Challenges in testing for platelet‐related adverse events

According to haemovigilance data from many Western countries, platelet concentrates (PCs) is the type of blood component most frequently associated with transfusion reactions [1]. In a prospective study by Heddle et al. adverse events were reported to occur in up to 31% of platelet transfusion [2], but more recently, the incidence of transfusions reactions to PCs has been reported to be around 2% or less [3, 4]. Although differences in the reporting practice make direct comparison between studies difficult, the frequency of adverse reactions to platelets has undoubtedly decreased considerably during the last two decades. The most important single cause that can explain this decreased number of transfusion reactions is various improvements in production and storage of PCs. Most platelet-related adverse events are mild and non-life threatening reactions. More rarely, transfusion of platelets is associated with potentially fatal reactions such as transfusion-related acute lung injury (TRALI), septicaemia and severe anaphylactic reactions. TRALI is a rare but serious and potentially fatal transfusion reaction that usually occurs after transfusion of plasma-containing blood products such as fresh frozen plasma or PC. From 2005 through 2009 TRALI was the leading cause of transfusion-associated death in the United States [5]. The incidence figures of TRALI vary widely, ranging from 1 of 432 to 1 of 88 000 per units of platelets [6]. The large variation in incidence rates is probably related to different definitions of TRALI, different methods of surveillance and different methods of tabulating the denominator data of blood products transfused across studies [6]. During the last decade the awareness of this rare transfusion reaction has increased considerably. More cases are now being recognised as TRALI, whereas previously the association with transfusion was overlooked, and cases were interpreted as acute lung injury with a different aetiology. The typical clinical features of TRALI are respiratory distress, hypotension, hypoxemia, tachycardia, bilateral pulmonary infiltrates and fever appearing within 6 h after initiation of the transfusion. Circulatory overload and other obvious causes of acute lung injury should be ruled out. Treatment of TRALI patients is mainly supportive and in mild cases oxygen support is usually sufficient. In more severe TRALI cases, however, artificial ventilation may be required. The mortality in the severe cases is in the range of 6–20% [7-9], but patients who survive usually recover within 48 h. For decades it has been known that anti-leucocyte antibodies in the blood component are associated with TRALI. Antibodies against HLA class I and class II antigens have been implicated as well as neutrophil specific alloantibodies. HLA class I antigens are expressed on all nucleated cells whereas HLA class II are only constitutively expressed on monocytes, macrophages, dendritic cells and B cells. For unknown reasons some antibody specificities (anti-HLA A2, anti-HLA B12 and anti-HNA3a) seem to be more often involved with the severe cases than others. In the large majority of cases it is antibodies present in the blood component that are implicated in TRALI, whereas antibodies in the donor are only rarely a causative factor. In some cases of TRALI there are neither anti-leucocyte antibodies in the transfused blood component nor in the patient’s blood. In these cases it has been suggested that neutrophil priming lipids, such as lysophosphatidylcholines, released from platelets or red blood cells during storage, may be a crucial pathogenic factor [10]. The key cells involved in the pathogenesis are the neutrophil granulocytes. Alloantibodies reacting with neutrophils (anti-HLA class I or anti-HNA antibodies) and/or neutrophil priming lipids lower the threshold for activation of the patient’s neutrophils. Strong neutrophil-reactive antibodies may be sufficient to induce TRALI in a patient without any predisposing factors while weak antibodies or neutrophil-priming lipids may be harmless unless the recipient is severely ill, suffering from a concurrent infection or inflammatory disease. Under such conditions the patient’s neutrophils, pulmonary endothelial cells and/or platelets may already be primed, and transfusion of a blood component containing anti-leucocyte antibodies and/or neutrophil-priming lipids will further activate the recipient’s neutrophils. Consequently, these hyper-reactive neutrophils will be activated intravascularly, they will become rigid and trapped in the pulmonary capillaries where they release pro-inflammatory mediators such as CXCL8 (IL8), cytotoxic reactive oxygen species (O2- and H2O2) and toxic enzymes [11]. The pulmonary endothelium will be damaged giving rise to increased vascular permeability which in turn will lead to exudation and development of non-cardiogenic pulmonary oedema. In the case of anti-HLA class II antibodies, recent studies suggest that the antibodies bind to and stimulate monocytes to release pro-inflammatory cytokines, which in turn activate the patient’s neutrophils [12]. During the last decade several transfusion centres have implemented preventive measures to reduce the risk of TRALI, such as limiting the collection of plasma or single donor platelets to male donors, or female donors without a history of pregnancy, or to donors who have been shown not to have anti-leucocyte antibodies [13]. Another approach to reduce the risk of TRALI is to use solvent detergent-treated pooled plasma (SD plasma), which is devoid of anti-leucocyte antibodies [14] instead of fresh frozen plasma (FFP). This transfusion practice has been used in Norway for nearly two decades and during this period not a single case of TRALI after transfusion of SD plasma has been reported to the Norwegian haemovigilance system [15]. Although TRALI is a clinical diagnosis, laboratory investigations are required to explore whether any of the donors of the suspected blood components have anti-leucocyte antibodies. The laboratory case workup varies considerably from centre to centre [16]. The lymphocytotoxicity test, enzyme-linked immunosorbent assay (ELISA), flow cytometry and bead array assays are techniques that are frequently used to identify anti-HLA antibodies [17]. The granulocyte immunoflourescence test (by flow cytometry and/or microscopy), granulocyte agglutination test (GAT) and monoclonal antibody immobilization of granulocyte antigen (MAIGA) test are used for the detection of antibodies against granulocytes [17]. There are numerous challenges for the laboratory examining TRALI cases. Ideally, at least two different methods for antibody identification should be used, as some methods are more reliable for identification of certain antibody specificities. As an example, GAT is the best method for identification of anti-HNA-3a, which is known to be involved in many of the most severe cases of TRALI [17]. The analyses are labour-intensive and in many cases there are a number of blood donors under suspicion. If the patient had received two units of PCs produced from buffy coats (PC-BCs) a total of 8–10 donors should be investigated, but obtaining a new blood sample from all 8 to 10 implicated donors can be difficult. How should we interpret the results if anti-leucocyte antibodies are detected in one of the donors? Does this mean that we have a laboratory confirmation of the diagnosis? Not necessarily, because anti-HLA antibodies can be detected in 25% of female donors [18, 19]. Thus, the presence of anti-HLA antibodies may just be a coincidence and these antibodies will do no harm in the recipient unless the patient’s neutrophils carry the cognate antigen. Likewise, anti-HNA-1a will not cause TRALI if the recipient is HNA-1a negative. Hence, a laboratory confirmation of the diagnosis will require a sample from the patient, from which neutrophils and monocytes can be isolated and tested against plasma samples from the donors. Since the patient’s neutrophils must be fresh in order to make the cross matches with the donors’ plasma, it is quite often logistically difficult to obtain a sample from the patient once all of the samples from the donors have eventually been collected. Consequently, this important test is often not carried out. What consequences should be drawn from the analyses for anti-leucocyte antibodies? Both the AABB bulletin [20] and the Canadian Consensus Panel [6] recommended that a donor implicated in a TRALI case, where the donor has antibodies against leucocyte antigens of the recipient, should be deferred from future donations (or have their donations restricted to the further manufacture of washed or frozen deglycerolized red cells). Despite these recommendations a recent survey has demonstrated that the donor management policies vary considerable in the United Stated [16]. If anti-leucocyte antibodies cannot be detected in plasma from the implicated donors deferral should not be necessary. The transmission of viruses is a risk that is not only restricted to transfusion of PCs but is associated with all blood products that have not been subjected to pathogen reduction treatment. In the Western countries the risk of viral transmission is very low and will not be further discussed in the present paper. Transmission of bacteria, however, is a significant risk associated with platelet transfusions and transfusion-associated septicaemia represents a considerable proportion of transfusion-associated fatalities [13]. Many blood banks have implemented pathogen reduction technology to reduce the risk of septicaemia in recipients of PCs. Without this technology the frequency of PCs contaminated with bacteria varies from 0·03% [21] to 0·7% [22]. The risk of bacterial contamination of PCs can be minimized by strictly adhering to a chain of procedures that involves: (1) carefully interviewing the blood donors to exclude donors with possible bacteraemia, (2) careful disinfection of the donors’ skin before venipuncture, (3) usage of a diversion pouch for collection of the first 30 ml of blood that may contain a skin plug, (4) sampling 10 ml of each PC the day after blood collection for an automatic bacterial culture systems, (5) visual inspection of each unit before the PC is issued for transfusion and discarding all PCs without swirling, and finally (6) emphasizing for the donor the importance of reporting to the blood bank if he or she gets ill within a few days after donation. If a patient develops a febrile reaction and bacterial contamination of the PC is suspected, it is essential that samples are collected from the patient for blood cultures and that the transfusion set and the platelet storage container is examined for bacterial contamination. However, the later part often gives rise to problems because the transfusion set and the platelet storage container have not been kept sterile. In order to conclude that the patient’s transfusion reaction is caused by bacterial contamination of the PC, the same bacterial strain found in the remains of the PC should also be present in the patient’s blood. Allergic transfusion reactions (ATRs) are the most frequently reported adverse event in transfusion [23]. PCs are involved in around 1/3 of these types of reactions, and together with plasma, PCs are associated with the more severe reactions [24], suggesting that blood components containing large amounts of plasma may be associated with more severe allergic transfusion reactions. Accordingly, replacement of plasma in the PCs with various platelet additive solutions have been shown to significantly reduced the frequency of allergic transfusion reactions [25, 26]. In most cases, ATRs are associated with IgE or IgG antibodies in the recipient’s serum reacting with drugs, chemicals (e.g. ethylene oxide) or allotypic serum proteins in the transfused blood. Although life-threatening reactions can occur in patients with IgA deficiency and anti-IgA due to previous immunization, anti-IgA is in fact only rarely the cause of allergic transfusion reactions. Complement-derived anaphylatoxins (C3a and C5a), cytokines, chemokines (such as CCL5 or RANTES), bradykinin, histamine and other biological response modifiers may accumulate in the blood component during storage, and these substances have also been implicated in ATRs [24, 27]. Measurement of tryptase, an enzyme released to serum during mast cells activation, is used at some centres as a diagnostic marker for anaphylaxis [28]. Checking the recipient’s serum for anti-IgA is usually also a part of the laboratory case workup. However, as the causes of allergic transfusion reactions are vast, the reason for an allergic adverse event usually remains unknown. Febrile non-haemolytic transfusion reaction (FNHTR) is suspected if the patient gets a fever (a rise of body temperature of more than 1°C) and complains of chills, rigors, and/or cold sensations, during or shortly after transfusion of a PC, and there are no other obvious explanations for these symptoms. The temperature increase, however, may be masked by antipyretics. With symptomatic treatment the patients usually recover rather quickly. FNHTR is seen more often after platelet than plasma or red cells transfusions. Before the implementation of universal leucoreduction of PCs, FNHTR was associated with around 1/3 of platelet transfusions [2]. Since pre-storage leucoreduction has been universally adopted there has been a significant reduction of the frequency of FNHTR [29, 30]. Recent studies suggest that the frequency of FNHTR can be further reduced by implementing pathogen reduction technology [4, 31]. The pathophysiology of FNHTR is complex and only partly known. Although it is well-known that a large number of biological response modifiers such as sCD154 (sCD40L), IL-1β, TNFα, CXCL4 (PF4), CXCL8 (IL-8), IL-6, CCL3 (MCP-1), complement activation products, and many others, accumulate in the PCs during storage [32, 33], the clinical role of each of these is not clarified. After implementation of pre-storage leucoreduction of PCs, those substances primarily produced by leucocytes are probably not of any clinical importance. Most laboratory investigations of FNHTR have been carried out in a research context, and because of the multitude of agents that have been suggested to be involved there is no consensus regarding which analyses should be included in a routine laboratory case workup for FNHTR. The presence of anti-platelet alloantibodies is a well-known cause of refractoriness to platelet transfusions. Such antibodies can also be present in donors who have previously been immunized through pregnancy or transfusion. However, most centres do not screen their platelet donors for anti-platelet antibodies, and therefore PCs containing anti-platelet antibodies can be transfused to patients. If the PC is given prophylactically, one will probably only notice an unsatisfactory post-transfusion platelet increment, but if given to a bleeding patient severe thrombocytopenia may occur [34], which may further increase the patient’s bleeding. Apart from causing thrombocytopenia, most anti-platelet antibodies are usually considered as clinically silent. However, both anti-HLA antibodies [35] and platelet specific antibodies [34, 36, 37] have been associated with allergic reactions [34, 35, 37] and FNHTR [34, 36]. How often are anti-platelet antibodies transfused to patients? We know that approximately 25% of female donors are HLA-immunized [18, 19] and if around half of the donors are females we can expect that anti-HLA antibodies are present in more than 10% of the collected units. The number of donors with platelet specific antibodies is much smaller. Assuming that half of the female donors have a history of a previous pregnancy, that 2% are HPA-1a negative and that 10% of these women develop anti-HPA-1a [38], then not more than 1 of 2000 PCs will contain anti-HPA-1a if produced by platelet apheresis. Moreover, due to the recent years’ focus on the association between anti-HLA-antibodies and TRALI, many centres have changed their policy and defer all female platelet apheresis donors with histories of prior pregnancies. Thus, for those centres, the number of apheresis PCs containing anti-platelet antibodies will not represent a major problem. How is the situation at centres that primarily produce PC-BC? Female donors with a history of prior pregnancies are not excluded from whole blood donations and many of these buffy coats will be used for production of PCs. Given this scenario we can expect that anti-HLA antibodies will be present in up to 40% of the PC-BC and anti-HPA-1a in one of 500 PCs. There is an apparent discrepancy between this high number of PC-BCs containing anti-HLA antibodies and how often transfusion of PC-BCs are associated with TRALI. There may be two reasons for this discrepancy. First, the volume of plasma from one HLA-immunized donor in a PC-BC is less than 20 ml and this volume is probably too small to elicit a case of full-blown TRALI. Secondly, in many cases anti-HLA class I antibodies from one donor will bind to HLA class I molecules expressed on platelets from one or more of the other donors. This will also be the case when anti-HPA-1a is present in one of the buffy coats. What will the consequences be if antibodies from one donor bind to platelets of one or more of the other donors in a PC-BC? First, it is conceivable that antibody-sensitized platelets will have reduced survival after transfusion, resulting in suboptimal post-transfusion platelet increment. Second, the antibodies by themselves may affect platelet function. Both anti-HLA and anti-HPA-1a antibodies can activate platelets [39-41] and anti-HPA-1a has been shown to induce release of the chemokine CCL5 (RANTES) from platelets [40], a pro-inflammatory chemokine that has been implicated in allergic transfusion reactions [42]. Thus, when anti-platelet antibodies are present in PC-BCs, it is possible that during storage these antibodies can increase the concentration of platelet-derived cytokines and chemokines in the PC-BC to levels that clinically will result in adverse events when the PCs are transfused. At our hospital we have recently had a case of full blown FNHTR in association with transfusion of a 3-day old PC produced from four buffy coats, where the laboratory case workup revealed that one of the donors had high level of anti-HPA-1a. It is, however, not known how often adverse events related to transfusion of PC-BCs are associated with the presence of anti-platelet antibodies in the PC. Platelet-related adverse events are a significant challenge in transfusion medicine. Core features of the pathophysiology of TRALI have been disclosed and there is an increasing consensus regarding preventive measures, laboratory case workup and donor management policies. The pathophysiology of ATR and FNHTR is multifaceted and complex and this is an impediment for standardization of laboratory examinations of such cases. Although anti-platelet antibodies usually are clinically silent, apart from causing thrombocytopenia, they may be of clinical importance in some cases where transfusion of PC-BCs are associated with adverse events. The author declares that there are no potential conflicts of interest.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.