zyxwv zyxwvuts zyxwvut International Forum Vox Sang 199;62:5744 Should All Platelet Concentrates Issued Be Leukocyte-Poor? zyxwvutsrqpon zyxwvutsr H . W Reesink, U.E. Nydegger Alloimunization against HLA antigens and subsequent refractoriness to platelet transfusions is one of the major problems in transfusion medicine today. When leukocytes are not removed from cellular blood components (red cell concentrates, RBC, and platelet concentrates, PC), about half of the multitransfused patients will develop HLA antibodies [l-3). Refractoriness to platelet transfusion due to HLA antibodies will lead to a considerable bleeding risk for the patientsanddoeslimit the possibility for adequate treatment. Blood banks faced with refractory patients are requested to provide HLA-compatible PC which requires the maintenance of a large pool of HLA-typed donors. These donors should consent to, often repeated, cytapheresis within a short period of time. Besides the emotional stress for the patients, clinicians, donors and the blood bank staff, providing HLA-matched platelets is a costly therapy regarding the labor and the equipment for cytapheresis. It therefore appears more efficient to prevent HLA immunization rather than to cope with problems of refractoriness to PC transfusions. In several studies it has been demonstrated that the use of leukocyte-poor blood products (leukocyte-depleted RBC and PC) reduces the risk of HLA alloimmunization from 40-50 to 10-20% [3-51. The majority of patients who formed HLA antibodies de- spite the use of leukocyte-poor products had a history of pregnancies or transfusions of nonleukocyte-depleted RBC and PC. It is likely that such patients have a secondary (booster) immune response to minimal amounts of HLA antigens, which is unavoidable, not even by leukocyte depletion of blood products. The majority of contributors to the Forum do share the opinion that it is worthwhile to implement leukocyte depletion of PC as a standard procedure in blood banks. Besides the prevention of HLA immunization, the removal of leukocytes also prevents febrile transfusion reactions and transmission of viruses (such as cytomegalovirus and human T cell leukemia I and 2viruses) and will provide better storage conditions. It is still a matter of debate as to what extent leukocytes should be removed from blood products to prevent primary HLA immunization. At present the minimal number of allogeneic leukocytes in PC to elicit a primary HLA antibody response in patients is not well known and is estimated to be about 1 x 1O6IeukocytesperdonorunitofPCresulting in a maximum of lo7leukocytes per transfusion [6]. If the amount of leukocytes is reduced to 10-15 X 10' per pool of PC HLA immunization will drastically be reduced (3-51. Some authors of the Forum advocate to remove leucocytes from PC as soon as possible i.e. within24 hafterthe collectionof the whole blood to avoid fragmentation of granulocytes and the release of harmful enzymes from leukocytes. Several methods to prepare leukocytepoor PC are described in the present Forum i.e. the use of platelet-rich plasma or buffy coats derived from units of whole blood as source for PC with subsequent filtration and centrifugation for removal of leukocytes. Our contributors emphasize quality assurance of the final product with regard to the number of remaining leukocytes, the number and viability of the platelets (swirling effect) and the pH. A special problem is imposed by the fact that leukocytes below levels of 108/1cannot be determined accurately with current methodology (flow cytometers, electronic particle counters and cell chambers). However, electronicparticle counting isat present probably the easiest and quickest method to check the upper limit of leukocytes present in leukocyte-poor blood products. Finally, althoughcumbersome, studies in patients should prove that the chosen method of preparation of leukocyte-poor PC indeed was efficient to prevent primary immunization against HLA antigens in recipients and the cost effectiveness to provide leukocyte-poor PC needs to be determined. One author states that leukocyte-poor PC prepared from single buffy coats by centrifugation requires trained staff without additional expense, whereas other authors who prepare these products by filtration are faced with the extra costs of filters, wafers for the sterile connection device and special large bags for storage of the PC. However, most authors judge that these extra costs are acceptable for the prevention of HLA immunization, compared to the costs and efforts to provide the refractory patient with HLAmatched PC. zyxwvuts zyxwvut Q 1992 S. Karger AG. Basel 0012-9007/92/0621-MI57 $2.75/0 zyxwvut zyxwvutsrq UV-B irradiation of PC to inactivate HLA-class I1 bearing cells is regarded to be not yet applicable by the members of the Foruni. It may be concluded from the present contributors that all patients who may need platelet transfusions should be transfused with leukocyte-poor RBC and PC from the moment they become transfusion-dependent to avoid HLA immunization and subsequent refractoriness to random platelet transfusions. H. W. Reesink, MD, PhD Medical Director Red Cross Blood Bank Amsterdam NL-Amsterdam (The Netherlands) of 10% immunization, mainly in women with previous pregnancies [4]. Standard platelet transfusions result in about 50% immunization necessitating HLA-matched platelet In the past decade several controlled and transfusion. This requires costs for single doalso well-documented uncontrolled studies nor procedures (US$ 30/unit) and platelet demonstrated a reduction of HLA immuni- cross-matching (US$ 10) and in addition an organization is necessary to select and call-in zation by leukocyte depletion of platelet transfusions [for review, see I]. From these HLA-matched donors anticipating to the instudies the lowest immunizing dose of white dividual needs of a particular patient, often cells can only be estimated. In a controlled including cito-virology testing. These costs study, comparing two methods of leukocyte are exclusive the costs to maintain an HLAdepletion, Marwijk Kooy et a]. [ 2 ] found a typed donor file which has to be available 57% incidence of HLA immunization in pa- anyhow independent of the number of altients who received platelet transfusions con- loimmunized patients. The cost-benefit of filtration is only pretaining > 5 x lo7white cells (>8 x 10' leukocytesldonor unit). This incidence of immuni- sent in those patients for whom alloimmunization is also found in patients who received zation increases costs of support such as paunlimited standard blood transfusions. In tients eligible for platelet supportive care, contrast, patients who were transfused with bone marrow or organ transplantation. Beplatelet concentrates containing < 5 x 107 sides, morbidity and in some instances morleukocytes developed HLA antibodies in tality is reduced when by transfusion of leu8%. The amount of contaminating leuko- kocyte-depleted blood cytomegalovirus cytes in the platelet transfusions in the latter transmission is prevented, for instance in the group, however, was in 96% of the trans- case of neonatal transfusions. Although leukocyte depletion also offers other advantagfusions below 5 X lo', still leaving a 'grey' zone between 5 x 10' and 5 X lo7leukocytes es with regard to reduction of' microaggrefrom which the immunogenicity is not exactly gates and febrile transfusion reactions, less known, albeit definitely reduced. With the expensive methods such as differential cenavailable knowledge, platelet transfusions trifugation, are effective as well. An alternative to leukocyte depletion is should aim at < 5 X 10' leukocytes (< lo6leukocytesldonor unit) and not exceed an upper nor offered by UV-B irradiation inactivating limit of lo7 above which immunization was HLA class I1 antigens. UV-B irradiation has incidently shown [3]. To reach this goal by a many drawbacks: Quality control in routine reproducible procedure, suitable for routine blood banking is impossible, the formationof blood banking, filtration offers the best ap- microaggregates is enhanced, the metaboproach. lism of leukocytes is deleterious for an optiFor the moment no studies are available mal environment for platelets during stor(yet) proving that the new generation of cel- age, the transmission of cytomegalovirus is lulose acetate and polyester filters remove presumably maintained, UV-B induces white cell fragments effectively with regard DNA damage and the immunosuppressive to immunization and transmission of cytome- effects of blood transfusions -- a subject of galovirus, implicating that filtration should concern in oncology - may be potentiated. not be postponed for several days. UV-B irradiation may have a place in the An optimal approach appears to store future in combination with leukocyte deplewhole blood for at least 8 h at 20°C to allow tion in an attempt to further reduce allo-imgranulocytes and monocytes to kill residual munization in high risk recipients with previbacteria which might be present in 2% of all ous pregnancies. blood withdrawn lege artis. Within 24 h, platelet suspensions should be pooled usingsterA. Brand ile connection devices, filtered and subseDepartment of Immunohaematology and quently stored. The shorter the duration of Blood Bank storage the better the quality of platelets, but University Hospital Leiden when < 10yplatelets/mlare stored in 1,000-ml Rijnsburgerweg 10 2333 AA Leiden (The Netherlands) bags providing surface area while rotating, PVC bags are suitable for several days' storage. The additional costs of such filtration procedure is approximately fl. 20 (US$ 10) per random donor unit (using volunteer donors) and this results in an overall percentage A. Brand zyxw zyxwvutsrqpon zyxwvutsrqpo zyxwvutsrqponm U.E. Nydegger, MD Head. Division of Transfusion Medicine. Central Laboratory of Hematology University of Berne CH-3010 Berne (Switzerland) .............................................. References Howard JE, Perkins HA: The natural history of alloimmunization to platelets. Transfusion 1978: 18:496S03. Dutcher JP, Schiffer CA, Aisner J , Wiernik PH: Alloimmunization following platelet transfusion: The absence of a dose response relationship. Blood 1981;57:395-398. 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. Murphy MF, Metcalfe P. Thomas H. Eve J, Ord J. Lister TA, Waters AH: Use of leukocyte-poor blood products and HLA-matched donors to prevent allo-immunization by platelet transfusions. Br J Haematol 1988529-534. Brand A. Claas FMJ. Voogt PJ. Wasser MNJM, Eernisse JG: Alloimmunization after leukocyte depleted multiple random donor platelet transfusions. Vox Sang 1988;M: 160166. Fisher M. Chapman JR, Ting A, Morris PJ: Alloimmunization to HLA antigens following transfusion with leukocyte-poor and purified platelet suspensions. Vox Sang 1985:49:331335. zyxwvutsrqpo 58 Reesink/Nydegger/Brand/Pietersz/Andreul International Forum Gmur/Murphy/Schiffer/Kickler zyxw zyxwvut zyxwvutsrqp zyxwvutsrq ............................................... References 1 Brand A: White cell depletion. Why and how? in Nance ST (ed): Blood Transfusion in the 1990's. Arlington, AABB, 1990; pp 35-61. 2 Manvijk Kooy M. van Prooijen HC, Moes M, Bosman-Stants I, Akkerman JWN: The use of leukocyte-depleted platelet concentrates for the prevention of refractoriness and primary HLA allo-immunization: A prospective randomized trial. Blood, in press. 3 Fisher M, Chapman JR, Ting A, Morris PJ: Alloimmunization against MHC antigens after platelet transfusions is due to contaminating leukocytes in the platelet suspension. Vox Sang 1985;49:331-335. 4 Brand A, Claas FMJ. Voogt PJ, et al: Alloimmunization after leukocyte depleted multiple random donor platelet transfusions. Vox Sang 1988;54:16Ck166. R. N . I. Pietersz Transfusion of leukocyte-depleted blood products will reduce HLA sensitization resulting in fewer febrile transfusion reactions and less immune-mediated refractoriness to platelet therapy [l]. Furthermore, transmission of leukocyte-associated viruses will be diminished. The immunogenic dose of leukocytes, however, is still unknown and dependson the type of leukocytes, the immunogenicity of the HLA antigens and the responsiveness of the recipient. Moreover, the commonly available methods for counting leukocytes do not have a great accuracy if the leukocyte concentration drops below 1 cell per PI. If we assume that the immunogenic dose is 50 x lo6 leukocytes as recently suggested by van Manvijk Kooy et al. [2], how can this be achieved in blood bank routine. For the filtration of red cell concentrates various filters are available to reduce the leukocyte number to below 50 x 10' and even 10 x 10'. However, most filters have a maximum capacity to retain leukocytes and perform better if the initial leukocyte number is diminished to approximately 1,000x lo', which can be accomplished by first removing the buffy coat from the red cells (quality control data from 1990in Amsterdam; Cellselect filter, NPBI; n = 14.113, leukocytes < 2 x lo6 in 100% of the filtered unit). For platelet concentrates (PC) it is not easy to meet a criterion of 50 x 10' leukocytes in a pool of 4-6 donor units. Very careful preparation of PC with the platelet-rich plasma method results in leukocyte contaminations of 20 x lo6or higher per donor unit. In PCpreparedfrom singledonor buffy coats routinely fewer than 10 x 10' leukocytes are present and in more than 90% of these PC fewer than 5 x 10' (quality control data from 1990 in Amsterdam: n=5.216 pools of 6 PC, leukocytes 5 k 4, mean k SD, per donor unit). However, this low leukocyte contamination can only be achieved at the cost of a decrease in platelet number to 4045 x 10' per donor unit [3], (quality control data from 1990 in Amsterdam: n=4.178 pools of 6 PC: platelets 47 k 9, mean t SD, per donor units). In another method of PC preparation from buffy coats the buffy coats are pooled and the pool is centrifuged for a second time. A platelet-rich plasma with a high platelet count (75 X 10') and a low leukocyte contamination, i.e. below 50 x 106per 4 donor units could be obtained routinely [4]. For PC obtained by cytapheresis it depends on the machine used, how high the leukocyte contamination will be. With some machines leukocyte numbers per PC are below 50 x lo6, while with others the PC may contain 1,OOO x 10' or more leukocytes. To remove leukocytes from PC an extra centrifugation can be applied or the PC can befiltered. Bothmethods will induce platelet loss ranging from 10 to 20%. Another drawback of the leukocyte depletion filters for platelets is the same as for red cell filters: the capacity of the filter to remove leukocytes is not always known. Since the initial number of leukocytes in the PC may vary from approximately 50 to more than 1,000 x lo6,the leukocytes in every unit of filtered PC should be checked. An initially low number of leukocytes in the PC will improve the final result. For every method used a final check before issuing the leukocyte-depleted PC should include: (1) the swirling effect in the PC; (2) a leukocyte count with an upper limit of 50 X 10' leukocytes per PC, and (3) a platelet count. Considering the possibilities for preparation and the costs, leukocyte-poor PC from single buffy coats can be manufactured routinely only requiring trained staff and special inserts for centrifugation [3]. Pooling of buffy coats is easy to perform but requires the use of asterile connection device (SCD) and special large platelet storage bags, which increases the costs [4]. Cytapheresis is expensive in donor time and software. Leukocyte depletion by filtration of the PC will increase the costs with the price of the filter(US$30), andifthefilteredPCaretobe stored with the price of the wafers (US$1.5 each) for the SCD connections and special large platelet storage bag (US$ 15). Moreover, storage of nonfiltered PC has been assessed more thoroughly than of filtered PC which are often mixed with solutions used to prime or rinse the filter. There is sufficient evidence that if PC contain < 50 x lo6leukocytes, cytomegalovirus transmission will be prevented. UV-B irradiation may alter the function of class I1 bearing cells but as yet there is too little evidence that the method is safe for clinical use to prevent alloimmunization. UV-B may induce unknown late effects on DNA synthesis, the effect of irradiation on platelet function and survival is not yet known, there are still practical problems to be solved concerning the irradiation and type of plastic for the bags. Finally, UV-B irradiation will not prevent transmission of leukocyte-associated viruses. zyxwvutsrqponm zyxwv zyxwvut 59 zyxwvutsrqp zy In conclusion, for patients with malignant hematological disorders andor chemotherapy. patients on hemodialysis and organ transplant candidates, PC should be leukocyte-depleted. Since in most large centers the majority of patients belong to this group all PC should be leukocyte-depleted. G. Andreu The French standards for leukocyte-poor red cell and platelet concentrates have been recently modified. For platelet concentrates, three different productsare to beconsidered: (1) platelet concentrates prepared from whole blood donation (= standard platelet concentrates): they contain at least SO x 10' plateletsinavolumeof40-70 ml. Leukocytepoor platelet concentrates contain less than 5 x 10" leukocytes; (2) platelet concentrates prepared by apheresis along with plasma collection: they are defined as multiple of standardplatelet concentrates, accordingto their platelet and leukocyte content; (3) platelet concentrates prepared by apheresis (= single donor platelet concentrates); they contain at least 400 X 10' platelets in less than 500 ml. Leukocyte-poor single donor platelet concentrates contain less than 5 x 10" leukocytes. Actually, it is already possible to produce leukocyte-poor platelet concentrates containing less than 0.1x lo6 leukocytes in platelet concentrates prepared from whole blood, and 10'Ieukocytes in single donor platelet Concentrates. However, standards must take account of all the techniques and the materials in use in the country, and the values finally adopted should be accessible for every blood center in France. The fact that leukocyte contamination is harmful to platelets during storage has been studies in experiments involving add-back of controlled quantities of leukocytes in otherwise identical platelet preparations [l].In addition, spontaneous lysis of leukocytes during storage is not negligible. The major part of cytoplasmicas well as membrane materials released cannot be blocked by the filters used for leukodepletion. Moreover, it has been shown in an animal experiment [2] that filtration after storage was less efficient than early filtration to prevent alloimmunization against major histocompatibility complex antigens. These three arguments are in favor of anearly leukocyte depletion, at the time of preparation of platelet concentrates. Platelet integrity immediately after the preparation of platelet concentrates is highly dependent on the method used: if we consider only standard platelet concentrates, the buffy coat methods are less detrimental to platelets than the conventional platelet-rich plasma procedure [3]. All the above information leads to the conclusion that the best way to prepare and store leukocyte-poor concentrates is the combination of the buffy coat methodology and an early filtration before storage. This association has already been developed in somecentres[4]. TheriskofHLAimmunization in patients multitransfused with red cell and platelet concentrates when standard products are used varies from 28 to 71% (mean = 41%) as analyzed in 19published and personal data from 1979 to 1989, involving overall 1,393 patients. Standard blood components used in these studies were almost all prepared with the conventional plateletrich method, therefore highly contaminated with leukocytes. In contrast, 0-28% (mean = lY%)ofpatientstransfusedwith leukocyte-poor blood components form antiHLA antibodies, in 11 studies involving 663 patients. In almost all these studies, leukocyte-poor blood components were prepared at the time of transfusion, and the number of leukocytes remaining in the final products was higher than the values we can obtain today with the association of the buffy coat methodology and filtration. However, leukocyte-poor blood components do not seem to be able to completely prevent HLA alloimmunization: they are an efficient means to prevent primary HLA immunization, while secondary response i n previously HLA-sensitized patients, either by transfusion or pregnancy is poorly prevented. In our recent experience at the HBtel-Dieu in Paris. 10 out of 54 patients transfused with leukocyte-poor blood components had HLA antibodies, a secondary response being most probable for 8 patients. UV-B irradiation is an interesting means to expect a further reduction of transfusioninduced HLA immunization: there is a body of experiments suggesting that the failure of mononuclear cells to stimulate allogeneic cells in vitro and in vivo after UV-B irradiation is not a simple absence of response, but rather the development of an immune tolerance in the recipient. The difficulties to organize a UV-B trial today are less technical [S]specific blood bags and irradiator are needed -than conceptual: as UV-B irradiation cannot be performed on red cell concentrates, patients must receive the best leukocytepoor red cell concentrates available. In the French trial, we decided as acceptable a maximumof2 x 1O'leukocytecontent perredcell concentrate, which needs the systematic combination of buffy coat methodology and filtration, and a rigourous quality control. The comparison with a group of patients receiving leukocyte-poor blood components is mandatory. More questionable is the necessity of a control group receiving standard pla- zyxwvutsrqpo zyxwvutsrqp R.N. I. Pietersz, MD PhD Red Cross Blood Bank Amsterdam Plesmanlaan 125 1066 CX Amsterdam (The Netherlands) .............................................. References Brand A, Claas FMJ, Voogt PJ, Wasser MNJM, Eernisse JG: Alloimmunization after leukocyte depleted multiple random donor platelet transfusions. Vox Sang 1988;54:160166. Van Marwijk Kooy M, van Prooijen HC, Moes M, Bosma-Stants I , Akkerman JWN: The use of leukocyte-depleted platelet concentrates for the prevention of refractoriness and primary HLA alloimmunization: A prospective randomized trial. Blood 1991:77:1-15. Pietersz RNI, Reesink HW, Dekker WJA. Fijen FJ: Preparation of leukocyte-poor platelet concentrates from buffy coats. I Special inserts for centrifuge cups. Vox Sang 1987:53:203207. Eriksson L, Hogman CF: Platelet concentrates in an additive solution prepared from buffy coats. Vox Sang 1990;59:140-1-15. zyxwvu zyxwvutsrqpo 60 Reesink/Nydegger/Brand/Pie tersz/Andreu/ Gmiir/Murphy/Schiffer/Kickler International Forum zyxwvutsrqpo telet concentrates. We found this group useful, since the only one randomized clinical trial comparing leukocyte-poor and standard platelet concentrates, while all patients receive leukocyte-poor red cell concentrates, was not conclusive [ 6 ] . In case UV-B irradiation efficiency is demonstrated to reduce secondary response, it could not replace the use of leukocyte-poor blood components: leukodepletion has many other advantages apart from HLA immunization prevention. It would be an additional technique. If the efficiency is well established in the next few years by clinical trials, it will be a benefit for the patients at medical as well as economical point of vue: patients receive always less platelet concentrates when they have no HLA immunization, nor refractoriness to platelet transfusion. ml. Neutrophils (constituting up to 10%)nor- Jurgl? Gmiir Leukocyte depletion of blood products by use of *thirdgeneration’ (adsorption) filters is being used with increasing frequency in an effort to prevent HLA alloimmunization. Several clinical studies suggest that there is a clear trend with the incidence of reported alloimmunization reaching zero when the number of leukocytes falls below 5 X lo6per transfusion [I. 21. To obtain such ‘clean’ products the use of ‘third generation’ filters is inevitable which adds considerable costs and additional work to the preparation of blood components and leads to some cell loss. Therefore. the question arises wether such ‘clean’ products are of proven clinical advantage compared to red blood cell (RBC) or platelet transfusions (PLT) leukocyte depleted by a simple centrifugation step and containing 1-2 log higher leukocyte counts. This question is not easy to resolve. It is well accepted that modern conventional counting techniques by electronic inpedance (Coulter Counter S Plus 1V)or light scatter (TechniconH*l)cannotbe utilized to monitor white cell counts below 0.1 x 10’A corresponding to 2-4 x 10’ per single donor apheresis product (SD-PC) or 5 x loh per standard platelet concentrate (PC). A modified microscopic counting method(1:2(PC)andl:lO(RBC)ratherthan 1:lO and 1:lOO dilution, respectively: 36 rather than 9 squares counted) seems to increase the sensitivity by 2 log (PC) and 1log (RBC), respectively. However, a comparison of counting methods and fluorescein-activated flow cytometric techniques has revealed that > 2 log leukocyte depletion cannot be accurately monitored by the former [3]. At our institution analyses of serially diluted PC by FACS showed that the correlation of predicted and measured cell counts are no longer linear below 0.02 X 10’11 [unpubl. observations]. This corresponds to 4 X 10‘ per SD-PC or 1 x 10’per PC. Thus. for the present leukocyte-poor blood components have to be defined by an upper limit of leukocyte content which seems to be in the order of 5 X lo6 per platelet transfusion and lo7 per RBC transfusion. Due to these technical limitations the term ’leukocyte-free’blood components is inappropriate. Moreover, clinical studies comparing standard and ‘virtually leukocyte-free’ RBC and/or PLT transfusions should be carefully checked for their leukocyte counting methodology [l, 21. Leukocytes are usually present in PC in numbers varying from 1x 10’ to 3 x 10’ per mally disintegrate during the first 2 days of storage and thereby may release hydrolytic enzymes from their granules into the plasma. It has been shown that these enzymes affect membrane composition and platelet function [4]. These observations point to the importance of removing leukocytes from PC before storage. On the other hand, caution is indicated because there is no published experience documenting that storage life is not altered by preceding filtration. For more than 10 years it has been our policy to use exclusively random SD-PC in patients who will need platelet support for more than 2-3 weeks. This policy is substantiated by the low incidence of alloimmunization observed after random SD-PC in leukemiapatientswithoutpriorsensitization[5].In our hands, the rateof antibody formation and refractoriness is comparably low whether leukocyte-rich (-4 x 10y/PLT) or leukocytepoor (-3 x 10s/PLT)random SD-PCare used [5]. Hence, it would be difficult to prove that an additional filtration of SD-PC would add any advantage to the already low incidence of alloimmunization (2/49) in nonpresensitized recipients of standard SD apheresis platelets. By which measure the higher incidence in patients with prior pregnancies could be reduced is not yet established. zyxw zyxwv zyxwvuts zyxwvutsrqp zyxwvutsrqpo G. Andreu, MD Poste de Transfusion Sanguine HBtel-Dieu de Paris 1 place du Parvis Notre-Dame 75181 Paris Cedex 01 (France) .............................................. References I Pietersz RNI. de Korte D, Reesink HW, van Den Ende A, Dekker WJA. Roos D : Preparation of leukocyte-poor platelet concentrates from buffy coats. Vox Sang 1988:55: 11-20. 2 Blajchman M: Communication presented at the Research and Progress session on leukocyte-poor blood components. ISBTlAABB meeting. Los Angeles. November 10-15. 1990. 3 Fijnheer R, Pietersz RNI. de Korte D, Gouwerok CWN. Dekker WJA, Reesink HW, Roos D: Platelet activation during preparation of platelet concentrates: A comparison of the platelet-rich plasma and the buffy coat methods. Transfusion 1990:30:634-638. 4 Angue M. Chatelain P. Domy M. Guignier F, Richaud P: Preparation de concentres des plaquettes humaines deleucocytees par centrifugation et filtration d’un pool de buffy-coats connectes stkrilement. Revue Fr Transfus Hemobiol, in press. 5 Andreu G. Boccaccio C. Lecrubier Ch, Fretault J. Coursaget J, Leguen JP, Oleggini M, Fournel JJ, Samama M: Ultra violet irradiation of platelet concentrates: Feasibility in transfusion practice. Transfusion 1990:30:401406. 6 Schiffer CA, Dutcher JP, Aisner J. Hogge D. Wiernik PH, Reilly JP: A randomized trial of leukocyte-depleted platelet transfusion to modify alloimmunisation in patients with leukemia. Blood 1983:62:815. Jurg P. Gmur, MD Universitatsspital Zurich Departement fiir Innere Medizin Abteilung fiir Hamatologie Ramistrasse 100 8091 Zurich (Switzerland) .............................................. zyxwvu References 1 Brand A, Claas F, Voogt P, et al: Alloimmunization after leukocyte-depleted multiple random donor platelet transfusions. Vox Sang 198854: 160. 2 Saarinen U, Kekomaki R, Siimes M, Myllyla G: Effective prophylaxis against platelet refractoriness in multitransfused patients by use of leukocyte-free blood components. Blood 1990;75:512. 3 Bodensteiner DC: Leukocyte depletion filters: a comparison of efficiency. Am J Hemato1 1990:35:184. 4 Sloand EM, Klein HG: Effects of white cells on platelets during storage. Transfusion 1990: 30:333. 5 Gmur J, Burger J, Sauter Chr, et al: Alloimmunization by leukocyte-rich or leukocytepoor random single donor platelets. Progr Clin Biol Res 1990;337:45. zyxwvutsrqponm 61 zyxwvutsr zyxwvu Scott Murphy Several studies strongly suggest that leukodepletion of blood products prior to infusion reduces the incidence of subsequent alloimmunization in recipients [I]. Alternative approaches such as ultraviolet irradiation show promise but have not yet been adequately tested clinically. There are no established disadvantages of leukodepletion of platelet concentrates other than the increased cost and the loss of 10% of the platelets processed. Therefore, current data support leukodepletion of all platelet concentrates issued to patients with diagnoses in which the need for long-term plateletsupport can be predicted. In fact, some of the most difficult patients to support are those who have been alloimmunized by transfusion priortotheestablishmentofsuchadiagnosis. In such presensitized individuals, the dose of leukocytes necessary to elicit a secondaryimmune response may be very small and beyond the capacity of any leukodepletion procedure. Therefore, in principle, all blood products issued should be leukopoor. Like all clinical recommendations, this recommendation is a best estimate as optimal management for an individual patient at a point in time when the data available is still incomplete. It has been calcualted that the number of patients who will benefit significantly from leukodepletion is rather small [ 2 ) . Therefore, it is particularly important that the cost of leukodepletion be low. The extent of leukodepletion required is not yet known. It seems obvious that the requirement will vary frompatienttopatient. Forexample,inacute leukemia, some patients have very stormy courses requiring frequent infusions of many units of platelets while others go through induction rather easily with relatively few infusions. We assume that the risk for alloimmunization varies with the total number of contaminating leukocytes infused within a given time period. Current data suggests that we must achieve at least a 2 log reduction from 1IeukocytellO' platelets, which is characteristic for random donor concentrates prepared from platelet-rich plasma to 1leukocytell0' platelets. Perhaps another one or two logs of reduction will ultimately be required. In the long run. it will probably be preferable to develop methods which will render platelet concentrates relatively free of leukocytes at the time of their preparation. Some devices for plateletpheresis [3] and the buffy coat method [4] of processing donations of whole blood represent steps in this direction. However, a second step, probably a filtration step, will be required to achieve optimal leukodepletion in most cases. Filtration at the bedside duringinfusion iseffective, at least in theory IS]. However, it is difficult to assure use of consistent techniques and establish adequate quality control in this setting. Furthermore, there may be some disintegration of leukocytes during storage prior to infusion. Leukocyte fragments may be immunogenic but not filtrable [6]. Therefore it seems preferable to complete leukodepletion in the blood center at some point during the first 24 h of storage. This step has to be compatible with maintenance of platelet integrity during subsequent storage for 4-5 days. There is not yet adequate data showing that any leukodepletion procedure meets this requirement. However, it seems highly likely that such procedures will be developed and verified in the near future. Scott Murphy, MD Professor of Medicine Associate Director for Clinical Programs Cardeza Foundation for Hematologic Research 1015 Walnut Street Philadelphia, PA 19107 (USA) .............................................. zyxwvu zyxwvuts zyx References 1 Meryman HT: Transfusion-induced alloimmunization and immunosuppression and the effects of leukocyte depletion. Transfusion Med Rev 1989;3:180. 2 Schiffer CA, Dutcher JP, Aisner J, Hogge D. Wiernik PH, Reilly JP: A randomized trial of leukocyte depleted platelet transfusion to modify alloimmunization in patients with leukemia. Blood 1983;62:815. 3 Hester JP, Ventura GJ, Bouzher T: Platelet concentrate collection in a dud-stage channel using computer-generated algorithms for collection and prediction of yield. Plas Ther Transf Techno1 1987;8:377. 4 Pietersz RNI, Reesink HW, Dekker WJA. Fijen FJ: Preparation of leukocyte-poor platelet concentrates from buffy coats. Vox Sang 1987; S3:203. 5 Kickier 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. 6 Engelfriet CP, Diepenhorst P. Vande Giessen M, Von Riesz E: Removal of leukocytes from whole blood and erythrocyte suspensions by filtration through cottonwool. Vox Sang 1975: 28:81. zyxwv 62 Reesink/N ydeggerlBrandlPieterszl Andreul GmiirMurphylSchifferlKickler International Forum Charles A . Schifler zyxwvutsrq The development of antibody against class I HLA antigens is the major cause of immune-mediated refractoriness in recipients of multiple platelet transfusions. Alloimmunization develops in approximately one third of patients with acute myeloid leukemia (AML) undergoing induction chemotherapy, in a smaller fraction of patients with acute lymphocytic leukemia and in a higher proportion of individuals not receiving concurrent cytotoxic chemotherapy [I]. Although most attention is appropriately focused on the clinical problem of the management of alloimmunized patients, perhaps the biologically most interesting point is that the majority of individuals exposed to HLA antigens from multiple donors do not form antibody against these antigens and indeed behave as if they were immune ’tolerant’. There is compelling preclinical evidence to suggest that it is the leukocytes suspended in platelet transfusions which are responsible for the initial recognition by the recipient’s immune system with the subsequent development of alloantibody [reviewed in 21. In addition. there is evidence that even modest leukocyte depletion by avariety of meanscan attenuate or eliminate febrile transfusion reactions in patients who have had repeated reactionsof this type [3].In recent years more effective blood filters have become available which can results in a > 3 3 log reduction of leukocytes from platelet suspension and packed red blood cells to a level of contamination which may be ‘nonimmunogenic’. A number of clinical trials have been performed, most of which have been interpreted as demonstrating a reduction in alloimmunization when filtered blood products are utilized [reviewedin4].There aremanymethodologic issues, however, which prevent general extrapolation of these results. There are also sufficient differences amongst the trials to preclude the application of the now fashionable ‘metaanalysis’technique. Some of these problems include: (1) small numbers of patients (largest trial includes 69 evaluable patients); (2) variability in patient diagnoses and treatment: (3) exclusion of potentially ’presensitized’ patients with prior pregnancies or transfusions from some studies; (4) inconsistencies in the frequency ofalloimmunization noted with comparable amounts of leucocyte contamination; ( 5 ) variable quality control of leucocyte numbers in transfused products. Table 1. Potential effect of leukocytedepleted platelets for 100 patients with Initial total AM1 Alloimmunized on admission or anamnestic response Alloimmunization rate CR rate Intensive post .CR treatment ‘Effectiveness’ of leukocyte depletion Granulocyte transfusions Protocol errors Patients, % Patients ‘benefitting’ zyxw 100 10 3040 75 15 67 5 10 90 36 27 21 14 13 12 CR = Complete remission. The available filters are quite expensive and represent a considerable increase in blood banking administrative and technologist costs, particularly since all of the red blood cell transfusions must also be leukocyte-depleted. In addition, all filters produce appreciable (10-20%) loss of platelets, potentially increasing the number of units which must be transfused. There is little experience with storage of filtered platelets (most, if not all studies filtered immediately before administration), and limited information about the reproducibility of leukocyte depletion when these techniques are used in blood centers of different size and sophistication or at the bedside. The latter point is of particular relevance because of data inferring that transfusions should contain < lo6 leukocytes in order to reduce immunogenic- isof particular interest because it is potentially extremely simple to administer. Because of the cost considerations, the inconclusive (albeit very tantalizing) results of the available clinical trials, and our current ability to successfully support the fraction of patients who do become alloimmunized, platelet leukocyte filtration to prevent alloimmunization in multiple transfused recipients cannot be recommended for routine use at this time. zyxwvutsrqp ence of a variety of factors suggest that only 10-15% of patients might actually ‘benefit’ ry using standard platelet transfusions and the approximate 10% of patients alloimmunized at oresentation would not derive benefit from leukocyte-depleted blood products (table 1). Because of these considerations, any technique employed to reduce alloimmunization must be inexpensive and produce sufficient and reproducible leukocyte depletion or modification. A large, randomized multiinstitutional study (Tryal to Reduce Alloimmunization to Platelets), which should address many of the questions raised by the studies which have been done to date, has been begun in the United States. Leukocyte filtration, as well as UV-B irradiation of platelets, will be evaluated. The latter approach ~ $ ~ ~ ~ o ~ Oncology $ ~ e ~ ~ Universityof Maryland Cancer Center universitv of Mawland School of Medicine 22 S. Grekne Street Baltimore, MD 21201 (USA) ~ n d References zyxw munized patients. Blood 1987;70:1727-1729. 2 Meryman HT Transfusion-induced alloimmunization and immunosuppression and the effects of leukocyte depletion. Transfusion Med Rev 1989;3:18&193. 3 Schiffer CA, Patten E., Reilly J , Patel S: Effective leukocyte removal from platelet preparations by centrifugation in a new pooling bag. Transfusion 1987;27:162-164. 4 Schiffer CA: Prevention of alloimmunization against platelets. Blood. in press. zyxwvutsrqp zyxwvutsrqpo zyxwvutsrqp 63 zyxwvutsrq zyxwvutsrq zyxwvut Thomas S. Kickler Clinical and experimental data suggest that leukocyte depletion from units of platelets may serve two main purposes first prevention of alloimmunization and second improvement in the function and viability of stored platelets [l,21. If it can be shown that there is significant improvement of platelet quality by leukocyte depleton prior to storage, prestorage leukocyte depletion should become a standard procedure for all platelet concentrates. If the prevention of alloimmunization is the only concern, not all patients require leukocyte-depleted platelet and leukocyte depletion would only be required for selected patients. The definition of leukocyte-poor platelets is dependent upon the purpose of the leukocyte depletion. If one wishes to reduce alloimmunization, several clinical studies suggest that residual leukocyte contamination of 10' may be sufficient [l].However it should be noted that these studies were performed in immunosuppressed patients. It is unknown whether this degree of leukodepletion is sufficient for other patient groups not heavily immunosuppressed such as those with aplastic anemia. If the goal of leukocyte depletion is the improvement of platelet quality and viability, it isnot known whether reducing residual leukocytes to 10'will be sufficient to enhance the quality of stored platelets [2]. Little is known about the number of leukocytes that are important in mediating injury to the platelet during storage. From personal observations, removal of 3-4 logsof leukocytes prior to storage may be sufficient to improve the in vivo and in vitro recovery of platelets after 5 days of storage. If leukodepletion is to be performed it should be done prior to storage to gain the benefit of improving the quality of stored platelets as well as reducing the risk of alloimmunization. Contaminating leukocytes may cause impaired platelet viability and function independent of their influencing any drop in pH. Neutrophils may become disrupted during centrifugation or during storage leading to the release of hydrolytic enzymes. These enzymes, specifically Cathepsin G and elastase, can alter glycoprotein IB/IX expression and possibly other important membrane receptors. Lymphocytes, the predominant contaminating leukocyte in platelets, mayalso inhibit platelet function [3]. Leukocyte depletion may be efficiently performed by filtration leading to 2-3 log reduction in leukocyte contamination. If one prepares platelets by the buffy coat method, very low levels of leukocyte contamination will also result [4].It may even be possible to combine filtration with platelet preparation by the buffy coat method to make a more leukocyte-poor product. It is not known whether ultraviolet irradiation will prove superior to leukocyte depletion in preventing alloimmunication to platelet transfusions. The simplicity of ultraviolet irradiation makes it an appealing alternative to filtration. However, there are only limited data on the efficacy of ultraviolet irradiation in preventing the alloimmune response to platelet transfusions in humans [ 5 ] . Numerous clinical studies have been performed using leukocyte removal as a modality in preventing or delaying alloimmunization to HLA antigens. Although there is considerably more experience with removal of leukocytes than with ultraviolet irradiation, the studies are relatively small and the criteria of refractoriness or alloimmunization inadequate. Only patients requiring relatively long-term platelet transfusions are at risk for alloimmunization. Surgical patients would not benefit from the expense of using leukocyte-depleted platelets. It should also be noted that up to 10% of platelets may be lost by filtration. This may mean that if routine filtration removal of leukocytes is instituted, more platelets may need to be prepared to supply the platelet needs of patients. ................................................ References 1 Kickler TS: Platelet alloimmunization. Transfusion Med Rev 1990;4:8-18. 2 Sloand EM. Klein HG: Effect o f white cells on platelets during storage. Transfusion 1990;30: 333-338. 3 Nicolini FA, Wilson AC, Metha P, Metha JL: Comparative platelet inhibitory effects of human neutrophils and lymphocytes. J Lab Clin Med 1990:116:147-152. 4 Fijnheer R, Pietersz RNI, Korte D . Gouwerok CWN, Dekker WJA, Ree\ink HW. Roos D: Platelet activation during preparation of platelet concentrates: A comparison of the platelet rich plasma and the buffy coat methods. Transfusion 1990;30:634438. 5 Menitove JE, Kagen LR, Aster RH, et al: Alloimmunization is decreased in patients receiving UV-B irradiation platelet concentrates and leucocyte depleted red cells. Blood 1990; 761404. zyxwv zyxwvutsrqpo zyxwvutsrq 64 Thomas S. Kickler MD Associate Director The Johns Hopkins Hospital Department of Laboratory Medicine Immunohematology Division 600 North Wolfe Street Baltimore, MD 21205 (USA) Reesink/Nydegger/Brand/Pietersz/Andreu/ International Forum Gmiir/Murphy/Schiffer/Kickler