red blood cell products in the immediate postoperative period. The decision to transfuse a patient with blood postoperatively was made by the clinician who was responsible for the patient's postoperative care, and who was not involved with the study. The clinical criteria used to determine the need for transfusion consisted of the following: systolic blood pressure, less than 80 mm Hg; m-ean arterial pressure, less than 50 mm Hg; central venous pressure, less than 5 mm Hg; pulmonary capillary wedge pressure, less than 5 mm Hg, cardiac index, less than 2.0 L/min/m2; evidence of inadequate end-organ perfusion (ie, urine output, less than 20 ml/hr); or anemia (hematocrit, less than 25 ~0%). Any patient who bled more than 400 ml in the first 4 hours after operation and who met any of these criteria underwent transfusion. Once the decision to transfuse was made, the patient received either the volume of mediastinal shed blood that had collected up to that point, or unit of packed red blood cells. Autologous packed cells were used if available; otherwise, homologous packed cells were transfused. If a second or third transfusion was given in the first 24 hours after surgery, those patients who had initially received shed mediastinal blood received additional shed blood, and those patients who had initially received banked blood received additional units of packed red blood cells. Patients randomized to receive shed blood whose blood requirements exceeded the amount of shed blood available or who required blood transfusions beyond 48 hours postoperatively were given transfusions of homologous packed red blood cells. Procedures Anesthesia was induced with fentanyl, and maintained with a combination of fentanyl, muscle relaxants, and either halothane or isoflurane. Operations were performed using a standard cardiopulmonary bypass, with a crystalloid prime delivered at a rate of 2.4 L * minp3 * m-' using a single two-staged venous cannula inserted through the atrial appendage and an 8-mm ascending aortic cannula. Patients were cooled to a temperature range of 25" to 30 C, depending on the complexity of the surgical procedure. Myocardial protection was achieved with an antegradely delivered cold (4"C), hyperkalemic crystalloid cardioplegic solution and topical ice slush. Heparin was administered in an initial dose of 3 mg/kg of body weight, with additional doses given to maintain the ACT above 480 seconds during extracorporeal circulation. At the end of bypass, the effects of heparin were neutralized with protamine sulfate, given in a ratio of 0.5 mg of protamine to.0 mg of heparin for the initial heparin dose, and.0 mg of protamine to.0 mg of heparin for all subsequent heparin doses. Systemic temperature was measured intraoperatively and postoperatively using a thermistor placed in the bladder. Blood Loss Measurement of the postoperative blood loss was started intraoperatively when the ACT had normalized after the administration of protamine. This was done by collecting all the blood aspirated from the surgical field and by weighing all blood-soaked sponges. Postoperatively, an accurate hourly record was kept of all mediastinal drainage until the mediastinal drainage tubes were removed. Mediastinal shed blood was collected using the Pleur-evac Autotransfusion System (model A-5005-ATS; Deknatel) with a model A-200-ATS (Deknatel) polyvinyl chloride blood collection bag containing an inline 200-pm nylon mesh filter by means of a closed system with -20 cm H,O suction applied. This collection system contained no anticoagulant and none was added. Mediastinal shed blood was transfused without washing by detaching the autotransfusion replacement bag and reinfusing the blood through a standard 40-pm screen blood filter (Model SQ4OS; Pall, Fajardo, Puerto Rico) via a peripheral intravenous line. For those patients randomized to receive banked blood, standard citrate-phosphate-dextrose ADSOGpreserved cross-matched packed red blood cell units stored at 4 C for up to 42 days were used. Platelets were transfused by pooling platelets isolated from four to ten units of ABO-compatible blood before transfusion. Laboratory Procedures Before transfusion, blood samples for analysis were drawn from either the unit of shed mediastinal blood collected or the unit of packed red blood cells to be transfused. Patient blood samples were obtained before induction of anesthesia, 20 minutes after the institution of cardiopulmonary bypass, at the end of cardiopulmonary bypass, and 0 minutes, 2 hours, and 24 hours after transfusion with either shed mediastinal or banked blood. Additional blood samples were also drawn on postoperative days 2, 3, and 7. Blood samples were collected in K,EDTA anticoagulant for measurement of the hemoglobin concentration (in grams per deciliter), hematocrit value (in volume percent), white blood cell count (x l@/pl), platelet count ( x 03/pL, using phase microscopy), and mean platelet volume (pl) as detailed previously [7]. Serum creatinine levels were determined using a Beckman Astra 8 serum electrolyte analyzer (Beckman Instruments, Brea, CA). Total protein and albumin levels were measured using the method of Kingsley (8. Factor V (percentage of

3 994; AXFORD ET AL 67 normal), factor VIII clotting protein (factor VIIIc; percentage of normal), and fibrinogen (in milligrams per deciliter) levels were measured by clotting assays using blood collected in 3.8% sodium citrate [9]. The antithrombin level (percentage of normal) was measured by a heparin cofactor assay with a chromogenic substrate [lo]. The protein C level (percentage of normal) was measured according to the method of Nicham and associates [ll] using the chromogenic substrate CBS. Fibronectin (in micrograms per milliliter) was measured by an immunoturbidometric assay [2]. Plasminogen activity (percentage of normal) was measured using the chromogenic substrate S-225 according to the method of Friberger [3]. The in vivo activation of fibrinolysis (ie, plasmin activation) was assessed indirectly by measuring the consumption of plasma antiplasmin activity (percentage of normal), as described by Gallimore and associates [4]. Fibrin degradation products (in micrograms per milliliter) were measured using a Trombowell-cotest Kit (Burroughs Wellcome Diagnostics, Greenville, NC) [ 5. D-Dimer levels were measured by latex immunoassay using a monoclonal antibody, as described by Mirshahi and colleagues [6]. Whole blood microaggregates (in millimeters mercury per gram of hemoglobin) were assessed using the method of Swank [7]. Standard template bleeding times, uncorrected for skin temperature, were performed in duplicate using the Simplate I bleeding time module (General Diagnostics, Durham, NC) according to the procedure of Babson and Babson [HI. Bleeding times were determined before the induction of anesthesia, during and at the completion of cardiopulmonary bypass, immediately before transfusion, and 0 minutes, 2 hours, and 24 hours after transfusion in the intensive care unit. Additional bleeding times were also determined on postoperative days 2, 3, and 7. Clinical Parameters Hospital mortality included all deaths occurring within 30 days of the operation. All patients in this study underwent technetium-99m pertechnetate angiography preoperatively and week postoperatively to quantitate the left ventricular ejection fraction [9]. Diagnosis of perioperative myocardial infarction was based on the appearance of new Q waves or loss of R waves on the electrocardiogram. Ventilatory support was quantitated arbitrarily as the number of hours of postoperative intubation. The postoperative low cardiac output syndrome was defined as the need for inotropic agents for more than 48 hours or the need for an intraaortic balloon pump for postoperative support. A posttransfusion febrile reaction was defined as a maximum temperature of greater than (38.5"C) within the first 6 hours after transfusion. Data Analysis Data are expressed as the mean 2 the standard error of the mean. Two-way analysis of variance with a repeated measures design (MANOVA) was used to detect significant changes in variables with respect to the time and type of transfusion (shed blood versus banked blood) throughout the course of the study. When a significant change was identified by MANOVA, the paired t test was used to identify significant differences between specific time points within a group, and either univariate analysis of variance or the Student's t test was used to detect significant differences between groups (shed blood versus banked blood) at discrete time points. Categoric variables were analyzed between the type of transfusion using either the 2 test or the Fisher's exact test, when appropriate. A p value of 0.05 was considered significant for all analyses. All the statistical analyses were performed with the SAS statistical package (Cary, NC). Results Of the initial 03 patients, 7 were excluded from the study. Twenty-nine patients did not bleed the required 400 ml in the first 4 hours after protamine reversal to warrant transfusion, and none of these patients subsequently received a transfusion for the indication of acute blood volume expansion within 2 hours of their operation. Thirty patients who bled 400 ml were not considered in need of a transfusion based on their hemodynamic data, and were thus excluded. Five patients had received aspirin within 24 hours of their operation as part of an unrelated study, and were excluded. In 2 patients who required reoperation for bleeding within the first 24 hours, bleeding in both, which was controlled, was from saphenous vein graft branches. Both of these patients were excluded. One patient required a postoperative left ventricular assist device with anticoagulation, and, in another patient, the right ventricle was injured during sternotomy and he suffered a large associated blood loss. Three patients died from low cardiac output within 24 hours of their operations. Thus, 32 patients ultimately completed the study and were included in the data analysis. Sixteen patients were randomized to receive nonwashed shed mediastinal blood (group ) and 6 patients received banked blood (group 2) for their initial transfusion therapy. The procedures in these patients were performed for the treatment of isolated coronary artery disease (n = 23), valvular disease (n = 4), and combined disease (n = 5). One patient in each group had to undergo emergent coronary artery bypass grafting for the control of unstable angina that had been treated with an intraaortic balloon pump. The patient characteristics are summarized in Table. There were no significant differences between group and group 2 with respect to patient age, the type of procedure performed, the preoperative or postoperative left ventricular ejection fraction, the duration of cardiopulmonary bypass, the duration of aortic crossclamping, the frequency of postoperative myocardial infarction, the need for postoperative ventilatory support, the frequency of the low cardiac output syndrome, or the frequency of the posttransfusion febrile reaction. In addition, there were no significant differences in the serum creatinine levels (in milligrams per deciliter) between the two groups preoperatively (group,.2 * 0.; group 2,. 2 O.l), 24 hours after transfusion (group, ;

4 68 AXFORD ET AL 994; Table. Preoperative and Postoperative Characteristics of Pa tien tsa Group lb Group 2' p Characteristics (n = 6) (n = 6) Value Age (Y) Procedure (No. of patients) CABG Valved Valved and CABG RVG ejection fraction Preoperative Postoperative Cardiopulmonary bypass Total bypass time (min) Cross-clamp time (min) Postoperative MI Duration of intubation (h) Low cardiac output syndrome (No. of patients) Posttransfusion febrile reaction (No. of patients) 60 t t f f 9 86 f 2 29 t t f f f 0 67 f f 2 2 a Values expressed as mean? standard error of mean. Transfused with nonwashed shed mediastinal blood. Transfused with banked red blood cells. Aortic or mitral valve replacement. CABG = coronary artery bypass grafting; MI = myocardial infarction; = not significant; RVG = radioventriculogram. group 2,.2 & O.l), or 7 days after transfusion (group,.2 * 0.; group 2,.2 rt 0.). Analysis of Blood Units Transfused The average time to transfusion after the onset of collection of the first unit of shed mediastinal blood in group was 2.4 rt 0.2 hours. The average length of storage at 4 C of the first unit of packed red blood cells transfused in each patient in group 2 was 7 k 2 days. The average volume of the first shed mediastinal blood transfusion given to patients in group was 453 & 34 ml (range, 300 to 840 ml). The average volume of total shed mediastinal blood received in group patients was 70 * 90 ml (range, 300 to,700 ml). Table 2 summarizes the results of the in vitro analysis of each unit of shed mediastinal blood and standard banked blood before transfusion. These data demonstrate that, compared to nonwashed shed mediastinal blood, the hematocrit value, cellular hemoglobin content, and platelet count (all p < 0.005) are all significantly greater in banked blood. Shed blood possesses greater factor VIIIc activity, antithrombin I activity, protein C levels, plasminogen activity, and antiplasmin activity (all p < 0.0). Shed and banked blood contain comparable levels of plasma hemoglobin, fibrinogen, and fibronectin. In addition, the amount of particulate microaggregates is the same for both. The levels of fibrin degradation products in shed and banked units of blood are shown in Figure. All measured units of banked blood had low levels of fibrin degradation prod- Table 2. In Vitro Blood Unit Analysis" Shed Banked P Hematologic Variables Bloodb Blood' Value Hemoglobin (g/dl) Hematocrit (~0%) Plasma hemoglobin (mg/dl) White blood cells ( x 03/pL) Platelet count ( x ~O~/~L) Factor VIIIc (%) Fibrinogen (mg/dl) Antithrombin I (%) Protein C (%) Fibronectin (pg/ml) Plasminogen (%) Antiplasmin (%) White blood aggregates (Swank; mm Hg/g Hb) 7.9 t t 32? t t 9 25 f,4 26 f 33? 4 7 t 7 I2 f 6 57 f 3 39 t 2 6 f t f 3 98 t ? t 29 Of t 3 89 f 3 26 f 20 f 2 64 f a Values expressed as mean? standard error of mean. Nonwashed shed mediastinal blood. Liquid preserved packed red blood cells. ucts (less than 20 pg/ml; n = 9), whereas all measured units of nonwashed shed mediastinal blood exhibited elevated levels (20 to 80 pg/ml; n = 5; 80 to 320 pg/ml, n = 8; Fisher's exact test, p < 0.000). The levels of D-dimer in shed and banked units of blood are shown in Figure 2. All measured units of banked blood had low levels of D-dimer (less than 0.5 pg/ml; n = ll), while all measured units of nonwashed shed blood had elevated levels (greater than 2.0 &ml; n = 2; Fisher's exact test, p < 0.000). Hematologic Variables There were no significant differences between group and group 2 in any of the hematologic variables, including the hematocrit, hemoglobin level, plasma hemoglobin = t m.30 ~ 3 " < > 320 Fibrin Degradation Product * p <0.000 by Fisher's Exact Test Fig I, Level of fibrin degradation product in shed nonwashed autologous mediastinal blood (filled bars) and liquid preserved autologous and homologous blood (hatched bar).

5 994; AXFORD ET AL 69 c > 2.0 D-Dimer * p ~0.000 by Fisher's Exact Test Fig 2. Level of D-dimer in shed nonwashed autologous mediastinal blood (filled bar) and liquid preserved autologous and homologous blood (hatched bar). level, white blood cell count, platelet count, and plasma microaggregate level either at baseline or at any subsequent time point in the study. The plasma hemoglobin level increased significantly ( p < 0.05) in both groups while on cardiopulmonary bypass, from between 7 to 8 mg/dl preoperatively to approximately 60 to 70 mg/dl at the end of bypass, and then fell rapidly ( p < 0.05) in the immediate postoperative period to 27.5 f 3.2 mg/dl in group and 8.3 f 4. mg/dl in group 2 by 0 minutes after the first transfusion. The plasma hemoglobin level then returned to baseline levels by postoperative day. The plasma microaggregate levels never increased significantly above the baseline value in either group for the duration of the study. Plasma Protein Variables Throughout the study, there were no significant differences between group and group 2 in the plasma protein variables, including the factor V, factor VIIIc, and antithrombin I activity, as well as the fibrinogen and fibronectin levels. Protein C levels fell significantly in both groups (p < 0.05) while on cardiopulmonary bypass (group, 2% f 7% to 65% f 5%; group 2, 02% * 5% to 67% f 5%). However, in contrast to the other proteins, the level of protein C was significantly greater (p < 0.05) in group (72% f 4%) than in group 2 (63% f 2%) 2 hours after transfusion. These differences were no longer significant by the first postoperative day, and the level of protein C remained below baseline in both groups until the final measurement on the seventh postoperative day. Except for protein C, all plasma protein levels measured were not significantly changed by transfusion with either nonwashed shed mediastinal blood or liquid preserved packed cells during this study. Bleeding Time and Fibrinolytic Variables Table 3 summarizes the bleeding times, not corrected for skin temperature, and plasma fibrinolytic variables assessed during the study. The bleeding time was not significantly different between the two groups at baseline. It increased in both groups with the initiation of cardiopulmonary bypass and remained significantly elevated (p < 0.05) in both groups at the end of bypass (group, 7.5? 0.5 minutes to 4.7 f.3 minutes; group 2,8.4 f 0.6 minutes to 4.3 f.3 minutes). The bleeding time gradually returned to the baseline level by the first postoperative day; however, it remained elevated (p < 0.05) in both groups 2 hours after transfusion (group, 0.0 *. minutes; group 2, 0.7 f 0.7 minutes). The bleeding time was never significantly different between the two groups at any time point throughout the study and was not affected by transfusion with either nonwashed shed mediastinal or banked blood (see Table 3). The plasminogen levels were not significantly different at baseline between the two groups, fell significantly ( p < 0.05) in both groups with the initiation of cardiopulmonary bypass, and gradually rose in the postoperative period, returning to baseline levels by the seventh postoperative day. The plasminogen levels were never significantly different between the Table 3. Bleeding Time and Fibrinolutic Variables" Variable Preoperative values On CPB End CPB 0 min after TXN 2 h after TXN Postop day Postop day 2 Postop day 3 Postop day 7 Bleeding Time (min) Plasminogen (%) Antiplasmin (%) Group lb Group 2' Group lb Group 2' Group lb Group 2' 7.5 f 0.5 >20 f f f f. 9.3 f f f * 0.6 >20 f f.3. f. 0.7 f f f * f 5 56 f 3 52 f 3 50 f 4 66 f 7 77 f 5 8 f 4 49 f 7 56 f 4 55 f 2 53 f 3 48 f 3 48 f 3 60 f 5 84 f 5 78 f 3d 7 f 2 57? 4d 58 f 3 58 f 2 70 f 3 85 f f 8 45 f 2 5 f 3 49 f 3 6 f 3 78 f 2 90 f a Values expressed as mean 2 standard error of mean. cells. p < CPB = cardiopulmonary bypass; TXN = transfusion. Transfused with nonwashed shed mediastinal blood. ' Transfused with banked red blood

6 620 AXFORD ET AL 994;57:65-22 Table 4. Blood Loss and Blood Product Use" Blood Loss and Product P Use Group lb Group 2' Value Blood loss (ml) From protamine to TXN First 2 hr after TXN Total postoperative Total colloid volume replacement (ml) First 24 h postoperative 7 days postoperative Specific blood product replacement (units) Packed red blood cells Fresh frozen plasma Platelets 888 f ,06 * 50,66 * 340, * * 2.7,00 t , ,694 f 54,803 f k f.8 a Values expressed as mean 2 standard error of mean. Transfused with nonwashed shed mediastinal blood. Transfused with banked red blood cells. CABG = comq artery bypass grafhng = transfusion. = not sipfimt; TXN two groups at any time point during the study, including 2 hours after transfusion (see Table 3). The baseline plasma antiplasmin levels did differ significantly (p < 0.05) between group (78% 2 3%) and group 2 (7% k 2%), but this probably represents an experimental population sampling error rather than any real intergroup difference. The antiplasmin levels decreased significantly (p < 0.05) in both groups with the initiation of cardiopulmonary bypass and remained significantly lower in group 2 (45% 2 2%; p < 0.05) than in group (57% * 4%) at the end of bypass. The antiplasmin levels in both groups were unaffected by transfusion, and returned to preoperative levels by the second postoperative day, at which point they were no longer significantly different from one another. The antiplasmin levels then increased significantly ( p < 0.05) in a similar fashion in both groups for the remainder of the study and were not significantly different between the two groups (see Table 3). Blood Loss and Blood Product Use Table 4 shows the data for blood loss and blood product utilization during the study. There were no significant differences in blood loss between the two groups, either from the end of protamine administration until the start of the first transfusion with shed mediastinal or banked blood (group, ml; group 2,,00 * 98 ml) or during the first 2 hours after the first transfusion (group, 84 * 40 ml; group 2, 304 * 07 ml); there were also no significant differences in the total blood loss in the postoperative period between the two groups (group, 2, ml; group 2, 2, ml). There were no significant differences between the two groups in the total blood volume replacement with colloid solutions in either the first 24 hours postoperatively (group,,66 * 340 ml; group 2,,694 * 54 ml) or the first 7 days postoperatively (group,,89 * 362 ml; group 2,,803 k 522 ml). There were no significant differences in terms of the specific blood product replacement between the two groups, including fresh frozen plasma and platelets (see Table 4); however, those patients who also received shed mediastinal blood tended to need fewer units of red blood cells (group, 2.0 * 0.5 units; group 2, units; p < 0.). The distribution of the different types of red blood cell products received by the patients in this study is depicted in Figure 3. Only patient in group received autologous liquid preserved blood when additional transfusion therapy proved necessary, and this patient also subsequently received a unit of homologous liquid preserved blood. Five patients in group 2 received autologous liquid preserved blood for their initial transfusion therapy, and 3 of these patients required additional transfusion with homologous liquid preserved blood. In all, 6 of the 6 patients (38%) who received shed mediastinal blood (group ) did not subsequently receive transfusions of homologous liquid preserved blood, whereas only 2 of the 6 patients (3%) who received autologous liquid preserved blood (group 2) did not receive a transfusion of homologous liquid preserved blood. Again, this tended toward, but did not reach, statistical significance (2 analysis, p < 0.). Comment This prospective study was designed to examine the respective effects of nonwashed shed mediastinal blood transfusion and standard banked blood transfusion on in vivo platelet and related clotting and fibrinolytic functions after cardiopulmonary bypass for open heart operations. Methods to minimize the need for homologous blood replacement postoperatively in the cardiac surgery patient population are being actively investigated because of the increased concern surrounding the risks of blood transfusion [2&28]. The postoperative reinfusion of shed mediastinal blood has been extensively studied to determine the utility of this readily available source of autologous blood, and has been shown to lead to a reduced need for homologous blood, but not an increased blood loss, after a cardiac operation [5, 6, 27, 29. The results of this study are consistent with those of previous studies that have shown the composition of nonwashed shed mediastinal blood consists of a hematocrit of 20 vol%, an elevated plasma hemoglobin level of 32 mg/dl, low levels of platelets (22,OOO/pL), extremely low levels of fibrinogen, high levels of fibrin degradation products, and decreased factor VIII clotting activity [5, 24, 27, 30. In addition, the present study has demonstrated that nonwashed shed mediastinal blood contains increased levels of D-dimer protein, and decreased levels of antithrombin, protein C, and plasminogen, as well as reduced antiplasmin activity, although these values are greater for shed mediastinal blood than for banked red blood cells. These data show that shed mediastinal blood undergoes extensive coagulation and clot lysis within the

7 994; AXFORD ET AL E W 2500 H Autologous Shed Blood Autoogous Liquid Preserved Blood Homologous Liquid Preserved Blood Fig 3. Volume of autologous shed rnediastinal blood and autologous and homologous liquid preserved blood transfused into patients during the first week after open heart operation GROUP 2 PATIENT NUMBER GROUP mediastinum and pleural spaces before collection, consistent with the observations in recent reports [3]. The data from this study are consistent with data reported by Schaff [5] and Johnson [6] and their colleagues, who studied the effects of transfusion with shed mediastinal blood versus those of homologous banked blood and found no difference in terms of the postoperative blood loss, consumption of plasma coagulation factors, or evidence of bleeding complications attributable to the use of mediastinal blood. These authors noted a reduction in the homologous blood requirement in the group receiving shed mediastinal blood, which our data did not show. Hartz and associates [3] observed no significant increase in the by-products of fibrinogen consumption or fibrinolysis after transfusion with shed mediastinal blood, indicating no evidence for the induction of either disseminated intravascular coagulation or increased fibrinolysis. New data from the present study also support this hypothesis, as we did not witness activation of fibrinolytic pathways in vivo, as evidenced by the consumption of clotting factors, plasminogen, or antiplasmin in response to transfusion with nonwashed shed mediastinal blood. Our study also showed that protein C levels (a natural anticoagulant and a fibrinolytic agent) are increased in patients receiving nonwashed shed mediastinal blood; however, the functional importance of this finding is unknown, and this did not precipitate any apparent bleeding complications. Most importantly, our data demonstrated that transfusion with nonwashed shed mediastinal blood produces no additional loss of platelet function in excess of that normally present after cardiopulmonary bypass [ Specifically, there was no evidence of increased in vivo platelet dysfunction, as assessed by the template bleeding time in the group that received nonwashed shed mediastinal blood that con- tained high levels of fibrin degradation products and D-dimer protein. The results of this study indicate that the transfusion of approximately 700 ml of nonwashed shed mediastinal blood is safe. However, our findings do not apply to transfusion with massive amounts of shed mediastinal blood (>2,000 ml), which may be injurious. However, modest amounts of mediastinal blood should be routinely used as the simplest and first source of blood for transfusion and blood volume support in the postoperative management of all cardiac surgery patients. This work was supported by the US Navy (Office of Naval Research contracts N C-068 and N C-08, with the funds provided by the Naval Medical Research and Development Command), and by the Richard Warren Surgical Research and Educational Fund. References. Wenger RK, Lukasiewicz H, Mikuta BS, Niewiarowski S, Edmunds LH. Loss of platelet fibrinogen receptors during clinical cardiouulmonarv bvuass. I Thorac Cardiovasc Surg I II 989; ' 2. Harker LA, Maluass TW. Branson HE, Hessel EA. Slichter Sl. I Mechanism of abnormal bleeding in patients undergoing cardiopulmonary bypass: acquired transient platelet dysfunction associated with selective alpha-granule release. Blood 980;56: Edmunds LH, Ellison N, Colman RW, et al. Platelet function during cardiac operation: comparison of membrane and bubble oxygenators. J Thorac Cardiovac Surg 982;83: Gimple LW, Gold HK, Leinbach RC, et al. Correlation between template bleeding times and spontaneous bleeding during treatment of acute myocardial infarction with recombinant tissue-type plasminogen activator. Circulation 989; 80:58-8.

8 622 AXFORD ET AL 994; Schaff HV, Hauer JM, Bell WR, et al. Autotransfusion of shed mediastinal blood after cardiac surgery. J Thorac Cardiovasc Surg 978;75: Johnson RG, Rosenkrantz KR, Preston RA, Hopkins C, Daggett WM. The efficacy of postoperative autotransfusion in patients undergoing cardiac operations. 983;36: Thompson CB, Eaton KA, Princiotta SM, Rushin CA, Valeri CR. Size dependent platelet subpopulations: relationship of platelet volume to ultrastructure, enzymatic activity, and function. Br J Haematol 982;50: Kingsley GR. The determination of serum total protein, albumin, and globulin by the Biuret reactions. Br J Haematol 982;50: Feingold HM, Pivacek LE, Melaragno AJ, Valeri CR. Coagulation assays and platelet aggregation patterns in human, baboon, and canine blood. Am J Vet Res 986; Abildgaard U, Lie M, Odegard OR. Antithrombin (heparin cofactor) assay with new chromogenic substrate ( and Chromozym TH). Thromb Res 977;: Nicham F, Guichiaoua JF, Contant G, Martinoli JL. Rapid determination of protein C activity. Ann Biol Clin 988;46: Saba TM, Albert WH, Blumenstock FA, Evanega G, Staehler F, Cho F. Evaluation of a rapid immunoturbidometric assay for opsonic fibronectin in surgical and trauma patients. J Lab Clin Med 98;98: Friberger P. Methods for the determination of plasmin, antiplasmin, and plasminogen by means of the substrate S-225. Haemostasis 978; Gallimore MJ, Amundsen E, Aasen AO, Larsbraaten M, Lyngaas K, Svendsen L. Studies on plasma antiplasmin activity using a new plasmip specific chromogenic tripeptide substrate. Thromb Res 979;4: Arocha-Pinango CL. A comparison for the TRCII and latex particle tests for the titration of Fr antigen. J Clin Pathol 972;25: Mirshahi M, Soria J, Soria C, Perrot JY, Boucheix C. A latex immunoassay of fibridfibrinogen degradation products in plasma using a monoclonal antibody. Thromb Res 986;44: Swank RL. Alteration of blood on storage: measurement of adhesiveness by "aging" platelets and leukocytes and their removal by filtration. N Engl J Med 96;265: Babson SR, Babson AL. Development and evaluation of a disposable device for performing simultaneous duplicate bleeding time determinations. Am J Clin Pathol 978;70: Schoolman M, Bianco JA, Khuri SF, et al. The radionuclide evaluation of septa wall motion following coronary bypass surgery. Nucl Med Commun 985;6: Bidstrup BP, Royston D, Sapsford RN, Taylor KM, Cosgrove DM. Reduction in blood loss and blood use after cardiopulmonary bypass with high dose aprotinin (Trasylol). J Thorac Cardiovasc Surg 989; Cohen ND, Munoz A, Reitz BA, et al. Transmission of retroviruses by transfusion of screened blood in patients undergoing cardiac surgery. N Engl J Med 989;320: Goodnough LT, Johnston MFM, Ramsey G, et al. Guidelines for transfusion support in patients undergoing coronary artery bypass grafting. 990;50: Collins JD, Bassendine MF, Codd AA, Collins A, Ferner RE, James OFW. Prospective study of post-transfusion hepatitis after cardiac surgery in a British centre. Br Med J 983; Thurer RL, Hauer JM. Autotransfusion and blood conservation. Curr Probl Surg 982;9: Weisel RD, Charlesworth DC, Mickleborough LL, et al. Limitations of blood conservation. J Thorac Cardiovasc Surg 984;88: Love TR, Hendren WG, OKeefe DD, Daggett WM. Transfusion of predonated autologous blood in elective cardiac surgery. 987;43: Thurer RL, Lytle BW, Cosgrove DM, Loop FD. Autotransfusion following cardiac operations: a randomized, prospective study. 979;275OC Mayer ED, Welsch M, Tanzeem A. Reduction of postoperative donor blood requirement by use of the cell separator. Scand J Thorac Cardiovasc Surg 985;9: Griffith LD, Billman GF, Daily PO, Lane TA. Apparent coagulopathy caused by infusion of shed mediastinal blood and its prevention by washing of the infusate. Ann Thorac Surg 989;47: Carter RF, McArdle B, Morritt GM. Autologous transfusion of mediastinal drainage blood: a report of its use following open-heart surgery. Anaesthesia 98;36: Hartz RS, Smith JA, Green D. Autotransfusion after cardiac operation. J Thorac Cardiovasc Surg 988;96: George JN, Pickett EB, Saucerman S, et al. Platelet surface glycoproteins: studies on resting and activated platelets and platelet membrane microparticles in normal subjects, and observations in patients during adult respiratory distress syndrome and cardiac surgery. J Clin Invest 986;78:34C-S. 33. Musial J, Niewiarowski S, Hershock D, Morinelli TA, Colman RW, Edmunds LH. Loss of fibrinogen receptors from the platelet surface during simulated extracorporeal circulation. J Lab Clin Med 985;05: Van Oeveren W, Eijsman L, Roozendaal KJ, Wildevuur CRH. On the mechanism of platelet preservation during cardiopulmonary bypass by aprotinin. Lancet 988;: van Oeveren W, Jansen NJG, Bidstrup BP, et al. Effects of aprotinin on hemostatic mechanisms during cardiopulmonary bypass. 987;44:64C-45.