Risk Factors for Transfusion-Related Acute Lung Injury

Discussion

To our knowledge, this is the first meta-analysis of risk factors of TRALI in critically ill patients. We summarized and analyzed the potential risk factors of TRALI for critical patients in 13 studies. The number of transfusions and units of fresh frozen plasma were risk factors related to blood transfusion. Age, female sex, tobacco use, chronic alcohol abuse, positive fluid balance, shock before transfusion, ASA score, and mechanical ventilation may be potential risk factors for TRALI, although < 5 studies were included. More studies are expected to be included for analysis. Tidal volume, liver disease, sepsis, blood tumor disease, blood products from female blood donors, red blood cell units, platelet units, pumping time during extracorporeal circulation, surgery, and liver transplantation were not identified as risk factors of TRALI. Some of these conclusions seem to be logical, but our research findings strengthen previous evidence. However, some evidence was contrary to previous conclusions.

In 1951, Barnard³¹ reported that a patient with acute leukemia died because of the acute lung reaction caused by an allergic reaction to blood transfusion, which was the first report of lung injury due to blood transfusion. After that, several scientists reported similar cases. Until the 1980s, TRALI was officially recognized as an independent disease.³² In 2004, the consensus meeting held by the CBS Agency standardized 2 entities for the first time: TRALI and possible TRALI (pTRALI), in which the cause of acute lung injury is not clear (ie, it could be ARDS or blood transfusion).^3,33 In 2005, the definition of the CBS was expanded to include patients with acute lung injury deterioration due to blood transfusion, and physicians were expected to determine whether a patient had TRALI or ARDS according to clinical standards and judgments.³³ In 2019, researchers used the Delphi Methodology to re-classify TRALI into 2 types: TRALI I (without risk factors of ARDS) and TRALI II (with risk factors of ARDS or existing mild ARDS).^34,35 In the 14 years after the CBS published the guideline, a large amount of information from prospective and retrospective research related to TRALI has been accumulated. A careful reading of the literature will show that previous definitions have not been strictly applied, and there is a large degree of heterogeneity in the incidence rates and mortality rates. In conclusion, there are several problems with the current body of evidence. First, the clinical manifestations and laboratory tests of TRALI were not independent pathological features, so it was difficult to differentiate diagnosis; in addition, the tools to perform the diagnostic analysis were not extensive enough.³⁶ Second, except for the measurement of blood oxygen saturation ( ), there was no objective laboratory test in the diagnostic criteria, so the diagnosis was rather subjective.² Next, some previous studies indicate that the risk factors of ARDS and TRALI may be similar, and some host-related factors were positively correlated with ARDS, such as mechanical ventilation, tobacco use, age, and shock. These led to unclear taxonomic boundaries between TRALI and ARDS and made it almost impossible to distinguish TRALI from other acute lung injuries in clinical practice.³⁴ Finally, according to the definition, TRALI appears temporarily within 6 h after blood transfusion, but a few cases have been reported to occur between 6 h and 48 h after blood transfusion, emphasizing the need for a standardized definition of TRALI.³⁷ To date, although there is a relatively large number of studies on TRALI, the inclusion criteria for the TRALI population were not uniform among the studies; in addition, the sample size in some studies is relatively small, the pathogenesis is unclear, and the research results are different.

Because the diagnostic criteria of TRALI in previous studies are not uniform, and to understand the risk factors of TRALI comprehensively, this study included studies on TRALI and pTRALI as per the definition of the CBS. According to the Berlin standard published in 2012, the term “acute lung injury” has been eliminated, and, according to the degree of hypoxia, ARDS was divided into mild, moderate, and severe.³⁵ For this reason, we also included cases of transfusion-related ARDS in the case group.

In terms of host-related risk factors, our results indicate that as patients aged, the risk of TRALI increased. Some studies have reported that most subjects in the elderly population exhibited chronic low-grade inflammation, which is manifested by high baseline levels of pro-inflammatory cytokines in the body, which are likely to cause TRALI.³⁸ In addition, women who receive blood transfusions are more likely than men to develop TRALI. Previous studies have reported that human leukocyte antigen I (HLA I), HLA II, and granulocyte antibodies are common in women who have given birth. When these antibodies encounter homologous antigens from the transfusion plasma, it can lead to neutrophil activation and oxidative substance release, which can damage the lung endothelium.^39,40 This shows that previous pregnancy experience is a very important factor leading to TRALI. Chronic alcohol abuse may also increase the risk, possibly because the level of glutathione, an antioxidant in the lungs, decreases, which in turn reduces the phagocytic function of the apoptotic cells, resulting in increased inflammation in the lungs.²⁷ Patients with a history of smoking have varying degrees of respiratory problems, which may increase the risk of disease.⁴¹ Shock can cause tissue damage and may induce TRALI by activating the patient's neutrophils.³⁴ The ASA score reflects the physical condition of the patient, and a high score may increase the risk of illness.⁴²

In terms of transfusion-related factors, our results indicate that the amount of transfusion and number of fresh frozen plasma units were significantly related to the occurrence of TRALI, while female sex of the donors and number of red blood cell units and platelet units did not seem to significantly increase the risk of TRALI. Since 2003, the British government mandated that female plasma be used in manufacturing to reduce the incidence of TRALI. Since the implementation of this approach, the number of TRALI cases in the United Kingdom has indeed decreased.¹⁸ The only test that can be used to diagnose TRALI is to monitor the content of HLA and granulocyte antibodies in the plasma of the recipient or donor. The principle is that the relevant antigens and antibodies contained in the donor and recipient will induce related inflammatory reactions, which will lead to lung injury. The discovery of antigen-antibody reactions strongly supports the diagnosis of TRALI. However, in one study, we did not find consistency between antigens and antibodies in 15% of TRALI cases, suggesting that a simple antigen-antibody reaction does not necessarily lead to clinical TRALI.⁴³ In a retrospective study, Toy et al⁴⁴ investigated a group of female donors with a history of pregnancy and had multiple antibodies related to HLA antigens. TRALI did not occur when the donated blood was transfused to 55 subjects with non-neutropenia of known HLA type.⁴⁴ In this regard, some researchers believe that the occurrence of TRALI may require 2 different events: one related to the clinical situation of the recipient (such as active infection), and the other related to the infusion of blood products containing HLA antibodies, granulocyte antibodies, or other biological response regulators.⁴³

In this study, we found that most host-related risk factors are characterized by a pro-inflammatory state, which may stimulate lung neutrophils as the “first attack” in the development of TRALI.⁴⁵ Therefore, the clinical state of the host may lead to a decrease in lung compliance before blood transfusion, as well as a deterioration in lung compliance after blood transfusion, which may lead to lung injury.⁴⁶ More studies indicate that host-related risk factors may play a more critical role in the development of TRALI than transfusion-related risk factors. This conclusion highlights the importance of the “first attack” mentioned above.⁹

In terms of device-related risk factors, our results indicate that mechanical ventilation may be related to the occurrence of TRALI. A previous study reported that mechanical ventilation accelerates and aggravates lung injury.⁴⁷ When the peak airway pressure is > 30 cm H₂O, the risk of acute lung injury increases significantly.^48,49 The above conclusions suggest that mechanical ventilation also plays an important role in triggering TRALI in the early stage and a synergistic role in the later stage with blood transfusion in the development of lung injury.

In terms of the mitigation strategies for disease prevention, through the analysis and summary of the included studies, we have drawn the following conclusions. First, in the future, it is necessary to have a more reasonable definition and classification of TRALI. Previous studies have reported that TRALI and pTRALI seem to have different etiologies, strategies, and clinical outcomes.²⁷ For TRALI, a restrictive transfusion policy was still strongly recommended.⁶ Compared with TRALI, ARDS risk factors and not blood transfusion risk factors were important driving factors of clinical outcomes in patients with pTRALI.²⁷ In the future, we need to further study the role of blood transfusion in the pathophysiology of pTRALI, and more rigorous prospective studies are needed for these patients. Second, many studies have shown the importance of host-related factors in the prevention of TRALI. For patients in poor condition, doctors may need to reconsider the restrictive transfusion policy.¹³ Studies have noted that, in critically ill patients, even if the restrictive transfusion strategy was well implemented, it may also lead to the occurrence of TRALI.² According to the 2 events model, most critical patients are characterized by a pro-inflammatory state, which plays an important role in the pathogenesis of TRALI, stimulating pulmonary neutrophils and acting as a “first event” in the occurrence of TRALI. The severity of their condition may contribute to a lower threshold to develop TRALI.^7-9 It is suggested that for patients in better condition, the mitigation strategy is biased in favor of the restrictive transfusion policy, but for patients in poor condition, the mitigation strategy may be biased in favor of the control of host-related factors.

Our study has some advantages and limitations. By formulating a systematic research plan, a comprehensive search strategy, and strict inclusion and exclusion criteria, as well as by evaluating the research quality, the objectivity and consistency of this study can be considered high. The findings could help both clinicians and patients interpret the current evidence.

Although the research is innovative, some limitations should be recognized. First, the majority of the studies included in the analysis were retrospective studies, and the diversity and complexity of TRALI-related influencing factors were affected to a certain extent by selection bias. Second, there were very few studies on some risk factors. In this case, the power of these factors is too weak to find a significant association with a relatively small impact. We believe that as the number of powerful cohort studies increases, clearer and higher levels of evidence will become available. Third, because some data are lacking (such as APACHE II scores), some potential risk factors could not be analyzed in groups. Finally, because the diagnostic criteria of TRALI have not been completely unified, there is variability in the diagnosis of TRALI in some studies. In this regard, we formulated the inclusion criteria for meta-analysis according to the needs of this study to provide guidance for clinicians as soon as possible.