Hitoshi Okazaki, Research and Development Department, Central Blood Institute, Blood Service Headquarters, Japanese Red Cross Society, 2-1-67 Tatsumi, Koto-ku, Tokyo 135-8521, Japan E-mail: firstname.lastname@example.org
Transfusion-related acute lung injury (TRALI) is one of the most serious complications of blood transfusion. We have been gathering information about TRALl in Japan since 1997 through our national haemovigilance system. Recently, we have 20–40 TRALI cases per year (five million bags issued per year). Although TRALI is a rare complication of transfusion, it may become severe and may sometimes be fatal. A substantial number of TRALI cases have been caused by donor HLA or HNA antibodies, which were generated in alloimmunized, i.e. parous females. Considering this mechanism, TRALI mitigation strategies such as the use of male-only-plasma have been implemented in many blood centres or countries. Because of the availability of male plasma and the time limit for the production of fresh frozen plasma (FFP), we first tried to prepare FFP-LR2 derived from 400 ml whole blood only from male donors. FFP-LR2 accounts for 75% of FFP in Japan. (Six per cent of FFP products are from 200 ml whole blood and 19% from apheresis plasma.) A preliminary project on male-predominant plasma was started at several core blood centres. To minimize the burden on nurses regarding marking bags collected from males or females, which might lead to serious errors in collection sites, we updated the computer system that allows the staff members in the production department to sort male and female blood automatically. In a month after the system implementation nationwide, we successfully increased the production rate of male plasma for FFP-LR2 product by up to more than 98%. Regarding FFP-LR1 derived from 200 ml whole blood, the percentage of female donors is approximately 75%. One of the core blood centres has successfully implemented the same strategy for producing FFP-LR1, but others have not done so yet. Although male-only-plasma strategies achieved good results without any expensive screening tests to reduce TRALI incidence in some countries, the standard measures for other plasma rich blood products such as platelet concentrates or apheresis plasma have not been established yet. There are some measures to consider reducing the risk of TRALI caused by other plasma-rich products – recruiting only males as apheresis platelet donors, testing a certain number of donors for HLA antibodies, replacing the supernatant of a platelet pool with male plasma and excluding female donors from apheresis donations. Some blood centres adopted some of these measures, but so far there is little evidence for the efficacy of these measures for reducing the risk of TRALI. We have recently conducted research about the relationship between strength of HLA antibodies and TRALI development. On the basis of the results of the research, we can set the cut-off level in HLA antibody tests, which can reduce the risk of TRALI by eliminating strong HLA antibodies that presumably cause TRALI development if the specificity of an antibody matches the patient’s antigen. Preliminary screening for donor HLA antibodies is ongoing in three core blood centres in Japan, results of which enable us to establish effective and feasible screening strategies for HLA antibodies in the near future.
Transfusion-related acute lung injury (TRALI) is one of the most serious complications of blood transfusion. We have been gathering information about TRALl in Japan since 1997 through our national haemovigilance system. Recently, we have had 20–40 TRALI cases per year (5 million bags issued per year). Although TRALI is a rare complication of transfusion, it may become severe and may sometimes be fatal. It has been the most common cause of transfusion-related mortality reported to the U.S. Food and Drug Administration in recent years .
Male-only plasma strategy
A substantial number of TRALI cases have been caused by donor HLA and/or HNA antibodies generated in alloimmunized, i.e., parous females. Considering this mechanism, TRALI mitigation strategies such as the use of male-only plasma have been implemented in many blood centres or countries. The Serious Hazards of Transfusion report from the United Kingdom first showed that the male-predominant plasma strategy for fresh frozen plasma (FFP) and platelet pools has been effective for TRALI mitigation . The TRIP Dutch National Haemovigilance Office reported that there was a 33% decrease in the number of TRALI cases after implementing the male-only quarantine FFP strategy as a preventive measure . The American Red Cross also reported that as the percentage of male plasma increased from 55% in 2006 to 95% in 2008, there was a significant decrease in the number of TRALI cases involving only plasma transfusion in 2008 compared with 2006 (32 cases vs. 7). Because of the availability of male plasma and the time limit for the production of FFP in the United States, the production of FFP was shifted to that of FP24, which was frozen within 24 h from blood collection, to prepare male-predominant plasma .
The manufacturing and inventory management problems concerning the time limit for freezing plasma have been raised along with the establishment of the project for consolidating blood centres in Japan. However, we do not have the license for producing FP24 and we have three different plasma products, namely, FFP-LR1 prepared from 200 ml whole blood, FFP-LR2 prepared from 400 ml whole blood, and FFP-LR Ap prepared by plasmapheresis, which make the above-mentioned problem more complex.
Among these three plasma products, FFP-LR2 is the most common product, which accounts for 75% of the total FFP supply in Japan; in addition, 80% of FFP-LR2 has already been prepared from male donors. We first attempted to prepare FFP-LR2 only from male donors. A preliminary project of preparing male-predominant plasma started at several core blood centres. In order for this project to precede, a hand-written marking for male or female blood bag at collection sites was required, which might entail some extra work for blood bank nurses. This problem was subsequently solved with minimal extra work for blood bank nurses. To minimize the burden on nurses regarding marking bags collected from male or female, which might cause serious errors in the collection sites, we updated the computer system that allows the staff members in the production department to sort male and female blood donations automatically. In a month after system implementation nationwide, we successfully increased the production rate of FFP-LR2 from male plasma up to more than 98% (Fig. 1)
The male-predominant-plasma strategy can be relatively easily implemented in countries where the amount of plasma used is relatively small. In Japan, the amount of plasma used is relatively large, moreover, as we mentioned above, we have three different plasma products so far; this situation makes implementation of the strategy more complex. Regarding FFP-LR1 derived from 200 ml whole blood, the percentage of female donors is approximately 75%. One of the core blood centres has successfully implemented the same strategy for producing FFP-LR1 as that of FFP-LR2, but other core blood centres have not done so yet. Moreover, we continuously increase the rate of 400 ml whole-blood donation instead of 200 ml whole-blood donation, which will subsequently raise the percentage of FFP prepared from male donors.
HLA antibody screening
Although the male-only-plasma strategy achieved good results without any expensive screening tests for reducing TRALI incidence in some countries, standard measures for other blood products have not been established yet. If we apply the same measures to red blood cells or apheresis platelet products, there will be shortage of these products. Regarding plasma-rich products, there are some measures to be considered for reducing the risk of TRALI: recruiting only males as apheresis platelet donors, testing certain donors for HLA antibodies, replacing the supernatant of a platelet pool with male plasma, and excluding female donors from apheresis donations. Some blood centres adopted some of these measures, but so far, there has been little evidence for these measures to be effective for reducing the risk of TRALI.
The ISBT working party on granulocyte immunobiology published the recommendations for leucocyte antibody screening in the investigation and prevention of antibody-mediated transfusion-related acute lung injury in 2009 . In addition to the conventional lymphocyte toxicity (LCT) assay or lymphocyte immunofluorescence test (LIFT), highly sensitive assays using enzyme immunoassays or flow cytometry with microbeads were recommended in this report.
There have been several studies of HLA antibody screening. A study targeting 96 apheresis platelet donors was performed by Fadeyi et al. in 2008 . In this study, they compared enzyme-linked immunosorbent assay (ELISA) with high-throughput microbead-flow assay and showed that microbead-flow assay was more sensitive than ELISA. Powers et al. conducted a study of the efficacy and safety of different TRALI-risk-reduction strategies . In this study, they examined pregnancy and transfusion histories in relation to the presence of HLA alloantibodies detected by flow-cytometry-based screening. Moreover, they estimated the predicted cost using different screening strategies. They concluded that examining donor history, which is a reliable predictor of alloimmunization, and testing only alloexposed donors were cost-effective TRALI prevention strategies. Vassallo et al. compared the efficacy and safety of different TRALI-risk-reduction strategies employing more than 2000 plateletpheresis donors using either ELISA (DonorScreen, GTIdiagnostics, WI, USA) or a prototype automated platform, microbeads, and a Luminex-based flow analyzer (One Lambda, Inc., CA, USA) . In this study, alloexposed females showed the highest reactivity on both platforms, and a much lower rate for nonexposed donors. They concluded that a strategy of only screening parous donors is reasonable. Several studies were conducted by the NHLBI Retrovirus Epidemiology in Donor Study II (REDS-II) group in the United States. The studies showed that the prevalence of HLA antibodies increases with more pregnancies . The group compared among five different HLA antibody tests . The complete concordance among different assays was not observed, but there was a reasonable agreement among assays. The group also showed the rough equivalent cutoffs by different assays. The RED-II study group showed different cutoff levels with their resulting donation losses, and provided the guidelines for the decision on TRALI-risk-reduction strategies at blood centres, because one of the major concerns is a balance between the cutoff level of an HLA antibody test and the percentage of donor loss .
Despite these extensive investigations about HLA antibody testing, whether such HLA antibody screening could really reduce the risk of TRALI remains uncertain. Recent clinical studies have already shown that the presence of HLA or HNA antibodies in associated blood products is a risk factor for developing TRALI in cardiac surgery patients . This finding was consistent with that of another previous study by Gajic et al., in which transfusion risk factors were compared between patients who did and did not develop acute lung injury . Although the presence of HLA and/or HNA antibodies could be a risk factor for TRALI, the positivity rate of an HLA antibody test depends on the cutoff level of the assay used. Therefore, we should be very careful when interpreting these kinds of data. More recently, a large multicentre retrospective cohort study has been conducted to evaluate whether HLA antibody screening of donors would reduce the risk of TRALI or possible TRALI . In that study, TRALI incidence was compared between the recipients who received a plasma-rich component from HLA- antibody-positive donors defined by relatively high cutoff levels (NBG values greater than 10·8 for Class I and greater than 6·8 for Class II) and negative donors (NBG values of less than 2·2). There was no significant difference, but the incidence in recipients who received components from HLA-antibody-positive donors was relatively higher.
The concept of preventing TRALI by HLA antibody screening is based on the prediction that the presence of highly reactive HLA antibodies poses a risk of TRALI development. Although the threshold model for TRALI was developed by Bux et al. , this prediction was not scientifically supported until our recent study was published. We conducted a retrospective observational study concerning the strength of donor HLA antibodies and TRALI development using recipients with transfusion complications including TRALI, possible TRALI and other non-haemolytic transfusion reactions . The specificities of HLA antibodies were determined in donors associated with transfusion complications, and recipient antigens were also determined. Of 1038 cases of transfusion complication, we detected cognate antigens specific to donor HLA antibodies in 18 cases of TRALI, four cases of possible TRALI and three cases of NHTRs. The strength of HLA antibodies was estimated on the basis of mean fluorescence intensity determined by single-bead assay, and sum of MFIs was calculated. The sum of MFIs in TRALI cases was significantly higher than that in NHTR cases. The associated donors’ samples were subjected to ELISA for detection of HLA antibodies, and the same tendency was observed. The ROC curve for the prediction of TRALI development was drawn and the optimal cutoff for ELISA screening was proposed for preventing TRALI. Although weak HLA antibodies may provoke TRALI reactions in certain critically ill patients, the observation that strong antibodies associated with TRALI confirmed the appropriateness of using HLA antibody screening as a TRALI-risk-reduction strategy.
Among several HNA antibodies, the HNA-3a antibody is considered to pose a high risk. A German study group reported seven cases of HNA-3a-antibody-associated TRALI cases within only 2 years between 2006 and 2007 . Considering the similar frequency of the HNA-3a antigen among different ethnicities , few HNA-3a-antibody-associated TRALI cases were detected in Japan (so far, only two cases in 15 years) and other countries. This low frequency of HNA-3a antibody detection in TRALI cases or a blood donor population should be examined if a sensitive high-throughput assay of HNA-3a is to be developed. Other HNA antibodies such as 1a, 1b, 1c, and 2a can be tested by several methods including beads assay (LABScreen MULTI, OneLambda, Inc., CA, USA). The REDS-II group in the US reported the low frequency (0·7%) of HNA antibodies in the donor population of over 1000, and concluded that no urgent implementation of donor HNA antibody testing is needed at present .
There were a substantial number of TRALI cases in which no HLA/HNA antibodies were detected in associated donors. Before implementing prestorage leucocyte reduction, the combination between patient antibodies and donor leucocytes may be related to the development of TRALI. However, after implementing prestorage leucocyte reduction, the possibility of TRALI development owing to this combination was thought to be considerably reduced. A recent Korean report of two cases of TRALI related to patient antibody warned of the possibility that even after leucoreduction, donor leucocytes bound to patient antibodies may initiate TRALI reactions . If such cases are detected, antigen-matched blood products should be provided.
Antibodies other than HLA/HNA
One of the other possible triggers of TRALI is a donor antibody other than HLA/HNA antibodies. We have recently found a high-titer CD36 antibody in a donor that caused respiratory complication after transfusion, the mechanism of which may be similar to that induced by HLA class II antibodies, because CD36 antigens are usually on the surface of platelets and monocytes, not on granulocytes . Another antibody we have recently found to be associated with TRALI is the siglec-14 antibody . This antibody can stimulate neutrophils in vitro. Thus, antibodies other than the known HLA/HNA antibodies in donors associated with TRALI cases should be scrutinized hereafter. Screening policies for such antibodies will be determined by their frequency and the estimation of increased risk of TRALI.
A recent study by Food and Drug Administration (FDA) researchers revealed that UV-treated platelet concentrates cause TRALI in lipopolysaccharide pretreated animal models even after washing platelets . This finding raises the caution about using activated platelets for patients with underlying diseases, and also about the possibility that platelets themselves cause acute lung injury. This observation was consistent with the study using a mouse model of TRALI, in which platelets are essential for TRALI development .
For so-called nonimmune TRALI, there have been no preventive measures so far. However, the shelf lives are 21 days for packed red blood cells (PRBCs) and 4 days for platelet concentrates in Japan. These short shelf lives probably reduce the risk of excess accumulation of any cell-derived biological response modifiers such as lysophosphatidylcholine (LPC). A study by Silliman et al. revealed that using 42-day-old RBCs or 5-day-old PCs increases the risk of TRALI [25,26]. One clinical study of critically ill patients revealed that the accumulation of LPC was found as a risk factor for the development of TRALI . Moreover, a recent study using a trauma-haemorrhage rat model of TRALI showed that using more than 28 days old PRBCs was related to an increased mortality . The relationship between the age of blood products and the possibility of increased risk of respiratory complications should be evaluated in a large-scale prospective study in the future.
At this point, the complete prevention of TRALI has not been achieved yet. Although the mechanism underlying the development of TRALI has been gradually elucidated, no strategies other than the use of male-only plasma have been evaluated so far. The screening for HLA antibodies has been adopted by many blood centres, considering the balance between safety and stable supply without reliable evidence. In the near future, we should verify whether this preventive measure indeed reduces the risk of TRALI.