Prevalence of Platelet Transfusion Reactions Before and After Implementation of Leukocyte-Depleted Platelet Concentrates by Filtration

zyxwvu zyxwvutsr Original Paper Vox Sang 1993;65:103-107 Lawrence Tim Goodnougha, James Riddell N a Hillard Lazarma Theresa L. Chafelb Greg Prince Donna Hendrix Roslyn Yomtovian a a Departments of Medicine and Pathology, Case Western Reserve University, and the Ireland Cancer Center of University Hospitals of Cleveland, Ohio, USA zy zy Prevalence of Platelet Transfusion Reactions Before and After Implementation of leukocyte-Depleted Platelet Concentrates by Filtration ................................................................................................. Abstract To determine the impact of platelet leukodepletion by filtration on the overall prevalence of reported transfusion reactions associated with platelet concentrates, we audited platelet transfusion reactions after infusion of platelet concentrates reported at University Hospitals of Cleveland over 6 months before (interval 1, July 1,1989to December 31,1989) and after (interval 2, July 1,1990 to December 31,1990) implementation of the Pall PL 50 filter on our adult Hematology-Oncology inpatient unit (Division 60). Thirty-two (1.7%) of 1,901 random, pooled platelet transfusion events resulted in blood bank transfusion reaction workups in interval 1, compared to 90 (5.3%) of 1,704 in interval 2 (p<O.OOl). The Division 60 service accounted for more of our hospital-wide platelet reactions after implementation of the filter in interval 2 (84%) than before filtration in interval 1 (42%), p=0.002. The prevalence of reaction workups for Division 60 was 0.6% for intervall, compared to 4.3% €orinterval 2 (p<O.OOl). No differences were found between interval 1and interval 2 for the rate of discontinuation of platelet transfusion (36 vs. 32%, p = 0.14), rate of premedication for platelet transfusion (72 vs. 65%, p = O h ) , percentage of direct antiglobulin test-positive reactions (17 vs. 5.4%, p = 0.09), percentage showing icteric/hemoiyzed serum (15 vs. 4.4%, p = 0.09), or reactions believed to be due to red blood cell incompatibility (8.8 vs. 1.1%, p = 0.1). We conclude that the use of expensive platelet filtration devices has not decreased the morbidity of random, pooled platelet transfusions, nor the prevalence of timeconsuming blood bank evaluation of platelet transfusion reactions in this setting. zyxw Supported in part by the Transfusion Medicine Academic Award (K07-HL01625) from the National Heart, Lung, and Blood Institute of the NIH, and by a grant from the NIH, NCI, USPHS No. P30CA43703. Received: August 12,1992 Accepted: Dec. 18, 1992 Lawrence Tim Goodnough, MD Division of Laboratory Medicine Washington University School of Medicine Box 8118,660 South Euclid Avenue St. Louis, MO 63110 (USA) zyxw D 1993 S.Karger AG, Basel CW2-Y007/93/0652-0103 $?.75/0 Introduction zyxwvu Table 1. Criteria for blood bank workup of platelet-associated transfusion reactions zyxwvutsrq zyxwvut zyxw Recent advances in the technology of leukocyte depletion have made it possible to achieve significant reductions in leukocyte contamination ob blood components. Leukocyte depletion of platelets by centrifugal methods [1,2], filtration [3,4], and modern apheresis technology [5]techniques are now available for the attempted treatment or prevention of leukocyte-mediated transfusion reactions, including platelet alloimmunization [6], nonhemolytic febrile transfusion reactions [7], transfusion-associated graft-versus-host disease [8]; transmission of cytomegalovirus [9], and transfusion-induced immunomodulation [lo]. Platelet-associated transfusion reactions occur commonly; a successful intervention to reduce the prevalence of this complication has important consequences in reducing the morbidity and expense related to platelet transfusion reactions and their laboratory evaluation. We recently instituted routine use of the commercially available Pall PL 50 filter to leukocyte-depleted random-donor platelet concentrates. To investigate whether leukodepletion of platelets reduced platelet concentrateassociated transfusion reactions, we reviewed platelet transfusion reaction evaluations on 1specialized inpatient unit at the Ireland Cancer Center of the University Hospitals of Cleveland before and after Implementation of the Pall PL 50 filter. 1 Patient temperature increase' 2 Chills and/or rigors 3 Diastolic blood pressure change equal to or greater than 20 mm Hg 4 Respiratory distress 5 Red or brown discoloration of urine 6 Progressive signs and symptoms of shock 7 Backpain 8 Significant change in clinical status 9 Urticaria1 reaction (hives, itching, rash) ' Platelet transfusion could be continued if temperature elevation <l.O°C or if reaction was urticaria1 in nature. Table 2. Patients undergoing in-patient treatment' at the Ireland Cancer Center Interval 1 n Bone marrow Transplantation autologous allogeneic Chemotherapy only Supportive Care Total 78 17 46 36 72 8 46 47 177 173 ' Methods With admission diagnoses of leukemia, myelodysplasia, aplastic anemia, or solid tumor (including lymphoma). Use of the Pall PL 50 filter (Pall Biomedical Products Corporation, East Hills, N.Y., USA) was implemented on March l, 1990, to provide leukodepleted random-donor platelet concentrates as the primary source of platelet transfusion support for adult patients undergoing inpatient therapy on 1specialized inpatient unit (Division 60) at the Ireland Cancer Center of the University Hospitals of Cleveland. Nursing personnel were trained to implement the filter at the bedside during platelet transfusion, according to the package insert instructions. Platelet-associated transfusion reactions were reviewed over two &month intervals: before (interval 1, July 1,1989 to December 31,1989) and after (interval 2, July 1,1990 to December 31,1990) implementation of the Pall PL 50 filter. Platelet concentrates during both audit intervals consisted of 6 pooled random-donor units purchased from one regional blood center source. No changesin administrative policy were made between the two audit intervals with respect to distribution or inventory management that would affect platelet age at time of transfusion. Platelet-associated transfusion reactions were defined based on recommendations by the American Association of Blood Bank Standards for Blood Banks and Transfusion Services [11]. Reactions were reported for blood bank evaluation by criteria shown in table 1, according to a written description in the Division 60 nursing procedure manual, a document in use for several years on the unit. For any platelet-associated transfusion reaction 104 Interval 2 n (e.g. fever, hives) patientswere routinelypremedicatedwith acetaminophen 650 mg and diphenhydramine 50 mg PO, 15-30 min before subsequent platelet transfusions. Hydrocortisone 25-50 mg i.v. was added at the discretion of the physician. According to the procedure manual, platelet transfusions were discontinued and evaluation requested for symptoms and signs including temperature elevation 239"C, rigors, change in patient vital signs, or any change in patient clinical condition regarded as significant (table 1). Temperature elevations and change in patient vital signs considered due to factors independent of platelet transfusion (infection and/or sepsis), drug administration (e.g. amphotericin B), volume overload, or red cell transfusion did not require or routinely result in a platelet-associated transfusion reaction evaluation. Under the guidelines of the procedure manual, platelet transfusions could be continued if temperature elevation did not exceed 1°C or if the reactions were felt to be urticarial in nature. Patient treatment categories for intervals 1and 2 were not different and are characterized in table 2. Quality control evaluation for leukopoor platelets by centrifugation for this period indicated that t he Goodnough/Riddell/Lazarus/Chafel/ Prince/Hendrix/Yomtovian Leukocyte Depletion of Platelets zyxwvu zyxwvutsrqpo zyxwvutsrqpon zyxwvutsr mean leukocyte content for 6 random, pooled platelet concentrates was 1.2 x lo8.After Pall PL 50 filtration, the mean leukocyte content in leukodepleted platelets was less than 0.1 X lo7(below the range of linearity by routine electronic counting, Sysmex Instrument, TOA Corporation. Kobe, Japan), indicating a 2 log depletion of leukocytes. During this interval, red cell transfusions were leukodepleted with Pall RC-so filters forpatients whohad twopreviouslydocumented febrile episodes associated with red cell transfusion. Statistical methods were carried out using the independent group t test to determine statistical significance within variables. Results During interval 1,36platelet-associated transfusion reactions at University Hospitals were reported. Four of these were associated with single donor apheresis platelet products and were excluded from subsequent analysis; 32 (1.7%) of 1,901random-donor platelet concentrate transfusion events resulted in blood bank workups in interval 1 (table 3). Of these, 10 (0.6%) of 1,780 platelet concentrates were transfused to Division 60 patients. During audit interval 2, 91 platelet-associated transfusion reactions at UHC were reported. One of these was associated with a single donor apheresis platelet product and was excluded from subsequent analysis; 90 (5.3%) of 1,704platelet concentrate transfusion events resulted in blood bank evaluations in interval 2 (p <0.001 compared to interval 1). Of these, 71 (4.3%) of 1,620platelet concentrates were transfused to Division 60 patients (p<O.Ol compared to interval 1). Thus, the Division 60 patients accounted for more platelet-associated transfusion reaction workups in interval 2 [71 (84%) of 901 than in interval 1[lo (42%) of 321,p = 0.002. As shown in table 3, no differences were found between interval 1and interval 2 when analyzed for clinical characteristics such as patient gender, hematologic versus solid tumor malignancy, the rate of discontinuation of platelet transfusions, or the rate of patient premedication. Similarly, no differences were found in evaluation results when analyzed for evidence of red blood cell incompatibility: the percentage of positive direct antiglobulin tests, percentage showing icteric/hemolyzed serum, and workup conclusions were not different between the two intervals. Thus, the prevalence of nonhemolytic plateletassociated transfusion reactions was 91.2% in interval 1 and 98.9% in interval 2 (p = 0.10). To determine the effect of leukodepletion of platelet concentrates in patients with a previous history of platelet-associated transfusion reaction, a group of 12 intensively transfused Division 60 patients in the 2 months before and the 2 months following routine application of the Table 3. Platelet concentrate-associated transfusion reactions' Interval 1 Interval 2 P % Yo Prevalence of transfusion reactions University Hospitals 1.7 0.6 adult, inpatient Oncology Unit Clinical characteristics sex (female) 50 69 hematologic malignancy platelet transfusion 36 discontinued 72 patient premedicated Blood bank evaluation results 17 positive direct antiglobulin test 15 serum icterichemolyzed interpretation of red blood 8.8 cell incompatibility 5.3 4.3 <0.001 <0.001 49 51 32 0.89 0.08 0.14 65 0.60 5.4 0.09 4.4 1.1 0.09 0.10 See Methods. Pall PL 50 filter was analyzed further. Patients reviewed had a prior history of one or more reactions to unfiltered platelet concentrates. In 304 platelet concentrate transfusion events within this cohort, platelet-associated transfusion reactions were documented in 20 (13%) of the 152 unfiltered compared to 15 (loo/) of the 152 filtered random, pooled platelet concentrate transfusions (p = 0.534). To determine whether ABO platelet transfusion mismatches could have increased the likelihood of plateletassociated transfusion reactions, we subsequently reviewed all such events over a 6-month interval. Of 111 studied, the prevalence of reactions was 86 (78%) with ABO-matched platelets and 25 (23%) with ABO-mismatched platelets. This compares with frequencies of ABO-matched and ABO-mismatched platelet transfusions of 72 and 28%, respectively, in our Transfusion Service in a survey of 481 platelet transfusions during a representative month. These data do not support an effect of ABO match or ABO mismatch on the prevalence of platelet-associated transfusion reactions in our platelet transfusion recipients. 105 Discussion zyxwvutsrqpo The potential benefits of leukocyte depletion at the time of blood component transfusion must be weighed against the added costs for rendering platelets leukocyte depleted. For example, the efficacy of leukocyte depletion in delaying or preventing platelet alloimmunization has been reported previously [4,6,12]. Despite these results, studies to date have not demonstrated clinical patient benefits such as reduced platelet and/or red blood cell transfusion needs or reduced hemorrhagic complications in patients undergoing intensive chemotherapy programs; for these reasons, prevention of alloimmunization against platelets is currently the focus of a multicenter, randomized trial in which leukocyte depletion by filtration [13] is to be compared to other technologic interventions (single donor apheresis platelets [59] and ultraviolet platelet irradiation [14]) and standard platelet concentrates [15]. Similarly, the use of leukocyte-depleted platelets has been recommended as a method to prevent or significantly diminish recurrent febrile reactions associated with platelet transfusions [13]; a recent report on the use of leukocyte-depleted platelets by filtration of platelet concentrates in preventing febrile transfusion reactions showed limited efficacy (20 vs. 14% in unmodified vs. filtered platelet concentrates, respectively) in patients who had previously documented febrile transfusion reactions to unmodified platelets [16]. The patients in this report and in our study, moreover, are commonly alloimmunized, so that filtration to diminish febrile reactions would not be expected to modify the posttransfusion platelet increment [16-181. This has led to speculation that in this setting, leukocyte reduction of platelets may be of value only in patients who have febrile reactions who have satisfactory posttransfusion platelet counts [17-201. Since platelet-associated transfusion reactions that are not leukocyte related, such as allergic, hemolytic, or bacterial, occur in this setting, Oncology Services use guidelines [ll] that determine when a platelet transfusion reaction requires evaluation. These platelet-associated transfusion reactions that are not ameliorated by leukodepletion must therefore be included in an analysis of filtration efficacy under “intention to treat”. The additional costs of platelet filtration must be compared to the potential benefits (reduced patient morbidity, fewer discontinued platelet transfusions, reduced transfusion service workload for evaluation of reactions). Alternatively, the routine practice of testing for HLA antibodies in patients with platelet-associated transfusion reactions may identify a subgroup of patients for whom HLA-matched 106 platelets, rather than leukodepletion by filtration, may be a more effective intervention. In this report, we found no benefit to a policy in which leukocyte-depleted platelet concentrates were used routinely in an attempt to prevent platelet-associated transfusion reactions in the transfusion support of adult patients undergoing inpatient treatment for complex hematologic and oncologic disorders. The prevalence of platelet-associated transfusion reactions was not reduced; the blood bank transfusion service workup results of these reactions were not found to be different; and the percentage of platelet concentrate transfusion events that were successfully completed was no different before and after implementation of this policy. In addition, leukodepletion of platelet concentrates in 12 patients with previously documented platelet-associated transfusion reactions did not reduce the prevalence of reactions during 152 subsequent platelet transfusions in this group, in which reactions reoccurred in only 13% of the platelet transfusion events in patients receiving unfiltered platelets. These results confirm a previously published report indicating limited efficacy in leukodepletion by filtration in patients who were clinically alloimmunized [16]. Our study therefore demonstrated that the routine use of platelet concentrate leukodepletion by expensive filtration devices at the time of transfusion resulted in no benefit, when defined by the morbidity and expense associated with platelet-associated reactions to pooled platelet concentrates, along with medical and laboratory resource utilization associated with platelet-associated transfusion reaction evaluations. While reactions due to causes other than those related to transfusion (e.g., amphotericin B) were not excluded, the prevalence of those over two 6-month audit intervals would be expected to be no different. Similarly, we found no evidence that other factors such as ABO-mismatched platelet transfusions were associated with prevalence of platelet-associated transfusion reactions. We conclude that randomized, prospective trials of leukocyte-depleted platelet products are needed before leukocyte depletion technology is used routinely at the time of transfusion of platelet-pooled concentrates for the purpose of preventing platelet-associated transfusion reactions. Filtration of platelets to generate a leukodepleted product at the time of blood donation [21] may better address leukocyte-specific factors that are important in the pathogenesis of platelet-associated transfusion reactions. GoodnougWRiddelllLazarusiChafeV Prince/Hendrix/Yomtovian Leukocyte Depletion of Platelets zy References 1 Schiffer CA, Patten E, Reilly J , Patel S: Effec- 2 3 4 5 6 7 z zyxwvutsrqpo ................................................................................................................................................... tive leukocyte removal from platelet preparations by centrifugation in a new pooling bag. Transfusion, 1987;27:162-164. Kalmin ND, Ore11JE, Villarreal IG: An effective method for the preparation of leukocytepoor platelets. Transfusion 1987;27:281-283. Sniciewski I, O’Donnell MR, Nowicki B, Hill LR: Prevention of refractoriness and HLA alloimmunization using filtered blood products. Blood 1988;71:1402-1407. van Marwijk Kooy M, van Prooijen HC, Moes M, Bosma-Stants I, Akkerman JWN: Use of leukocyte-depleted platelet concentrates for the prevention of refractoriness and primary HLA alloimmunization: A prospective randomized trial. Blood 1991;77:201-205. Anderson KC, Gorgone BC, Wahlers E , Cook J, Barrett B, Anderson J: Preparation and utilization of leukocyte pool apheresis platelets. Transfus Sci, in press. Claus FHJ, Smeenk RJT, Schmidt R, Van Steenbrugge GJ, Eernisse JG: Alloimmunization against the MHC antigens after platelet transfusions is due to contaminating leukocytes in the platelet suspension. Exp Hematol 1981;9:84-89. Perkins HA, Payne R, Ferguson J, Wood M: Non-hemolytic febrile transfusion reactions. Vox Sang 1966;11:578-600. 8 Anderson KC, Weinstein HJ: Transfusion-associated graft-versus-host disease. N Engl J Med 1990;323:315-321. 9 Bowden RA, Slichter SJ, Sayers MH, Mori M, Cays MJ; Meyers .ID: Use of leukocyte-depleted platelets and cytomegalovirus-seronegative red blood cells for prevention of primary cytomegalovirus infection after marrow transplantation. Blood 1991;78:246-250. 10 Merryman HT Transfusion-induced alloimmunization and immunosuppression and the effects of leukocyte depletion. Transfus Med Rev 1989;3:180-193. 11 Widmann FK: American Association of Blood Bank Standards for Blood Banks and Transfusion Services, ed 14. Arlington, American Association of Blood Banks, 1991, p 41. 12 Eernisse JG, Brand A: Prevention of platelet refractoriness due to HLA antibodies by administration of leukocyte-poor blood components. Exp Hematol 1981;9:77-83. 13 Kickler TS, Bell W, Ness PM, Drew H , Pall D: Depletion of white cells from platelet concentrates with a new absorption filter. Transfusion 1989;29:411-414. 14 Andrew G, Boccaccio C, Lecrubier C, Fretault J, Courseget J , GeGuen JP, et al: Ultraviolet irradiation of platelet concentrates. Feasibility in transfusion practice. Transfusion 1990;30:401-406. 15 Schiffer CA: Prevention of alloimmunization against platelets. Blood 1991;77:1 4 . 16 Mangano MM, Chambers CA, Kruskall MS: Limited efficacy of leukopoor platelets for prevention of febrile transfusion reactions. Am J Clin Pathol 1991;95:733-738. 17 Schiffer CA, Patten E, Reilly J, Patel S: Effective leukocyte removal from platelet preparations by centrifugation in a new pooling bag. Transfusion 1990;30:30-33. 18 Brand A, Claas FHJ, Voogt PJ. Wasser MNJM, Eernisse JG: Alloimmunization after leukocyte-reduced multiple random donor platelet transfusions. Vox Sang 198834: 160166. 19 Mintz PD: Febrile reactions to platelet transfusions. Am J Clin Pathol 1991;95:609-611. 20 Lane TA, Anderson KC, Goodnough LT, et al: Leukocyte reduction in blood component therapy. Ann Intern Med 1992;117:152-161. 21 van Marwijk KM, van Prooijen HC, Borghuis L, Moes M, Akkerman WN: A method to prepare white cell-poor platelet concentrates with optimal preservation of platelet viability. Transfusion 1990:30:34-38. zyxwvutsr 107