Universal leukoreduction and acute transfusion reactions: putting the puzzle together

Significant progress has been made over the past decade to prevent transfusion-transmitted infections; however, acute reactions to blood transfusion continue to occur. Acute reactions are defined as adverse events occurring during or within a few hours of the transfusion. Based on our understanding of the eitology of these events, they have been classified as hemolytic, febrile nonhemolytic, allergic, anaphylactic, TRALI, hypotensive, or sepsis owing to bacterial contamination. A variety of metabolic complications can also occur.1 Historically, febrile nonhemolytic transfusion reactions (FNHTRs) are the adverse events most commonly encountered with reported frequencies ranging from 0.5 to 38 percent depending on the product and the type of patient being transfused.2-4 FNHTRs were first characterized in the late 1950s and early 1960s and were named after the signs and symptoms which typically occurred: a fever (>1°C increase in temperature) that was often accompanied by chills, cold, rigors, and discomfort.5,6 Because our understanding of these events has improved over the past decade, it may be time to revisit the original terminology that has been used for more than four decades. Recent studies have shown that symptoms of chills, cold, and rigors often occur in the absence of fever.7,8 If these afebrile reactions are an alternative manifestation of febrile nonhemolytic reaction, then the term FNHTR is not an appropriate descriptor of these events. Use of the current classification could also lead to underreporting or variability in reporting, depending on whether the afebrile events are included or excluded. Perhaps it is time to give FNHTRs a new name, allowing for afebrile reactions with similar symptoms to be included in the classification. A standardized classification of these events would provide reporting consistency and facilitate interstudy comparisons. FNHTR were first described following RBC transfusions;5,6 however, when technology allowed for PLT production and storage for up to 5 days, these reactions were reported even more frequently.9-11 WBC antibodies in the recipient's plasma were identified as the cause of most FNHTRs after RBC transfusions.6,12-14 These antibodies reacted with WBCs present in the transfused blood product resulting in the release of pyrogens that caused fever. This theory was supported by observations that the removal of WBCs from the RBC product would prevent most reactions.6,13,14 The PLT scenario was different. Reactions still occurred after removal of WBCs and men sometimes reacted to their first PLT transfusion, raising questions about the validity of the antibody mechanism.11,15 These observations along with a series of experimental studies have confirmed that most FNHTRs to PLTs are not mediated by antibody but are caused by WBC-derived cytokines that accumulated in the PLT product during storage.16-18 When WBCs are removed from PLTs before storage, cytokines are not generated and the frequency of reactions is greatly reduced.8,18 Nevertheless, residual reactions have still been reported in approximately 2 percent of prestorage leukoreduced PLT products transfused to hematology-oncology patients.8 When storage-generated cytokines were implicated as the cause of PLT reactions, the question arose as to whether or not they also played a causative role in RBC reactions. The cytokine theory was supported by the fact that an association had been shown between the age of the RBC product and the likelihood of a reaction.11,19 The cytokine theory was not supported by the observations that leukoreduction after storage effectively prevented most RBC reactions12-14 and that storage-generated cytokines were not readily detectable in RBC products.20,21 Regardless of the mechanism, leukoreduction of PLT and RBC products before storage should decrease the frequency of FNHTRs; however, there has been controversy as to the clinical benefit that would be observed if leukoreduction before storage was implemented universally. There has been one randomized controlled trial that has shown leukoreduction of PLTs before storage to be more effective at preventing FNHTRs than leukoreduction or plasma removal after storage; however, this study focused on a selective patient population and was not designed to assess the impact of universal leukoreduction.8 The data to assess the impact of universal leukoreduction on RBC reactions are also controversial. One US retrospective observational study concluded that universal leukoreduction of RBCs before storage had no impact on the frequency of acute reactions.22 In contrast to these results, an observational study from the UK demonstrated a significant benefit to universal leukoreduction with a 50 percent relative risk reduction.23 The only randomized controlled trial assessing the effectiveness of prestorage leukoreduction to prevent FNHTRs to RBCs showed a trend (not significant) toward fewer acute reactions in the group receiving products that were universally leukoreduced before storage.24 In this study, reactions were reported as a secondary outcome measure. The reason for the lack of randomized controlled trials to investigate the effectiveness of universal leukoreduction on acute transfusion reactions is obvious. With a reaction frequency of 1 percent or less, the sample size required for an appropriately powered study would be huge, making such a study costly and logistically difficult—perhaps impossible—to perform. Nevertheless, there are other methods that can be used to provide evidence of effectiveness when dealing with rare events. In this issue of TRANSFUSION, three different groups have used retrospective cohort study designs to assess the impact of universal leukoreduction before storage on preventing acute reactions to RBCs and/or PLTs. All three studies had the same objective: to compare the frequencies of acute reactions to RBC and/or PLT transfusion during a period when nonleukoreduced products were routinely given to a period when universal leukoreduction had been implemented. All three investigative groups were able to address this question because they had a hemovigilance system in place to ensure that reactions were systematically documented and classified. The study by Yazer and colleagues25 investigated both RBC and PLT reactions occurring in seven Canadian health-care facilities during a period before and after implementation of universal leukoreduction. In the two American studies by Paglino and coworkers26 and King and coworkers,27 the implementation of universal leukoreduction before storage was gradual, allowing also for a trend analysis during the transition phase. Paglino and associates26 assessed reaction frequencies to both RBCs and PLTs, whereas King and collegues27 focused only on RBCs. In spite of the practice differences that undoubtedly occurred among the three groups, the study results are surprisingly similar. RBC reactions before leukoreduction ranged from 0.33 to 0.37 percent. After leukoreduction, the frequency of FNHTRs to RBCs was reduced, ranging from 0.15 to 0.19 percent. Yazer and colleagues25 and Paglino and colleagues26 also assessed the impact of universal leukoreduction on PLT reactions. Although the baseline reaction frequencies were more variable (0.45 and 2.18 percent, respectively), the reaction frequencies after implementation of universal leukoreduction were similar (0.11 and 0.15 percent, respectively). These reported frequencies to PLTs that were leukoreduced before storage are lower than the frequency reported by our group, but this discrepancy can be explained. Our studies were performed in the context of a randomized controlled trial using a selective patient population (hematology and/or oncology patients) and employing rigorous follow-up on each patient who received transfusion.7,8 This type of study design is aimed at assessing efficacy and is often referred to as an explanatory study. The three studies in this issue of TRANSFUSION would be classified as effectiveness or “pragmatic” studies, because they are asking a much broader question.28 Does leukoreduction before storage reduce the frequency of acute reactions to blood products if applied to all transfused patients (i.e., universal application)? The consistent results of the three studies provide good evidence that universal prestorage leukoreduction is effective in reducing the frequency of FNHTRs to both RBCs and PLTs, although one could argue the cost effectiveness of such as approach. Indeed, the decision to implement universal leukoreduction is complex and in most situations has been based on multiple benefits including a reduction in alloimmunization, prevention of CMV transmission, and fewer acute adverse events.29 So where do we go from here? The primary scientific issue remaining to be resolved is to understand why the occasional patient continues to react to the leukoreduced product. All three of the groups have speculated that PLT-derived CD40L or other soluble mediators (vascular endothelial growth factor or transforming growth factor-β1) could be responsible for these residual reactions.30-32 Identifying the causative agents is challenging, but identifying effective interventions to prevent these reactions is within the realm of evidence-based medicine. The information-rich patients who are worthy of additional studies are those individuals who have repeated reactions to leukoreduced blood products. A specific study design, termed an N of 1 study, could be used to identify an effective intervention for a patient who consistently reacts to blood products that are leukoreduced before storage.33 This design involves a randomized controlled crossover trial in just one patient. For example, the patient who has repeated reactions to leukoreduced RBC products could be given either fresh RBCs or RBCs that have been washed to remove any residual substances that may be generated during storage. The intervention selected would be paired with the standard product and the transfusion order within each pair randomly selected. The patient would agree to undergo a series of transfusions that were blocked in this manner and would be carefully observed to determine if there was a difference in the frequency of reactions to the two product types. In most situations, a significant answer can be obtained after a maximum of five to seven transfusion pairs. This type of study design is applicable only if the effect of the intervention is short-lived and if the baseline state for the patient is similar for each transfusion. The patient also must be willing to collaborate and the transfusion service must be able to provide the appropriate product as needed. By applying this type of method, we will gain additional evidence-based information about effective ways to manage these residual reactions and possibly gain insight into their cause. Finally, not only do the three studies in this issue of TRANSFUSION add to our knowledge about the effectiveness of universal leukoreduction and adverse reactions, but they also illustrate an important methodologic concept: reporting of results and how the method of reporting could influence interpretation. The articles by Yazer et al.25 and King et al.27 report the results of their studies by providing event frequencies in the leukoreduced and nonleukoreduced periods, allowing for an easy calculation of the absolute risk reduction for RBCs of 0.14 and 0.22 percent, respectively. Paglino and colleagues26 had a similar absolute risk reduction (0.16%) but chose to express their results as a relative risk reduction of 47.1 percent. A third approach that could have been used to express the results is the concept of “number needed to treat.”34 Using the data from Paglino et al., 625 patients would have to be treated with products that were leukoreduced before to prevent one reaction. It can be seen from these summary statistics that different interpretations of clinical relevance could arise depending on the reporting option selected. There are no guidelines on the optimal summary statistic to use. Authors are free to choose their method of reporting but must clearly specify whether the risk reduction is relative or absolute.35 More importantly, readers should understand these summary statistics and their interpretation and remember to consider the baseline event rate when deciding on the clinical relevance of such results. In summary, the three studies in this issue of TRANSFUSION have applied appropriate methods for studying rare events and provide consistent and significant evidence that universal leukoreduction can prevent FNHTRs to both PLTs and RBCs. The clinical impact is limited because of the low baseline incidence of these events; however, these data should help to resolve the controversy of whether or not universal leukoreduction prevents FNHTRs to RBCs. Like putting together pieces of a puzzle, the picture is finally beginning to form. It is now time for the transfusion community to focus on improving our understanding of residual reactions and adopting a standardized classification to include afebrile events. Only then will the final pieces of the puzzle surrounding FNHTRs fall into place. This work is supported by a Canada Research Chair, Canadian Institute of Health Research, Ottawa, Canada and Canadian Blood Services, Ottawa, Canada. The clerical assistance of Jeanette Button is also acknowledged.

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.