Hemorrhage, coagulation, and transfusion: a risk-benefit analysis

Hemorrhage, Coagulation, and Transfusion: A Risk-Benefit Analysis Bruce

D. Spiess,

MD

Transfusion risks include the possibility of ABO/Rh incompatibility, sepsis, febrile reactions, immunosuppression, and viral transmission; incidences and consequences of these complications are reviewed. Predonation of autologous blood generally reduces the need for homologous blood by about 30% to 40%. but relatively few coronary artery bypass surgery (CABG) patients predonate blood. Drug products to decrease blood use include 1-deamino-8-o-arginine vasopressin (DDAVP), tranexamic acid, epsilon-aminocaproic acid, and aprotinin. A recent study suggests that a subgroup of patients with abnormal platelet function may benefit from a platelet therapy such as DDAVP. The prophylactic use of tranexamic acid reduces cardiac surgery postoperative blood loss, as measured by chest-tube output, by about 30%; unfortunately, data demonstrating a reduction in transfusion requirements are not available. Aprotinin use is associated

with major reductions in blood transfusion requirements. Aprotinin provides platelet protection during cardiopulmonary bypass. Duration of stay in the intensive care unit was not increased by use of aprotinin, thus alleviating some concerns that aprotinin might promote coronary thrombosis. A recent report cites early graft closure as a major concern with aprotinin therapy, but data from other studies show no significant differences in rates of graft closure between patients receiving and those not receiving aprotinin. Routine use of a thromboelastogram with all cardiopulmonary bypass surgery at the University of Washington Hospital has reduced use of blood products by 30%. Copyright o 1994 by W.B. Saunders Company

T

present with the use of any blood product, whether whole blood, platelets, packed red cells, FFP, or cryoprecipitate. Instances of ABO/Rh incompatibility are due to human error, either at the blood bank or in the operating room. The incidence of ABO/Rh incompatibility is variously estimated at from 1 in 2,000 to 1 in 6,000 units transfused.7-9 Fortunately, most instances do not result in fatal hemolytic transfusion reactions, which are estimated to occur in about 1 in 100,000 units transfused. Massive posttransfusion sepsisdue to contamination with Yersinia, Streptococcus, or Staphylococcus occurs with approximately the same frequency as fatal hemolytic transfusion reactions, with an incidence of about 1 in 100,000 units transfused. Febrile reactions generally do not pose a significant problem. However, the problems of long-term alloimmunization and immunosuppression have been demanding increasing attention, particularly in the orthopedic literature.*” Fairly well-controlled studies comparing the use of homologous blood with autologous blood suggest that immunosuppression after transfusion is related to the onset of secondary infections in patients who receive large quantities of homologous blood. In Japan, potentially fatal graftversus-host disease (GVHD) reportedly occurs in 1 in 300 to 1 in 400 units transfused. The risk of GVHD is higher with designated donors than with blood from the general population.

HE MEAN AGE of patients undergoing cardiopulmonary bypass (CPB) is increasing, and older patients bleed more. This factor is one of many fostering current efforts to find therapeutic modalities that decrease the use of blood products. Resulting changes in practice include not priming CPB machines with blood except in extreme situations where the patient’s oxygen-carrying capacity is a concern, the use of cell-salvageintraoperatively, and increasing use of postoperative salvage.1,2 Transfusion practices have also changed, largely in response to the pressure of lay concern about the acquired immunodeficiency syndrome (AIDS) epidemic. The trigger for transfusion has been lowered; most patients probably now enter the intensive care unit (ICU) with hematocrits of about 25% or even less. Current monitoring capabilities allow evaluation of oxygen supply and demand ratios that may make possible even more discriminating selection of patients who need to be transfused. A word of caution when caring for patients bleeding in the ICU and when perusing the literature on transfusion medicine: 450 mL of mediastinal blood loss does not equal 1 unit of blood lost from the patient; the red-cell mass is not the same.3The hematocrit of most chest-tube output is not only significantly lower than that of a unit of packed red blood cells, but also significantly lower than that of a unit of whole blood. Despite the considerable lay pressure to reduce use of blood products, transfusion practice in coronary artery bypass graft (CABG) surgery in the United States varies greatly from one institution to another. While most bypass patients receive red-cell transfusions, fresh-frozen plasma (FFP) and platelet transfusions are frequent in some institutions and rare in others.4 Clearly, considerations other than safety alone guide transfusion policies. TRANSFUSION

Cardiothoracic

Viral

transfusion,

hemorrhage,

coagulation,

DDAVP

Transmission

Although public attention has focused primarily on the risk of human immunodeficiency virus (HIV) transmission through blood transfusion, the risk of HIV transmission, estimated to be only between 1 in 100,000 and 1 in l,OOO,OOO

RISKS

The current blood supply is the safest ever available in the history of transfusion medicine.5,6Nevertheless, transfusion does not now and never will entail zero risk. The risks involved include ABO/Rh incompatibility, sepsis, febrile reactions, immunosuppression, and viral transmission. With the exception of ABO/Rh incompatibility, these risks are Journalof

KEY WORDS:

From the Division of Cardiothoracic Anesthesia, University of Washington School of Medicine, Seattle, WA. Address reprint requests to Bruce D. Spiess, MD, Division of Cardiothoracic Anesthesia, University of Washington, 1959 NE Pacific St, Seattle, WA 98195. Copyright 0 1994 by l%B. Saunders Company 1053-0770/9410801-0206$03.00/0

and VascularAnesthesia, Vol8, No 1, Suppl 1 (February), Sponsored by Miles, Inc, Pharmaceutical Division

1994: pp 19-22

19

BRUCE

20

units transfused, is not the major challenge to making the blood supply safer overall.” The risk of transmitting hepatitis B or C through transfusion poses a more pressing challenge. Available data indicate that about 1 in every 300 units transfused is possibly transmitting hepatitis B and 1 in every 300 to 500 units transfused is possibly transmitting hepatitis C. A recent encouraging report from Johns Hopkins, however, indicates that, at least at its blood bank, the incidence of viral transmission has been reduced to below 1 in 300 to 500 and may be as low as 1 in 3,000 units transfused.5 A hospitalized patient who is transfused receives on average about 6.2 units. This, of course, represents 6.2 donor exposures, and the risk involved is additive. The cost to society of viral transmission by transfusion is high. Ten percent of patients infected with hepatitis B or C develop chronic active hepatitis. Ten percent become cirrhotic and eventually require either a liver transplant or terminal cirrhotic care. Hepatic transplantation is notorious for being extremely expensive, but terminal cirrhotic care is equally costly with its requirements for the treatment of gastrointestinal (GI) bleeding, renal failure, multiple organ failures, infections, encephalopathy, and other possible complications. Thus, the cost of a blood transfusion should be seen as entailing more than just the cost of obtaining blood from the blood bank and transfusing it into a patient. A more encompassing view must include the amortized cost of caring for those patients whose livers are destroyed by a virus transmitted by the blood they receive. TRANSFUSION

Autologous

ALTERNATIVES

Blood

In an effort to reduce the use of homologous blood, considerable attention has been given to predonation of autologous blood before cardiac surgery.12-16Various studies have shown that such predonation generally reduces the need for homologous blood by about 30% to 40%; in one report, the use of homologous blood was required in only 27% of cases of cardiac surgery with autologous blood compared with 83% of caseswithout patient predonation.13 Nevertheless, a multicenter audit of transfusion practice in CABG surgery showed that only 3.3% of patients donate autologous blood.’ One reason for this low percentage may simply reflect the difficulty of getting patients to make multiple visits to the blood bank to predonate their blood. Another reason is that many patients undergo CABG surgery semi-emergently, within a week or so of their catheterization, which does not allow time to harvest much autologous blood. Although most studies indicate that it is safe to routinely take blood from someone with coronary artery disease before cardiac surgery and then to reinfuse that blood into the patient during surgery,17*‘sone study showed that such patients may experience significant ST-segment depression.14 Also, 20% of these patients become hypotensive during the period in which they are predonating their blood. The safety of this practice, therefore, remains uncertain. Practices entailing the use of platelet-rich plasma or

D. SPIESS

euvolemic hemodilution preoperatively or intraoperatively still remain fairly controversial, lacking substantial amounts of outcome data. Drug Products

There are four compounds that either have been used or are under investigation for the purpose of decreasing blood use: l-D-deamino-arginine vasopressin (DDAVP), tranexamic acid, r-aminocaproic acid, and aprotinin. A 1988 paper created much excitement by demonstrating a significant decrease in transfusion requirements with DDAVP.19 However, subsequent studies failed to show similar positive results, and interest in this compound faded rapidly. This situation may have changed with the recent publication of a study that identified a subgroup of patients who indeed did benefit from DDAVP during CABG.*O In this study, the thromboelastograph (TEG) was used to identify a subgroup of patients with abnormal platelet function. Although multiple factors contribute to the maximum amplitude demonstrated on a TEG, platelet function is certainly one of the significant factors, so that amplitudes of less than 50 mm strongly suggest some abnormality of platelet function. In the subgroup of patients with maximum amplitudes of less than 50 mm, patients who received DDAVP during cardiac surgery showed a highly significant decrease in chest-tube output compared with those who did not. The study thus suggeststhat it may be useful to monitor patients to detect those particularly at risk for platelet dysfunction who may benefit from available platelet therapies such as DDAVP. Prophylactic use of tranexamic acid has repeatedly been shown to reduce cardiac surgery postoperative blood loss, as measured by chest-tube output, by about 3O%.*i This agent, as a fibrinolytic inhibitor, is useful only if used prophylactically as the patient goes on bypass, to help prevent the end stage of contact activation, ie, the activation of plasminogen to plasmin. Unfortunately, the investigators have not shown major differences in transfusion requirements between patients treated prophylactically with tranexamic acid and those who were not. It would seem that transfusion requirements data are critical to the evaluation of the usefulness of any practice designed to reduce the risk of transfusion-transmitted disease. Epsilon-aminocaproic acid acts identically to tranexamic acid, with the data indicating very similar results. The main difference between the two compounds is in potency. Aprotinin has been widely used in Europe. It is a 58 amino acid polypeptide that acts as a nonspecific serine protease inhibitor. It is a naturally occurring trypsin inhibitor and kallikrein inactivator. It is thought to act by some mechanism that inhibits either kallikrein or plasmin or both, although its exact mechanism of action remains unclear despite studies that have been reported.22-27Aprotinin is distinguished from some of the other therapies designed to reduce blood loss in that its use has been associated with reductions in blood transfusion requirements. Aprotinin somehow provides platelet protection against

RISK-BENEFIT

OF TRANSFUSION

21

the initial effect of CPB.22 An electron microscopy study showed significant platelet preservation with aprotinin therapy. 28 The benefit appears to be possibly due to the better preservation of glycoprotein Ib and IIb/IIIa binding sites on platelet surfaces. van Oeveren and colleagues looked at a number of the effects of aprotinin on hemostatic mechanisms during CPB, in an effort to elucidate its mechanisms of action.27 They found levels of tissue plasminogen activator (tPA) during CPB to be lower with aprotinin than without. They also found levels of thromboxane B2 (TxB2), which normally peak during bypass due to thromboxane release from platelets, to be lower with aprotinin, indicating platelet preservation. Marx et a129conducted a study that compared levels of both procoagulant and fibrinolytic system proteins during cardiac surgery in patients receiving aprotinin and in those who did not. Among their findings was that the elevation in levels of tPA that normally occurs during CPB either did not occur or occurred to a lesser extent in patients treated with aprotinin. They found, however, that patients treated with aprotinin showed increased levels of plasminogen activator inhibitor (PAI) compared with patients who did not receive aprotinin (Fig 1). These findings were in accord with earlier observations by Tanaka et a1.30The explanation for these aprotinin effects remains unclear. Dietrich et al, reporting on 3 years of experience with 1,784 patients undergoing cardiac surgery, found, as other investigators had previously, that aprotinin was associated with savings in blood loss, as measured both by chest-tube output and transfusion requirements.31 They also found that the duration of stay in the ICU was no longer for patients who received aprotinin than for those who did not. This latter finding tends to alleviate the concerns expressed by some that aprotinin, by increasing levels of the procoagulant PAI, may promote coronary thrombosis. Clearly, if a large number of patients receiving aprotinin were thrombosing, they would be staying in the ICU longer than those not receiving aprotinin. Mortality rates also did not differ between the aprotinin patients and control patients. The issue of whether or not aprotinin has any effect on

lu” x I2 10

8 B 4 2 0I

i

i

i

! I

Ti.1 i'

i f

/

i

'-5"'

1

i

f .1.-..-.-.-.;.! T P “?.. ii ! “?i.. ! ! 1 i .~..~““” item 1 2

3

4

s

Table

1. Incidence

of Patency

Among

21 Grafts

from

7 Postmortem

Examinations Graft Source

Aprotinin

Placebo

Veins

6 (50%)

5 (100%)

Internal thoracic artery Gastroepiploic artery

3 (100%) 1 (100%)

0 0

P Value 0.10 NA NA

Abbreviation: NA, not applicable. Data from Cosgrove et aL3*

early graft closure after revascularization is the subject of considerable current discussion. The discussion was prompted by Cosgrove et al who, in the abstract and conclusion of their recently published paper reporting on their study of aprotinin therapy for reoperative myocardial revascularization, stated that early graft closure is a major concern with aprotinin therapy.32This claim is controversial and warrants critical attention. Although the conclusion to the Cosgrove paper includes the claim that aprotinin causes early graft closure, the actual data presented in the paper show no statistically significant differences in the incidence of myocardial infarction between patients receiving low-dose aprotinin, highdose aprotinin, and placebo. Furthermore, the data on myocardial infarction rates referred to in the abstract differ from the data on rates reported in the paper itself. The abstract only refers to findings in patients who died, with grafts examined at necroscopy (Table 1). Among these patients, 6 of 12 vein grafts were found to be closed in patients who had received aprotinin compared with none of 5 vein grafts examined in patients who did not receive aprotinin. Two problems exist with respect to interpretation of these data. First, the difference reported is not statistically significant (P = 0.10) at the usual accepted level (P < 0.05). Second, the number of patients accounting for the examined vein grafts is not given. Conceivably, as many as 6 vein grafts could have been present in one patient whose death after widespread thrombosis could have accounted for all findings of graft closure. Certainly, it would not be reasonable to conclude that aprotinin therapy causes graft closure from the evidence of only one or a very few patients. Other data are available from studies in which either computed tomography (CT) scanning or recatheterization in the early postoperative period has been used to ascertain graft patency.33 These series each comprise about 100 patients, with 150 grafts in each group. They show no significant differences in graft-closure rates between patients receiving and not receiving aprotinin therapy. As a coagulationist, I feel strongly that coagulation has much to do with who has a perioperative myocardial infarction and who does not. Nevertheless, on the basis of available data, I see no reason to conclude at this point that aprotinin therapy promotes hypercoagulability. THROMBOELASTOGRAPHY

Fig 1. t-PA:C, Statistical details (.-.-.-.-.I and lower

t-PA:Ag, and PAI at different sampling times. are at bottom of figure. Dashed lines indicate upper (-1 limits of normal range. (Data from Marx et al.29)

With respect to coagulation, experience at the University of Washington Hospital may be instructive. A TEG is

22

BRUCE

routinely used with all CPB surgery at this institution. Retrospective data indicate that its routine use has reduced utilization of blood products by 30%. Even more important, by requiring that the TEG be normalized before a patient is returned to the operating room (unless the patient is hemodynamically unstable), the reoperation rate has been reduced from 5.7% to 1.2%.

D. SPIESS

CONCLUSION

There is much yet to learn about coagulation and bypass. Patient variability is the norm and certain populations are probably at risk for hemorrhage and certain for early graft thrombosis. Means of identifying those persons at risk so that therapy can be properly targeted must be improved.

REFERENCES

1. Goodnough LT, Johnston MF, Shah T, et al: A twoinstitution study of transfusion practices in 78 consecutive adult elective open-heart procedures. Am J Clin Path01 91:468-472,1989 2. McCarthy PM, Popovsky MA, Schaff HV, et al: Effects of blood conservation efforts in cardiac operations at the Mayo Clinic. Mayo Clinic Proc 63:225-229,1988 3. Lane TA: Blood salvage in cardiovascular surgery, in Stehling L (ed): Perioperative Autologous Transfusion. Arlington, VA, American Association of Blood Banks, 1991, p 71 4. Goodnough LT, Johnston MFM, Toy PTCY, et al: The variability of transfusion practice in coronary artery bypass surgery. JAMA 265:86-90,199l 5. Donahue JG, Munoz A, Ness PM, et al: The declining risk of post-transfusion hepatitis C virus infection. N Engl J Med 327:369373,1992 6. Sirchia G, Giovanetti AM, Parravicini A, et al: Prospective evaluation of posttransfusion hepatitis. Transfusion 31:299-302,199l 7. Binder LS, Ginsberg V, Harmel MH: A six-year study of incompatible blood transfusions. Surg Gynecol Obstet 108:19-34,1959 8. Murphy WG, McClelland DB: Deceptively low morbidity from failure to practice safe blood transfusion: An analysis of serious blood transfusion errors. VOX Sang 57~59-62, 1989 9. Gaydos JC, Cowan DN, Polk AJ, et al: Blood typing errors on U.S. Army identification cards and tags. Milit Med 153:618-620, 1988 10. Fernandez MC, Gottlieb M, Menitove JE: Blood transfusions and postoperative infection in orthopedic patients. Transfusion 32:318-322,1992 11. Warner MA, Faust RJ: Risks of transfusion. Anesthesiology Clinics of North America: Hemorrhagic Disorders. 8:501-517,199O 12. Love TR, Hendren WG, O’Keefe DD, et al: Transfusion of predonated autologous blood in elective cardiac surgery. Ann Thorac Surg 43:508-512,1987 13. Gwings DV, Kruskall MS, Thurer BL, et ah Autologous blood donation prior to elective cardiac surgery: Safety and effect on subsequent blood use. JAMA 262:1963-1968,1989 14. Spiess BD, Sassetti R, McCarthy RS, et al: Autologous blood donation: Hemodynamics in a high-risk patient population. Transfusion 32:17-22,1992 15. Britton LW, Eastlund DT, Dziuban SW, et al: Predonated autologous blood use in elective cardiac surgery. Ann Thorac Surg 47~529-532, 1989 16. Cove H, Matloff J, Sacks HJ, et al: Autologous blood transfusion in coronary artery bypass surgery. Transfusion 16:245248,1976 17. Robblee JA: Pro: Blood should be harvested immediately before cardiopulmonary bypass and infused after protamine reversal to decrease blood loss following cardiopulmonary bypass. J Cardiothorac Anesth 4:519-521,199O

18. Starr NJ: Con: Blood should not be harvested immediately before cardiopulmonary bypass and infused after protamine reversal to decrease blood loss following cardiopulmonary bypass. J Cardiothorac Anesth 4:522-525,199O 19. Salzman EW, Weinstein MJ, Weintraub RM, et al: Treatment with desmopressin acetate to reduce blood loss after cardiac surgery. N Engl J Med 314:1402-1406,1986 20. Mongan PD, Hosking MP: The role of desmopressin acetate in patients undergoing coronary artery bypass surgery. Anesthesiology 77:38-46,1992 21. Horrow JC, Hlavacek J, Strong MD, et al: Prophylactic tranexamic acid decreases bleeding after cardiac operation. J Thorac Cardiovasc Surg 99:70-74,199O 22. van Oeveren W, Harder MP, Roozendal KJ, et al: Aprotinin protects platelets against the initial effect of cardiopulmonary bypass. J Thorac Cardiovasc Surg 99:788-796,199O 23. Lu H, Soria C, Commin PL, et al: Hemostasis in patients undergoing extracorporeal circulation: The effect of aprotinin (Trasylol). Thromb Haemost 66:633-637,199l 24. Nagaoka H, Innami R, Murayama F, et al: Effects of aprotinin on prostaglandin and platelet function in open heart surgery. J Cardiovasc Surg 32:31-37,199l 25. Royston D: Aprotinin prevents bleeding and has effects on platelets and fibrinolysis. J Cardiothorac Vast Anesth 6:18-23,199l 26. Dietrich W, Spannagl M, Jochum M, et al: Influence of high-dose aprotinin treatment on blood loss and coagulation patterns in patients undergoing myocardial revascularization. Anesthesiology 73:1119-1126,199O 27. van Oeveren W, Jansen NJ, Bidstrup BP, et al: Effects of aprotinin on hemostatic mechanisms during cardiopulmonary bypass. Ann Thorac Surg 44:640-645,1987 28. Lavee J, Savion N, SmolinsQ A, et al: Platelet protection by aprotinin in cardiopulmonary bypass: Electron microscopic study. Ann Thorac Surg 53:477-481,1992 29. Marx G, Pokar H, Reuter H, et al: The effects of aprotinin on hemostatic function during cardiac surgery. J Cardiothorac Vast Anesth 51467-474, 1991 30. Tanaka K, Takao M, Yada I, et al: Alterations in coagulation and fibrinolysis associated with cardiopulmonary bypass during open heart surgery. J Cardiothorac Anesth 3:181-188,1989 31. Dietrich W, Barankay A, Hahnel CH, et al: High-dose aprotinin in cardiac surgery: Three years’ experience in 1,784 patients. J Cardiothorac Vast Anesth 6:324-327, 1992 32. Cosgrove DM, Heric B, Lytle BW, et al: Aprotinin therapy for reoperative myocardial revascularization: A placebo-controlled study. Ann Thorac Surg 54:1031-1038,1992 33. Bidstrup BP, Underwood SR, Sapsford RN: Effect of aprotinin (Trasylol) on aorta-coronary bypass graft patency. J Thorac Cardiovasc Surg 105:147-152,1993