Transfusion-related acute lung injury (TRALI)

INTRODUCTION — Transfusion-related acute lung injury (TRALI) is a serious and potentially fatal complication of blood product transfusion in which a patient develops rapid onset lung injury and noncardiogenic pulmonary edema due to activation of immune cells in the lungs. It can be severe enough to be life-threatening, with fever, chills, and hypoxemic respiratory failure. Because it is a clinically diagnosed syndrome, understanding of its pathophysiology and diagnostic criteria continue to evolve.

The epidemiology, pathogenesis, risk factors, clinical features, management, and prevention strategies for TRALI are presented here. Other transfusion reactions and general issues related to acute lung injury are discussed separately.

●Overview of acute transfusion reactions – (See "Approach to the patient with a suspected acute transfusion reaction".)

●Allergic and anaphylactic reactions – (See "Immunologic transfusion reactions".)

●Hemolysis – (See "Hemolytic transfusion reactions".)

●Circulatory overload – (See "Transfusion-associated circulatory overload (TACO)".)

●Pulmonary edema – (See "Noncardiogenic pulmonary edema" and "Acute respiratory distress syndrome: Clinical features, diagnosis, and complications in adults".)

TERMINOLOGY AND DIAGNOSTIC CRITERIA — The first extensive series of cases using the term TRALI was published in 1985 [1]; cases continued being reported for the next two decades, but criteria for making the diagnosis varied. In 2004 and 2005, these criteria were standardized with proposed modifications to these criteria recommended in 2019. The major distinction between the two sets of criteria relates to case reporting in hemovigilance systems and appropriate classification of cases for research studies. Direct patient management is not affected by nor dependent on which classification scheme is used [2].

●NHLBI and CCC definition – In 2004 and 2005, the first uniform (consensus) definitions for TRALI were published by a National Heart, Lung, and Blood Institute (NHLBI) working group as well as a Canadian Consensus Conference (CCC), in which TRALI was defined as new acute lung injury (ALI)/acute respiratory distress syndrome (ARDS) occurring during or within six hours after blood product administration (table 1) [2-4].

Both definitions indicated that when a clear temporal relationship to an alternative risk factor for ALI/ARDS coexisted, a formal diagnosis of TRALI could not be made. In these circumstances, the preferred terminology has generally been "possible TRALI". Though the phrase "transfused ARDS" has also been suggested, "possible TRALI" remains the more prevalent descriptor [5]. This CCC definition is still in widespread use.

●Revised Delphi panel definition – In 2019, a modified classification scheme was proposed based on new knowledge gained since the 2004 CCC (table 2) [6]. This classification, which was developed by a Delphi panel composed of international TRALI experts, reaffirmed that TRALI remains a clinical diagnosis and does not require any specific laboratory results such as detection of cognate leukocyte antibodies (see 'Clinical diagnosis' below). The modified system is anticipated to improve the consistency and accuracy of reporting of transfusion-associated respiratory complications and to benefit hemovigilance systems and clinical research.

The main modifications to the 2004 CCC TRALI definition include the following:

•A new terminology of TRALI type I and TRALI type II is recommended (table 2).

-TRALI type I occurs in patients with no concomitant risk factors for acute respiratory distress syndrome (ARDS), which is concordant with the 2004 CCC definition of TRALI.

-TRALI type II occurs in patients who either have concomitant risk factors for ARDS or who have prevalent mild ARDS but whose acute deterioration in respiratory status is believed to be due to the transfusion event. Thus, the presence of mild pre-transfusion ARDS no longer excludes the diagnosis of TRALI.

•The term "possible TRALI" was removed because of its ambiguity.

•Patients who develop inflammatory lung edema and meet ARDS criteria following transfusion should be classified as ARDS and not TRALI if there was evidence of respiratory deterioration in the 12 hours preceding the transfusion episode.

•The updated 2012 ARDS consensus definition (referred to as the BERLIN definition) has been evaluated for its relevance to TRALI, and essential updates have been incorporated in the new TRALI definition [7]. This includes an updated list of commonly accepted ARDS risk factors as well as a more comprehensive approach to the evaluation of lung edema (eg, including chest computed tomography [CT] scan and lung ultrasound).

•Although the time frame for the onset of clinical symptoms remains within six hours of the transfusion, the abnormal chest imaging results may occur up to 24 hours after TRALI onset. This revised time frame allows for delays in obtaining chest imaging results.

When comparing the differing criteria, the concept of TRALI type I remains consistent with prior descriptions of TRALI. In contrast, possible TRALI cases using the old system are now divided into two groups (either TRALI type II or ARDS) depending on the clinical course in the 12 hour interval preceding transfusion. If the clinical course was stable, the new criteria allow for a diagnosis of TRALI type II, even if mild ARDS was present prior to transfusion. Alternatively, if the patient had been experiencing respiratory deterioration prior to the transfusion event, the new criteria would lead to a diagnosis of ARDS. In light of the above, the new diagnostic criteria require a more thorough clinical evaluation of the transfusion recipient's status in the 12-hour interval preceding transfusion.

Some experts have begun to use the 2019 updates [8,9]; however, most hemovigilance networks continue to use the 2004 CCC definition of TRALI, sometimes with slight modifications due to limitations in obtaining precise case reporting data. An example of this can be found in the criteria used in the United States [10].

Most published studies have also used the 2004 CCC classification system (sometimes with some slight modification), and we generally use the 2004 CCC definition for the remainder of this chapter. A 2023 study suggested similar case capture rates when comparing the revised TRALI criteria to the criteria proposed by the CCC [11].

EPIDEMIOLOGY — The true incidence of TRALI is difficult to determine for many reasons including poor syndrome recognition, differing TRALI case definitions, and reliance of hemovigilance systems on passive reporting rather than active surveillance strategies [12-16].

It has been well established that TRALI incidence decreased dramatically following the institution of TRALI mitigation strategies for transfused plasma and platelet components that were instituted in the mid to late 2000s, especially avoidance of plasma components from multiparous female donors. Supporting evidence is discussed below. (See 'Prevention' below.)

A 2024 meta-analysis attempted to address some of these issues by screening hundreds of published articles and selecting 80 studies (encompassing 176 million transfused blood components) [17]. As determined from active surveillance studies, the meta-analysis reported incidence rates of TRALI on a per component basis as follows:

●Red blood cells (RBCs) – 0.17 per 10,000 (95% CI, 0.03-0.43); equivalent to 0.0017 percent or 1 in 58,824

●Platelets – 0.31 per 10,000 (95% CI, 0.22-0.42); equivalent to 0.0031 percent or 1 in 32,258

●Plasma – 3.19 per 10,000 (95% CI, 0.09-10.66); equivalent to 0.032 percent or 1 in 3134. When two outlier studies were omitted, the incidence was 0.2 per 10,000 which is more congruent with results from RBC and platelet transfusion.

These rates were thirty fold lower in studies using passive surveillance (0.02 to 0.10 per 10,000) [17]. TRALI incidence from platelet or plasma transfusion was only based on data that were generated after TRALI risk mitigation procedures were instituted. (See 'Prevention' below.)

Historically, prior studies had reported higher TRALI incidence rates. As an example, prior to risk mitigation implementation, TRALI incidence was reported to be approximately 0.04 to 0.1 percent per transfused patient or approximately 0.02 percent of transfused components (1 in 5000) [1,18-20].

Additionally, due to the contribution of specific recipient risk factors, TRALI incidence is also likely to depend on the population of interest. As an example, estimates suggest that the rate of TRALI in critically ill populations may reach 5 to 8 percent [21,22]. (See 'Recipient risk factors' below.)

TRALI was the leading cause of transfusion-related mortality in the United States until approximately 2016, when it was overtaken by transfusion-associated circulatory overload (TACO) as the most common cause of transfusion-associated mortality [23,24]. It still remains a leading cause of transfusion-associated mortality [25]. (See "Approach to the patient with a suspected acute transfusion reaction", section on 'Mortality'.)

RISK FACTORS — Specific risk factors for TRALI can be divided conceptually into recipient risk factors and blood component risk factors.

Recipient risk factors — TRALI has been reported to occur in essentially all age groups and equally in male and female transfusion recipients [26-28]. An underlying condition such as recent surgery, cytokine treatment, massive blood transfusion, and active infection have been implicated in some but not all studies [19,27,29-36].

Critically ill patients, particularly those with evidence of systemic inflammation, appear to be at highest risk for TRALI [2,21,22,37,38].

Low levels of the antiinflammatory cytokine interleukin-10 (IL-10) and elevation of C-reactive protein (CRP) have been associated with increased risk for TRALI [39,40].

Studies evaluating TRALI risk factors are limited by a relatively small number of TRALI cases available for analysis and the use of varied diagnostic criteria. The following studies are illustrative of the available data:

A multicenter prospective cohort investigation using active surveillance strategies for TRALI detection identified recipient risk factors for TRALI in 89 cases compared with 164 matched transfused controls [41]. The following pre-transfusion TRALI risk factors were identified:

●Liver transplantation surgery

●Chronic alcohol abuse

●Shock

●Higher peak airway pressure while being mechanically ventilated

●Current smoking

●Higher interleukin (IL)-8 levels

●Positive fluid balance

In a nested case-control study evaluating patients in an intensive care unit (ICU) for more than 48 hours, risk factors for TRALI/possible TRALI included emergency cardiac surgery, hematologic malignancy, massive transfusion, sepsis, mechanical ventilation, and a high Acute Physiology and Chronic Health Evaluation II (APACHE II) score [21]. In another study evaluating critically ill patients in a medical ICU, sepsis, liver disease, and a history of alcohol abuse were more common in transfusion recipients who developed TRALI/possible TRALI than in transfused controls without respiratory compromise [22].

A Medicare database review (over 11 million patients, 2556 with a TRALI diagnosis code) identified modestly higher rates of TRALI in recipients of platelet or plasma-containing products rather than red blood cells; females versus males; White individuals; and individuals with postinflammatory pulmonary fibrosis, smoking, or other medical illnesses in the prior six months [36]. Many of these recipient factors (chronic alcohol abuse, smoking, shock prior to transfusion, positive fluid balance prior to transfusion) were confirmed to be associated with possible TRALI in a subsequent nested case-control study specifically evaluating risk factors for possible TRALI [42].

Blood component risk factors — Virtually all blood components have been associated with TRALI. This includes transfused whole blood derived platelets, cryoprecipitate [19], and granulocytes [43], as well as intravenous immune globulin preparations [44-46], and allogeneic stem cells [4,31,47,48].

Prior to the institution of TRALI risk mitigation strategies, plasma components and apheresis platelet concentrates conferred the highest risk of TRALI per component. Subsequently, due to transfusion of a much larger number of red cell units compared with plasma and platelets, the largest number of TRALI-related deaths in the United States and other developed countries have occurred with red blood cell transfusion [23,25]. (See 'Prevention' below.)

●Donor sex and high-plasma-volume blood components — Though TRALI has been associated with virtually all blood products, components with larger volumes of plasma, such as plasma, apheresis platelet concentrates, and whole blood, have been consistently shown to carry the greatest risk per component or per transfusion episode [12,15,17,30,49]. Despite this association with higher volumes of plasma, even components that contain small amounts of plasma can cause TRALI. For example, TRALI can occur after the transfusion of RBCs suspended in additive solution; this component contains only 10 to 40 mL of plasma [50].

Multiple studies have demonstrated a role for female sex and increased parity of the donor in the risk for TRALI [41,51-53]. The role of these factors was illustrated in a multicenter prospective cohort investigation utilizing active TRALI surveillance strategies, which identified the following donor/blood component risk factors for TRALI [41]:

•Plasma or whole blood from female donors.

•Increased volume of highly reactive transfused anti-human leukocyte antigen (HLA) Class II antibody with specificity for a cognate recipient HLA antigen (ie, antibodies for which the recipient had the corresponding HLA antigen).

•Increased volume of transfused anti-human neutrophil antigen (HNA) antibody [41].

The presence of non-cognate anti-HLA Class II antibodies, weaker cognate anti-HLA Class II antibodies, or any anti-HLA Class I antibody were not associated with TRALI in this study. However, other case reports suggest that anti-HLA Class I antibodies can cause TRALI, though the risk is likely to be lower than with Class II antibodies [12,54,55].

●Red blood cell storage duration — Although a longer duration of red blood cell storage has been suggested to increase the risk of TRALI, multiple randomized clinical trials and large observational studies have not confirmed this association [41,56-60]. In aggregate, the available evidence suggests that the duration of red blood cell storage is not a major risk factor in the development and/or severity of TRALI.

PATHOGENESIS — The generally accepted theory for TRALI pathogenesis is that it occurs via a two-hit mechanism [19,61,62].

●Neutrophil sequestration and priming – The first hit involves neutrophil sequestration and priming in the lung microvasculature, due to recipient factors such as endothelial injury. Priming refers to shifting of neutrophils to a state where they will respond to an otherwise innocuous or weak signal [63]. Endothelial cells are thought to be responsible for both the neutrophil sequestration (through adhesion molecules) and priming (through cytokine release). Generally these events are coupled and exist prior to the transfusion, although there may be circumstances in which they can occur as a result of the transfusion. (See 'Recipient risk factors' above.)

●Neutrophil activation – The second hit is activation of recipient neutrophils by a factor in the blood product. Activation is associated with the release from neutrophils of cytokines, reactive oxygen species, oxidases, and proteases that damage the pulmonary capillary endothelium. This damage causes inflammatory (non-hydrostatic) pulmonary edema. Transfused factors responsible for host neutrophil activation can include antibodies in the blood component directed against recipient antigens, or soluble factors such as bioactive lipids that can activate neutrophils. Donor anti-leukocyte antibodies can bind to antigens on recipient neutrophils or possibly to other cells such as monocytes or pulmonary endothelial cells; this is referred to by some authors as immune TRALI [64-68]. Bioactive lipids and other soluble factors in the transfused blood component can act as biologic response modifiers (BRMs); TRALI resulting from these non-antibody BRMs is sometimes referred to as non-immune TRALI [62]. (See 'Blood component risk factors' above.)

Clinical observations supporting the two-hit theory come from retrospective studies demonstrating that most transfused blood products containing HLA antibody do not cause TRALI, even if a cognate recipient antigen is present [69]. In retrospective studies, TRALI was detected in only 2.9 percent of individuals who received blood products from donors previously implicated in causing TRALI [54,69-72]. Despite this small percentage, deferring donors implicated in a particular case of TRALI from future donation is recommended, as there have been examples of a single donor causing TRALI in multiple different recipients. (See 'Prevention' below.)

In one case of fatal TRALI, the implicated donor was a multiparous female who had made 290 previous plasma apheresis donations [13]. Of 36 patients who had received her plasma in the prior two years, 15 experienced transfusion reactions with pulmonary symptoms, and 21 (58 percent) did not, again highlighting that factors from the transfusion alone are insufficient to cause TRALI. In this case, the donor had antibodies against a leukocyte antigen (HNA-3a) present in more than 95 percent of the general population that has been associated with severe often fatal TRALI cases [73]. In another case, TRALI occurred in a recipient of a single lung transplant only in the transplanted lung, and one of the blood products the recipient received contained anti-HLA Class I antibodies that matched an HLA antigen present only in the transplanted lung [74].

The proportion of TRALI caused by antibodies versus BRMs remains undetermined and is likely to vary by the type of component transfused, with antibody-mediated mechanisms explaining the majority of the cases due to plasma, and non-immune BRMs responsible for most cases from red blood cell transfusion [30,75,76]. The widely stated statistic that 80 to 85 percent of TRALI cases are due to the antibody-mediated mechanism is likely influenced by publication bias [77]. Of note, a case series comparing BRM-mediated TRALI with antibody-mediated TRALI suggests that BRM-mediated TRALI is a less severe condition [78].

A second model related to the two-hit theory is the threshold model. This model agrees that two hits are usually necessary for TRALI, but allows for the possibility that in some cases the second hit is so strong that an initial priming event is not required [78-81]. This theory explains TRALI cases that have occurred in otherwise healthy individuals who have received fresh frozen plasma (FFP) as a treatment for reversing warfarin anticoagulation.

Several observations support the role of neutrophils as the major effector cells in TRALI [62]. Neutrophil activation occurs in animal models of TRALI [82-87]. Transient neutropenia indicative of pulmonary sequestration has been seen in patients in the early phase of TRALI, whereas TRALI is rarely seen in neutropenic patients [54,88-92]. In an autopsy case, postmortem examination within two hours of TRALI onset revealed aggregates of neutrophils within the pulmonary capillaries associated with inflammatory edema and endothelial injury [93].

Multiple mechanisms of neutrophil activation in TRALI have been proposed. As an example, antibodies to human leukocyte antigen (HLA) Class I and human neutrophil antigens (HNA) may bind to recipient neutrophils and trigger their activation [79,80,94,95]. It has been known for years that female donors, due to exposure to fetal alloantigens during pregnancy, have a much higher prevalence of anti-HLA antibodies than do male donors [96,97]. While anti-HNA antibodies account for only a small percentage of TRALI cases from most countries (<5 percent), in Germany these constitute a higher percentage (28 percent in one series) [98]. Antibodies directed against the HNA 3a antigen have been associated with a substantial number of severe or fatal TRALI cases [99,100].

TRALI can also be associated with donor antibodies directed against HLA Class II, which is primarily expressed on antigen-presenting cells [63,80]. These antibodies may bind to HLA Class II antigens on monocytes, causing release of cytokines that in turn activate primed neutrophils [95,101-103].

Another much less common mechanism for antibody mediated TRALI development appears to occur when leukocyte antibodies are present in the recipient (more commonly in multiparous female recipients) and react with antigens on leukocytes in the transfused blood product. This is sometimes referred to as "reverse TRALI"; several cases potentially attributable to this mechanism have been reported, even with leukocyte reduced components [62,104,105].

When the second hit for neutrophil activation is a BRM rather than an antibody, the BRM is believed to activate neutrophils directly. Most of the work implicating these substances has come from in vitro and animal studies, and has implicated lysophosphatidylcholines (from white blood cells and platelets), neutral lipids (from the breakdown of red cell membranes), ceramide, soluble CD40 ligand (which accumulates in stored platelet concentrates), and both platelet- and red blood cell-derived microparticles in neutrophil activation [19,32,61,62,85,86,106-112].

CLINICAL PRESENTATION — The characteristic clinical presentation of TRALI is the sudden onset of hypoxemic respiratory insufficiency during or shortly after the transfusion of a blood product (table 1) [3,4]. Symptoms may be delayed as long as six hours, but usually begin within one to two hours of initiating the blood component infusion [3,31]. Indeed, the majority of cases occur within minutes of initiating a transfusion [19]. A number of additional signs and symptoms associated with noncardiogenic pulmonary edema and inflammation have been reported [3,30,31,91,92,113,114]. As an example, in a retrospective study of 49 TRALI cases, the most common signs and symptoms and their approximate frequencies were as follows [30]:

●Hypoxemia: in an intubated patient this could manifest as a change in oxygenation or increased oxygen requirements (100 percent, by definition)

●Pulmonary infiltrates on chest imaging (100 percent, by definition); the cardiac silhouette is classically normal

●If previously intubated, pink frothy airway secretions from the endotracheal tube (56 percent)

●Fever (33 percent)

●Hypotension (32 percent)

●Cyanosis (25 percent)

Other reports have also noted tachypnea, tachycardia, and elevated peak and plateau airway pressures in intubated patients. An acute, transient drop in the peripheral neutrophil count (consistent with sequestration of large numbers of neutrophils in the lungs) has also been reported.

DIAGNOSIS

Clinical diagnosis — TRALI should be considered whenever a patient develops hypoxemic respiratory insufficiency during or shortly after transfusion of any blood product [3]. TRALI is a clinical diagnosis made using the criteria outlined by the NHLBI's working group on TRALI or the Canadian Consensus Conference (CCC) on TRALI (table 1) or the 2019 Delphi panel [3,4,6]. (See 'Terminology and diagnostic criteria' above.)

These criteria require the presence of new acute respiratory distress syndrome (ARDS) occurring during or within six hours after blood product administration, documented by hypoxemia and abnormal chest imaging. Hypoxemia is documented when oxygen saturation is ≤90 percent on room air or the PaO2/FIO2 ratio is <300 mmHg, although other signs of hypoxia can also satisfy this criterion. Chest imaging must demonstrate bilateral pulmonary infiltrates. (See "Noncardiogenic pulmonary edema", section on 'Definition of noncardiogenic pulmonary edema' and "Acute respiratory distress syndrome: Clinical features, diagnosis, and complications in adults".)

When these conditions are met, but a clear temporal relationship to an alternative risk factor for ARDS coexists, the CCC criteria indicate that a formal diagnosis of TRALI cannot be made. In these circumstances, "possible TRALI" is the more appropriate diagnosis [5]. Use of these separate diagnostic categories (ie, TRALI and possible TRALI) allows for their separate reporting in surveillance systems, which may facilitate differential approaches to donor investigation and management, as well as targeting research programs to either (or both) groups of patients. The newer proposed classification system (TRALI type I and TRALI type II) permits the same kind of differential approach. (See 'Terminology and diagnostic criteria' above and 'Differential diagnosis' below.)

When TRALI is suspected, the transfusion medicine service should be contacted. The treating physician should evaluate the recipient’s vital signs, assess the extent of hypoxemia, and obtain chest imaging. Pulse oximetry is often sufficient, but arterial blood gas analysis may be warranted in more severe cases. Consideration of the likelihood of other potential causes for the respiratory distress (eg, cardiovascular compromise, anaphylaxis, sepsis, exacerbation of underlying lung disease or atelectasis) should guide appropriate clinician-initiated laboratory testing. (See 'Differential diagnosis' below.)

Immediate actions are listed below. (See 'Treatment' below.)

DIFFERENTIAL DIAGNOSIS — The differential diagnosis of TRALI includes other conditions that can manifest with respiratory distress following transfusion. This includes other causes of acute respiratory distress syndrome (ARDS) as well as other transfusion reactions. These conditions are described in more detail in the following sections.

Possible TRALI — A formal diagnosis of TRALI requires the absence of temporal relationships with additional risk factors for ARDS, such as aspiration, pneumonia, sepsis, toxic inhalation, lung contusion, trauma, burn injury, or pancreatitis. Using the Canadian Consensus Conference (CCC) criteria, a diagnosis of TRALI cannot be made when clinical criteria for TRALI are met but additional risk factors for ARDS are also present (table 1). In these circumstances, the terms "possible TRALI," "TRALI with other ALI risk factors," and "transfused ARDS" have previously been used. The term "transfused ARDS" was based in part on observations that clinical outcomes in this group of patients tend to resemble those in non-transfused patients with ARDS. Studies suggest that in this setting, recipient factors may be more closely associated with the development of lung injury than factors directly related to the blood component(s) [5]. (See 'Terminology and diagnostic criteria' above.)

Although massive transfusion has historically been recognized as a risk factor for ARDS, a diagnosis of TRALI (rather than possible TRALI) should be made in this circumstance provided that no other ARDS risk factors were present.

TACO — Transfusion-associated circulatory overload (TACO) is another cause of transfusion-related respiratory insufficiency. Though historically reported most commonly in older individuals and small children, TACO can occur in all age ranges. Additional risk factors for TACO include compromised cardiac function, positive fluid balance, renal dysfunction, emergency surgery, and rapid blood product administration [115]. TACO appears to occur more frequently in surgical or intensive care settings, where large fluid volumes and blood are administered. Characteristic features and differentiating factors for TRALI versus TACO are shown in the table (table 3) [116]. (See "Transfusion-associated circulatory overload (TACO)".)

As a general rule, TRALI is more likely to be associated with fever, hypotension, and exudative pulmonary infiltrates, and less likely to respond to diuresis. In contrast, TACO is more likely to be associated with findings suggesting volume overload (eg, positive fluid balance, elevated jugular venous pressure, elevated pulmonary artery occlusion pressure) or poor cardiac function (eg, history of congestive heart failure, reduced left ventricular ejection fraction, or diastolic dysfunction). Similarly, elevated systolic blood pressures near the time of dyspnea onset, a widened pulmonary vascular pedicle width or increased cardiothoracic ratio on chest radiography, and/or increased circulating levels of brain natriuretic peptide (BNP) or N-terminal (NT)-Pro-BNP suggest a diagnosis of TACO rather than TRALI [117-123]. (See "Natriuretic peptide measurement in heart failure".)

Differentiating TRALI from TACO can be a significant challenge, particularly as both may coexist [117,124,125]. In such cases, pulmonary findings may be due to a combination of hydrostatic (eg, TACO) and non-hydrostatic (eg, TRALI) lung edema. Indeed, studies show that approximately 30 percent of ARDS patients have at least mild evidence of left atrial hypertension [126]. In these circumstances, ARDS is diagnosed in the presence of fluid overload based on clinical judgment that the degree of fluid overload is not sufficient to explain the extent of hypoxemia and pulmonary infiltrates. This type of mixed picture may be more likely to occur in intensive care unit (ICU) patients.

ARDS — Milder manifestations of inflammatory lung edema such as a PaO₂/FiO₂ <300 but >200 were historically referred to as acute lung injury (ALI). Subsequent consensus conference recommendations from pulmonary experts have transitioned this terminology to "mild ARDS" [7].

If cardiopulmonary deterioration or findings consistent with moderate to severe ARDS are present in the 12 hours prior to the initiation of transfusion or arise more than six hours after a transfusion episode, then ARDS is the appropriate diagnosis. In a subset of mild ARDS cases, where the clinical course in the 12 hours preceding transfusion has been stable, it is recognized that the administration of blood products may in fact contribute to ARDS development. In such circumstances, the revised 2019 definitions for TRALI would recommend a diagnosis of TRALI type II. (See 'Terminology and diagnostic criteria' above and "Acute respiratory distress syndrome: Clinical features, diagnosis, and complications in adults".)

Other transfusion reactions

●Hemolytic transfusion reactions – Hemolytic transfusion reactions can cause respiratory distress, but fever and chills tend to predominate in hemolytic transfusion reactions more than in TRALI. Hemolytic reactions are typically due to ABO incompatibility, and pink serum or hemoglobinuria may be seen due to intravascular hemolysis. The direct antiglobulin (Coombs) test will be positive in hemolytic reactions, but not in TRALI. (See "Hemolytic transfusion reactions", section on 'Acute hemolytic transfusion reactions'.)

●Severe allergic transfusion reactions (anaphylaxis) – Anaphylaxis can cause respiratory distress similar to TRALI, but anaphylaxis is more often associated with airway signs and symptoms such as stridor, cough, wheezing, nasal congestion, and bronchospasm. Anaphylaxis is also more often associated with a new rash, gastrointestinal symptoms such as nausea and/or vomiting, and shock. Notably, however, the latter may be seen in TRALI as well. Most severe allergic transfusion reactions are due to both donor and recipient factors. One recognized etiology is transfusion of IgA-containing products to an IgA deficient recipient who has antibodies to IgA. (See "Immunologic transfusion reactions", section on 'Anaphylactic transfusion reactions' and "Anaphylaxis: Emergency treatment".)

●Sepsis – Sepsis and septic shock are often associated with respiratory distress, fever, and hypotension. Evidence for an active infectious process (leukocytosis, fever, positive microbiology) suggests the diagnosis of sepsis or septic shock rather than TRALI. Transfusion-associated sepsis is generally associated with the administration of bacterially contaminated platelets. In this case, evaluation of the transfusion by the transfusion medicine service will typically identify a contaminating microorganism. (See "Platelet transfusion: Indications, ordering, and associated risks", section on 'Complications'.)

TREATMENT

Overview — If either TRALI or possible TRALI is suspected, the transfusion should be discontinued immediately. Physicians should alert the transfusion medicine service and initiate an evaluation for a transfusion reaction. This is important for the protection of future recipients as well as for TRALI laboratory testing and work-up. This initial evaluation is described in more detail above. (See 'Diagnosis' above.)

Management of the patient with either TRALI or possible TRALI is supportive, with oxygen supplementation for the correction of hypoxemia being the cornerstone of treatment. Non-invasive respiratory support with continuous positive airway pressure (CPAP) or bilevel positive airway pressure (BiPAP) may be sufficient in less severe cases, but endotracheal intubation with invasive mechanical ventilation is often required [22]. The available evidence suggests that most patients who develop either TRALI or possible TRALI will require ventilatory support (approximately 70 to 80 percent) [1,21,22,127].

Ventilation — There have been no prospective clinical investigations on strategies for managing mechanical ventilation in patients with TRALI, and it is generally believed that ventilator management should be guided by the same principles used in patients with other forms of ARDS [128]. The successful application of more extreme life support interventions, such as extracorporeal membrane oxygenation, has also been described in TRALI [129]. (See "Acute respiratory distress syndrome: Fluid management, pharmacotherapy, and supportive care in adults" and "Acute respiratory distress syndrome: Ventilator management strategies for adults" and "Extracorporeal life support in adults in the intensive care unit: Overview".)

Hemodynamic support — Patients with TRALI often present with hypovolemia and associated hypotension [130]. The initial goal of hemodynamic management is to ensure adequate end-organ perfusion. This may be achieved with fluid resuscitation and/or vasoactive support. Caution should be taken with early empiric administration of diuretic therapy, as it may result in hypotension in those who were initially hemodynamically stable [47,131]. However, for patients with sustained hypoxemia and demonstrated hemodynamic stability, administration of diuretic therapy may be an acceptable intervention. (See "Treatment of severe hypovolemia or hypovolemic shock in adults" and "Evaluation and management of suspected sepsis and septic shock in adults", section on 'Initial therapy'.)

Steroids — Intravenous corticosteroids have been extensively studied in the setting of ARDS with mixed results [47,132,133]. In regards to TRALI, there are isolated case reports supporting high-dose corticosteroids in the treatment of this syndrome [47,134]. However, the efficacy of corticosteroids has not been tested in prospective clinical studies, and the limited anecdotal evidence is unconvincing. Thus, we do not recommend the routine use of corticosteroids when TRALI is suspected. Moreover, due to evidence suggesting harm when initiating corticosteroids late in the course of ARDS (>14 days after syndrome onset) [47,131,135], we advise against the use of corticosteroid therapy when the lung injury is fully established and has been present for more than two weeks. (See "Acute respiratory distress syndrome: Fluid management, pharmacotherapy, and supportive care in adults", section on 'Glucocorticoids'.)

Investigational strategies — In addition to the therapies described above, a number of additional ARDS and TRALI treatment and prevention strategies have been proposed and are under various stages of investigation. Examples include:

●HMG-CoA reductase inhibitors (statins) [136].

●Aspirin [136,137].

●Ascorbic acid [138].

●Alternatives to allogeneic blood products (eg, RBC substitutes, prothrombin complex concentrates, fibrinogen concentrates, and activated factor VII). (See "Approach to the patient who declines blood transfusion", section on 'Other interventions'.)

However, at present, none of these therapies has sufficient evidence to justify its use as a routine TRALI prevention or treatment measure. Interest in IL-10 as a therapeutic option for TRALI will require additional clinical validation [139].

Observations that suggest that neutrophil extracellular traps (NETs) may play a role in TRALI pathogenesis have raised the potential for compounds that reduce the development of NETs to mitigate risk or severity of TRALI [140]. (See "An overview of the innate immune system", section on 'Neutrophil extracellular traps'.)

Interventions aiming to impact the primary effector cells involved in TRALI pathogenesis (eg, neutrophils, endothelial cells, macrophages, platelets) have shown promise in preclinical studies [140]. However, none of these therapies has sufficient evidence to justify its use as a routine TRALI prevention or treatment measure.

Safety of additional transfusions — Patients who recover from TRALI do not appear to be at increased risk for recurrent episodes following transfusions from other donors. However, published experience is limited. Survivors of TRALI can receive additional blood products in the future, and transfusion of needed blood products should not be withheld [31,47]. However, individuals who recover from TRALI should not receive plasma-containing blood products from the implicated donor [141-144].

REPORTING TO THE TRANSFUSION SERVICE — When a pulmonary transfusion reaction such as TRALI is suspected, the case should be reported to the hospital transfusion medicine service as promptly as possible [6]. Based on the details of the case, the transfusion medicine service will initiate a transfusion reaction workup, which will usually include drawing a sample from the patient for appropriate laboratory tests. These may include complete blood count, bilirubin, haptoglobin, direct antiglobulin (Coombs) test, brain natriuretic peptide (BNP) or N-terminal (NT)-Pro-BNP, and HLA antigen typing. (See "Natriuretic peptide measurement in heart failure".)

As noted above, the classification of TRALI is evolving, and either the older or the newer classification scheme can be used, depending on the existing procedures used by the hospital transfusion service. (See 'Terminology and diagnostic criteria' above.)

●If the older classification is used, the categories would be TRALI, possible TRALI, transfusion-associated circulatory overload (TACO), or other.

●If the newer classification is used, the categories would be TRALI (type I or type II), ARDS, TACO, TRALI/TACO cannot distinguish, or an alternate diagnosis.

If TRALI is suspected, the transfusion medicine service will compile a list of blood products transfused within the previous six hours and forward this list to the blood supplier. The blood supplier will then follow its procedures to recall these donors and initiate testing for HLA and possibly HNA antibodies. However, these results will not be available for several weeks or months. If a donor with an HLA or HNA antibody that matches a cognate antigen in the recipient is identified, some authors have proposed further classifying the case as "TRALI, antibody-mediated"; however, such classification requires HLA typing of the recipient, which is often not performed [5]. Identifying the implicated donor(s) and deferring that individual from future high plasma volume donation requires cooperation between the clinician (to provide accurate clinical information), the hospital transfusion medicine service, and the blood supplier. (See 'Prevention' below.)

Due to the difficulty in determining the underlying etiology of pulmonary edema after transfusion, particularly in very ill patients, the transfusion medicine service may benefit from collaboration with critical care or pulmonary physicians who may be able to assist in the classification of reported cases [2].

PROGNOSIS

Typical course — Initial descriptions detailing the clinical course for TRALI suggested quick resolution of hypoxemia, generally within 24 to 48 hours of symptom onset [1]. However, the majority of patients who develop TRALI will require admission to an intensive care unit and ventilator support [21,22].

For those requiring mechanical ventilation, these early reports described a mean duration of ventilatory support lasting approximately 40 hours [1]. In contrast, subsequent evidence suggests that most cases will require respiratory support for a longer period (approximately 3 to 10 days) [21,22].

More contemporary data frequently report composite outcomes from cases of TRALI and possible TRALI and limit the study population to the critically ill. Such cases are likely to represent the more severe end of the clinical spectrum. In contrast, milder cases of TRALI and/or possible TRALI may go undiagnosed or unreported. As such, the clinical presentation, need for life-support interventions, and outcomes for patients with TRALI and/or possible TRALI remain incompletely defined.

Although limited information is available, it appears that most survivors will recover to their baseline pulmonary function, and they can safely receive additional blood products in the future [3,31,47]. (See "Acute respiratory distress syndrome: Fluid management, pharmacotherapy, and supportive care in adults".)

Mortality — Mortality rates for TRALI vary according to the patient population and the time period (eg, before or after institution of TRALI mitigation strategies). (See 'Prevention' below.)

Historical estimates for TRALI–associated mortality have been reported in different ranges (5 to 8 percent; 13 to 21 percent; or, from a prospective study, 17 percent) [12,29,30,47,98,145].

In case series of critically ill patients, in which outcomes in those with TRALI or possible TRALI were combined, mortality rates were higher, ranging from 41 to 67 percent [20-22,127,145].

PREVENTION — In all cases of TRALI and in some cases of possible TRALI, the transfusion medicine service and the blood collection facility should investigate all of the associated donors for the presence of anti-human leukocyte antigen (HLA) and possibly anti-human neutrophil antigen (HNA) antibodies [146], with the goal of identifying donors who should be deferred from future donations. In reality, the number of donors investigated may be more limited and the laboratory work-up performed by transfusion services may vary depending upon the number of donors in a case; the availability of donor samples; the availability of neutrophil antibody testing, which in the United States is restricted to a few specialty laboratories; and the availability of a recipient sample for HLA antigen typing, which is region-specific (routinely done in some countries but not others) [3,147,148]. Some laboratories in Europe also perform leukocyte crossmatching as part of the evaluation [146].

A donor who is shown to have leukocyte antibodies (anti-HLA or anti-HNA) that match or are likely to match a recipient's leukocyte antigens, is classified as an implicated donor and is, at minimum, deferred from future plasma apheresis or platelet apheresis donations. In some countries, the recipient is also tested to see if the donor's anti-HLA or anti-HNA antibodies match the recipient's HLA or HNA antigens. Most transfusion medicine services will defer an implicated donor from any type of blood donation (not just apheresis platelets or plasma); this is particularly true if the donor has anti-HNA 3a [3].

In addition, several general blood donor management strategies are in use to reduce the incidence of TRALI [3,12,15,24,49,147,149-160]:

●Adherence to current guidelines of blood component use, especially for plasma, to decrease recipient exposure to transfused units.

●Deferral of donors implicated in a TRALI reaction.

●For high plasma volume components, selection of male donors or female donors who have never been pregnant, as these individuals are less likely to be alloimmunized to leukocyte antigens. This applies to FFP, plasma frozen within 24 hours of phlebotomy (FP-24), cryo-reduced (cryo-poor) plasma, apheresis platelets, buffy coat derived platelet pools resuspended in plasma from a single donor, and whole blood. (See 'Mitigating TRALI risk from plasma via deferral of multiparous female donors' below.)

●Use of pooled solvent detergent plasma as an alternative to FFP.

●Testing of parous apheresis donors of platelets or plasma for anti-HLA antibodies.

Mitigating TRALI risk from plasma via deferral of multiparous female donors — It has been well established that donors implicated in TRALI cases are more likely to be female and more likely to be multiparous [41,52,53]. These factors resulted in the implementation of TRALI risk mitigation policies during the mid to late 2000s throughout most of Europe and the United States, in which transfusable plasma units were predominantly obtained from male donors, thereby avoiding the transfusion of plasma units from female donors [12,15,151-153,161]. This policy is feasible for plasma transfusion because the number of plasma units derived from whole blood or apheresis collections is in excess of demand. In contrast, this policy is not feasible for apheresis platelet transfusion where restriction of units to male donors would seriously jeopardize the platelet supply.

The importance of these policies is illustrated by a large multicenter study conducted under the auspices of the National Heart, Lung, and Blood Institute (NHLBI) Retrovirology and Donor Epidemiology Program II (REDS-II) [159]. This study, known as the Leukocyte Antibody Prevalence Study (LAPS), used modern flow cytometry techniques to assess the prevalence of HLA antibodies in almost 8000 volunteer blood donors. There was a dose-response increase in the frequency of anti-HLA antibodies according to parity, from 1.7 percent for never pregnant females to 32.2 percent for four or more pregnancies. Donors who themselves had received a blood transfusion, males, and never pregnant female donors all showed very low frequency of anti-HLA antibodies (in the range of 1 to 2 percent) [159,162]. These data suggest that TRALI risk mitigation programs do not need to exclude plasma from never pregnant females or from donors with a previous transfusion history [153]. Nevertheless, some European countries do not transfuse plasma obtained from donors who have themselves received a transfusion [151]. These data also have helped to guide policies for HLA antibody testing of selected subpopulations of female plasma and platelet apheresis donors.

The following observations support the effectiveness of sex-based risk reduction policies for reducing TRALI incidence from plasma transfusion:

●A comparison of TRALI rates from 2006 (before mitigation measures were introduced) versus 2009 (after mitigation measures were introduced) using the same case detection strategies showed a three-fold reduction in TRALI incidence [41,163]:

•Incidence rate in 2006 – 1 in 3891 (0.0257 percent)

•Incidence rate in 2009 – 1 in 12,345 (0.0081 percent)

●The United Kingdom National Blood Service, starting in late 2003, provided 80 to 90 percent of FFP from male donors. This was associated with a significant decrease in the risk of highly likely/probable TRALI due to FFP (from 15.5 per million units issued from 1999 to 2004 to 3.2 per million from 2005 to 2006) [12]. There were 23 highly likely/probable TRALI cases from all components in 2003, compared with three in 2010 [152].

●In a four-year nested case-control study conducted from 2006 to 2009 at the University of California San Francisco and Mayo Clinic, TRALI incidence in 2006 (prior to implementation of risk reduction methods) was 0.0257 percent (1 in 3891); in 2009, after implementation of risk reduction policies for plasma products, TRALI incidence decreased to 0.0081 percent (1 in 12,345) [41].

●In the US, annual TRALI fatalities attributed to plasma transfusion and reported to the FDA declined from a peak of 23 cases in 2006 to three cases in 2011 [16,164]. In the three years prior to introduction of the sex-based TRALI risk reduction policies, plasma accounted for 48 percent of the fatal cases reported to Food and Drug Administration (FDA), compared with 27 percent in the four years following their implementation (2008 to 2011). Similarly, cases of suspected TRALI from plasma reported to the American Red Cross National Hemovigilance Program decreased from a rate of approximately 20 probable TRALI cases per million plasma components in 2006 to approximately four per million plasma components in 2008. Moreover, in contrast to six TRALI fatalities related to plasma transfusion in 2006, there were none in 2008 [15].

●In Germany, antibody-mediated TRALI from FFP transfusion decreased following the implementation of TRALI risk reduction policies. The rate fell from 12.71 per million units in 2006 to 2007 (pre-implementation of risk reduction policies) to 6.81 per million units in 2008 to 2009, during which there was partial implementation of risk reduction policies. With full implementation in 2010, no cases were reported [165].

Data from these and other studies were aggregated in a meta-analysis and a systematic review, both of which reported a significant reduction in TRALI risk from plasma products after the adoption of risk reduction measures (odds ratio [OR] 0.62; 95% CI 0.42 to 0.92 in the meta-analysis when comparing rates before and after risk reduction, and OR 0.27; 95% CI 0.20 to 0.38 in the systematic review) [166,167].

Use of solvent/detergent-treated plasma — Another strategy for TRALI risk reduction is to transfuse a pooled solvent/detergent-treated plasma product (S/D plasma) in place of FFP. Each unit of S/D plasma is prepared from multiple donors, which would be likely to dilute out any anti-HLA antibodies capable of causing TRALI and/or to bind anti-HLA antibodies by soluble HLA antigens [154,156]. An S/D plasma product was approved by the United States Food and Drug Administration in 2013, and S/D plasma has been widely used in many European countries for many years. (See "Pathogen inactivation of blood products", section on 'Plasma/FFP'.)

If TRALI can occur with S/D plasma, it is likely to be exceedingly rare:

●Review of 212,000 S/D plasma transfusions from a controlled hemovigilance dataset collected in France in 2007 and 2008 found no TRALI cases associated solely with S/D plasma; in contrast, the TRALI incidence rate from FFP transfusion was 1 in 31,000 units [49].

●Observational data from over 10 million S/D plasma unit transfusions did not document any cases of TRALI [155].

●A case report described three cases of possible TRALI associated with S/D plasma; upon further review, one of the cases was deemed unlikely and one was lacking data, leaving a single case attributable to the S/D plasma product over a four-year time frame, during which an estimated 220,000 units of S/D plasma were transfused [8].

●A retrospective study over a six month observation period in patients in the intensive care unit in which 2008 units of S/D plasma were transfused identified a single case of probable TRALI in which the sole transfusion product was S/D plasma [9].

●Studies on 32 batches of S/D plasma were unable to detect anti-HLA antibodies [154].

Mitigating TRALI risk from platelets — Some blood centers have implemented a policy of testing selected populations of platelet apheresis donors (eg, previously pregnant females, depending on the number of pregnancies) for anti-HLA antibodies and then redirecting anti-HLA antibody positive donors away from all high-plasma-volume donations, including platelet apheresis [153]. In late 2016, this strategy became an Association for the Advancement of Blood & Biotherapies (AABB) Standard, meaning that blood collectors accredited by the AABB are required to conform with the policy [168].

Another strategy considered for platelets is to resuspend the platelets in platelet additive solution (PAS) [151]. This reduces the plasma volume of the product by two-thirds. A report from the French Hemovigilance Network database covering the period from 2007 to 2013 documented that PAS was used in less than half of whole-blood-derived pooled buffy coat platelet products transfused from 2008 to 2009 and in over 98 percent of platelet products transfused during 2012 to 2013 [169]. This resulted in an approximately fivefold reduction in TRALI incidence (95% CI 1.6-13.4) from buffy coat platelets. A similar analysis for apheresis platelets collected prior to implementation of HLA antibody qualification testing (implemented in 2010 in France) did not show a protective effect of PAS. The amelioration of TRALI with whole-blood-derived platelets but not apheresis platelets may relate to the residual volume of plasma in the transfused platelet preparation, which is approximately 25 mL for whole-blood buffy coat platelets and approximately 150 mL for apheresis platelets.

Mitigating TRALI risk from RBC units — Due to the success of TRALI risk reduction measures for plasma, the relative percentage of TRALI cases from red blood cell (RBC) units has increased, such that the majority of current TRALI cases are associated with RBC transfusion. TRALI risk reduction measures appear to have had no effect on the frequency of TRALI cases associated with RBC transfusion [16,154,170].

Data suggest that a non-antibody mediated mechanism, presumably a soluble biologic response modifier, may account for the majority of TRALI cases caused by RBC transfusion [30,75,76]. In one case series, donor antibodies directed against cognate recipient antigens were identified less frequently in TRALI cases involving only RBC transfusions (18 percent) than in cases involving FFP (82 percent) [30]. Two subsequent analyses found that TRALI cases associated with RBC transfusion were not correlated with female or other alloimmunized donors. In contrast, these studies did associate high-plasma volume blood components from female and alloimmunized plasma donors with risk for TRALI [30,75].

Although RBC mediated TRALI continues to occur, no effective and practical risk reduction measures are available, with the exception of conservative RBC transfusion practices.

One retrospective report suggested that leukocyte reduction may have decreased TRALI incidence at a single institution; however, a randomized clinical trial comparing pre-storage leukocyte reduced RBC transfusion with standard allogeneic RBC transfusion found no evidence for differing rates of either early or late ARDS in transfused trauma patients [171,172]. Furthermore, TRALI has been reported in many countries that practice universal leukoreduction [151].

Based on the mechanism for red blood cell-induced TRALI, it has been suggested that the use of younger red blood cells or washed red blood cells would prevent some TRALI cases [151,173]. However, the data on red blood cell age and TRALI are contradictory and unconvincing. Moreover, most data concerning prevention of TRALI by RBC washing are retrospective and observational [57-59,86,151]. Additionally, a multicenter clinical trial did not find that RBC washing was effective in preventing end-organ injury in the setting of cardiac surgery, including markers of lung injury [174]. Furthermore, washing large number of RBC units is not logistically feasible for most institutions.

A novel approach to TRALI mitigation utilizing a prototype pre-storage leukoreduction filter has also been described [68]. This filter removes platelets and plasma proteins such as IgG from the filtered units in addition to leukocytes. An initial evaluation in an animal model system suggested that TRALI incidence from transfused RBC units was reduced with this filter. However, much more work is needed before this approach is ready for clinical testing.

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Transfusion and patient blood management".)

SUMMARY AND RECOMMENDATIONS

●Incidence – Transfusion-related acute lung injury (TRALI) is a clinically diagnosed transfusion complication involving a syndrome of acute lung injury with noncardiogenic pulmonary edema. TRALI is a leading cause of transfusion-related mortality in the United States. A revised classification (table 2) may improve reporting and allow better estimates of incidence. (See 'Terminology and diagnostic criteria' above and 'Epidemiology' above.)

●Risk factors – TRALI can occur with any blood component in any patient. Historically, risk correlated with a high volume of donor plasma, but following institution of risk mitigation strategies, most cases now occur with red blood cells (RBCs). (See 'Risk factors' above.)

●Pathogenesis – A "two-hit" hypothesis proposes that recipient neutrophils are primed for activation by the patient's underlying clinical condition. The second hit involves activation of these neutrophils by substances in the transfused product (antibodies to human leukocyte antigens [HLA] or biologic response modifiers). Rarely, the transfused product may be the sole factor. (See 'Pathogenesis' above.)

●Presentation – TRALI characteristically presents with sudden onset hypoxemic respiratory insufficiency during or shortly after transfusion. Frothy airway secretions may be seen (if intubated), along with fever, cyanosis, and hypotension. (See 'Clinical presentation' above.)

●Evaluation – TRALI should be considered when hypoxemic respiratory distress occurs during or within six hours after transfusion. An additional required diagnostic criterion is the presence of acute bilateral infiltrates on chest imaging that are not believed to be primarily the result of left atrial hypertension. While the infiltrates generally present within six hours of transfusion, a TRALI diagnosis may be made if the abnormal imaging is documented within 24 hours of transfusion. Diagnostic criteria are provided in the tables (table 1 and table 2). (See 'Diagnosis' above.)

●Differential diagnosis – Other conditions with respiratory distress following transfusion include other causes of pulmonary edema and other transfusion reactions. Distinguishing TRALI from transfusion-associated circulatory overload (TACO) is important for management (table 3). (See 'Differential diagnosis' above.)

●Management – When TRALI is suspected, the transfusion should be stopped immediately, vital signs and respiratory status should be assessed, and chest imaging should be obtained. The transfusion medicine service should be notified that TRALI is suspected. Therapy is supportive with supplemental oxygen and ventilatory support, with lung protective strategies when clinically indicated. Mortality is significant, but surviving patients generally recover completely. (See 'Treatment' above and 'Prognosis' above.)

●Prevention – Donors implicated in a case of TRALI are permanently deferred from future apheresis donation of platelets or plasma and possibly whole blood donation. Multiparous women are most likely to have formed anti-HLA antibodies, and most resource-rich countries supply products with high plasma volume (plasma, apheresis platelets, whole blood) exclusively, or at least predominantly, from male donors, female donors with no prior pregnancy, or donors who test negative for HLA antibodies. (See 'Prevention' above.)

ACKNOWLEDGMENT — We are saddened by the death of Arthur J Silvergleid, MD, who passed away in April 2024. The UpToDate editorial staff gratefully acknowledges the extensive contributions of Dr. Silvergleid to earlier versions of this and many other UpToDate topics.