sign in sign up
<<
Home , Transfusion medicine

Use of in transfusion

DANIEL B. BRUBAKER, DO

Platelet transfusions are bene- mittee on Hospital Transfusion Practices pub- ficial to treat or prevent bleeding in the lished guidelines for the improvement of thrombocytopenic patient. They are fre- transfusion practices.4 At the same time, the quently used in patients with hypoplastic Joint Commission on Accreditation of Hospi- bone marrow, in cardiovascular tals5 mandated an evaluation of the appropri- patients, and those involved in trauma. Be- ateness of all blood component transfusions cause platelets have short survival, large that their inspections closely evaluate. More numbers of units are required. recently, a National Institutes of Health (NIH) Also, platelet transfusions are expensive Consensus Conference convened to address vari- and not without complications. They can ous issues in platelet . 6 The wide dis- cause alloimmunization, provoke transfu- semination that the AABB guidelines and con- sion reactions, or transmit infectious dis- sensus report have received will lead to im- ease, of which hepatitis C (non-A, non-B proved platelet use. However, there is a con- hepatitis) is of greatest concern. Therefore, tinuing lack of consensus in many areas of documented indications and close moni- platelet therapy because of a lack of under- toring of the transfused platelets are nec- standing of platelet function. essary. Platelets are cell fragments with disk-like structures that prevent bleeding. Their func- tion in hemostasis is to adhere to the suben- The availability of platelet concentrate trans- dothelial collagen microfibrils during vascu- fusions in the 1960s led to a considerable re- lar injury, secrete adhesive proteins, activate duction in death due to hemorrhage in leuke- other platelets to aggregate, and induce the mia patients. From 1971 to 1980, platelet us- intrinsic pathway. age increased nearly 600% as new indications Normal adults have 150,000 to 450,000 plate- were explored. 2 However, recent audits have lets per cubic millimeter in the circulation. If shown that 40% of all platelet transfusions either the number of platelets is reduced or were used inappropriately. 3 In 1985, the Ameri- the function is abnormal, bleeding may result. can Association of Blood Banks (AABB) Com- In vivo platelet function can be assessed by measuring the in vivo bleeding time. Patients are not very compliant with the test, and there is a large coefficient of variation with a nor- Dr Brubaker is assistant professor, UCLA School of Medi- mal bleeding time between 2 and minutes; cine, Los Angeles, and head of transfusion medicine labo- 8 ratory and diagnostic , department of pa- therefore, a standardized test of platelet func- thology, Harbor-UCLA Medical Center, County of Los tion is needed. Bleeding, however, does not usu- Angeles, Torrance. ally occur unless the bleeding time is twice normal. In general, a platelet count of 50,000/ The views expressed herein are those of the authors and do not necessarily reflect the views of the county ,LI, is adequate for hemostasis. Spontaneous of Los Angeles Department of Health and Human Serv- hemorrhage does not occur unless the platelet ices or the University of California. count is less than 10,000/RL. Thrombocytopenic bleeding is unique in Reprint requests to Daniel B. Brubaker, DO, depart- ment of , Harbor-UCLA Medical Center, 1000 that bleeding arises from small vessels and cap- W Carson St, Torrance, CA 90509. illaries of the skin and mucosa. The severest

Clinical practice • Brubaker JAOA • Vol 89 • No 12 • December 1989 • 1553

Table 1 such as induction chemotherapy for acute leu- Indications for Platelet Transfusions kemia, but not to those with chronic myelo- suppression, that is, aplastic .6 Need for Diagnosis platelet transfusions Prophylactic platelet transfusions are given to decrease patient morbidity with bleeding epi- • due to Useful to treat or pre- sodes. They do not decrease mortality due to decreased production: vent bleeding Leukemia—postinduction of bleeding and require twice the number of trans- chemotherapy fused platelets for therapeutic results. 89 This Bone marrow transplantation Aplastic anemia issue will remain controversial until reliable, Myelofibrosis/chronic myelog- well-controlled studies solve the dilemma. enous leukemia If surgery or invasive procedures such as lum- • Dilutional thrombocytopenia Needed when platelet bar puncture or liver biopsy are necessary, (as in massive transfusion) count < 50,000 or in platelet transfusions are required to raise the excessive bleeding platelet count above 50,000/11L. Platelets are • Thrombocytopenia due to in- Not indicated unless not required if the platelet count is above creased destruction: bleeding is life- Idiopathic thrombocytopenic threatenting 50,000/p L, and the bleeding time is less than purpura twice the upper limit of normal (12 minutes).6 Thrombotic thrombocytopenic purpura Adjunctive therapy may also be beneficial Hemolytic uremic syndrome and may decrease the number of platelets trans- • Platelet function abnormality Only after corrected fused. Epsilon aminocaproic acid (Amicar), an (as in uremia and plasma functional defect antifibrinolytic agent, may decrease sponta- cell dyscrasias) neous bleeding. 10 A preparation of conjugated estrogens (Premarin) can also be used in women to suppress menstruation and menor- form of thrombocytopenic bleeding is intra- rhagia. Vitamin K is useful to prevent coagu- cerebral bleeding, which may be rapidly fatal. lation factor deficiency, particularly in patients Factors that exacerbate bleeding in thrombocy- on antibiotics who are not eating. 11 Bleeding topenic patients include a rapid fall in plate- in uremic patients responds to let count, fever, sepsis, certain medication such or desmopressin acetate (DDAVP), so that the as aspirin, vomiting, uremia, concomitant co- use of platelet transfusions can be eliminated agulation factor deficiency, and trauma. in such cases. Dilutional thrombocytopenia from massive Indications for platelet transfusions blood transfusions, as in trauma, does not pre- Platelet transfusions are indicated for bleed- sent a problem for two reasons: (1) microvas- ing patients with thrombocytopenia due to de- cular bleeding does not develop; and (2) plate- creased marrow production (Table 1). A di- let counts usually do not drop below 50,000/ lemma occurs in the prophylactic use of plate- pi.L. More than 30 units of blood must be trans- let transfusion. Some authorities 6 believe that fused before the platelet count drops below no prophylactic platelet transfusions are indi- 50,000/RL, or there must be coexisting dissemi- cated, whereas others give platelets routinely nated intravascular (DIC). In the when the patients count drops below 20,000/ majority of patients receiving 10 to 20 units 1.11. This cutoff number originated from the of blood, bleeding problems do not develop; how- 1960s and 1970s when most patients were ever, a platelet count should be taken after treated for fever with aspirin. Because ace- every 12 to 15 units of blood, with platelet trans- taminophen is currently used, the suggested fusions reserved for thrombocytopenic patients level may be lowered to 10,000 7 or even 5000/ and patients with microvascular bleeding.2 LAL (I recommend 10,000/g,L). Prophylactic plate- Patients who have thrombocytopenia due to let transfusions have been recommended only peripheral destruction, such as in idiopathic to patients with transient myelosuppression (immune) thrombocytopenic purpura (ITP) or

1554 • JAOA • Vol 89 • No 12 • December 1989 Clinical practice • Brubaker thrombotic thrombocytopenic purpura (TTP) Platelet products, dosage, and do not benefit from platelet transfusions.13 administration Transfusion of platelets in TTP accentuates Platelets are prepared and concentrated by two the and can cause rapid demise. Be- methods. Random donor or unrelated platelet cause hemolytic uremic syndrome (HUS) is concentrates are prepared at the community very similar to TTP, results would be similar. blood center or Red Cross from individual units Idiopathic thrombocytopenic purpura rapidly of ." The platelet concentrates destroys platelets because of an autoimmune have about 0.5-1.0 x 1011 platelets in 50 ± antibody toward platelets. The only indication 10 mL of plasma. These units are frequently for platelet transfusions in ITP and TTP pa- combined as 6-, 8-, 10-, or 12-unit pools. tients is in life-threatening bleeding such as The second platelet product is prepared from CNS bleed. 12 Disseminated intravascular co- blood drawn from single donors connected to agulation is more controversial. In my opin- an instrument, such as Fenwal CS- ion, the underlying cause for DIC should be 3000, Cobe Spectrum III, or Haemonetics S-30 treated and the patient supported with plate- instruments. The platelets are separated from let and cryoprecipitate transfusions. the rest of the blood elements and concentrated presents further problems to give approximately 5 x 10 11 platelets in one in the use of platelet transfusions. One study14 bag. Such products are called single donor plate- has shown no correlation between the platelet let concentrates, single donor units, or count and intraoperative or postoperative bleed- units, and each is equivalent ing in patients undergoing cardiac surgery. to eight random donor units. The advantages However, other studies15,16 have shown signifi- of this product is one-donor exposure to HLA cant platelet functional defects and significant antigens and only one risk of hepatitis (hepa- postoperative bleeding without platelet trans- titis B or hepatitis C [non-A, non-B]) or ac- fusions. Functional platelet defects result from quired immunodeficiency syndrome (AIDS) ver- cardiopulmonary bypass. 16 The results of yet sus ten risks in the random donor units. 13 The another trial" suggest that desmopression problem is getting enough donors to meet the may cut platelet transfusions in half in pa- need. tients undergoing cardiopulmonary bypass. All platelet products are stored at room tem- However, many cardiovascular surgeons have perature and constantly rotated to have good not experienced these results. gaseous exchange and prevent aggregates. Cur- The effect of aspirin therapy in cardiac sur- rently, they can be stored for 5 days. Platelets gery patients is also controversial. One study18 should never be stored in a refrigerator. Both reported that patients undergoing cardiac sur- platelet products also contain WBCs, usually gery who took aspirin were not more likely about 1 x 109. to bleed than those who did not take aspirin. The number of units of platelet concentrates Other studies 19 showed prolonged, profuse to be transfused will depend on the starting bleeding in these patients. In my experience platelet count and the desired platelet count. and that of others,20 as well as the experience One random donor unit is given per 10 kg of of several cardiovascular surgeons, platelet adult body weight. Each unit will raise the transfusions are indicated postoperatively in platelet count in an adult by 5000 to 10,000/ cardiac surgery patients until proof shows oth- 111. 13 A 10-unit pool should raise the platelet erwise. count in an adult to 70,000 to 100,000 above Other platelet function defects are found in baseline. One single-donor unit should give a uremia and the dysproteinemias, such as Wald- similar incremental rise in the platelet count. ensttoms macroglobulinemia. Platelet trans- The survival of transfused platelets is 5 to 9 fusions should not be given to uremic patients days; therefore, platelets are given more fre- until after dialysis or to the patient with dys- quently than RBCs. Patients should not re- proteinemia until after plasma exchange. quire more than 12 units every 3 to 4 days.

Clinical practice • Brubaker JAOA • Vol 89 • No 12 • December 1989 • 1555 Those who require at Almost any size needle can be used for plate- shorter intervals have increased platelet de- let administration; however, a 19-gauge works struction of their transfused platelets. best. Pooled platelets have a final volume of It is desirable to give platelets of the same 400 to 600 mL, which would take too long to ABO type as the patients; however, this is not run in with a small needle. Small-bore nee- always possible nor is it essential. ABO anti- dles are acceptable for pediatric patients. To gens on platelets are extremely weak; there- transfuse the maximum number of platelets, fore, platelets are not destroyed through natu- the bag and the tubing should be rinsed with ral ABO isohemagglutinins. There is some 0.9% saline solution. Gravity infusion is the RBC contamination in platelet concentrates simplest and most common means of admini- (usually less then 0.5 mL packed cell volume) stration. If an emergency situation requires which can be hemolyzed, but because the vol- intravenous push, Cutter Biologicals, a divi- ume is so small, the patient experiences no ad- sion of Miles Laboratories, has made an excel- verse effects. However, there is a risk in pedi- lent device to push platelets via a 50-mL syr- atric patients or women of child-bearing age, inge (22" Y Blood Component Set with auto- because Rh-positive RBCs could sensitize an matic check valves and filter, Code 812-51, Cut- Rh-negative patient to produce anti-RhoD.21 ter Biologicals, Berkeley, Calif). Platelets may This sensitization occurs infrequently in pa- also be given with an infusion pump. tients with leukemia because they are immuno- The rate of infusion is 5 to 10 mL per min- suppressed by their disease and by chemother- ute, but it may run as fast as tolerated except apy. However, I insist on giving group-specific in small children, in whom a slow infusion rate platelets to these two groups of patients. If Rh- is desirable (1 to 3 mL per minute). As with positive platelets are given to a child or to a all blood infusions, vital signs should be moni- woman of child-bearing age, Rh immune globu- tored before and after platelet transfusions. lin can be given. One dose will protect against the RBCs in 30 units of Rh-positive platelets. Monitoring platelet tranfusions If enough ABO mismatched platelets are trans- Platelet counts are essential in platelet ther- fused to a patient, a positive direct antiglobu- apy to assess the outcome. Certainly stoppage lin test may occur, but significant of bleeding after platelet transfusion serves as is rare.22 the best indication of outcome. However, pa- The requisite number of random donor plate- tients frequently are not bleeding at the time let concentrates are pooled in the of platelet transfusion so other methods are before transfusion. This provides a convenient required. A pretransfusion (pre-TFN) platelet and safe product for administration. The plate- count and a 1- to 2-hour posttransfusion (post- let concentrates can be stored at room tem- TFN) platelet count is desirable. 13,25 The cor- perature for 5 days, but after pooling, they rected count increment (CCI) is used to moni- must be administered within 4 hours (Food tor patients at Harbor–UCLA Medical Center. and Drug Administration requirement). 23 Sin- gle-donor (plateletpheresis) units can be stored in the blood bank and directly transfused. CCI = Post-TFN — pre-TFN platelet counts/platelet count in bag ( x 10 11) x Platelets must be administered through the 0.6 x body surface area (BSA) m2. standard 170 blood filter. Fresh platelets tend to aggregate and get trapped in the fil- ter; however, this usually is not a problem be- The platelet count is obtained from the cause the platelets are stored and rotated so pooled bag of platelets before transfusion. At as to prevent aggregation. In fact, stored plate- Harbor–UCLA, 0.6 is now used as the splenic lets can be given through a microaggregate pooling factor. Indium 111-labeled transfused filter; however, this procedure is not recom- platelets have been used to compare platelet mended for routine administration.24 counts with platelet survival, and it was found

1556 • JAOA • Vol 89 • No 12 • December 1989 Clinical practice • Brubaker that this factor must be used at 1- to 2-hour tion, and intravenous administration of counts, but not 18- to 24-hour counts. As an gamma globulins. 25 All these decrease trans- example, suppose the pre-TFN count is 10,000/ fused platelet survival. However, in many of tiL and the post-TFN count was 50,000/p.L. The these , it is not known when the trans- platelet count in the bag is 8.4 x 10 11 and the fused platelet counts decrease; in some cases, patients BSA is 1.6 m2, the CCI is: it may be within 2 hours of transfusion; in oth- ers 12 hours; and in still others, 24 to 36 hours. This is an area for further research. 40,000/5.0 [8.4 x 0.6] x 1.6 = 12,800. The method to evaluate appropriateness of all transfusions must be an ongoing, compre- If the CCI is less than 7500, increased de- hensive approach to quality assurance of clini- struction of the transfused platelets is occur- cal transfusion practice. The guidelines set in ring.25 For convenience, 10 units of random- Harbor–UCLA Medical Center include imme- donor platelets or one single donor plateletpher- diate consultation of the pathologist or trans- esis should give an incremental rise of more fusion laboratory with the request- than 30,000/RL. A smaller rise indicates in- ing physician when platelet transfusion is or- creased destruction. There are two reasons for dered under the following conditions: poor posttransfusion increments: antibody de- • Platelet count is > 100,000/RL and there is struction and pathophysiologic destruction.26 no platelet function defect. Platelets have strong HLA antigens and • Platelet count in a preoperative patient is platelet-specific antigens on their surface. > 50,0004L or the bleeding time is less than There is a strong correlation between the pres- 12 minutes. ence of HLA antibodies and poor transfused- • Platelet count in a patient with decreased platelet survival. 26 Because of the polymor- platelet production is > 20,000/KL and the phism of the HLA system, the contaminating patient is not actively bleeding. lymphocytes present in the platelet concen- • Patients tested bleeding time is > 12 minutes. trate and the large number of donors the pa- • Patient does not have a platelet count at the tient is exposed to, alloimmunization is a ma- time of request. jor problem. 27 Nearly 50% of leukemia patients • Patient who has received multiple platelet and between 80% and 100% of aplastic ane- transfusions does not have a platelet count mia patients become refractory to random- increment of at least 30,000/RL (adult receiv- donor platelet transfusions. 25 Many of these ing 10 random donor units). patients respond to HLA-matched platelets or • Special requests, for example, HLA-matched platelet crossmatch-compatible platelets col- platelets, white blood cell–depleted platelets. lected from single donors by apheresis.26 Use • Patient has a diagnosis of ITP, TTP, HUS, of platelets filtered through a cotton wool fil- or DIC. ter (Imugard IG 500 filter, Terumo Corp, Pis- • If the physician requests > 15 units or < 4 cataway, NJ) also prevents refractoriness.28 units (unless for a neonatal or pediatric pa- More recently, many are using un- tient) of random-donor pooled platelets. related non-HLA–matched single-donor plate- • Patients receiving more than 12 units within lets at the onset of platelet therapy to prevent 48 hours. alloimmunization. The problem is finding These same guidelines can be used by a hos- enough donors willing to spend 2 hours on an pital transfusion committee to review inappro- apheresis instrument. priate use in the hospital. There are several Pathophysiologic causes of poor posttrans- different approaches to monitoring blood prod- fusion increments include sepsis, , fe- ucts, and each institution will usually individu- ver, , splenomegaly, DIC, he- alize its own approach. However, the criteria patic disease, dysproteinemias, ITP, collagen as outlined here for platelet transfusions , acute renal transplant rejec- should be uniform.

Clinical practice • Brubaker JAOA • Vol 89 • No 12 • December 1989 • 1557 Table 2 edema, or perilaryngeal edema. 29 Usually, 25 Infectious Agents Transmitted to 50 mg of diphenhydramine hydrochloride by Platelet Concentrate Transfusions (Benadryl) ends the symptoms. More severe Human immunodeficient virus (AIDS) reactions such as perilaryngeal edema should Hepatits-B be treated with epinephrine and steroids. Hepatitis C (non-A. non-B hepatitis) Cytomegalic inclusion virus Hemolytic transfusion reactions from the con- Bacteria taminating RBCs in platelet concentrates are Brucellosis usually not noticed. The small quantity of Babesiasis Filariosis RBCs in platelet concentrates, usually < 0.5 Rocky Mountain spotted fever mL/unit or 5 mL/10 units, makes hemolysis Malaria Syphilis of donor RBCs insignificant. However, as has Toxoplasmosis already been mentioned, passive transfer of do- Trypanosomiasis nor anti-A or anti-B may cause the develop- Source: Gilcher RO: Adverse effects of transfusion mediated by infec- ment of a positive direct antiglobulin test re- tious agents. Plasma Titer Transfus Technol 1985;6:81-93. sult and, rarely, hemolysis as well in the re- cipient. Volume overload can be another complica- tion of platelet transfusions. Although each Complications of platelet transfusions platelet concentrate contains 50 to 70 mL, af- Because platelet concentrates also contain ter pooling the volume can be as high as 800 WBCs, RBCs, and plasma, recipients may be- mL. If volume is a problem for the patient, the come sensitized to platelet, HLA, granulocyte, platelet pool can be concentrated by centrifu- RBC, and plasma antigens. As previously men- gation in the blood bank with as much as 400 tioned, sensitization to HLA or platelet anti- to 500 mL of plasma expelled from the bag. gens or both leads to failure to respond to ran- One must keep in mind that these platelets dom-donor platelet transfusions (alloimmuni- must lie flat with no rotation for 1 to 2 hours zation).24,25 to resolve the preparation defect. The two most common transfusion reactions Lymphocytes in the platelets cause not only are febrile-nonhemolytic transfusion reaction febrile transfusion reactions and platelet al- and allergic transfusion reaction. 29 Febrile- loimmunization, but also graft-versus-host nonhemolytic transfusion reaction is charac- (GVH) disease. 3° This occurs in immunodefi- terized by chills, fever, and flulike symptoms. cient patients (with the exception of AIDS pa- It is caused by WBC antigen—preformed WBC tients). Posttransfusion GVH disease occurs antibody reaction. Acetaminophen prevents or mostly in patients with acute leukemia, diminishes the symptoms, and premedication Hodgkins disease, non-Hodgkins lymphomas, before future transfusions usually prevents and severe combined immunodeficiency. To pre- symptoms. If symptoms persist, WBC-depleted vent GVH in these patients, RBCs and plate- blood products must be given.29 At Harbor— lets are irradiated with 3000 rad at Harbor— UCLA Medical Center, these products are pre- UCLA Medical Center. pared in the blood bank by use of a cellulose- Any infectious disease that can be transmit- acetate filter (Erypur Filter, Organon-Ibknika ted by can be transmitted Corp, Durham, NC) for RBCs and cotton wool by platelet concentrates (Table 2). Because filter (Imugard IG 500) for platelets. I find that platelet transfusions expose the patient to mul- these filters are the most cost-effective means tiple donors, the risk goes up accordingly. Bac- of removing leukocytes with very little of the terial is of foremost concern with plate- technologists time. The technique is easy to lets because they are stored at room tempera- learn and does not require special equipment. ture.12 For this reason, all blood banks are re- Allergic reactions are characterized by ur- quired by Food and Drug Administration regu- ticaria, pruritus, skin rashes, periorbital lations to store platelet concentrates for 5 days.

1558 • JAOA • Vol 89 • No 12 • December 1989 Clinical practice • Brubaker Comment 15.Mohr R, Golan M, Martinowitz U, et al: Effect of cardiac operation on platelets. J Thorac Cardiouasc Surg 1986:92:434- In this paper, I have attempted to outline an 441. approach to the use of platelets in transfusion. 16.Mezzano D, Aranda E, Urzila J, et al: Changes in platelet Indications for transfusions should be docu- a-thromboglobulin, fibrinogen, albumin, 5-hydroxytryptamine, mented clearly, and monitoring of platelet ATP, and ADP during and after surgery with extracorporeal circulation in man. Am J Hematol 1986;22:133-142. transfusions is essential. Because transfusion 17.Salzman EW, Weinstein MJ, Weintraub RM, et al: Treat- reactions may occur, platelets should be admini- ment with desmopressin acetate to reduce blood loss after car- diac surgery: A double-blind randomized trial. N Engl J Med stered the least number of times possible to 1986;314:1402-1406. minimize risks and complications. 18.Weksler BB, Pett SB, Alonso D, et al: Differential inhibi- tion by aspirin of vascular and platelet prostaglandin synthe- sis in atherosclerotic patients. N Engl J Med 1983;308:800- 1. Hersh EM, Bodey GP, Nies BA, et al: Causes of death in 805. acute leukemia: A ten-year study of 414 patients from 1954- 1963. JAMA 1965;193:99-103. 19.Chesebro JH, Fuster V, Elveback LR, et al: Trial of com- bined Warfarin plus dipyridamole or aspirin therapy in pros- 2, Surgenor DM, Schmitzer SS: The nations blood resource. A thetic heart valve replacement: Danger of aspirin compared with summary report. NIH publication No. 85-2028. Bethesda, Md, dipyridamole. Am J Cardiol 1983;51:1537-1541. National Institutes of Health, March 1985. 20. Harker LA, Malpass TW, Branson HE, et al: Mechanism 3. Simpson MB: Prospective-concurrent audits and medical con- of abnormal bleeding in patients undergoing cardiopulmonary sultation for platelet transfusions. Transfusion 1987;27:192- 195. bypass: Acquired transient platelet dysfunction associated with selective a-granule release. Blood 1980;56:824-834. 4. Grindon AJ, Tomasulo PS, Bergin JJ , et al: The hospital trans- 21.Goldfinger D, McGinnis MH: Rh-incompatible platelet trans- fusion committee: Guidelines for improving practice. JAMA fusions-risks and consequences of sensitizing immunosup- 1985;253:540-543. pressed patients. N Engl J Med 1971;284:942-944. 5.Joint Commission on Accreditation of Hospitals. Accredita- 22.Pierce RN, Reich LM, Mayer K: Hemolysis following plate- tion Manual for Hospitals. Chicago, The Joint Commission, 1985. let transfusions from ABO-incompatible donors. Transfusion 6. Consensus Conference: Platelet transfusion therapy. JAMA 1980;25:60-62. 1987;257:1777-1780. 23. Code of Federal Regulations. 21CFR 640.26. Washington, 7. Aderka D, Praff G, Santo M, et al: Bleeding due to throm- DC, Office of the Federal Register, 1986, p 127. bocytopenia in acute leukemias and reevaluation of the pro- 24. Snyder EL, Hezzey A, Cooper-Smith M, et al: Effect of mi- phylactic platelet transfusion policy. Am J Med Sci croaggregate blood filtration on platelet concentrates in vitro. 1986;291:147-151. Transfusion 1981;21:427-434. 8. Soloman J, Bofenkamp T, Fahey JL, et al: Platelet prophy- 25. Brubaker DB, Duke JC, Romine M: Predictive value of en- laxis in acute non-lymphoblastic leukaemia. Lancet 1978;1:267. zyme-linked immunoassay platelet crossmatching for transfu- 9. Murphy S, Litwin S, Herring LM, et al: Indications for plate- sion of platelet concentrates to alloimmunized recipients. Am let transfusion in children with acute leukemia. Am J Hematol J Hematol 1987;24:375-387. 1982;12:347-356. 26.Brubaker DB, Romine M: Relationship of HLA and platelet- 10.Gardner FH, Helmer RE III: Aminocaproic acid: Use in con- reactive antibodies In alloimmunized patients refractory to plate- trol of hemorrhage in patients with amegakaryocytic throm- let therapy. Am J Hematol 1987;26:341-352. bocytopenia. JAMA 1980;243:35-37. 27. Duquesnoy RJ: Donor selection in platelet transfusion ther- 11.Alperin JB: Coagulopathy caused by vitamin K deficiency apy of alloimmunized thrombocytopenic patients, in Greenwalt in critically ill, hospitalized patients. JAMA 1987;258:1916- TJ, Jamieson GA (eds): The Blood Platelet in Transfusion Ther- 1919. apy. New York, AR Liss Inc, 1978, pp 229-243. 12.Reed RL II, Ciavarell D, Heimbach DM, et al: Prophylactic 28. Brubaker DB, Romine CM: The in vitro evaluation of two platelet administration during massive transfusion. A prospec- filters (Erypur and Imugard IG 500) for white cell-poor platelet tive, randomized, double-blind clinical study. Ann Surg concentrates. Transfusion 1988;28:383-385. 1986;203:40-48. 29.Brubaker DB: Immunologically mediated immediate adverse 13.Eisenstaedt: Blood component therapy in the treatment of effects of blood transfusions (allergic, febrile nonhemolytic, and platelet disorders. Semin Hematol 1986;23:1-17. noncardiogenic pulmonary edema). Plasma Vier Transfus Tech- 14.Simon TL, Akl BF, Murphy W: Controlled trial of routine nol 1985;6:19-31. administration of platelet concentrates in cardiopulmonary by- 30.Brubaker DB: Transfusion-associated graft-versus-host dis- pass surgery. Ann Thorac Surg 1984;37:359-364. ease. Human Pathology 1986;17:1085-1088.

Clinical practice • Brubaker JAOA • Vol 89 • No 12 • December 1989 • 1559