Coagulation and Transfusion Medicine / TRANSFUSION-RELATED ACUTE LUNG INJURY

Transfusion-Related Acute Lung Injury

Past, Present, and Future
Tad Cherry, MD,1 Mark Steciuk, MD, PhD,2 Vishnu V.B. Reddy, MD,2 and Marisa B. Marques, MD2
Key Words: Transfusion-related acute lung injury; TRALI; Transfusion reaction; HLA antibodies; Neutrophilic infiltrates
DOI: 10.1309/D3F7BXH466AE3G0P

Abstract
Noncardiogenic pulmonary edema caused by
transfusion has been observed for almost 60 years.
Today, we know this entity as transfusion-related acute
lung injury (TRALI). TRALI is an uncommon but
potentially fatal adverse reaction to transfusion of
plasma-containing blood components. It is typified by
dyspnea, cough, hypoxemia, and pulmonary edema
within 6 hours of transfusion. Most commonly, it is
caused by donor HLA antibodies that react with
recipient antigens. It may also be caused by
biologically active compounds accumulated during
storage of blood products, which are capable of
priming neutrophils. Without a gold standard, the
diagnosis of TRALI relies on a high index of suspicion
and on excluding other types of transfusion reactions.
Although current definitions of TRALI depend on
symptoms, laboratory parameters can aid in the
diagnosis and frequently identify the causative donor
unit. As our understanding of TRALI deepens, risk
reduction or prevention may become possible.

Recently, a 72-year-old man with peripheral vascular disease was admitted to our institution with an occluded right
lower extremity arterial bypass graft and cellulitis of the right
foot. On the second hospital day, 2 U of fresh frozen plasma
(FFP) were ordered to correct an international normalized
ratio of 1.73 before surgery.
After transfusion of approximately 100 mL of the first
unit of FFP, the patient became acutely short of breath, hypoxemic, and hypotensive. The transfusion was stopped, and the
patient was given supplemental oxygen. A chest radiograph
taken shortly after the onset of symptoms revealed bilateral
pulmonary infiltrates. Intubation and mechanical ventilation
were required to maintain adequate oxygenation, and vasopressors were needed to maintain adequate perfusion. A CBC
count drawn approximately 1 hour after the onset of symptoms revealed a WBC count of 7,500/L (7.5 109/L), which
was markedly decreased from 10,560/L (10.6 109/L) earlier that day. During the ensuing hours, the patient became
anuric, and after several hours of maximal support, the family
decided against further resuscitative measures. Shortly thereafter, the patient became asystolic and was pronounced dead.
Autopsy revealed edematous lungs with a combined
weight of 2,780 g. Histologic examination revealed neutrophilic aggregates throughout the pulmonary, central nervous system, hepatic, and renal vasculature. Investigation of the
implicated unit of FFP revealed that it was from a multiparous
female donor with class I and II histocompatibility antibodies.

The Pre-TRALI Era

In 1950, Lanman and colleagues1 described what may be
the first published account of TRALI. To our knowledge, this
Am J Clin Pathol 2008;129:287-297

American Society for Clinical Pathology

287

DOI: 10.1309/D3F7BXH466AE3G0P

287
287

Cherry et al / TRANSFUSION-RELATED ACUTE LUNG INJURY

report has thus far gone unnoticed in the TRALI literature.

They described an unusual reaction to a transfusion of whole
blood from a 68-year-old donor with chronic lymphocytic
leukemia to a 54-year-old man with squamous cell carcinoma
of the esophagus. The recipient had been transfused with
whole blood from the same donor 18 days prior to the reaction. Immediately after beginning the second transfusion, noncardiogenic pulmonary edema, evidenced by dyspnea,
cyanosis, asthmatic breathing, rales, chills, and fever, developed in the recipient. Within 5 minutes, his total WBC count
fell from 18,430/L (18.4 109/L) to 1,700/L (1.7 109/L),
and the relative neutrophil count fell to 8% of pretransfusion
levels. The authors speculated that this reaction was anaphylactoid and due to transfusion of blood from a leukemic person causing WBC margination in the recipients lungs.
One year later, Barnard2 described a connection between
transfusion of allogeneic blood and pulmonary compromise.
The patient had acute leukemia and sought care because of
epistaxis. After having his nose packed, he was given a transfusion of whole blood, during which acute pulmonary edema
developed.2 Shortly thereafter, a second transfusion commenced, and the patient died. Barnard2 ascribed the pulmonary edema to hypersensitivity, not volume overload.
Bittingham3 was able to induce a similar set of symptoms by
infusing whole blood with known leukoagglutinins into a
healthy volunteer. Approximately 45 minutes after infusion of
50 mL of blood, the test subject began experiencing fever,
chills, tachypnea, dyspnea, cyanosis, hypotension, and
leukopenia followed by leukocytosis. By the next day, the
patients symptoms had resolved.
Years later, Phillips and Fleischner4 described the first
series of cases in which pulmonary edema, thought not to be
attributable to volume overload, developed in patients who
had received transfusions. The first patient was a 23-year-old
parturient who was given 500 mL of whole blood for anemia.
Fifteen minutes into the transfusion, a cough developed, and
her chest radiograph revealed bilateral pulmonary edema.
Aggressive diuresis was carried out, but the pulmonary edema
showed only mild initial improvement. By the third day, the
edema had resolved. In 2 other patients, fever and cyanosis
developed after they received whole blood, and they were
noted to have pulmonary edema that resolved within 1 to 5
days. Phillips and Fleischner4 postulated that the pulmonary
edema was due to incompatibility of undetermined nature.
In 1968, Ward et al5 described the case of a 19-year-old
man with melena in whom a shaking chill and fever developed
during infusion of a third unit of blood. Although pulmonary
edema was diagnosed radiographically, circulatory overload
was dismissed based on the patients orthostatic symptoms.
He was diagnosed with a pulmonary hypersensitivity reaction,
and leukoagglutinins were later found in his serum. Ward6
added 3 similar cases to the literature in 1970.
288
288

Am J Clin Pathol 2008;129:287-297

DOI: 10.1309/D3F7BXH466AE3G0P

Thompson et al7 contributed to the literature in 1971 with

the report of 2 cases of pulmonary edema attributable to whole
blood in normovolemic patients. In both patients, chills, fever,
dyspnea, cough, tachycardia, and hypotension developed
abruptly, but the pulmonary edema resolved quickly during
the ensuing days. Further investigation into these reactions
revealed erythrocyte incompatibility was not the culprit, and
cultures from donor and recipient blood samples were negative. Instead, antibodies from donor serum reacting with recipient leukocytes were found in both cases. Extensive testing of
the antibody from the first donor revealed a specificity that
was probably not directly related to the human leukocyte antigens (HLA) A locus, the neutrophil antigens NA1, NB1, NC1,
or the 5a-5b locus.7 In both cases, the donors were multiparous women.

A New Disease
Numerous case reports of pulmonary reactions due to
transfusions were published after the initial description by
Lanman et al.1 Anaphylactoid,1 hypersensitivity,2,8 unknown
incompatibility,4 allergic,8,9 and noncardiogenic pulmonary
edema10 are some of the terms used to describe this pattern of
pulmonary deterioration following transfusion of blood products. In 1983, Popovsky et al11 recognized a pattern of acute
pulmonary compromise in a series of 5 patients and termed it
transfusion-related acute lung injury (TRALI). All patients
had received whole blood or packed RBCs (PRBCs). Analysis
of the recipient samples and donor units revealed lymphocytotoxic antibodies in the latter. HLA antibodies against recipient antigens were found in 3 of 5 cases.11 Based on the total
number of patients and units transfused during the observed
period, they estimated an incidence of TRALI of 0.02% per
unit transfused and 0.16% per recipient.11 Although rare,
TRALI has become the leading cause of transfusion-related
deaths reported to the US Food and Drug Administration
(FDA) since 2004 (J.C. Goldsmith, written communication,
May 4, 2007).
In an effort to confirm their initial findings, Popovsky and
Moore12 identified 36 patients with acute respiratory distress
occurring within 4 hours of transfusion from 22,292 patients
who had received 194,715 units. In 89% of the cases, granulocyte antibodies were detected in the serum of the donors.
Lymphocytotoxic antibodies were found in 72% of the cases.
The lack of antibodies in recipients pretransfusion serum led
the authors to believe that passive transfer of antibodies is
associated with the occurrence of TRALI and likely a factor in
its pathogenesis.12 A similar incidence of 0.02% per unit transfused and 0.16% per patient transfused was demonstrated in
this study. Of the 36 patients, 2 died, resulting in a 6% mortality risk.12 In the reports by Popovsky et al11 and Popovsky and
American Society for Clinical Pathology

Coagulation and Transfusion Medicine / REVIEW ARTICLE

Moore,12 whole blood, PRBCs, and FFP led to the development

of TRALI. All blood components have now been implicated in
TRALI, including whole bloodderived platelets,13 apheresis
platelets,14 cryoprecipitate,15 granulocytes,16 stem cells,17
immune globulin concentrates,18 and even autologous RBCs.19
Seeger and colleagues20 reproduced this antibody-mediated lung injury in an animal model. Isolated rabbit lungs were
perfused with plasma containing anti-5b antibody, 5b-positive
human granulocytes, and rabbit plasma as a complement
source. Significant increases in pulmonary arterial pressure
and vascular permeability occurred 3 to 6 hours following the
infusion. When the experiments were repeated without 1 of
the 3 components, no pulmonary edema occurred.20 By using
rat lungs, Sachs et al21 demonstrated pulmonary edema without the addition of complement. Human neutrophil antigen
(HNA)-2a monoclonal antibodies (mAb) and neutrophils with
a high density of the equivalent antigen were added to the perfusate of isolated rat lungs, and an increase in vascular permeability was observed. In a separate experiment, the addition of
formyl-Met-Leu-Phe, a component of bacterial cell walls,
accelerated the increase in vascular permeability.21
In an effort to further understand the pathogenesis of
TRALI, Looney and colleagues22 developed an in vivo mouse
model. Passive transfusion of a major histocompatibility complex (MHC)-I mAb (H2Kd) to cognate mice produced increases in excess lung water, vascular and epithelial permeability to
protein, and decreased alveolar fluid clearance. Severe pulmonary sequestration of neutrophils and peripheral neutropenia were also observed.22 In a separate experiment, mice were
first treated with Gr-1 mAb to selectively induce neutropenia.
When the animals were challenged with MHC-I mAb, ALI
failed to develop. Next, FcR/ mice were challenged with
MHC-I mAb and did not experience ALI. Recognizing that
FcRI and FcRIII are also present on lymphocytes, monocytes, mast cells, and dendritic cells, the authors injected wildtype neutrophils into the FcR/ mice and challenged them
with MHC-I mAb.22 Untreated mice showed no evidence of
ALI, whereas mice transfused with wild-type neutrophils
challenged with MHC-I mAb did. Histologic examination
revealed the presence of the mAb in the microvasculature of
the lungs, kidneys, and liver, suggesting that neutrophils were
trapped in various organs besides the pulmonary vasculature.
The authors concluded that neutrophils and their Fc receptors were essential to the pathogenesis of ALI in this mouse
model of TRALI.22
Popovsky and Moore12 were unable to identify granulocyte antibodies in 11% of their patients with TRALI. Silliman
and colleagues23 studied 10 patients with TRALI whose unit
donors did not have significant titers of HLA antibodies and
only half expressed weakly positive granulocyte-specific antibodies. However, they observed that all patients with TRALI
had significant underlying medical conditions (eg, infection,

recent surgery, massive transfusion, or cytokine administration) and hypothesized that 2 events are necessary to induce
TRALI.23 Thus, predisposing clinical factors would prime
neutrophils and are necessary but not sufficient to cause
TRALI. The second insult would be the transfusion or, more
specifically, biologically active lipids that accumulate during
storage of cellular components.23 Subsequently, Silliman et
al14,24 identified lysophosphatidylcholine as a component of
these biologically active lipids and showed that it can prime
neutrophils. In 2003, the same investigators prospectively
determined that the lipid-priming activity was higher in units
of whole blood platelets implicated in TRALI than in control
units.25 In a case report of recurrent TRALI, Win and colleagues26 also noted that biologically active lipids may have
had a role in the second TRALI event.
Silliman and colleagues27 reproduced TRALI with biologically active lipids in animal models. In keeping with the 2insult model, the authors first pretreated rats with endotoxin.
The rat lungs were then isolated, ventilated, and perfused with
saline, 5% fresh human plasma, plasma from stored blood
from the day of isolation, plasma from stored blood from the
day of outdate, lipid extracts from day of outdated plasma, or
purified lysophosphatidylcholine.27 Rat lungs pretreated with
endotoxin and perfused with day-of-outdate plasma developed
ALI. Significant pulmonary edema was also seen in the endotoxin-pretreated lungs perfused with lipid extracts from dayof-outdate plasma and lysophosphatidylcholine.27 A similar
study in rats with plasma from whole bloodderived and
apheresis platelets was done by Silliman et al.28 Again, plasma from both sources collected at day 5 induced ALI in endotoxin-pretreated rats, but plasma from day 0 did not.28
In a retrospective case control study, Khan et al29 identified the CD40 ligand (CD40L) as a cofactor in the development of TRALI. Soluble CD40L levels were 76% higher in
implicated platelet concentrates than in nonimplicated units.29
The highest levels were found in apheresis products. The
authors also found that CD40L accumulates during the storage of PRBCs and whole blood.29 Furthermore, CD40L
increased in the posttransfusion plasma of 8 of 12 patients
with TRALI compared with pretransfusion levels. These findings suggested an association among soluble CD40L in
platelet concentrates, activation of the innate immune system,
and development of TRALI.29

Differentiating TRALI
Because transfusions typically are given to sick patients,
it is important to consider all entities that might cause acute
respiratory distress. Transfusion-associated circulatory overload (TACO) is chief among these entities. Reports of the incidence of TACO vary from less than 1% to 11% in critically ill
Am J Clin Pathol 2008;129:287-297

American Society for Clinical Pathology

289

DOI: 10.1309/D3F7BXH466AE3G0P

289
289

Cherry et al / TRANSFUSION-RELATED ACUTE LUNG INJURY

medical patients.30,31 Mortality from TACO has been estimated from 3.6%30 to 20%.31 Clinically, patients with TACO have
tachypnea, dyspnea, cyanosis, tachycardia, and hypertension.32 They also have signs of circulatory overload, such as
jugular venous distension and an elevated pulmonary artery
occlusion pressure.32 Such signs may be present before the
initiation of transfusion, and review of the patients intake and
output will likely add evidence for the diagnosis.33 TACO usually responds to diuresis and ventilatory support.34
Respiratory distress is a major symptom in anaphylactic
transfusion reactions as well. These typically include tachypnea, cyanosis, and wheezing.35 Hypotension is also frequently
a component.35 Theses symptoms typically arise from laryngeal and bronchial edema instead of interstitial pulmonary
involvement.35 Skin manifestations include urticaria, erythema, and edema of the face and trunk.35 Bacterial contamination
of transfused blood products should also be considered in a
patient with respiratory distress. Sepsis usually manifests as
hypotension, fever, and even circulatory collapse, often accompanied by respiratory distress.32 Culture of the blood bag is crucial in the evaluation. Finally, an acute hemolytic transfusion
reaction must also be considered and ruled out because the
signs and symptoms may include respiratory distress.

Suspecting and Diagnosing TRALI

A high index of suspicion is necessary to accurately identify TRALI. In a look-back study of patients who received
blood components from a donor implicated in TRALI, Kopko
et al36 determined that TRALI was underrecognized and
underreported. Anyone experiencing dyspnea, hypoxemia,
pulmonary edema, hypotension, and fever temporally related
to transfusion should be suspected of having TRALI. To aid in
the diagnosis and management of TRALI, the hospital transfusion service should be notified immediately whenever
TRALI is suspected.
In their original article, Popovsky et al11 studied patients
in whom unexplained pulmonary edema developed following
transfusions. Later, they added respiratory distress, hypoxemia, and hypotension in the absence of volume overload or
underlying pulmonary disease, occurring usually within 1 to 2
hours of transfusion of plasma-containing products.12
Silliman et al25 defined TRALI as follows: (1) respiratory
insufficiency (tachypnea, shortness of breath, increased work
of breathing, and cyanosis) accompanied by significant oxygen desaturation as the predominant presenting symptom; (2)
respiratory compromise that required immediate intervention;
(3) onset of symptoms temporally related to transfusion (within 4 hours; most occurring within 10 to 30 minutes); and (4)
no other clinical cause (eg, volume overload, allergic manifestation, or sepsis) evident for the pulmonary compromise.
290
290

Am J Clin Pathol 2008;129:287-297

DOI: 10.1309/D3F7BXH466AE3G0P

Several definitions of TRALI have been published in the

United States and abroad Table 1.37-40 The American-European
Consensus Conference (AECC) definition differs most notably
in its inclusion of possible TRALI cases.39 It is noteworthy that
the National Heart, Lung, and Blood Institute and the AECC definitions are based solely on clinical symptoms. Therefore, careful observation of the patient is crucial in diagnosing TRALI.
The use of brain natriuretic peptide (BNP) levels has
recently been postulated as a laboratory adjunct in the differentiation of TRALI from TACO. BNP is a polypeptide
released by the ventricles and atria in response to volume or
pressure overload.41 A study by Maisel et al42 showed that
BNP levels were more accurate than any historical or physical findings or laboratory values in identifying congestive
heart failure as the cause of dyspnea. In a case reported by
Burgher et al,41 a patient suspected of having TRALI was
diagnosed with TACO based on normal pretransfusion BNP
and elevated posttransfusion BNP levels. Further work has
confirmed the usefulness of BNP in predicting TACO. Zhou et
al43 demonstrated 81% sensitivity, 89% specificity, 89% positive predictive value, 81% negative predictive value, and 87%
accuracy of BNP in diagnosing TACO.
Atrial natriuretic peptide is another major peptide that has
been postulated as a superior marker for TACO because of its
rapid release into the blood following atrial stretch and its
shorter half-life of 2 to 4 minutes.44 Currently, however, atrial
natriuretic peptide testing is not widely available.
Despite availability of testing, BNP levels in the setting of
suspected TRALI have not been extensively studied. One case
report demonstrated a normal posttransfusion BNP level in a
patient with fatal TRALI.45 When interpreting BNP levels, it is
important to compare posttransfusion with pretransfusion levels. Because BNP can be accurately measured on refrigerated
samples, comparison is easily done with previously collected
blood samples. Although a normal BNP level may exclude

Table 1
Current Criteria for the Diagnosis of TRALI
American-European Consensus Conference Definition of ALI39
Acute onset
Bilateral pulmonary infiltrates evident on chest radiograph
Hypoxemia, defined as PaO2/FIO2 300
No evidence of left atrial hypertension (ie, no congestive heart
failure; or PAOP 18, if available)
National Heart, Lung, and Blood Institute Definition of TRALI40
No ALI before transfusion
Signs or symptoms of TRALI during or within 6 hours of transfusion
In patients with an alternative ALI risk factor, TRALI is still possible.
Massive transfusion should not exclude the possibility of TRALI.
European Haemovigilance Network Definition of TRALI37
Respiratory distress during or within 6 hours of transfusion
No signs of circulatory overload
Radiographic evidence of bilateral pulmonary infiltrates
ALI, acute lung injury; FIO2, fraction of inspired oxygen; PAOP, pulmonary artery
occlusion pressure; TRALI, transfusion-related ALI.

American Society for Clinical Pathology

Coagulation and Transfusion Medicine / REVIEW ARTICLE

TACO and posttransfusion increases in the BNP level favor

TACO, the role of BNP in TRALI remains to be determined.
Other laboratory parameters may also be useful.
Transient, acute neutropenia has been reported by many
authors,1,3,46-48 and the CBC count is a readily available, inexpensive tool to increase the likelihood of identifying TRALI.
Analysis of the pulmonary edema protein content can aid in
the exclusion of circulatory overload.31,49,50 A ratio of 0.75 or
more between the protein in the edema fluid and the plasma is
consistent with increased permeability, whereas a ratio of 0.65
or less is characteristic of hydrostatic edema.51 If the suspected case of TRALI proves fatal, gross examination of the lungs
will likely reveal diffuse edema, and microscopic examination
will identify an increased number of leukocytes in the
microvasculature and alveolar spaces Image 1.52 Finally, the
age of suspected blood products should also be determined
because cytokines and lipids accumulate during storage.27-28

prompt reporting of all transfusion-related deaths and encourages voluntary reporting of complications of transfusions.
Complete assessment requires specimens from the recipient, all blood components transfused within the preceding 6
hours, and the corresponding donor(s). The AECC suggests
testing the donor(s) for HLA class I and II and HNA antibodies and determining their specificity, if positive.39 Donors
implicated in TRALI are those with antibodies specific for a
recipient antigen or causing a positive WBC crossmatch.39
Although neutrophil priming activity has also been postulated
in TRALI,14,21 the AECC recognized the lack of widespread
availability for testing and recommended it only for research
settings.39 Because numerous studies have found that the
implicated unit was from a multiparous woman,7,12,58 serologic investigation of cases in which multiple transfusions were
given in the preceding 6 hours should begin with units from
such donors.

Treating TRALI

Predicting TRALI

The mainstay of treatment for TRALI remains supportive

care. If the suspected blood product is still being transfused, it
should be discontinued immediately.53 In the original series
studied by Popovsky and Moore,12 all 36 patients required
supplemental oxygen and 72% required mechanical ventilation. In 81% of patients, complete recovery occurred within
96 hours. Only 17% had persistent hypoxia and pulmonary
infiltrates up to 7 days.12 Although an optimal ventilation
strategy for TRALI has not been specifically studied, smaller
tidal volumes and optimization of positive end-expiratory
pressure seem to improve outcome in ALI.54
Proper diagnosis of TRALI and exclusion of TACO are
also important. Although examples of successful treatment of
TRALI with diuretics exist,5,11 Levy et al55 caution against
their routine use. Likewise, although various reports describe
corticosteroids used in TRALI,7,10,11 there exists no randomized, controlled study to support their need.56 Fluid replacement is crucial to treat the hypotension and respiratory signs
as evidenced by the example of immediate improvement in
oxygenation and hemodynamics observed after administration
of large volumes of 5% albumin.57

Although any blood component can cause TRALI, plasma-rich units are more likely to be the culprits.31 In addition,
Silliman et al25 found that patients with hematologic malignancies and those requiring coronary bypass surgery are at
particular risk.
Donor factors may also predict the possibility of developing TRALI. Since 1971, multiple previous pregnancies in the
donor have been considered a risk factor for inducing
TRALI.7 Densmore et al59 showed that the prevalence of HLA
sensitization is directly related to parity: 15% of women
reporting 1 to 2 pregnancies and 26% of women with 3 or
more. However, transfusion of HLA antibodies into patients
with corresponding antigens typically does not result in
TRALI.59,60 Palfi and colleagues58 conducted the first randomized, controlled trial of TRALI and multiparous donors.
They identified 100 critically ill patients who needed transfusion of 2 U of FFP and gave each one a unit from a nulliparous
and a unit from a multiparous donor. Patients were randomized to receive the nulliparous or the multiparous unit first, at
least 4 hours apart. Hemodynamics, cytokine levels, and
PaO2/fraction of inspired oxygen (FIO2) ratio were measured.
Although only 1 case of TRALI was diagnosed, a small but
statistically significant decrease in the PaO2/FIO2 ratio
occurred after transfusion of the multiparous units.58

Investigating TRALI
Although the diagnosis of TRALI relies on clinical symptoms and the exclusion of other causes, a thorough laboratory
investigation will help support the diagnosis. Any investigation into suspected TRALI should begin with immediate notification of the hospital transfusion service, which will ensure
proper evaluation and follow-up Table 2. The FDA requires

Preventing TRALI
Transfusion of any blood component poses a risk for
developing TRALI. Consistent, evidence-based transfusion
guidelines will likely decrease the number of transfusions
Am J Clin Pathol 2008;129:287-297

American Society for Clinical Pathology

291

DOI: 10.1309/D3F7BXH466AE3G0P

291
291

Cherry et al / TRANSFUSION-RELATED ACUTE LUNG INJURY

Image 1 A complete autopsy was performed on a patient who died with the clinical diagnosis of transfusion-related acute
lung injury. The most significant finding was pulmonary edema (combined lung weight of 2,780 g) (A) with neutrophilic
aggregates in the pulmonary vasculature (B, H&E, 180).

and the risk of associated morbidity and mortality.61-63 In a

landmark study, Hebert and colleagues64 compared a
restrictive RBC transfusion strategy with a liberal one, with
goal hemoglobin levels of 7 to 9 g/dL (70-90 g/L) and 10 to
12 g/dL (100-120 g/L), respectively, in patients admitted to
critical care units. The overall 30-day mortality between the
2 groups was not significantly different. Subgroup analysis
revealed that patients in the restrictive group who were
younger than 55 years and patients who were less severely
ill had lower 30-day mortality rates. These data suggest that
a restrictive RBC transfusion strategy is at least as effective
as a liberal one and that for certain subsets of patients it may
provide benefit.64
The use of plasma in the United States rose 70% from
1991 to 2001.67 Indeed, transfusion of plasma in the United
States far exceeds that in Europe.65 Most commonly, plasma
is used to correct coagulopathy in bleeding patients or patients
being prepared for invasive procedures. However, an extensive
review by Segal and Dzik66 concluded that there are insufficient data to support the assumption that abnormal laboratory
results are predictive of bleeding in patients undergoing invasive procedures. To assess the impact of FFP on the correction
of mild coagulation abnormalities, Abdel-Wahab and colleagues67 prospectively examined coagulation screening test
results before and after transfusion. The international normalized ratio normalized in only 0.8% of the patients and
decreased to at least halfway toward normalization in only
15%. Adoption of plasma transfusion strategies based on these
and other studies would undoubtedly reduce the amount of
FFP transfused in the United States.
292
292

Am J Clin Pathol 2008;129:287-297

DOI: 10.1309/D3F7BXH466AE3G0P

Prophylactic platelet transfusion in thrombocytopenic

patients has been a mainstay of therapy for many decades.
Historically, a platelet count of 20 103/L (20 109/L) was
used as the trigger prompting platelet transfusion in patients in
clinically stable condition.68 Research in the last decade has
indicated that a transfusion trigger of 10 103/L (10 109/L)
does not result in an increased risk of bleeding or RBC transfusion but does result in a reduction in the number of platelet
transfusions.69,70 More recently, the need for prophylactic
platelet transfusions has been questioned. Wandt and colleagues71 examined a therapeutic transfusion strategy in
patients following autologous peripheral stem cell transplantation. Platelet transfusions were given only for relevant bleeding. Only 19% of patients experienced minor or moderate
bleeding, and none had severe or life-threatening bleeding
episodes. Compared with historic control subjects, the number
of platelet transfusions was reduced by half.71 Clearly, adherence to lower platelet transfusion thresholds and reserving
them for treatment of clinical bleeding can reduce the number
of patient exposures.
Patients requiring surgery are one of the largest groups of
consumers of blood products. Multiple strategies have been
devised to reduce the number of transfusions required during
the perioperative period. In the weeks leading up to surgery,
Goodnough and Shander72 suggest a thorough review of a
patients preoperative CBC count to diagnose any anemia
that might be present and initiate appropriate therapy early
enough to correct the anemia. They also advocate careful
management of anticoagulation in the perioperative period,
including discontinuation of antiplatelet agents that patients
American Society for Clinical Pathology

Coagulation and Transfusion Medicine / REVIEW ARTICLE

Image 1 Also, marked congestion of most tissues, sludging of RBCs, and neutrophilic collections were found in the liver
sinusoids (C, H&E, 180), myocardium, glomeruli, and central nervous system.Several meningeal blood vessels of frontal
lobes and hippocampus contained RBCs and leukocytes (D, H&E, 50; E, H&E, 50), and pontine blood vessels had dense
neutrophilic aggregates (F, H&E, 180). Several small lacunar infarcts of the pons were noted (not shown).

might be taking.72 Administration of erythropoietin preoperatively has also been shown to decrease allogeneic transfusion requirements in certain populations.73 Intraoperatively,
reductions in transfusion rates of plasma and platelets are also
attainable. The use of antifibrinolytic drugs can decrease
transfusion requirements in a variety of surgical procedures,74
as can the avoidance of intraoperative hypothermia.75 The
retransfusion of postoperatively shed blood has also received
attention, and it is effective for reducing the number of allogeneic transfusions in total hip and knee arthroplasty.76
Despite maximal efforts to reduce the number of transfusions, patients will continue to require them for specific conditions. When transfusion of plasma is indicated, exclusion of

Table 2
Investigation of Suspected TRALI*
Discontinue transfusion
Notify hospital transfusion service
Repeat ABO typing and crossmatch
Obtain CBC count, levels of ABG, blood cultures, and chest radiograph
Return all component bags recently transfused
Test donor and recipient for HLA class I and II antibodies and human
neutrophil antigenspecific antibodies (antihuman globulin
complementdependent cytotoxicity or flow cytometry)
Determine specificity, if antibodies detected
Perform donor-recipient WBC crossmatch
Determine parity of donor(s)
Determine age of unit(s) transfused
Test for neutrophil-priming activity when available in research settings
ABG, arterial blood gases; TRALI, transfusion-related acute lung injury.
* Based, in part, on American-European Consensus Conference recommendations.39

Am J Clin Pathol 2008;129:287-297

American Society for Clinical Pathology

293

DOI: 10.1309/D3F7BXH466AE3G0P

293
293

Cherry et al / TRANSFUSION-RELATED ACUTE LUNG INJURY

units from multiparous women has been shown to decrease

the incidence of TRALI.77 If there is a choice, different preparations of plasma also seem to differ in their risk of causing
TRALI. Sinnott et al78 compared the presence of HLA antibodies in single-donor FFP units with pooled, solvent detergenttreated plasma. Antibodies were detected in 9% of FFP
units but in none of the solvent detergenttreated units.78
Although the latter units may test negative for HLA antibodies, performing FDA-mandated look-back investigations on
pooled components is time-consuming, cumbersome, and
costly. Therefore, this product is no longer available in the
United States. The 2004 consensus conference suggested that
TRALI-implicated donors with demonstrable antibodies be
deferred from future donations.40
In 2006 the AABB adopted the following recommendations to decrease the incidence of TRALI79: (1) Blood collecting facilities should implement interventions to minimize the
preparation of high plasma volume components from donors
known to have or be at risk for leukocyte alloimmunization. (2)
Blood transfusion facilities should work toward implementing
appropriate evidence-based hemotherapy practices to minimize
unnecessary transfusion. (3) Blood collection and blood transfusion facilities should monitor the incidence of reported
TRALI and TRALI-related mortality.
Strategies to minimize the risk of transfusing biologically active lipids and cytokines have also been presented.
Because these substances accumulate in cellular components
during storage,14,24 the use of PRBCs less than 14 days old24
and platelet concentrates less than 2 days old14 would mitigate
the effects of these compounds. However, given current
requirements regarding serologic and bacterial detection testing, providing platelets less than 2 days old is impractical, if

not impossible, in a clinical setting. Washing of cellular components reduces neutrophil priming activity and biologically
active substances80 but has negative effects on shelf life.81
Prestorage leukocyte reduction may reduce the risk of TRALI
by limiting complement-mediated processes, which have been
shown to be a contributor to TRALI in a rat model.20

What Is to Come?
We hope the near future will produce more answers
about the pathogenesis and prevention of TRALI. Large,
prospective studies of the incidence of TRALI might provide
more data about which patients are most at risk and which
donors are most likely associated with TRALI. From there,
various prevention strategies could be prospectively implemented to test their efficacy. In addition, donor management
could be honed to exclude donors whose blood is most likely to cause TRALI, while allowing others with less risk to
continue contributing to the already small donor pool.
Screening for HLA and HNA antibodies may eventually
become feasible, but cost and interpretation of results must be
taken into consideration before widespread implementation.
Finally, as Wallis82 suggested, the question of whether the
lungs are the only affected organ is still open. During the
autopsy of the patient who died of TRALI described in the
introduction, we identified microvascular neutrophil infiltration in various organs besides the lungs (Image 1) and Image
2. Although the pulmonary vasculature is the first capillary
bed encountered by transfusions, the effects of TRALI may
indeed be systemic, and further study to elucidate these
effects is warranted.

Image 2 Myeloperoxidase (MPO) stain confirmed that most of the cells in the microvasculature of the lungs and pons were
neutrophilic/MPO+ collections (A, 180; B, 100).

294
294

Am J Clin Pathol 2008;129:287-297

DOI: 10.1309/D3F7BXH466AE3G0P

American Society for Clinical Pathology

Coagulation and Transfusion Medicine / REVIEW ARTICLE

Conclusion
TRALI is a rare but serious complication of transfusion
therapy. Since the first description, a great deal of insight has
been gained into its pathogenesis, yet many questions remain
unanswered. There is clear evidence that passively transfused
antibodies have a role in many cases of TRALI. However, the
case for a role for biologically active lipids as causative agents
continues to mount. Whether these 2 causes are separate entities or represent different points along the same continuum is
unknown. Rapid recognition when cases of TRALI occur
remains vital to the further understanding and proper treatment of this complication. Notification of the transfusion service is crucial to ensure that a proper investigation is carried out
and, if necessary, the case reported to regulatory authorities.
Perhaps with a broader and more detailed understanding into
the pathogenesis of TRALI, at-risk donors and recipients can
be identified and measures taken to lessen the recipients risk.
From the Departments of 1Anesthesiology and 2Pathology,
Division of Laboratory Medicine, University of Alabama at
Birmingham.
Address reprint requests to Dr Cherry: Dept of Anesthesiology,
JT 964; 619 19th St S, Birmingham, AL 35249-6810.

References
1. Lanman JT, Bierman HR, Byron RL Jr. Transfusion of
leukemic leukocytes in man: hematologic and physiologic
changes. Blood. 1950;5:1099-1113.
2. Barnard RD. Indiscriminate transfusion: a critique of case
reports illustrating hypersensitivity reactions. N Y State J Med.
1951;51:2399-2402.
3. Bittingham TE. Immunologic studies on leukocytes. Vox Sang.
1957;2:242-248.
4. Phillips E, Fleischner FG. Pulmonary edema in the course
of a blood transfusion without overloading the circulation.
Dis Chest. 1966;50:619-623.
5. Ward HN, Lipscomb TS, Cawley LP. Pulmonary
hypersensitivity reaction after blood transfusion. Arch Intern
Med. 1968;122:362-366.
6. Ward HN. Pulmonary infiltrates associated with
leukoagglutinin transfusion reactions. Ann Intern Med.
1970;73:689-694.
7. Thompson JS, Severson BS, Parmely BS, et al. Pulmonary
hypersensitivity reactions induced by transfusion of non
HL-A leukoagglutinins. N Engl J Med. 1971;284:1120-1125.
8. Wolf CFW, Canale VC. Fatal pulmonary hypersensitivity
reaction to HL-A incompatible blood transfusion: report of a
case and review of the literature. Transfusion. 1976;16:135-140.
9. Kernoff PBA, Durrant IJ, Rizza CR, et al. Severe allergic
pulmonary oedema after plasma transfusion. Br J Haematol.
1972;23:777-781.
10. Culliford AR, Thomas S, Spencer FC. Fulminating
noncardiogenic pulmonary edema: a newly recognized hazard
during cardiac operations. J Thorac Cardiovasc Surg.
1980;80:868-875.

11. Popovsky MA, Abel MD, Moore SB. Transfusion-related acute

lung injury associated with passive transfer of antileukocyte
antibodies. Am Rev Respir Dis. 1983;128:185-189.
12. Popovsky MA, Moore SB. Diagnostic and pathogenetic
considerations in transfusion-related acute lung injury.
Transfusion. 1985;25:573-577.
13. Ramanathan RK, Triulzi DJ, Logan TF. Transfusion-related
acute lung injury following random donor platelet transfusion:
a report of two cases. Vox Sang. 1997;73:43-45.
14. Silliman CC, Dickey WO, Patterson AJ, et al. Analysis of the
priming activity of lipids generated during routine storage of
platelet concentrates. Transfusion. 1996;36:133-139.
15. Reese EP, McCullough JJ, Craddock PR. An adverse
pulmonary reaction to cryoprecipitate in a hemophiliac.
Transfusion. 1975;15:583-588.
16. Leach M, Vora AJ, Jones DA, et al. Transfusion-related acute lung
injury (TRALI) following autologous stem cell transplantation
for relapsed acute myeloid leukaemia: a case report and review
of the literature. Transfus Med. 1998;8:333-337.
17. Urahama N, Tanosaki R, Masahiro K, et al. TRALI after the
infusion of marrow stem cells in a patient with acute
lymphoblastic leukemia. Transfusion. 2003;43:1553-1557.
18. Rizk A, Gorson K, Kenney L, Weinstein R. Transfusionrelated acute lung injury after the infusion of IVIG.
Transfusion. 2001;41:264-268.
19. Covin RB, Ambruso DR, England KM, et al. Hypotension
and acute pulmonary insufficiency following transfusion of
autologous red blood cells during surgery: a case report and
review of the literature. Transfus Med. 2004;14:375-383.
20. Seeger W, Schneider U, Kreusler B, et al. Reproduction of
transfusion-related acute lung injury in an ex vivo lung model.
Blood. 1990;76:1438-1444.
21. Sachs UJH, Hattar K, Weissmann N, et al. Antibody-induced
neutrophil activation as a trigger for transfusion-related acute
lung injury in an ex vivo rat lung model. Blood. 2006;107:
1217-1219.
22. Looney MR, Su X, Van Ziffle JA, et al. Neutrophils and their
Fc receptors are essential in a mouse model of transfusionrelated acute lung injury. J Clin Invest. 2006;116:1615-1623.
23. Silliman CC, Paterson AJ, Dickey WO, et al. The association
of biologically active lipids with the development of
transfusion-related acute lung injury: a retrospective study.
Transfusion. 1997;37:719-726.
24. Silliman CC, Clay KL, Thurman GW, et al. Partial
characterization of lipids that develop during the routine
storage of blood and prime the neutrophil NADPH oxidase.
J Lab Clin Med. 1994;124:684-694.
25. Silliman CC, Boshkov LK, Mehdizadehkashi Z, et al.
Transfusion-related acute lung injury: epidemiology and a
prospective analysis of etiologic factors. Blood. 2003;101:454-462.
26. Win N, Montgomery J, Sage D, et al. Recurrent transfusionrelated acute lung injury. Transfusion. 2001;41:1421-1425.
27. Silliman CC, Voekel NF, Allard JD, et al. Plasma and lipids
from stored packed red blood cells cause acute lung injury in
an animal model. J Clin Invest. 1998;101:1458-1467.
28. Silliman CC, Bjornsen AJ, Wyman, TH, et al. Plasma and
lipids from stored platelets cause acute lung injury in an
animal model. Transfusion. 2003;42:633-640.
29. Khan SY, Kelher MR, Heal JM, et al. Soluble CD40 ligand
accumulates in stored blood components, primes neutrophils
through CD40, and is a potential cofactor in the development
of transfusion-related acute lung injury. Blood. 2006;108:24552462.

Am J Clin Pathol 2008;129:287-297

American Society for Clinical Pathology

295

DOI: 10.1309/D3F7BXH466AE3G0P

295
295

Cherry et al / TRANSFUSION-RELATED ACUTE LUNG INJURY

30. Popovsky MA. Transfusion and the lung: circulatory overload

and acute lung injury. Vox Sang. 2004;87(suppl 2):62-65.
31. Rana R, Fernandez-Perez ER, Khan SA, et al. Transfusionrelated acute lung injury in critically ill patients: a
retrospective study. Transfusion. 2006;46:1478-1483.
32. Popovsky MA. Transfusion-related acute lung injury. In:
Popovsky MA, ed. Transfusion Reactions. Bethesda, MD:
American Association of Blood Banks; 1996:167-178.
33. Gajic O, Gropper MA, Hubmayr RD. Pulmonary edema after
transfusion: how to differentiate transfusion-associated
circulatory overload from transfusion-related acute lung injury.
Crit Care Med. 2006;34(5 suppl):S109-S113.
34. Popovsky MA. Breathlessness and blood: a combustible
combination. Vox Sang. 2002;83(1 suppl):147-150.
35. Vamvakas EC. Allergic and anaphylactic reactions. In:
Popovsky MA, ed. Transfusion Reactions. Bethesda, MD:
American Association of Blood Banks; 1996:81-124.
36. Kopko PM, Marshall CS, MacKenzie MR, et al. Transfusionrelated acute lung injury: report of a clinical look-back
investigation. JAMA. 2002;287:1968-1971.
37. European Haemovigilance Network (EHN): Definition
of adverse transfusion events. Available at http://www.ehnorg.net. Accessed February 12, 2007.
38. Toy P, Popovsky MA, Abraham E, et al. Transfusion-related
acute lung injury: definition and review. Crit Care Med.
2005;33:721-726.
39. Bernard CR, Artigas A, Brigham KL, et al. The AmericanEuropean Consensus Conference on ARDS: definitions,
mechanisms, relevant outcomes, and clinical trial
coordination. Am J Respir Crit Care Med. 1994;149:818-824.
40. Kleinman S, Caulfield T, Chan P, et al. Toward an
understanding of transfusion-related acute lung injury:
statement of a consensus panel. Transfusion. 2004;44:17741789.
41. Burgher AH, Aslan D, Laudi N, et al. Use of brain natriuretic
peptide to evaluate transfusion-related acute lung injury
[letter]. Transfusion. 2004;44:1533-1534.
42. Maisel AS, Krishnaswamy P, Nowak RM, et al. Rapid
measurement of B-type natriuretic peptide in the emergency
diagnosis of heart failure. N Engl J Med. 2002;347:161-167.
43. Zhou L, Giacherio D, Cooling L, et al. Use of B-natriuretic
peptide as a diagnostic marker in the differential diagnosis
of transfusion-associated circulatory overload. Transfusion.
2005;45:1056-1063.
44. Fiebig EW, Wu AHB, Kromback J, et al. Transfusion-related
acute lung injury and transfusion-associated circulatory
overload: mutually exclusive or coexisting entities [letter]?
Transfusion. 2007;47:171-172.
45. Edelson BT, Despotis GJ, Lublin DM. B-type natriuretic
peptide measured during transfusion-related acute lung injury
[letter]. Transfusion. 2006;46:1453-1454.
46. Yomtovian R, Kline W, Press C, et al. Severe pulmonary
hypersensitivity associated with passive transfusion of a
neutrophil-specific antibody. Lancet. 1984;1:244-246.
47. Nakagawa M, Toy P. Acute and transient decrease in
neutrophil count in transfusion-related acute lung injury:
cases at one hospital. Transfusion. 2004;44:1689-1694.
48. Marques MB, Tuncer HH, Divers SG, et al. Acute transient
leukopenia as a sign of TRALI [letter]. Am J Hematol.
2005;80:90-91.
49. Yost C, Matthay M, Gropper M. Etiology of acute pulmonary
edema during liver transplantation: a series of cases with
analysis of the edema fluid. Chest. 2001;119:219-223.

296
296

Am J Clin Pathol 2008;129:287-297

DOI: 10.1309/D3F7BXH466AE3G0P

50. Looney M, Gropper M, Matthay M. Transfusion-related acute

lung injury: a review. Chest. 2004;126:249-258.
51. Matthay MA. Pathophysiology of pulmonary edema. Clin
Chest Med. 1985; 6:301-314.
52. Dry SM, Bechard KM, Milford EL, et al. The pathology of
transfusion-related acute lung injury. Am J Clin Pathol.
1999;112:216-221.
53. Popovsky MA. Transfusion-related acute lung injury. Curr
Opin Hematol. 2000;7:402-407.
54. ARDS Network. Ventilation with lower tidal volumes as
compared with traditional tidal volumes for acute lung injury
and the acute respiratory distress syndrome. N Engl J Med.
2000;342:1301-1308.
55. Levy F, Shabot M, Hart M, et al. Transfusion-associated
noncardiogenic pulmonary edema: report of a case and a
warning regarding treatment. Transfusion. 1986;26:278-281.
56. Goldman M, Webert KR, Arnold DM, et al. Proceedings
of a consensus conference: towards an understanding of
transfusion-related acute lung injury. Transfus Med Rev.
2005;19:2-31.
57. Djalali AG, Moore KA, Kelly E. Report of a patient with
severe transfusion-related acute lung injury after multiple
transfusions, resuscitated with albumin. Resuscitation.
2005;66:225-230.
58. Palfi M, Berg S, Ernerudh J, et al. A randomized controlled
trial of transfusion-related acute lung injury: is plasma from
multiparous blood donors dangerous? Transfusion.
2001;41:317-322.
59. Densmore TL, Goodnough LT, Ali S, et al. Prevalence of HLA
sensitization in female apheresis donors. Transfusion.
1999;39:103-106.
60. Toy P, Hollis-Perry KM, Jun J, et al. Recipients of blood from
a donor with multiple HLA antibodies: a lookback study of
transfusion-related acute lung injury. Transfusion. 2004;44:
1683-1688.
61. Spiess BD, Royston D, Levy JH, et al. Platelet transfusions
during coronary artery bypass graft surgery are associated with
serious adverse outcomes. Transfusion. 2004;44:1143-1148.
62. Goodnough LT, Soegiarso RW, Birkmeyer JD, et al. Economic
impact of inappropriate blood transfusions in coronary artery
bypass graft surgery. Am J Med. 1993;95:509-514.
63. Surgenor DM, Wallace EL, Churchill WH, et al. Red cell
transfusions in coronary artery bypass surgery (DRGs 106
and 107). Transfusion. 1992;32:458-464.
64. Hebert PC, Wells G, Blajchman M, et al, for the Transfusion
Requirements in Critical Care Investigators; Canadian
Critical Care Trials Group.. A multicenter, randomized,
controlled clinical trial of transfusion requirements in critical
care. N Engl J Med. 1999;340:409-417.
65. Wallis JP, Dzik WH. Is fresh frozen plasma overtransfused in
the United States [editorial]? Transfusion. 2004;44:1674-1675.
66. Segal J, Dzik WH, for the Transfusion Medicine/Hemostasis
Clinical Trials Network. Paucity of studies to support that
abnormal coagulation test results predict bleeding in the
setting of invasive procedures: an evidenced-based review.
Transfusion. 2005;45:1413-1425.
67. Abdel-Wahab OI, Healy B, Dzik WH. Effect of fresh-frozen
plasma transfusion on prothrombin time and bleeding in
patients with mild coagulation abnormalities. Transfusion.
2006;46:1279-1285.
68. Slichter SJ. Relationship between platelet count and bleeding
risk in thrombocytopenic patients. Transfus Med Rev.
2004;18:153-167.

American Society for Clinical Pathology

Coagulation and Transfusion Medicine / REVIEW ARTICLE

69. Wandt H, Frank M, Ehninger G, et al. Safety and cost

effectiveness of a 10 109/L trigger for prophylactic platelet
transfusions compared with the traditional 20 109/L trigger:
a prospective comparative trial in 105 patients with acute
myeloid leukemia. Blood. 1998;91:3601-3606.
70. Lawrence JB, Yomtovian R, Hammons T, et al. Lowering
the prophylactic platelet transfusion threshold: a prospective
analysis. Leuk Lymphoma. 2001;41:67-76.
71. Wandt H, Schaefer-Eckart K, Frank M, et al. A therapeutic
platelet transfusion strategy is safe and feasible in patients after
autologous peripheral blood stem cell transplantation. Bone
Marrow Transplant. 2006;37:387-392.
72. Goodnough LT, Shander A. Blood management. Arch Pathol
Lab Med. 2007;131:695-701.
73. Laupacis A, Fergusson D, for the International Study of Perioperative Transfusion (ISPOT) Investigators. Erythropoietin
to minimize perioperative blood transfusion: a systematic
review of randomized trials. Transfus Med. 1998;8:309-317.
74. Henry DA, Moxey AJ, Carless PA, et al. Antifibrinolytic
use for minimizing perioperative allogeneic blood transfusion.
Cochrane Database Syst Rev. 1999;(4):CD001886. doi:
10.1002/14651858.CD001886.
75. Spahn DR, Casutt M. Eliminating blood transfusions: new
aspects and perspectives. Anesthesiology. 2000;93:242-255.

76. Moonen AFCM, Knoors NT, van Os JJ, et al. Retransfusion of

filtered shed blood in primary total hip and knee arthroplasty:
a prospective randomized clinical trial. Transfusion.
2007;47:379-384.
77. Insunza A, Romon I, Gonzalez-Ponte M, et al.
Implementation of a strategy to prevent TRALI in a regional
blood centre. Transfus Med. 2004;14:157-164.
78. Sinnott P, Bodger S, Gupta A, et al. Presence of HLA
antibodies in single-donor-derived fresh frozen plasma
compared with pooled, solvent detergent-treated plasma
(Octaplas). Eur J Immunogenet. 2004;31:271-274.
79. Strong DM, Lipton KS. Transfusion-related acute lung injury.
Association Bulletin 06-07. Bethesda, MD: American
Association of Blood Banks; November 2006.
80. Silliman CC, Ambruso DR, Boshkov LK. Transfusion-related
acute lung injury. Blood. 2005;105:2266-2273.
81. Technical Manual Committee. Components from whole blood
donation. In: Brecher ME, ed. Technical Manual. 14th ed.
Bethesda, MD: American Association of Blood Banks Press;
2002.
82. Wallis JP. Transfusion-related acute lung injury (TRALI):
under-diagnosed and under-reported. Br J Anaesth.
2003;90:573-576.

Am J Clin Pathol 2008;129:287-297

American Society for Clinical Pathology

297

DOI: 10.1309/D3F7BXH466AE3G0P

297
297