A threshold model for the susceptibility to transfusion-related acute lung injury

Disponible en ligne sur

www.sciencedirect.com Transfusion Clinique et Biologique 19 (2012) 109–116

Review article

A threshold model for the susceptibility to transfusion-related acute lung injury Mécanisme de développement d’un œdème pulmonaire aigu lésionnel post-transfusionnel (TRALI) : le modèle du seuil de déclenchement U.J. Sachs Institute for Clinical Immunlogy and Transfusion Medicine, Justus Liebig University, Langhansstr. 7, 35392 Giessen, Germany Available online 5 June 2012

Abstract Transfusion-related acute lung injury (TRALI) is a serious, often life-threatening pulmonary transfusion reaction characterized by noncardiogenic lung oedema, hypoxemia and respiratory distress in temporal association with blood transfusion. The critical mechanism in TRALI is the sudden increase in permeability of the pulmonary endothelium and the subsequent, often extensive shift of fluid into the alveolae. The rapid clinical recovery seen in most patients makes it likely that this is a temporary phenomenon. Reactive oxygen species released by neutrophils or other cells are attractive candidate mediators of this process. There is experimental and clinical evidence that several pathways can induce barrier breakdown in TRALI, a concept known as the threshold model of TRALI. Surprisingly, neutrophils may not always be required. Other cells may play a role as multipliers or attenuators of TRALI, depending on recipient-related and transfusion-related factors involved. This review will summarize recent findings on pathophysiology, with a focus on newly discovered or disenchanted recipient-related and transfusion-related risk factors for TRALI and will present the threshold model of TRALI as a unifying concept on how TRALI develops. © 2012 Elsevier Masson SAS. All rights reserved. Keywords: TRALI; Pulmonary endothelium; Transfusion-related risk factors; Threshold model

Résumé L’œdème pulmonaire lésionnel post-transfusionnel, le transfusion-related acute lung injury (TRALI), est une réaction transfusionnelle sévère, mettant souvent en jeu la vie du malade, caractérisée par un œdème pulmonaire non cardiogénique, une hypoxémie et une détresse respiratoire associés dans le temps avec une transfusion sanguine. Le mécanisme central en est l’accroissement soudain de la perméabilité de l’endothélium pulmonaire dont découle une fuite extensive de liquide vers les alvéoles pulmonaires. La récupération clinique rapide, observée chez la plupart des patients, oriente vers un phénomène transitoire. Les dérivés réactifs de l’oxygène libérés par les neutrophiles ou par d’autres cellules constituent des candidats médiateurs attractifs de ce processus. Il existe des preuves expérimentales et cliniques que plusieurs voies physiopathologiques peuvent opérer, dans le TRALI, pour induire la rupture de cette barrière, cela a conduit à un concept connu comme le modèle du seuil de déclenchement. Il est surprenant de constater que les neutrophiles pourraient ne pas être constamment nécessaires. D’autres cellules pourraient, selon des facteurs pouvant dépendre du receveur ou de la transfusion elle-même, jouer un rôle amplificateur ou atténuateur dans la survenue du TRALI. Cette revue se concentrera sur les progrès récents concernant la physiopathologie du TRALI, certains facteurs de risques, associés au receveur ou à la transfusion, récemment mis en exergue ou contestés, cela nous conduira à présenter le modèle du seuil de déclenchement comme un concept unificateur du mécanisme de développement d’un TRALI. © 2012 Elsevier Masson SAS. Tous droits réservés. Mots clés : TRALI ; Œdème lésionnel post-transfusionnel ; Endothélium pulmonaire ; Facteurs de risques associés à la transfusion ; Modèle du seuil de déclenchement

1. Introduction

E-mail address: [email protected] 1246-7820/$ – see front matter © 2012 Elsevier Masson SAS. All rights reserved. http://dx.doi.org/10.1016/j.tracli.2012.03.003

Transfusion-related acute lung injury (TRALI) is a serious, often life-threatening pulmonary transfusion reaction characterized by non-cardiogenic lung oedema, hypoxemia

110

U.J. Sachs / Transfusion Clinique et Biologique 19 (2012) 109–116

and respiratory distress in temporal association with blood transfusion. Since Barnard’s initial description in 1951, noncardiogenic lung oedema related to transfusion has been widely reported using various designations, until in 1983, Popovsky et al. coined the term “transfusion-related acute lung injury” to emphasize the similarity in pathophysiology with the syndrome of acute lung injury seen in critically ill patients [1,2]. With the first analysis of a large series of patients reported in 1985 [3], TRALI was recognised as a distinct clinical entity. Broader recognition of TRALI, however, was gained only after publication of the results of the British “Serious Hazards of Transfusion” (SHOT) initiative which consistently demonstrated since 1996 that TRALI is one of the most common causes of transfusionassociated major morbidity and death [4]. Since our initial proposal of a threshold model of TRALI that considers both recipient-related and transfusion-related factors as relevant contributors to the pathomechanism of TRALI in 2007 [5], a number of important clinical and experimental findings were reported that further support our idea of a threshold that needs to be overcome before TRALI precipitates. These aspects will be summarized in this review. 2. Clinical features The critical mechanism in TRALI is the partial breakdown of the endothelial barrier separating the pulmonary capillary from the alveola; the subsequent (often extensive) shift of fluid into the alveolae accounts for the clinical picture of TRALI. Typical TRALI is characterised by respiratory distress, hypotension, fever and cyanosis appearing within minutes to just a few hours from the initiation of the relevant blood transfusion. In rare atypical cases, TRALI has been reported to occur later [6]. The conscious patient describes tightness in the chest, feels short of breath and develops a dry cough. He may also describe nausea, dizziness and rigours. On examination, the patient is hypoxic, often hypotensive and with tachypnoea and tachycardia. Widespread crepitations are heard on auscultation of the chest. In artificially ventilated patients, TRALI is indicated by hypoxemia, i.e. a sudden drop of the arterial oxygen tension, and sometimes copious frothy oedema can ooze from the endotracheal tube. The nature and quantity of this exudate are often remarked on by attending anaesthetists and it may be considered one of the hallmarks of severe TRALI; anaesthesiologists also often comment that the lungs “feel heavy” and are difficult to ventilate as the systemic blood pressure simultaneously begins to decrease, occasionally quite precipitously [7]. These initial symptoms are caused by the onset of pulmonary oedema. The subsequently performed chest radiograph shows bilateral generalized lung infiltrations. The florid radiologic picture is often described as “white lung”. There is sometimes a remarkable dichotomy between the florid radiologic picture accompanying significant oxygen desaturation and the paucity of auscultatory findings. Discrepancies in haemovigilance reports resulted in the development of a consensus view on which symptoms should be considered to be a sine qua non for a diagnosis of TRALI. The European Haemovigilance Network (EHN) suggested: 1) the occurrence of respiratory distress within 6 hours of initiation

of transfusion; 2) the absence of signs of circulatory overload, and 3) the radiographic evidence of new bilateral pulmonary infiltrates [8]. These are basically the criteria which Popovsky and Moore proposed in 1985 [3]. The Canadian Consensus Conference in Toronto in 2004 added the criteria 4) hypoxemia (PaO2 /FIO2 < 300 mmHg or pulse oximetry < 90% or other clinical evidence) and 5) no presence of other risk factors for acute lung injury [9]. If there is another acute lung injury risk factor present, “possible TRALI” should be diagnosed. Typical episodes of TRALI improve within 48 hours, although in a minority hypoxemia and pulmonary infiltrates can persist at least 7 days [3]. The severity of TRALI expands from mild, which can be treated by oxygen supply, to severe with necessity for mechanical ventilation. The reported mortality rate may be as high as 47%, but is between 10 and 20% in most reports [10,11]. In survivors, there seems to be no sustained pulmonary injury [3]. 3. Breakdown of the pulmonary endothelial barrier How can the pulmonary endothelial barrier breakdown in patients with TRALI be explained? First of all, histopathologic findings from fatal TRALI cases consistently demonstrate interstitial and intra-alveolar oedema [3,12–17]. Another consistent finding in TRALI is the presence of increased numbers of neutrophils within the pulmonary capillary vasculature and small pulmonary vessels [16,17]. On electron microscopy photographs, neutrophils were degranulated and focally in direct contact with denuded stretches of the capillary wall. A positive correlation has been reported between capillary leukostasis and the amount of proteinaceous fluid within the alveolar airspaces [17]. From these observations, it appears that the recipient’s neutrophil is central to the occurrence of endothelial barrier breakdown. In accordance with this observation, the indispensability of neutrophils was demonstrated in several experimental settings by different groups, although the actual mechanism by which neutrophils affect the barrier breakdown remained unanswered [18–21]. Some authors speculated that activated neutrophils may release their cytotoxic microbicidal arsenal and lead to apoptosis/necrosis of the endothelium [22,23]. The rapid clinical recovery seen in most patients, however, makes it more likely that the barrier is only temporarily disrupted. Reactive oxygen species are an attractive candidate mediator, because they are well known for their influence on endothelial permeability even if transfused as a pure substance in animal models [24–26]. Furthermore, interfering with reactive oxygen species production attenuated experimental TRALI in different studies [20,23,27]. Another finding is drawing our attention towards reactive oxygen species: there is new experimental evidence that neutrophils may not always be required to induce barrier breakdown in TRALI [28,29]. Depending on the transfusion-related event that leads to TRALI, other cell types such as monocytes, platelets, or pulmonary endothelial cells themselves may produce reactive oxygen species and lead to increased permeability of lung capillaries [28–30]. If this mechanism proves to be true, it would also help to explain reports on TRALI in neutropenic patients [31]. The initial idea of a simple model with two players

U.J. Sachs / Transfusion Clinique et Biologique 19 (2012) 109–116

Fig. 1. Players, multipliers, and attenuators in transfusion-related acute lung injury (TRALI). The initial mechanistic explanation that a substance contained in a blood transfusion (an antibody or a mediator) has an activating effect on the neutrophil, which in turn (after activation) will lead to endothelial barrier disturbance (pathway a), requires adaption. In the case of antibody-mediated TRALI, antibodies can activate neutrophils, as demonstrated for HLA (human leukocyte antigen) class I and human neutrophil antigen (HNA) antibodies, but there is evidence that some antibodies (namely, HNA-3a and possibly HLA class I) do primarily affect the endothelium itself (pathway b). Other cells can aid (multiply) or hinder (attenuate) the process. Monocytes have been described as multipliers for HLA class II (and, potentially, HLA class I) antibodies. Platelets may be involved as producers of reactive oxygen species that disturb the barrier function. Whether or not complement activation multiplies TRALI is not yet fully unravelled. Chemokine scavenger receptors on red blood cells may keep chemokines away and thus do possibly attenuate TRALI, as do T lymphocytes by a yet unknown mechanism. See text for details.

– neutrophils and the endothelium – is further shattered by the observation that besides monocytes and platelets, even red blood cells and T lymphocytes may possibly participate in the complex interplay as attenuators or multipliers of TRALI (Fig. 1) [32,33]. How and to which extent each cell type contributes to TRALI is unclear; however, considering the patient’s autologous cells as co-players in TRALI means to incorporate the patient into the TRALI model. This, at first glance, appears to lead to a difficult-to-dissect interplay, but at second glance will hopefully allow for (new) patient-focused strategies to decrease the risk of TRALI in the future. 4. Threshold model of TRALI At the bottom line, increased permeability of the pulmonary endothelial barrier is likely to be reached by several ways (see below). The threshold model of TRALI suggests that a certain threshold must be overcome to induce a TRALI reaction, i.e. to induce a barrier breakdown (Fig. 2) [5]. This threshold may be

Fig. 2. Threshold model of transfusion-related acute lung injury (TRALI) (see text for details). From [5], with permission.

111

lower for mild TRALI [34] (which can be treated with oxygen supply) and higher for severe TRALI (which requires mechanical ventilation and is associated with a considerable number of deaths). Overcoming of these thresholds depends on both recipient-related risk factors (the individual predisposition of the patient) and transfusion-related risk factors. It is very likely that in many TRALI cases, a number of both, recipient- and transfusion-related risk factors act together before TRALI precipitates. A number of recipient-related risk factors unravelled in thoroughly performed clinical trials are summarized in Table 1. Note that liver disease/alcoholism has been identified as a recipientrelated risk factor in three (adjusted OR 2.1–31.7), sepsis in two (adjusted OR, 2.5–2.6), and mechanical ventilation (adjusted OR 3–3.6) in two independent trials. Transfusion-related risk factors as identified in clinical studies are summarized in Table 2. Plasma transfusion is a relevant risk factor in all studies, especially the transfusion of units obtained from female donors (adjusted OR 1.51–4.5). Two studies identified antibodies against human leukocyte antigen (HLA) class II (adjusted OR 3.08–3.2) and against neutrophils (adjusted OR 1.71–4.85) as relevant transfusion-related risk factors, whereas antibodies against HLA class I were not a relevant risk factor in all three studies. Of note, red blood cell storage time was not associated with the occurrence of TRALI in two out of two studies, and lysophosphatidylcholines were associated with TRALI in one trial, but not associated in two other trials. Details will be discussed in the next two sections. 5. Antibody-induced (immune) TRALI TRALI appears within minutes to just a few hours from the initiation of the transfusion event. Although numerous other, recipient-related factors can contribute to the complex process of TRALI (Table 1), a relevant property of the blood component is a prerequisite for TRALI: the transfusion-related factor. In the majority of TRALI cases (65–89%), antibodies are present in the implicated blood component and antibodymediated (immune) TRALI has been reported as the leading cause in severe cases of TRALI [5,35,36]. An association between the induction of TRALI and the presence of antibodies in the transfused blood product is documented in several clinical studies and numerous experiments were performed in animal models proofing that antibodies are able to precipitate TRALI [18–21,27–30,32,33,37–43]. Because the majority of these antibodies are detected in blood components donated by females, the predominant use of male high-volume plasma components (fresh frozen plasma, platelets) was installed as a preventive measure in several countries [44]. This policy did indeed significantly reduce the incidence of TRALI both in large-scale surveillance studies and haemovigilance reports, proofing the overall importance of antibody-mediated TRALI [39,45–47]. We do have clinical and experimental evidence that antibodies can induce TRALI without any additional (obvious) co-stimulus (be it a transfusion- or recipient-related factor) [27,40,48]. However, several look-back studies identified recipients of blood components containing antibodies that induced

112

U.J. Sachs / Transfusion Clinique et Biologique 19 (2012) 109–116

Table 1 Recipient-related risk factors identified in clinical studies. Risk factor

OR

Author and year

Patient cohort

Number of patients with TRALI/number of enrolled patients

History of heavy alcoholism

2.7

Gajic et al., 2007 [37]

Intensive care unit patients

74/901 (8.2%)

Sepsis Liver disease

2.6 2.1

End-stage liver disease

31.7

Benson et al., 2010 [70]

22/225 (9.7%)

Emergency coronary artery bypass grafting

17.6

Vlaar et al., 2010 [11]

Patients with gastro-intestinal bleeding Intensive care unit patients

Hematologic malignancy Massive transfusion Mechanical ventilation Sepsis

13.1 4.5 3 2.5

Patient age

N/A

Time on cardiopulmonary bypass

N/A

Chronic alcohol abuse

5.9

Fluid balance before transfusion (increment per L) Peak airway pressure > 30 cm H2 O before transfusion Shock before transfusion Current smoker Liver surgery

1.15 3.6 4.2 3.4 6.7

109/2024 (5.4%)

Vlaar et al., 2011 [38]

Cardiac surgery patients

16/668 (2.4%)

Toy et al., 2012 [39]

Patients at tertiary care medical centers

Case-control-study: 89 TRALI, 164 control cases

TRALI: transfusion-related acute lung injury; N/A: not available. Table 2 Transfusion-related risk factors identified in clinical studies. Risk factor

Risk

Any high plasma volume component

Yes

2.78

Number of units from female donors Amount of plasma from female donors (L) Amount of plasma from female donors with at least one pregnancy (L) Number of HLA class II positive units Number of granulocyte immunofluorescence positive units Mean lysophosphatidylcholine Number of HLA class I positive units

Yes Yes Yes Yes Yes Yes No

1.51 5.09 9.48 3.08 4.85 1.61

Antibody-positive units

Yes

HLA class II positive units HLA class I positive units Granulocyte immunofluorescence positive units Red blood cells storage time Platelets storage time Lysophosphatidylcholine

No No No No No No

Female plasma

Yes

4.5

Quantity of cognate anti-HLA class II antibodies Volume of granulocyte immunofluorescence positive blood products HLA class I antibodies Amount of lysophosphatidylcholine Red blood cells storage time

Yes Yes No No No

3.2 1.71

HLA: human leukocyte antigen.

OR

14.2

Author and year

Patient cohort

Study, sample size, (TRALI cases)

Gajic et al., 2007 [37]

Intensive care unit patients

Prospective cohort study, n = 901 (74)

Vlaar et al., 2011 [38]

Cardiac surgery patients

Prospective case-control-study, n = 48 (16)

Toy et al., 2012 [39]

Tertiary care medical centre patients

Case-control-study, n = 253 (89)

U.J. Sachs / Transfusion Clinique et Biologique 19 (2012) 109–116

113

Table 3 Experimental studies (isolated rodent lungs or in vivo studies) on transfusion-related acute lung injury (TRALI). Agent used to induce TRALI

Experimental model

Stimulation

Study

Information gained on the (potential) relevance

Anti-HNA-3a (human) Anti-MHC class I (monoclonal antibody) Anti-HNA-2 (monoclonal antibody)

Rabbit lungs Mice Rat lungs

– Housinga –

Seeger et al., 1990 [40] Looney et al., 2006 [18]. Sachs et al., 2006 [27]

Anti-MHC class I (monoclonal antibody) Anti-MHC class I (monoclonal antibody) Anti-MHC class I (monoclonal antibody) Anti-MHC class I (monoclonal antibody) Anti-HLA class II (human) Anti-HNA-3a (human) Anti-MHC class I (monoclonal antibody) Non-immune Non-immune Non-immune Non-immune Non-immune

Mice Rats Mice Mice Rat lungs Mice Mice Rat lungs Rats Mice Sheep Rats

Lipopolysaccharides Lipopolysaccharides Ventilation Lipopolysaccharides Lipopolysaccharides Lipopolysaccharides – Lipopolysaccharides Lipopolysaccharides Lipopolysaccharides Lipopolysaccharides Trauma-hemorrhage

Looney et al., 2009 [30] Kelher et al., 2009 [19] Vlaar et al., 2010 [41] Fung et al., 2010 [33] Sachs et al., 2011 [20] Bayat et al., 2011 [28] Strait et al., 2011 [29] Silliman et al., 1998 [21] Kelher et al., 2009 [19] Mangalmurti et al., 2009 [32] Tung et al., 2011 [42] Nicholson et al., 2011 [43]

Complement Fc␥R Activation thresholds (antigen density, prestimulation) Antibody titer, platelets Antibody titer Lung injury by ventilation Antibody titer, T lymphocytes Monocytes Endothelial cells Monocytes, complement Storage Storage Storage Storage Storage

HLA: human leukocyte antigen; HNA: human neurophil antigen. a TRALI did only precipitate when animals were housed under non-sterile conditions.

TRALI in one, but not in the other patient [49–51]. Obviously, not all patients with antibody/antigen concordance will develop clinical signs of TRALI, i.e. additional stimuli are required to precipitate TRALI in many patients [52–54]. This clinical finding is also true in the experimental setting where most groups had to mimic a recipient-related risk factor by adding lipopolysaccharides as a prerequisite to precipitate TRALI in their model (Table 3). Two antibody entities have been identified as most relevant for the induction of TRALI, antibodies against HNA (human neurtrophil antigen)-3a and HLA class II antibodies [55]. Antibodies against HNA-3a are associated with a high mortality rate [55]. More than 80% of the Caucasian population express the HNA-3a epitope on their leukocytes. HNA-3bb woman exposed to HNA-3a positive blood cells during pregnancy can develop antibodies against this antigen. Such antibodies were shown to prime the respiratory burst of HNA-3a positive neutrophils which subsequently damaged pulmonary endothelial cells [23]. Since a recent study by Greinacher et al. who identified HNA-3 as the CTL2 protein on neutrophils failed to prove superoxide production in targeted neutrophils after exposure to anti-HNA3a antibodies, the exact mechanism by which these antibodies induce the breakdown of the endothelial barrier remained to be identified [56]. Our group has demonstrated that direct binding of HNA-3a antibodies to endothelial cells may, even in absence of neutrophils, induce increased endothelial permeability and TRALI in a murine model [28]. Such a direct mechanism could be one explanation for the severity of HNA-3a-antibody-induced TRALI. Case series have reported that cognate anti-HLA class II was the most frequent antibody implicated in TRALI [55,57]. The potential impact of HLA class II antibodies was first suspected by Kopko et al. [58]. They reported a TRALI case after transfusion of plasma which did not contain antibodies against HNA or HLA class I, but an antibody against HLA class II. HLA class II antigens are not expressed on the surface of neutrophils

or endothelial cells. Kopko et al. hypothesized that anti-HLA class II antibodies could bind to monocytes, which constitutively express HLA class II, and lead to neutrophil activation through the release of activating substance from monocytes [59]. A recent study in an ex vivo rat lung model precisely proved the efficiency of HLA class II antibodies to indirectly induce TRALI via monocytes [20]. Although the overall importance of HLA class I antibodies appears to be low, several animal studies have been conducted to identify their mechanism of action. The first in vivo study by Looney et al. showed abrogation of TRALI induction by MHC class I antibodies in the absence of Fc gamma receptors (using FcR␥–/– mice) [18]. Disregarding the concentration of antibodies used in these experiments (which was much higher than in a clinical setting), these observations suggest involvement of an indirect way of neutrophil activation by MHC class I antibodies. Neutrophil trapping and activation was most likely mediated through an interaction of the Fab part of the antibody with MHC class I antigens expressed on the endothelium and subsequent neutrophil recruitment through the interaction of the free Fc part of the antibody with the neutrophils’ Fc receptor [18]. These observations were followed by a study in which different monoclonal MHC class I antibodies were able to induce TRALI in rats by a direct interaction with neutrophils [19]. Only recently, in a complex study, Strait et al. found evidence that MHC class I antibodies, again in mice, might rather induce TRALI by a mechanism involving macrophages, complement activation, and binding of antibodies to the endothelium [29]. These findings are partially contradictive to previous results [18,30], but they are interesting as they shift the focus from the blood cell towards the endothelium itself. Unfortunately, these findings are of little (if any) clinical relevance, since HLA class I antibodies are only rarely involved in TRALI in humans; because they are often detected in blood components, most HLA class I antibodies are obviously incapable of producing TRALI, the reason for that being unclear (widespread antigen expression, antigen

114

U.J. Sachs / Transfusion Clinique et Biologique 19 (2012) 109–116

density, the antibodies’ epitope specificity, their ability to activate complement?). For a detailed discussion of other antibody specificities and specific aspects related to the type of blood component transfused, the reader is referred to recent reviews [35,60]. 6. Non-immune TRALI It is well established that TRALI can also develop after the transfusion of blood components that do not contain leukocyte antibodies – but, despite intense research during the last years, it is justified to state that the mechanism behind is still unclear. It is conceivable to assume that substances accumulated or produced during blood storage affect the recipient’s leukocytes, pulmonary endothelial cells, or even other cells, to finally precipitate the reaction, but results available up to now are conflicting and inconclusive. Since red blood cell concentrates are known to alter significantly during their shelf-life and contain only very little plasma that could carry antibodies, this blood component is a candidate mediator of antibody-negative TRALI. Administration of supernatant from stored (but not from fresh) human RBCs caused TRALI in an ex vivo and later also in an in vivo animal model of TRALI [19,21]. An attractive hypothesis was that biologically active breakdown products of phosphatidycholine, lysophosphatidycholine, are mediating this effect, as they were found to accumulate during storage [21]. More recently, other studies have demonstrated that lysophosphatidycholine may only be marginally involved in TRALI pathogenesis, if at all. First, three European groups showed independently that the total amount of lysophosphatidycholine or the lysophosphatidycholine pattern did not change during storage of human or rat red blood cell concentrates [61–63]. Second, one of these groups investigated the contribution of red blood cell concentrates in rat model [63]. Red blood cell concentrates were prepared from donor rats and transfused after storage. When rats were pretreated with lipopolysaccharides, there was no difference in the histopathology score or cytokine levels after the transfusion of fresh (day 0) and old (day 14) rat red blood cell concentrates. When red cells and supernatant were transfused separately, only the supernatant induced lung inflammation. Of note, initiation of lung oedema (a hallmark of TRALI) was not described. The authors demonstrate that lysophosphatidycholine was not involved in the inflammatory reaction [63]. A more recent study has now introduced non-polar lipids as a major derivative from the red blood cell membrane, which may potently activate neutrophils and cause increased endothelial permeability in an in vivo model [64]; these data await further confirmation. Currently available data from clinical trials, however, are also not supportive for a role of red blood cells storage time as a relevant risk factor for TRALI [14,38,39]. Nonetheless, the potential role of the aged red blood cell itself has also been investigated. Mangalmurti et al. found that the transfusion of aged red blood cell concentrates leads to lung inflammation in an in vivo model, even if the cells had been washed before transfusion, suggesting an effect related to the cell itself [32]. In the murine system, this effect was attributed to the loss of a chemokine scavenger receptor, known as the Duffy

antigen, from the red blood cell’s surface. It is well known that storage lesions differ between species; however, it should be noticed that washed red blood cell concentrates from rats could still elicit some inflammatory changes (though no TRALI) in healthy rat lungs [63]. It is possible that changes to the membrane indeed do contribute to TRALI (as a multiplier), but the molecular effect awaits further elucidation. Platelet transfusions are also involved in non-immune TRALI and accumulation of biologically active breakdown products has been investigated by several groups. In a cross-species study, supernatant of aged human platelets, but not of fresh platelets, caused acute lung injury in an ex vivo rat lung model, and lysophosphatidycholine was again suspected to cause this reaction [65]. In an advanced syngeneic rat model, whole aged rat platelet suspensions, but not fresh ones, led to neutrophil adherence and some oedema in lungs [66]. However, when lipopolysaccharide was used to mimic patient-related risk factors, stored platelets did no longer have any effect on histopathology scores, cytokine levels, or pulmonary wet weight when compared to fresh ones. Overall, only mild inflammatory and pro-coagulant effects were detected, but TRALI itself could not be induced with stored components [66]. Investigating the supernatant of stored rat platelet concentrates revealed that they did not contain higher levels of cytokines, but their lysophosphatidycholine content was increased – indicating that the relevance of lysophosphatidycholine must be questioned. There is now additional evidence that lysophosphatidycholine accumulation during storage is not the explanation for the capability of stored platelet concentrates to prime neutrophils [67]. Besides lysophosphatidycholine, CD40 ligand (CD40L), a proinflammatory mediator found in both cell-associated and soluble (sCD40L) forms, was suspected to be associated with non-immune mediated TRALI after platelet transfusion [68]. A murine study, however, revealed that using a substance which inhibits CD40L expression on platelets and lowers sCD40L serum levels, ciglitazone, did not interfere with TRALI when it was precipitated with MHC class I antibodies in mice [69]. In addition, antagonizing CD40L with anti-CD40L did not have an effect either. In line with this, patients with TRALI after cardiac surgery did not differ in their sCD40L levels from controls [69]. In contrast to the previous report, these results do not support a relevant role for CD40L in mediating TRALI. CD40L could also not be identified as a relevant risk factor in a recently published active surveillance study of the general transfusion population [53]. 7. Conclusion Contaminants present in blood components are capable of inducing TRALI in a transfusion recipient by interacting with the recipient’s leukocytes and/or endothelial cells. Whether or not such contaminants lead to the breakdown of the endothelial barrier depends on the patient’s individual predisposition and on the preactivation status of his leukocytes and endothelial cells. There is evidence that the presence or absence of attenuators and/or multipliers of activatory pathways may further modulate this interplay. It is conceivable that in most TRALI patients,

U.J. Sachs / Transfusion Clinique et Biologique 19 (2012) 109–116

several factors related to the transfusion event and related to the individual predisposition act together in order to overcome an (arbitrary) activation threshold, a pathological concept known as the threshold model of TRALI. Our knowledge about critical pathways determining the individual predisposition is still very limited; excellent clinical trials have identified critical medical conditions that can be integrated in the threshold model and may hopefully soon help us to identify patient-focused strategies to further decrease the incidence of TRALI. It is to be feared, however, that the mechanism behind many of these medical conditions will not be identified easily. The capability of antibodies to precipitate TRALI is generally acknowledged, even though how they induce the increase in endothelial permeability has not been pinpointed exactly so far. Whatsoever, it is one of the great achievements of transfusion medicine in the post-HIV era that transfusion-related risks mediated by antibodies have been identified and largely abandoned in many countries. It is without question that we do see cases of TRALI in the absence of antibodies (non-immune TRALI). Unfortunately, previously published promising results on mediators of non-immune TRALI could not be confirmed both in experiments and most clinical studies, and the mechanism of non-immune TRALI still awaits elucidation.

Disclosure of interest The author declares that he has no conflicts of interest concerning this article.

References [1] Barnard RD. Indiscriminate transfusion: a critique of case reports illustrating hypersensitivity reactions. NY State J Med 1951;51:2399–402. [2] Popovsky MA, Abel MP, Moore SB. Transfusion-related acute lung injury associated with passive transfer of antileukocyte antibodies. Am Rev Respir Dis 1983;128:185–9. [3] Popovsky MA, Moore SB. Diagnostic and pathogenetic considerations in transfusion-related acute lung injury. Transfusion 1985;25:573–7. [4] SHOT summaries and reports [homepage on the internet]. Available from: http://www.shotuk.org/home/ (last accessed on Feb 22, 2012). [5] Bux J, Sachs UJ. The pathogenesis of transfusion-related acute lung injury (TRALI). Br J Haematol 2007;136:788–99. [6] Zupanska B, Uhrynowska M, Konopka L. Transfusion-related acute lung injury due to granulocyte-agglutinating antibody in a patient with paroxysmale nocturnal hemoglobinuria. Transfusion 1999;39:944–7. [7] Moore SB. Transfusion-related acute lung injury (TRALI): clinical presentation, treatment, and prognosis. Crit Care Med 2006;34:S114–7. [8] European Haemovigllance network: Definition of adverse transfusion events. Available from http://www.ehn-org.net (last accessed on Nov 10, 2011). [9] Kleinman S, Caulfield T, Chan P, Davenport R, McFarland J, McPhedran S, et al. Towards an understanding of transfusion-related acute lung injury: statement of a consensus panel. Transfusion 2004;44:1774–89. [10] Bux J. Antibody-mediated (immune) transfusion-related acute lung injury. Vox Sang 2011;100:122–8. [11] Vlaar AP, Binnekade JM, Prins D, van Stein D, Hofstra JJ, Schultz MJ, et al. Risk factors and outcome of transfusion-related acute lung injury in the critically ill: a nested case-control study. Crit Care Med 2010;38:771–8. [12] Kernoff PB, Durrant IJ, Rizza CR, Wright FW. Severe allergic pulmonary oedema after plasma transfusion. Br J Haematol 1972;23:777–81.

115

[13] Wolf CF, Canale VC. Fatal pulmonary hypersensitivity reaction to HL-A incompatible blood transfusion: report of a case and review of the literature. Transfusion 1976;16:135–40. [14] Silliman CC, Paterson AJ, Dickey WO, Stroneck DF, Popovsky MA, Caldwell SA, et al. The association of biologically active lipids with the development of transfusion-related acute lung injury: a retrospective study. Transfusion 1997;37:719–26. [15] Flury R, Reutter F. Lethal anaphylactic shock during the transfusion of a thrombocyte concentrate. Schweiz Med Wochenschr 1966;96:918–20. [16] Felbo M, Jensen KG. Death in childbirth following transfusion of leukocyte-incompatible blood. Acta Haematol 1962;27:113–9. [17] Dry SM, Bechard KM, Milford EL, Churchill WH, Benjamin RJ. The pathology of transfusion-related acute lung injury. Am J Clin Pathol 1999;112:216–21. [18] Looney MR, Su X, Van Ziffle JA, Lowell CA, Matthay MA. Neutrophils and their Fc gamma receptors are essential in a mouse model of transfusionrelated acute lung injury. J Clin Invest 2006;116:1615–23. [19] Kelher MR, Masuno T, Moore EE, Damle S, Meng X, Song Y, et al. Plasma from stored packed red blood cells and MHC class I antibodies causes acute lung injury in a 2-event in vivo rat model. Blood 2009;113:2079–87. [20] Sachs UJ, Wasel W, Bayat B, Bohle RM, Hattar K, Berghöfer H, et al. Mechanism of transfusion-related acute lung injury induced by HLA class II antibodies. Blood 2011;117:669–77. [21] Silliman CC, Voelkel NF, Allard JD, Elzi DJ, Tuder RM, Johnson JL, et al. Plasma and lipids from stored packed red blood cells cause acute lung injury in an animal model. J Clin Invest 1998;101:1458–67. [22] Fung YL, Silliman CC. The role of neutrophils in the pathogenesis of transfusion-related acute lung injury. Transfus Med Rev 2009;23: 266–83. [23] Silliman CC, Curtis BR, Kopko PM, Khan SY, Kelher MR, Schuller RM, et al. Donor antibodies to HNA-3a implicated in TRALI reactions prime neutrophils and cause PMN-mediated damage to human pulmonary microvascular endothelial cells in a two-event in vitro model. Blood 2007;109:1752–5. [24] Johnson A, Phillips P, Hocking D, Tsan MF, Ferro T. Protein kinase inhibitor prevents pulmonary oedema in response to H2O2. Am J Physiol 1989;256:H1012–22. [25] Barnard ML, Matalon S. Mechanisms of extracellular reactive oxygen species injury to the pulmonary microvasculature. J Appl Physiol 1992;72:1724–9. [26] Seeger W, Hansen T, Rössig R, Schmehl T, Schütte H, Krämer HJ, et al. Hydrogen peroxide-induced increase in lung endothelial and epithelial permeability – effect of adenylate cyclase stimulation and phosphodiesterase inhibition. Microvasc Res 1995;50:1–17. [27] Sachs UJ, Hattar K, Weissmann N, Bohle RM, Siblelius U, Bux J. Antibody-induced neutrophil activation as a trigger for transfusion-related acute lung injury. Blood 2006;107:1217–9. [28] Bayat B, Tjahjono Y, Akylbek S, Hippenstiel S, Weissmann N, Sachs UJ, et al. Lung endothelial injury induced by HNA-3a antibodies in TRALI. Blood 2011;118:40 [ASH Annual Meeting Abstract]. [29] Strait RT, Hicks W, Barasa N, Mahler A, Khodoun M, Köhl J, et al. MHC class I-specific antibody binding to nonhematopoietic cells drives complement activation to induce transfusion-related acute lung injury in mice. J Exp Med 2011;208:2525–44. [30] Looney MR, Nguyen JX, Hu Y, Van Ziffle JA, Lowell CA, Matthay MA. Platelet depletion and aspirin treatment protect mice in a two-event model of transfusion-related acute lung injury. J Clin Invest 2009;119: 3450–61. [31] Finlayson J, Grey D, Kavangh L, Witt C. Transfusion-related acute lung injury in a neutropenic patient. Intern Med J 2011;41:638–41. [32] Mangalmurti NS, Xiong Z, Hulver M, Ranganathan M, Liu XH, Oriss T, et al. Loss of red cell chemokine scavenging promotes transfusion-related lung inflammation. Blood 2009;113:1158–66. [33] Fung YL, Kim M, Tabuchi A, Aslam R, Speck ER, Chow L, et al. Recipient T lymphocytes modulate the severity of antibody-mediated transfusionrelated acute lung injury. Blood 2010;116:3073–9. [34] Davis A, Mandal R, Johnson M, Makar R, Stowell C, Dzik S. A touch of TRALI. Transfusion 2008;48:541–5.

116

U.J. Sachs / Transfusion Clinique et Biologique 19 (2012) 109–116

[35] Flesch BK, Petershofen EK, Bux J. TRALI-new challenges for histocompatibility and immunogenetics in transfusion medicine. Tissue Antigens 2011;78:1–7. [36] Eder AF, Herron R, Strupp A, Dy B, Notari EP, Chambers LA, et al. Transfusion-related acute lung injury surveillance (2003–2005) and the potential impact of the selective use of plasma from male donors in the American Red Cross. Transfusion 2007;47:599–607. [37] Gajic O, Rana R, Winters JL, Yilmaz M, Mendez JL, Rickman OB, et al. Transfusion-related acute lung injury in the critically ill: prospective nested case–control study. Am J Respir Crit Care Med 2007;176:886–91. [38] Vlaar AP, Hofstra JJ, Determann RM, Veelo DP, Paulus F, Kulik W, et al. The incidence, risk factors, and outcome of transfusion-related acute lung injury in a cohort of cardiac surgery patients: a prospective nested casecontrol study. Blood 2011;117:4218–425. [39] Toy P, Gajic O, Bacchetti P, Looney MR, Gropper MA, Hubmayr R, et al. Transfusion-related acute lung injury: incidence and risk factors. Blood 2012;119:1757–67. [40] Seeger W, Schneider U, Kreusler B, von Witzleben E, Walmrath D, Grimminger F, et al. Reproduction of transfusion-related acute lung injury in an ex vivo lung model. Blood 1990;76:1438–44. [41] Vlaar AP, Wolthuis EK, Hofstra JJ, Roelofs JJ, Boon L, Schultz MJ, et al. Mechanical ventilation aggravates transfusion-related acute lung injury induced by MHC-I class antibodies. Intensive Care Med 2010;36:879–87. [42] Tung JP, Fung YL, Nataatmadja M, Colebourne KI, Esmaeel HM, Wilson K, et al. A novel in vivo ovine model of transfusion-related acute lung injury (TRALI). Vox Sang 2011;100:219–30. [43] Nicholson SE, Johnson RA, Craig T, Myers JG, Durante W, Stewart RM, et al. Transfusion-related acute lung injury in a rat model of traumahemorrhage. J Trauma 2011;70:466–71. [44] Middelburg RA, van Stein D, Zupanska B, Uhrynowska M, Gajic O, Mu˜nizDiaz E, et al. Female donors and transfusion-related acute lung injury: a case-referent study from the International TRALI Unisex Research Group. Transfusion 2010;50:2447–54. [45] Lucas G, Win N, Calvert A, Green A, Griffin E, Bendukidze N, et al. Reducing the incidence of TRALI in the UK: the results of screening for donor leukocyte antibodies and the development of national guidelines. Vox Sang 2011, http://dx.doi.org/10.1111/j.1423-0410.2011.01570.x (e-pub ahead of print). [46] Funk MB, Guenay S, Lohmann A, Henseler O, Heiden M, Hanschmann KM, et al. Benefit of transfusion-related acute lung injury risk-minimization measures – German haemovigilance data (2006–2010). Vox Sang 2012;102:317–23. [47] Lin Y, Saw CL, Hannach B, Goldman M. Transfusion-related acute lung injury preventive measures and their impact at Canadian blood services. Transfusion 2012;52:567–74. [48] Dooren MC, Ouwehand WH, Verhoeven AJ, von dem Borne AE, Kuijpers RW. Adult respiratory distress syndrome after experimental intravenous gamma-globulin concentrate and monocyte-reactive IgG antibodies. Lancet 1998;352:1601–2. [49] Yomtovian R, Kline W, Press C, Clay M, Engman H, Hammerschmidt D, et al. Severe pulmonary hypersensitivity associated with passive transfusion of a neutrophil-specific antibody. Lancet 1984;1:244–6. [50] Van Buren NL, Stroncek DF, Clay ME, McCullough J, Dalmasso AP. Transfusion-related acute lung injury caused by an NB2 granulocytespecific antibody in a patient with thrombotic thrombocytopenic purpura. Transfusion 1990;30:42–5. [51] Nicolle AL, Chapman CE, Carter V, Wallis JP. Transfusion-related acute lung injury caused by two donors with anti-human leucocyte antigen class II antibodies: a look-back investigation. Transfus Med 2004;14:225–30.

[52] Cooling L. Transfusion-related acute lung injury. JAMA 2002;17(288):315–6. [53] Toy P, Hollis-Perry KM, Jun J, Nakagawa M. Recipients of blood from a donor with multiple HLA antibodies: a lookback study of transfusionrelated acute lung injury. Transfusion 2004;44:1683–8. [54] Muniz M, Sheldon S, Schuller RM, Young NS, Klein HG, Leitman SF, et al. Patient-specific transfusion-related acute lung injury. Vox Sang 2008;94:70–3. [55] Reil A, Keller-Stanislawski B, Günay S, Bux J. Specificities of leucocyte alloantibodies in transfusion-related acute lung injury and results of leucocyte antibody screening of blood donors. Vox Sang 2008;95:313–7. [56] Greinacher A, Wesche J, Hammer E, Fürll B, Völker U, Reil A, et al. Characterization of the human neutrophil alloantigen-3a. Nat Med 2010;16:45–8. [57] Chapman CE, Stainsby D, Jones H, Love E, Massey E, Win N, et al. Serious hazards of transfusion steering group. Ten years of hemovigilance reports of transfusion-related acute lung injury in the United Kingdom and the impact of preferential use of male donor plasma. Transfusion 2009;49:440–52. [58] Kopko PM, Holland PV. Transfusion-related acute lung injury. Br J Haematol 1999;105:322–9. [59] Kopko PM, Paglieroni TG, Popovsky MA, Muto KN, MacKenzie MR, Holland PV. TRALI: correlation of antigen-antibody and monocyte activation in donor-recipient pairs. Transfusion 2003;43:177–84. [60] Sachs UJ. Recent insights into the mechanism of transfusion-related acute lung injury. Curr Opin Hematol 2011;18:436–42. [61] Ma´slanka K, Smole´nska-Sym G, Michur H, Wróbel A, Lachert E, Brojer E. Lysophosphatidylcholines: bioactive lipids generated during storage of blood components. Arch Immunol Ther Exp (Warsz) 2012;60:55–60. [62] Sachs UJ, Weissmann N, Wasel W, Bein G, Berghöfer H, Schilow W, et al. Supernatants from stored leukodepleted packed red blood cells do not regularly exhibit changes in their (lyso-)phosphatidylcholine composition and do not cause TRALI in an ex vivo rat lung model. Vox Sang 2010;99:456. [63] Vlaar AP, Hofstra JJ, Levi M, et al. Supernatant of aged erythrocytes causes lung inflammation and coagulopathy in a “two-hit” in vivo syngeneic transfusion model. Anesthesiology 2010;113:92–103. [64] Silliman CC, Moore EE, Kelher MR, Khan SY, Gellar L, Elzi DJ. Identification of lipids that accumulate during the routine storage of prestorage leukoreduced red blood cells and cause acute lung injury. Transfusion 2011;51:2549–54. [65] Silliman CC, Bjornsen AJ, Wyman TH, Kelher M, Allard J, Bieber S, et al. Plasma and lipids from stored platelets cause acute lung injury in an animal model. Transfusion 2003;43:633–40. [66] Vlaar AP, Hofstra JJ, Kulik W, van Lenthe H, Nieuwland R, Schultz MJ, et al. Supernatant of stored platelets causes lung inflammation and coagulopathy in a novel in vivo transfusion model. Blood 2010;116:1360–8. [67] Vlaar AP, Kulik W, Nieuwland R, Peters CP, Tool AT, van Bruggen R, et al. Accumulation of bioactive lipids during storage of blood products is not cell but plasma derived and temperature dependent. Transfusion 2011;51:2358–66. [68] Khan SY, Kelher MR, Heal JM, Blumberg N, Boshkov LK, Phipps R, et al. Soluble CD40 ligand accumulates in stored blood components, primes neutrophils through CD40, and is a potential cofactor in the development of transfusion-related acute lung injury. Blood 2006;108:2455–62. [69] Tuinman PR, Gerards MC, Jongsma G, Vlaar AP, Boon L, Juffermans NP. Lack of evidence of CD40 ligand involvement in transfusion-related acute lung injury. Clin Exp Immunol 2011;165:278–84. [70] Benson AB, Austin GL, Berg M, McFann KK, Thomas S, Ramirez G, et al. Transfusion-related acute lung injury in ICU patients admitted with gastrointestinal bleeding. Intensive Care Med 2010;36:1710–7.