Vox Sanguinis (2017) 112, 694–703

© 2017 International Society of Blood Transfusion

REVIEW ARTICLE DOI: 10.1111/vox.12573

Transfusion-related acute lung injury risk mitigation: an

update
Z. K. Otrock,1 C. Liu2 & B. J. Grossman2
1
Department of Pathology and Laboratory Medicine, Henry Ford Health System, Detroit, MI, USA
2
Department of Pathology and Immunology, Barnes-Jewish Hospital, Washington University, St Louis, MO, USA

Transfusion-related acute lung injury (TRALI) is a life-threatening complication

of transfusion. Greater understanding of the pathophysiology of this syndrome
has much improved during the last two decades. Plasma-containing components
from female donors with leucocyte antibodies were responsible for the majority
of TRALI fatalities before mitigation strategies were implemented. Over the past
15 years, measures to mitigate risk for TRALI have been implemented worldwide
and they continued to evolve with time. The AABB requires that all plasma con-
taining components and whole blood for transfusion must be collected from
men, women who have not been pregnant, or women who have tested negative
for human leucocyte antigen antibodies. Although the incidence of TRALI has
decreased following the institution of TRALI mitigation strategies, TRALI is still
the most common cause of transfusion-associated death in the United States. In
this review, we focus on TRALI risk mitigation strategies. We describe the mea-
sures taken by blood collection facilities to reduce the risk of TRALI in the United
Received: 3 February 2017,
States, Canada and European countries. We also review the literature for the
revised 11 July 2017,
accepted 11 August 2017,
effectiveness of these measures.
published online 25 September 2017 Key words: mitigation, review, risk, transfusion-related acute lung injury.

is a form of acute lung injury (ALI) which has been

Introduction
defined as acute hypoxemia with a ratio of partial
Transfusion-related acute lung injury (TRALI) is a life- pressure of oxygen to fractional inspired oxygen concen-
threatening complication of transfusion characterized by tration (PaO2/FiO2) of <300 mmHg associated with bilateral
the development of acute respiratory distress associated infiltrates on frontal chest radiograph and no evidence of
with non-cardiogenic pulmonary oedema occurring pulmonary vascular overload [10]. Patients can have a
within 6 h after receiving a blood transfusion [1, 2]. It is wide range of symptoms including fever, rigors, tachycar-
the leading cause of allogeneic blood transfusion-related dia, hypotension and less commonly hypertension [2]. The
mortality in the United States [3]. TRALI has been associ- most widely accepted definitions for TRALI were proposed
ated with virtually all blood components: plasma, plate- by the Canadian consensus conference in 2004 [1] and the
lets, cryoprecipitate [4], granulocytes [5], red blood cells NHLBI (National Heart, Lung and Blood Institute) in 2005
(RBC) [6], intravenous immune globulin [7], and allo- [11]. Both definitions require an acute onset of ALI within
geneic [8] and autologous [9] stem cells. 6 h after completion of blood transfusion; the presence of
Transfusion-related acute lung injury is a clinical diag- predisposing risk factors for the development of ALI differ-
nosis based on new respiratory symptoms and signs of entiates ‘suspected TRALI’ from ‘possible TRALI’. The Cana-
pulmonary oedema which are temporally related to a dian consensus, as compared to the NHLBI, introduced the
recent transfusion and without other underlying causes. It definition of ‘possible TRALI’ when there is an underlying
risk factor for ALI (Table 1) [11]. However, some investiga-
Correspondence: Brenda J. Grossman, Department of Pathology and tors have recently suggested dropping the term ‘possible
Immunology, Washington University School of Medicine, Campus TRALI’ and replacing it with ‘transfused ARDS (acute respi-
Box 8118, 660 South Euclid Avenue, Saint Louis, MO 63110, USA ratory distress syndrome)’ based on the fact that these
E-mail: [email protected]

694
 TRALI risk mitigation 695

Table 1 Criteria for TRALI diagnosis according to the Canadian consensus unit in 2009 [16, 19]. A very recent report from the Inter-
conference [1]a national Hemovigilance Network, which captures data
1. TRALI from seven countries (Australia, Thailand, Greece, Den-
mark, The Netherlands, France, and Canada) on all trans-
a. ALI fusion reactions, reported a TRALI death rate of 0
049 per
i Acute onset 100 000 blood components issued [20]. The most impor-
ii Hypoxemia tant risk factor for TRALI has been receiving plasma-con-
Research setting: PaO2/FiO2 ≤ 300 or SpO2 < 90% on room air. taining blood components from female donors [21, 22].
Non research setting: PaO2/FiO2 ≤ 300 or SpO2 < 90% on room air Other identified risk factors are recipient risk factors for
or other clinical evidence of hypoxemia ARDS; these include cardiac surgery, sepsis, chronic alco-
iii Bilateral infiltrates on frontal chest radiograph
hol abuse, smoking, positive fluid balance and shock
iv No evidence of left atrial hypertension (i.e. circulatory overload)
before transfusion [23–25].
b. No pre-existing ALI before transfusion
c. During or within 6 h of transfusion
In this review, we touch on the pathophysiology of
d. No temporal relationship to an alternative risk factor for ALI TRALI and focus on TRALI risk mitigation strategies. We
describe the measures taken by blood collection facilities
2. Possible TRALI to reduce the risk of TRALI in the United States, Canada
and European countries. We also review the literature for
the effectiveness of these measures.
a. ALI
b. No pre-existing ALI before transfusion
c. During or within 6 h of transfusion Pathophysiology
d. A clear temporal relationship to an alternative risk factor for ALI
The first evidence supporting the role of alloantibodies in
a
Used with permission from John Wiley and Sons. the pathogenesis of TRALI was published in 1983: five
patients with TRALI were found to have leukoagglutinat-
ing and lymphocytotoxic antibodies in the serum of the
patients had progression of their underlying ARDS follow- transfused blood products [26]. Later, many investigators
ing transfusion [12]. documented the association between TRALI and human
It can be challenging for clinicians to differentiate neutrophil antigen (HNA) or human leucocyte antigen
TRALI from transfusion-associated circulatory overload (HLA) antibodies in donor blood [18, 27, 28].
(TACO) especially in critically ill patients [13]. TACO also
causes transfusion-related respiratory insufficiency. TACO
‘Two-hit’ model
is more frequently seen in surgical patients or intensive
care settings where large fluid volumes are administered A ‘two-hit’ hypothesis was suggested stating that TRALI,
and in the elderly and in infants who may not tolerate like ALI and ARDS, results from two clinical insults: the
small increases in volume [14]. In general, TRALI is more first hit results from the clinical condition of the patient
likely to be associated with fever, hypotension and that causes activation of the pulmonary endothelium,
exudative pulmonary infiltrates [15]. In contrast, patients leading to the sequestration of polymorphonuclear leuco-
with TACO are more likely to present with findings sug- cytes (PMNs) to the activated pulmonary endothelium
gesting volume overload such as positive fluid balance, [29, 30]. These activated PMNs are responsible for the
elevated systemic blood pressures and pulmonary oedema release of cytokines and the expression of adhesion mole-
on chest radiograph. Some of these patients will have evi- cules. The second hit in antibody-mediated TRALI is the
dence of poor cardiac function such as history of conges- transfusion of plasma containing specific antibodies
tive heart failure, decreased left ventricular ejection directed against antigens on the PMN surface, which
fraction or diastolic dysfunction [15]. activates these PMNs, resulting in the release of proteases
Incidence of TRALI varies among countries because of that induce capillary leak and ALI [29]. In non-
its equivocal definition, failure to recognize the entity antibody-mediated TRALI, the second hit is caused by the
and different disciplines of reporting [16, 17]. The histori- accumulation of proinflammatory mediators and bioactive
cal reports suggest that TRALI occurred at a rate of 0
04– lipids in transfused blood [30, 31].
0
1% of transfused patients [4, 18]. The incidence of
TRALI has decreased following the institution of TRALI
Threshold model
mitigation strategies for transfused plasma and platelet
components. The reported incidence of TRALI has The threshold model has been proposed to explain TRALI
dropped from 1:3891 per unit in 2006 to 1:12 345 per in patients who are not high risk and without

© 2017 International Society of Blood Transfusion

Vox Sanguinis (2017) 112, 694–703
696 Z. K. Otrock et al.

predisposition. Thus, this model does not require the first had less pulmonary dysfunction and less death and hospi-
hit, and it assumes that patients can directly have the sec- talization. Although the findings raised some concerns
ond hit that is strong enough and does not need initial regarding the benefit of male-predominant plasma, the
priming. Therefore, TRALI can develop in a healthy indi- investigators acknowledged the limitations of their study
vidual once the neutrophil has surpassed a threshold of as being a retrospective and an underpowered study. They
activation, which can be seen in patients transfused with highlighted the need for additional studies to confirm
plasma containing a large quantity of antibodies [32]. these findings.
In the United Kingdom, interventions were imple-
mented by the National Blood Service (NBS) in 2003 to
Efforts to mitigate TRALI risk
prevent TRALI; these included the use of plasma from
For the past 15 years, efforts to reduce the risk for TRALI male donors for the production of FFP and for suspension
have mainly focused on antibody-mediated TRALI. It was of buffy coat-derived platelet pools. The NBS provided
clear that plasma-containing components from female 80–90% of plasma from male donors by late 2003. These
donors with leucocyte antibodies were responsible for the interventions were based on recommendations from the
majority of TRALI fatalities [21]. Many countries imple- UK hemovigilance scheme Serious Hazards of Transfusion
mented several strategies to reduce the incidence of (SHOT) to minimize TRALI risk [38]. Chapman et al. [39]
TRALI from these components. These strategies that relate reviewed SHOT data on reported cases of TRALI from
to donor deferral and management are not standardized 1996 through 2006. The male-predominant donor plasma
and they differ between individual blood banks [33, 34]. strategy has been associated with a gradual decrease in
In a survey conducted by the AABB TRALI Working the number of TRALI cases reported to SHOT; the risk of
Group to assess TRALI risk reduction practices in the Uni- highly likely/probable TRALI due to FFP has dropped
ted States, 41/47 (87%) blood centres had implemented from 15
5 per million units transfused during 1999
some platelet and plasma risk reduction by 2009 [34]. For through 2004 to 3
2 per million during 2005 through
groups A, B, and O plasma and whole blood collection, 2006 (P = 0
0079) and from 14 per million to 5
8 per
male-only donors, predominately male donors, never million for platelets (P = 0
068) [39]. The number of
pregnant female donors or a combination of these were reported fatalities decreased from nine in 2003 to one in
strategies used by most blood collecting facilities. As the 2006. This strategy was simple and low cost to implement
majority of whole blood-derived plasma is used for fur- in the United Kingdom, and it resulted in a significant
ther manufacturing, plasma from female donors could be decrease in TRALI cases. This decline in TRALI death has
diverted to manufacturing without impacting the plasma been maintained at the same low levels for the past
supply used for transfusion. This strategy could not be decade [40].
used for apheresis platelets, which would result in Other European countries have also taken measures to
availability issues. reduce the risk of TRALI from blood transfusions. In Ger-
The most commonly used policy to reduce the risk of many, antibody-mediated TRALI cases were reported
TRALI from apheresis platelets was increasing collections mainly after the administration of FFP, whereas platelet
from male donors (70%) followed by HLA antibody test- concentrates have not been identified as high-risk compo-
ing in selected plateletpheresis donors (43%) [34]. nents [41]. TRALI risk-reduction measures in 2008–2009
Table 2 summarizes the outcome of TRALI risk mitiga- focused exclusively on the manufacture of FFP; these
tion studies. measures consisted of excluding female donors with a
history of pregnancy or female donors who test positive
for HLA/HNA antibodies [42]. The comparison of the pre-
Exclusion of female donors
and the postimplementation period in Germany demon-
The original cohorts describing TRALI noted that in strated a continuously reduced frequency of antibody-
almost 90 percent of cases, either HLA or HNA antibodies mediated TRALI from plasma transfusion: from 12
71 per
could be detected in the plasma of donors [18, 26]. A million units in 2006/2007 to 6
81 per million units in
steep decline in TRALI incidence has been observed in the 2008/2009 and no case in 2010 [42]. In the Netherlands,
United Kingdom followed by other countries worldwide the male-only plasma for transfusion policy became
as the implementation of a predominantly male donor effective in July of 2007 and it was associated with a
fresh-frozen plasma (FFP) transfusion policy [35, 36]. 33% reduction of TRALI cases [43]. In Switzerland, there
However, one retrospective case–control study by Welsby was a considerable decrease of TRALI cases from 1/
et al., [37] comparing patients who underwent cardiac 21 000 to 1/86 000 plasma transfusions after the imple-
surgery and received either male or female donor plasma, mentation of only male-donor strategy for plasma in
found contrary outcomes: female donor plasma recipients 2007 [44].

© 2017 International Society of Blood Transfusion

Vox Sanguinis (2017) 112, 694–703
 Table 2 Effectiveness of TRALI risk mitigation strategies

Year of
References publication Country Study design Population Study period Risk reduction strategy Effect on TRALI risk

Wright et al., [77] 2008 United Kingdom Retrospective, 211 pts, ruptured 1998–2006 Male-only plasma for FFP ↓ ALI per patient: OR 0
39
university hospital AAA repair and BCDP PLTs (0
16–0
9)
Chapman et al., [39] 2009 United Kingdom Retrospective, national – 1996–2006 Male-only plasma for FFP ↓ TRALI: from 15
5 to 3
2 per

Vox Sanguinis (2017) 112, 694–703

and BCDP PLTs million units for FFP (P = 0
0079);
from 14 to 5
8 per million units
for PLTs (P = 0
068)
Nakazawa et al., [78] 2009 Japan Prospective, case–control, 82 surgical pts 2007–2008 Male-only plasma donor ↓ PD per patient: OR 0
219

© 2017 International Society of Blood Transfusion

university hospital (0
06–0
8), P = 0
022a
Eder et al., [48] 2010 United States Retrospective, national – 2006–2008 Male-only plasma donor ↓ TRALI per transfusion: OR 0
21
(0
08–0
45), P < 0
05
Vlaar et al., [79] 2010 The Netherlands Retrospective, 436 pts admitted 2004–2007 Male-only plasma donor ↓ TRALI per unit: RR 0
35
university hospital to ICU (0
14–0
88), P < 0
03
Wiersum-Osselton et al., [43] 2011 The Netherlands Retrospective, national – 2002–2009 Male-only plasma donor ↓ TRALI per transfusion: PAR 0
33
(0
09–0
51)b
Arinsburg et al., [80] 2012 United States Retrospective, three 461 598 transfusions 2005–2008 Male-only plasma donor Incidence of TRALI ↓ from 0
0084%
university hospitals to zero per transfusion (P = 0
052)c
Funk et al., [42] 2012 Germany Retrospective, national – 2006–2010 Combined strategyd ↓ TRALI per unit: OR 0
06 (0
01–0
41)
Lin et al., [49] 2012 Canada Retrospective, national – 2001–2009 Male-only plasma donor ↓ in the reported TRALI cases
Toy et al., [16] 2012 United States Prospective, case-control, 253 pts 2006–2009 Male-only plasma donor and Incidence of TRALI ↓ from 0
0257%
two university hospitals predominantly male-only PLTs to 0
0081% per unit (P = 0
002)
Jutzi and Amsler [44] 2013 Switzerland Retrospective, national – 2002–2013 Male-only plasma donor ↓ TRALI cases from 1/21 000
to 1/86 000 plasma transfusions
Clifford et al., [81] 2015 United States Retrospective, 3379 intraoperative 2004 and 2011 Male-only plasma donor No significant ↓ in incidence
university hospital transfusions of TRALI

TRALI, transfusion-related acute lung injury; pts, patients; AAA, abdominal aortic aneurysm; PLTs, platelets; ALI, acute lung injury; OR, odds ratio; BCDP, buffy coat-derived pooled; FFP, fresh-frozen plasma;
PD, pulmonary distress; RR, relative risk; ICU, intensive care unit; PAR, population-attributable risk.
a
PD was defined as PaO2/FiO2 ratio < 300.
b
These results accounted for ALL TRALI cases. When possible TRALI cases were excluded, the PAR was 0
37 (0
06–0
58).
c
These numbers pertain to transfusion of plasma. The incidence of TRALI did not change with red blood cell and platelet transfusion.
d
Male-only plasma donor and plasma from female donors without history of pregnancy or without HLA/HNA antibodies.
TRALI risk mitigation 697
698 Z. K. Otrock et al.

In the United States, progress towards implementing plasma for transfusion. The CBS experience with TRALI
TRALI risk mitigation strategies lagged behind the United from 2001 to 2009 was reported by Lin et al. [49]: in the
Kingdom. The AABB introduced, in a standard in 2006, a first 3 years, TRALI reports increased, most likely due to
requirement that blood banks shall evaluate donors impli- increased reporting and awareness that CBS was investi-
cated in TRALI regarding their continued eligibility to gating TRALI cases. From 2004 to 2007, a relatively
donate [45]. In November 2006, an AABB bulletin intro- stable period of reporting was observed (16, 14, 21 and
duced guidelines that recommended the implementation 36 TRALI cases for the years from 2004 to 2007, respec-
of measures to minimize the preparation of high plasma- tively) and accounted for the period when the TRALI risk
volume components from donors known to be leucocyte reduction measures were being implemented. The reported
alloimmunized or at increased risk for alloimmunization cases of definite and possible TRALI decreased in 2008
[46]. In 2007, another AABB bulletin elucidated that each (19 cases) and 2009 (12 cases) [49]. Although many
institution should design a TRALI risk reduction program TRALI risk reduction measures including those mentioned
to lower the risk of TRALI without compromising the above were implemented in Canada, the observed
availability of needed blood components [47]. The Ameri- decrease in plasma-associated cases was mostly attributed
can Red Cross (ARC) started distributing plasma collected to the change to predominantly male donor plasma for
from male donors (with the exception of Group AB transfusion. This was supported by laboratory investiga-
donors) only for transfusion to patients, while plasma tions showing that a female donor was implicated in 73%
from female donors was diverted to pharmaceuticals for of TRALI cases with detected donor antibodies [49].
manufacturing [48]. By November 2007, 95% of plasma
delivered for transfusion through the ARC was from male
Donor screening and testing
donors. The analysis of data from the ARC from 2006
through 2008 showed that TRALI involving only plasma Distributing plasma from male donors has significantly
transfusion was significantly reduced in 2008 compared reduced the risk of TRALI by 80% in the United States
to 2006 (32 vs. seven cases), to a level not different from [48]. Blood centres in Canada, United Kingdom and other
the rate of TRALI observed for RBC transfusion [48]. The European countries have also reported comparable benefit
ARC hemovigilance program revealed a continued reduc- of predominantly male-donor plasma in reducing TRALI
tion in the risk of TRALI with plasma transfusion in the [39, 42, 43, 49]. In spite of the success in implementing
4 years following the implementation of a predominantly predominantly male plasma, TRALI, including possible
male-donor plasma strategy: the risk of TRALI signifi- TRALI, remains the leading cause of transfusion-related
cantly decreased from 18
6 cases per million in 2006 to mortality reported to the Food and Drug Administration
4
2 cases per million in 2008 to 2011 (OR, 0
22; 95% CI, (FDA) in the last 7 years (2009–2015) [3]. The ISBT Work-
0
13–0
38; P < 0
0001) [48]. However, the risk from ing Party on Granulocyte Immunobiology suggested that
group AB plasma remained the same and was signifi- leucocyte antibody screening can identify alloimmunized
cantly greater than groups A, B, and O plasma in 2008 to blood donors, which subsequently can contribute to the
2011 reflecting continued reliance on female donors to prevention of antibody-mediated TRALI [50]. In April
meet increasing demand for AB plasma. 2009, the NHS Blood and Transplant (NHSBT) in the Uni-
Blood banks in Canada have also implemented mea- ted Kingdom began screening female apheresis platelet
sures to reduce TRALI. Measures taken progressively by donors for leucocyte antibodies, and only those who
Canadian Blood Services (CBS) to reduce the risk of tested negative were accepted for platelet donation. Over
TRALI included the use of plasma from predominantly a period of 12 months, 1157 female donors were evalu-
male donors, male donor plasma for suspension of whole ated; 336 (29%) donors tested positive for leucocyte anti-
blood-derived buffy coat platelets, and deferral of women bodies and were routed to RBC component donation [51].
with high risk of having HLA antibodies from apheresis This process, combined with other policies, further
platelet donation, particularly previously pregnant reduced TRALI incidence.
females. The CBS provides blood products to all provinces According to the Red Cross hemovigilance program,
in Canada, with the exception of the province of Qu
ebec. the residual TRALI risk from plasma was mostly attributed
The CBS TRALI Medical Review Group (TMRG) was estab- to group AB plasma as 40% of the distributed compo-
lished in 2006 and has developed national policies for nents were from female donors [52]. In addition, the risk
investigation of suspected TRALI cases reported to CBS. of TRALI from apheresis platelets was still present as
The group classifies all reported cases of potential TRALI women were still donating apheresis platelets; pooled pla-
based on the Canadian consensus conference definitions telets conferred lower risk for TRALI, presumably, because
[1]. In the Canadian production sites, there has been from of dilution effect of combining plasma from women with
93% to 96% compliance rate with using only male plasma from men. In addition to the recommendations on

© 2017 International Society of Blood Transfusion

Vox Sanguinis (2017) 112, 694–703
 TRALI risk mitigation 699

TRALI risk reduction released in 2006–2007, the AABB different assays [59]. Users may also take into considera-
announced more rigorous requirements for TRALI risk tion the throughput, cost-effectiveness and assay variation
mitigation and, as of 1 April 2014, all distributed plasma when choosing the method suitable for their own facilities.
and whole blood were required to be from men, women Second, work needs to be done to standardize the donor
who have never been pregnant, or women who have screening given the variety of assays in use. Blood estab-
tested negative for HLA antibodies since their most recent lishment may certainly choose to implement a single plat-
pregnancy [53]. Furthermore, the AABB extended this form across the system. But the precision of HLA antibody
requirement to include apheresis platelets with an imple- testing has been susceptible to relatively large interlabora-
mentation deadline of 1 October 2016 [54]. The additional tory, intralaboratory and lot-to-lot variations [60, 61].
time allowed for the implementation of these recommen- Eventually harmonization of the single antigen assay may
dations for apheresis platelets was based on concerns be achieved through interlaboratory collaborations and
about platelet availability. The effects of these standards vigorous quality improvement programs. Third, the opti-
will require time to measure their effectiveness. Nonethe- mal cut-off values for donor screening is yet to be deter-
less, a recent study evaluating the risk of TRALI caused mined. One of the most commonly referenced cut-off for
by apheresis platelets transfusions tried to estimate the donor HLA-antibody screen was based on statistical infer-
long-term benefit of testing female plateletpheresis ence using the negative background (NBG) ratios from
donors who have been pregnant for HLA antibodies [55]. non-sensitized males (LabScreen Mixed, ThermoFisher)
The rate of TRALI was significantly higher for apheresis [62]. The impact of a fixed cut-off may vary depending on
platelets compared to plasma or RBCs; 29 of the 41 a facility’s local demographics. Furthermore, commonly
apheresis platelets cases were donated by women, of used cut-off values (NBG ratio of 10
8 and 6
9 for class I
whom 28 had been pregnant, and one had not been preg- and II antibodies, respectively) have not been correlated
nant and was not tested. Twenty-five (61%) of the with the pathogenic potential of detected antibodies
apheresis platelets causing TRALI were donated by [58, 62, 63]. In the study by Toy et al., [16] components
women with HLA Class I/II antibodies. In five of six cases with Class II antibodies with NBG ratio above 27
5 was
that involved components containing HNA antibodies, the reported to increase the risk of TRALI (OR = 1
92), while
parous female donors also tested positive for multiple components with class I antibody were not associated with
HLA antibodies. The authors concluded that testing all TRALI even when the NGB ratio was above 59
3. The con-
parous female plateletpheresis donors for HLA antibodies troversy regarding the optimal cut-off for donor screening
might reduce the risk of TRALI by approximately 60% may persist until more high-quality evidence, correlating
and subsequently preclude some cases with coexisting the strength of antibody detected in vitro with the proba-
HNA antibodies [55]. bility of TRALI occurring in a patient, becomes available
Although the implementation of donor screening for [16].
alloantibodies will reduce the risk, it may not entirely
eliminate TRALI in the setting of apheresis platelet trans-
Pooled solvent/detergent-treated plasma
fusion as antibody-mediated TRALI makes up to 85% of
all reported cases [56]. In addition, anti-HLA antibodies Solvent/detergent treatment of pooled plasma has been
have been detected in 10
6% of women without a history used to reduce the risk of transmission of viruses from
of pregnancy or transfusion and in 4
3% of male donors multiple donor exposure. Some blood banks have used
without a history of transfusion [57]. However, these rates pooling of high plasma containing blood components to
of antibodies detected are much higher than the rates potentially mitigate antibody-mediated TRALI. Compo-
reported in blood donors in the United States [58]; this nents are prepared from a pool of multiple donors which
may be due to differences in assay cut-off. may dilute any leucocyte antibodies or neutralize them
Donor screening for TRALI mitigation also faces several by residual leucocytes or soluble HLA antigens in the
technical challenges. First, the optimal assay for this pur- pool, thus potentially reducing the risk of antibody-
pose is not unknown. Facilities in the USA may use either mediated TRALI [64].
an FDA cleared test for donor screening (Lifecodes In Norway, no cases of TRALI have been reported in
DonorScreen-HLA ELISA, Immucor) or other commercial more than 250 000 units of pooled plasma that have been
assays with laboratory validated cut-offs. These assays dif- transfused over a period of 12 years after the implemen-
fer in their design, sensitivity and specificity. Despite the tation of the solvent/detergent plasma in 1993 [65]. Con-
complexity in comparing the performance of different sistent with these clinical findings, Sachs et al. [66]
assays in detecting at-risk donors, it has been shown that screened 20 batches of solvent/detergent plasma that were
comparable cut-offs can be established to achieve broad found to be negative for granulocyte-specific as well as
agreement regarding positivity designation across HLA Class I and Class II antibodies. The same findings of

© 2017 International Society of Blood Transfusion

Vox Sanguinis (2017) 112, 694–703
700 Z. K. Otrock et al.

negative screening for HLA antibodies in pooled plasma relative to its benefit. In contrast, the implementation of
were reported in a study from the United Kingdom [67]. HLA antibody screening requires the addition of a labora-
In the United Kingdom, it was deemed more cost-effective tory screening test with known costs. Moreover, as this
to institute a predominantly male-plasma policy rather strategy has been implemented after predominantly male
than solvent/detergent plasma [68]. donor strategy implementation, its impact in reducing
residual TRALI risk is expected to be lower. Additional
potential strategies such as PAS, volume reduction and
Platelet additive solutions (PAS)
washing are also costly to implement and may result in
Reducing the volume of plasma in platelets has been lower count increments.
shown to decrease the incidence of some transfusion
reactions [69–71]. Platelet additive solutions (PAS) are
Conclusions
isotonic solutions commonly used as a substitute for
plasma to reduce the amount of plasma transfused with Since the adoption of a consensus definition for TRALI
platelets, and thus reduced other transfusion adverse and the better understanding of its pathogenesis, differ-
events, and to provide the possibility for pathogen inacti- ent strategies for TRALI risk mitigation have been intro-
vation using photochemical treatment [72]. PAS also duced. It is clear that there has not been a single
offers the possibility of adding certain components that comprehensive approach for TRALI risk mitigation. The
improve platelet storage conditions [73]. The use of PAS implementation of predominantly male plasma strategy
replaces 65% of plasma including plasma proteins and has significantly reduced the risk for antibody-mediated
isohemagglutinins. PAS is theoretically an available TRALI in plasma-containing blood components. However,
option for TRALI risk reduction in the setting of platelet TRALI remains the leading cause of transfusion-related
transfusion; however, evidence supporting its use is quite fatalities reported to the US FDA. The recently mandated
limited. Different formulations are available on the mar- HLA antibody testing for previously pregnant women is
ket and are extensively used in Europe; however, they are expected to further decrease the risk of TRALI for plasma;
less commonly used in the United States at this point in its impact on TRALI in platelets still needs to be deter-
time. mined. Additional options such as PAS, volume reduction
Another recent large multicenter study reported that and washing are being entertained [71, 76], but are
PAS-3 platelets were statistically superior to apheresis costly and may result in lower count increments. Further
platelets stored in 100% plasma with respect to the trans- studies are warranted to assess whether any of the new
fusion-related adverse reactions rate [74]. There were no approaches, namely HLA antibody screening, PAS and
TRALI cases observed in either study group; however, all washing, really reduce TRALI risk when applied as an
platelets were collected using additional TRALI mitigation addition to the existing mitigation strategies. The current
strategies. Although these findings might imply a plausi- risk mitigation focuses on donor- and product-related
ble reduction in TRALI risk, the use of PAS is less likely factors especially HLA antibodies; however, these factors
to completely eliminate the risk of TRALI from apheresis do not necessarily play a role in non-antibody-mediated
platelets. PAS platelets contain plasma, and there is evi- TRALI. Physicians should pay careful attention to the
dence to suggest that products containing a plasma vol- patient’s risk factors for ALI and implement evidence-
ume as little as 10–20 ml have been implicated in TRALI based transfusion practices to avoid transfusions when
cases [75]. This suggestion is supported by the fact that they are not indicated.
RBC with additive solutions, which theoretically contain
less plasma than platelets in PAS, are still associated with
Author contributions
TRALI [6].
ZO, CL, and BJG provided substantial contributions to
research design, or the acquisition, analysis or interpreta-
Cost
tion of data. ZO, CL and BJG involved in drafting the
Although the above-mentioned strategies have been manuscript or revising it critically. ZO, CL and BJG per-
implemented already in many countries, the cost of each formed approval of the submitted and final version.
must be considered and weighed against the potential
benefits. For example, the use of predominantly or exclu-
Conflict of interest
sively male donor strategy for plasma containing blood
products had a relatively low cost of implementation The authors declare no conflict of interests.

© 2017 International Society of Blood Transfusion

Vox Sanguinis (2017) 112, 694–703
 TRALI risk mitigation 701

References
1 Kleinman S, Caulfield T, Chan P, TRALI. Transfus Med Rev 2005; 19:2– plasma from male donors in the Amer-
et al.: Toward an understanding of 31 ican Red Cross. Transfusion 2007;
transfusion-related acute lung injury: 12 Toy P, Kleinman SH, Looney MR: Pro- 47:599–607
statement of a consensus panel. Trans- posed revised nomenclature for trans- 22 Middelburg RA, Van Stein D, Zupanska
fusion 2004; 44:1774–1789 fusion-related acute lung injury. B, et al.: Female donors and transfu-
2 Vlaar AP, Juffermans NP: Transfusion- Transfusion 2017; 57:709–713 sion-related acute lung injury: a
related acute lung injury: a clinical 13 Gajic O, Gropper MA, Hubmayr RD: case-referent study from the Interna-
review. Lancet 2013; 382:984–994 Pulmonary edema after transfusion: tional TRALI Unisex Research Group.
3 Center for Biologics Evaluation and how to differentiate transfusion- Transfusion 2010; 50:2447–2454
Research (CBER)/Food and Drug associated circulatory overload from 23 Gajic O, Rana R, Winters JL, et al.:
Administration (FDA). Fatalities repor- transfusion-related acute lung injury. Transfusion-related acute lung injury
ted to FDA following blood collection Crit Care Med 2006; 34(5 Suppl): in the critically ill: prospective nested
and transfusion. Annual summary for S109–S113 case-control study. Am J Respir Crit
fiscal year 2015. http://www.fda.gov/Bio 14 Shere-Wolfe RF, Galvagno SM Jr, Care Med 2007; 176:886–891
logicsBloodVaccines/SafetyAvailability/ Grissom TE: Critical care considera- 24 Vlaar AP, Hofstra JJ, Determann RM,
ReportaProblem/TransfusionDonationFa tions in the management of the et al.: The incidence, risk factors, and
talities/. [Last accessed 31 January 2017] trauma patient following initial resus- outcome of transfusion-related acute
4 Silliman CC, Boshkov LK, Mehdiza- citation. Scand J Trauma Resusc lung injury in a cohort of cardiac sur-
dehkashi Z, et al.: Transfusion-related Emerg Med 2012; 20:68 gery patients: a prospective nested
acute lung injury: epidemiology and a 15 Skeate RC, Eastlund T: Distinguishing case-control study. Blood 2011;
prospective analysis of etiologic fac- between transfusion related acute lung 117:4218–4225
tors. Blood 2003; 101:454–462 injury and transfusion associated cir- 25 Toy P, Bacchetti P, Grimes B, et al.:
5 Sachs UJ, Bux J: TRALI after the culatory overload. Curr Opin Hematol Recipient clinical risk factors predomi-
transfusion of cross-match-positive 2007; 14:682–687 nate in possible transfusion-related
granulocytes. Transfusion 2003; 43: 16 Toy P, Gajic O, Bacchetti P, et al.: acute lung injury. Transfusion 2015;
1683–1686 Transfusion-related acute lung injury: 55:947–952
6 Weber LL, Roberts LD, Sweeney JD: incidence and risk factors. Blood 2012; 26 Popovsky MA, Abel MD, Moore SB:
Residual plasma in red blood cells and 119:1757–1767 Transfusion-related acute lung injury
transfusion-related acute lung injury. 17 Vlaar AP, Wortel K, Binnekade JM, associated with passive transfer of
Transfusion 2014; 54:2425–2430 et al.: The practice of reporting trans- antileukocyte antibodies. Am Rev
7 Rizk A, Gorson KC, Kenney L, et al.: fusion-related acute lung injury: a Respir Dis 1983; 128:185–189
Transfusion-related acute lung injury national survey among clinical and 27 Kopko PM, Popovsky MA, MacKenzie
after the infusion of IVIG. Transfusion preclinical disciplines. Transfusion MR, et al.: HLA class II antibodies in
2001; 41:264–268 2010; 50:443–451 transfusion-related acute lung injury.
8 Ganguly S, Carrum G, Nizzi F, et al.: 18 Popovsky MA, Moore SB: Diagnostic Transfusion 2001; 41:1244–1248
Transfusion-related acute lung injury and pathogenetic considerations in 28 Davoren A, Curtis BR, Shulman IA,
(TRALI) following allogeneic stem cell transfusion-related acute lung injury. et al.: TRALI due to granulocyte-
transplant for acute myeloid leukemia. Transfusion 1985; 25:573–577 agglutinating human neutrophil anti-
Am J Hematol 2004; 75:48–51 19 Finlay HE, Cassorla L, Feiner J, et al.: gen-3a (5b) alloantibodies in donor
9 Leach M, Vora AJ, Jones DA, et al.: Designing and testing a computer- plasma: a report of 2 fatalities. Trans-
Transfusion-related acute lung injury based screening system for transfu- fusion 2003; 43:641–645
(TRALI) following autologous stem cell sion-related acute lung injury. Am J 29 Silliman CC: The two-event model of
transplant for relapsed acute myeloid Clin Pathol 2005; 124:601–609 transfusion-related acute lung injury.
leukaemia: a case report and review of 20 Politis C, Wiersum JC, Richardson C, Crit Care Med 2006; 34(5 Suppl):
the literature. Transfus Med 1998; et al.: The international haemovigi- S124–S131
8:333–337 lance network database for the 30 Silliman CC, Voelkel NF, Allard JD,
10 Bernard GR, Artigas A, Brigham KL, surveillance of adverse reactions and et al.: Plasma and lipids from stored
et al.: The American-European Consen- events in donors and recipients of packed red blood cells cause acute
sus Conference on ARDS. Definitions, blood components: technical issues lung injury in an animal model. J Clin
mechanisms, relevant outcomes, and and results. Vox Sang 2016; 111:409– Invest 1998; 101:1458–1467
clinical trial coordination. Am J Respir 417 31 Silliman CC, Bjornsen AJ, Wyman TH,
Crit Care Med 1994; 149:818–824 21 Eder AF, Herron R, Strupp A, et al.: et al.: Plasma and lipids from stored
11 Goldman M, Webert KE, Arnold DM, Transfusion-related acute lung injury platelets cause acute lung injury in an
et al.: Proceedings of a consensus con- surveillance (2003–2005) and the animal model. Transfusion 2003;
ference: towards an understanding of potential impact of the selective use of 43:633–640

© 2017 International Society of Blood Transfusion

Vox Sanguinis (2017) 112, 694–703
702 Z. K. Otrock et al.

32 Bux J, Sachs UJ: The pathogenesis of German haemovigilance data (2006– development of national guidelines.
transfusion-related acute lung injury 2010). Vox Sang 2012; 102:317–323 Vox Sang 2012; 103:10–17
(TRALI). Br J Haematol 2007; 136:788– 43 Wiersum-Osselton JC, Middelburg RA, 52 Eder AF, Dy BA, Perez JM, et al.: The
799 Beckers EA, et al.: Male-only fresh- residual risk of transfusion-related
33 Kopko P, Silva M, Shulman I, et al.: frozen plasma for transfusion-related acute lung injury at the American Red
AABB survey of transfusion-related acute lung injury prevention: before- Cross (2008–2011): limitations of a
acute lung injury policies and prac- and-after comparative cohort study. predominantly male-donor plasma
tices in the United States. Transfusion Transfusion 2011; 51:1278–1283 mitigation strategy. Transfusion 2013;
2007; 47:1679–1685 44 Jutzi M, Amsler L. SwissMedic Annual 53:1442–1449
34 Kleinman S, Grossman B, Kopko P: A Hemovigilance Report 2013, pp. 1–28, 53 Sher GM, Markowitz MA. AABB Asso-
national survey of transfusion-related 2014. Swiss-Medic, Bern. https:// ciation Bulletin #14-02 TRALI Risk
acute lung injury risk reduction poli- www.google.com/#q=Transfusion-rela Mitigation for Plasma and Whole
cies for platelets and plasma in the ted+acute+lung+injury+R%E2%82% Blood for Allogeneic http://www.aabb.
United States. Transfusion 2010; ACuesch%2C+M.+%26+Jutzi%2C+M.+ org/programs/publications/bulletins/d
50:1312–1321 (2013)+SwissMedic+Annual [Last accessed ocuments/ab14-02.pdf [Last accessed
35 Schmickl CN, Mastrobuoni S, Filippidis 31 January 2017] 31 January 2017]
FT, et al.: Male-predominant plasma 45 AABB Standard 5.4.2.1. Standards for 54 Sher G, Markowitz MA. AABB Associ-
transfusion strategy for preventing Blood Banks and Transfusion Services, ation Bulletin #14-07: Interim Stan-
transfusion-related acute lung injury: 24th edn. Bethesda (MD), AABB, 2006. dard to the 29th edition of Standards
a systematic review. Crit Care Med 46 Strong MD, Lipton KS. AABB Associa- for Blood Banks and Transfusion Ser-
2015; 43:205–225 tion Bulletin #06–07 Transfusion- vices. http://www.aabb.org/programs/
36 M€ uller MC, van Stein D, Binnekade Related Acute Lung Injury. http:// publications/bulletins/Documents/ab
JM, et al.: Low-risk transfusion-related www.bpro.or.jp/publication/pdf_jptrans/ 14-07.pdf. [Last accessed 31 January
acute lung injury donor strategies and us/us200611en.pdf.; http://www.aabb. 2017].
the impact on the onset of transfu- org/programs/publications/bulletins/Doc 55 Eder AF, Dy BA, O’Neill EM: Predicted
sion-related acute lung injury: a s/ab06-07.pdf [Last accessed 31 Jan- effect of selectively testing female
meta-analysis. Transfusion 2015; 55: uary 2017]. donors for HLA antibodies to mitigate
164–175 47 Connor JD, Lipton KS. AABB Associa- transfusion-related acute lung injury
37 Welsby IJ, Troughton M, Phillips-Bute tion Bulletin #07-03. Clarifications to risk from apheresis platelets. Transfu-
B, et al.: The relationship of plasma recommendations to reduce the risk of sion 2016; 56:1608–1615
transfusion from female and male TRALI. http://www.aabb.org/programs/ 56

Alvarez P, Carrasco R, Romero-
donors with outcome after cardiac sur- publications/bulletins/Pages/ab07-03.a Dapueto C, et al.: Transfusion-Related
gery. J Thorac Cardiovasc Surg 2010; spx [Last accessed 31 January 2017]. Acute Lung Injured (TRALI): current
140:1353–1360 48 Eder AF, Herron RM Jr, Strupp A, concepts. Open Respir Med J 2015;
38 Williamson LM, Heptonstall J, Soldan et al.: Effective reduction of transfu- 9:92–96
K: A SHOT in the arm for safer blood sion-related acute lung injury risk with 57 Sigle JP, Thierbach J, Infanti L, et al.:
transfusion. BMJ 1996; 313:1221– male-predominant plasma strategy in Anti-leucocyte antibodies in platelet
1222 the American Red Cross (2006–2008). apheresis donors with and without
39 Chapman CE, Stainsby D, Jones H, Transfusion 2010; 50:1732–1742 prior immunizing events: implications
et al.: Ten years of hemovigilance 49 Lin Y, Saw CL, Hannach B, et al.: for TRALI prevention. Vox Sang 2013;
reports of transfusion-related acute Transfusion-related acute lung injury 105:244–252
lung injury in the United Kingdom prevention measures and their impact 58 Triulzi DJ, Kleinman S, Kakaiya RM,
and the impact of preferential use of at Canadian Blood Services. Transfu- et al.: The effect of previous pregnancy
male donor plasma. Transfusion 2009; sion 2012; 52:567–574 and transfusion on HLA alloimmuniza-
49:440–452 50 ISBT Working Party on Granulocyte tion in blood donors: implications for a
40 Serious Hazards Of Transfusion, Annu- Immunobiology, Bierling P, Bux J, transfusion-related acute lung injury
als Reports and Summaries. https:// et al.: Recommendations of the ISBT risk reduction strategy. Transfusion
www.shotuk.org/shot-reports/ [Last Working Party on Granulocyte 2009; 49:1825–1835
accessed 9 July 2017]. Immunobiology for leucocyte antibody 59 Carrick DM, Norris PJ, Endres RO, et al.:
41 Keller-Stanislawski B, Reil A, G€
unay S, screening in the investigation and pre- Establishing assay cutoffs for HLA anti-
et al.: Frequency and severity of trans- vention of antibody-mediated transfu- body screening of apheresis donors.
fusion-related acute lung injury–Ger- sion-related acute lung injury. Vox Transfusion 2011; 51:2092–2101
man haemovigilance data (2006–2007). Sang 2009; 96:266–269 60 Gandhi MJ, Carrick DM, Jenkins S,
Vox Sang 2010; 98:70–77 51 Lucas G, Win N, Calvert A, et al.: et al.: Lot-to-lot variability in HLA
42 Funk MB, Guenay S, Lohmann A, et al.: Reducing the incidence of TRALI in antibody screening using a multiplexed
Benefit of transfusion-related acute the UK: the results of screening for bead-based assay. Transfusion 2013;
lung injury risk-minimization measures– donor leucocyte antibodies and the 53:1940–1947

© 2017 International Society of Blood Transfusion

Vox Sanguinis (2017) 112, 694–703
 TRALI risk mitigation 703

61 Reed EF, Rao P, Zhang Z, et al.: Com- 68 Riedler GF, Haycox AR, Duggan AK, cause TRALI? Transfus Med 2008;
prehensive assessment and standard- et al.: Cost-effectiveness of solvent/ 18:276–280
ization of solid phase multiplex-bead detergent-treated fresh-frozen plasma. 76 Dunbar NM: Current options for
arrays for the detection of antibodies Vox Sang 2003; 85:88–95 transfusion-related acute lung injury
to HLA. Am J Transplant 2013; 69 Heddle NM, Blajchman MA, Meyer risk mitigation in platelet transfu-
13:1859–1870 RM, et al.: A randomized controlled sions. Curr Opin Hematol 2015;
62 Kakaiya RM, Triulzi DJ, Wright DJ, trial comparing the frequency of 22:554–558
et al.: Prevalence of HLA antibodies in acute reactions to plasma-removed 77 Wright SE, Snowden CP, Athey SC,
remotely transfused or alloexposed platelets and prestorage WBC-reduced et al.: Acute lung injury after ruptured
volunteer blood donors. Transfusion platelets. Transfusion 2002; 42:556– abdominal aortic aneurysm repair: the
2010; 50:1328–1334 566 effect of excluding donations from
63 Endres RO, Kleinman SH, Carrick DM, 70 Azuma H, Hirayama J, Akino M, et al.: females from the production of fresh
et al.: Identification of specificities of Reduction in adverse reactions to pla- frozen plasma. Crit Care Med 2008;
antibodies against human leukocyte telets by the removal of plasma super- 36:1796–1802
antigens in blood donors. Transfusion natant and resuspension in a new 78 Nakazawa H, Ohnishi H, Okazaki H,
2010; 50:1749–1760 additive solution (M-sol). Transfusion et al.: Impact of fresh-frozen plasma
64 M}uller MC, Porcelijn L, Vlaar AP: 2009; 49:214–218 from male-only donors versus mixed-
Prevention of immune-mediated trans- 71 Tobian AA, Savage WJ, Tisch DJ, sex donors on postoperative respira-
fusion-related acute lung injury; from et al.: Prevention of allergic transfu- tory function in surgical patients: a
bloodbank to patient. Curr Pharm Des sion reactions to platelets and red prospective case-controlled study.
2012; 18:3241–3248 blood cells through plasma reduction. Transfusion 2009; 49:2434–2441
65 Flesland O, Seghatchian J, Solheim Transfusion 2011; 51:1676–1683 79 Vlaar AP, Binnekade JM, Prins D, et al.:
BG: The Norwegian Plasma Fractiona- 72 Murphy S: The efficacy of synthetic Risk factors and outcome of transfu-
tion Project–a 12 year clinical and media in the storage of human plate- sion-related acute lung injury in the
economic success story. Transfus lets for transfusion. Transfus Med Rev critically ill: a nested case-control
Apher Sci 2003; 28:93–100 1999; 13:153–163 study. Crit Care Med 2010; 38:771–778
66 Sachs UJ, Kauschat D, Bein G: White 73 Gulliksson H: Defining the optimal 80 Arinsburg SA, Skerrett DL, Karp JK,
blood cell-reactive antibodies are unde- storage conditions for the long-term et al.: Conversion to low transfusion-
tectable in solvent/detergent plasma. storage of platelets. Transfus Med Rev related acute lung injury (TRALI)-risk
Transfusion 2005; 45:1628–1631 2003; 17:209–215 plasma significantly reduces TRALI.
67 Sinnott P, Bodger S, Gupta A, et al.: 74 Cohn CS, Stubbs J, Schwartz J, et al.: Transfusion 2012; 52:946–952
Presence of HLA antibodies in single- A comparison of adverse reaction rates 81 Clifford L, Jia Q, Subramanian A,
donor-derived fresh frozen plasma for PAS C versus plasma platelet units. et al.: Characterizing the epidemiology
compared with pooled, solvent deter- Transfusion 2014; 54:1927–1934 of postoperative transfusion-related
gent-treated plasma (Octaplas). Eur J 75 Win N, Chapman CE, Bowles KM, acute lung injury. Anesthesiology
Immunogenet 2004; 31:271–274 et al.: How much residual plasma may 2015; 122:12–20

© 2017 International Society of Blood Transfusion

Vox Sanguinis (2017) 112, 694–703