Trali

Transfusion Related Acute Lung Injury

1. Transfusion Related
Acute Lung Injury
TRALI
Raul H. Morales-Borges , MD
Medical Director
American Red Cross Blood Services
Puerto Rico Region
What’s all this I
hear about fatal
Trolleys?
Never mind!
That’s TRALI !

2. First TRALI Reported
•1992
•Reported to the Center for Biologics
Evaluation and research (CBER)
•Popovski MA, Chaplin HC, Moore SB:
Transfusion-related lung injury: a neglected
serious complication of hemotherapy.
Transfusion 1992; 32:589-592.

3. Statistics: Reports of TRALI
•15% of Transfusion-related fatalities on 1976-1985 &
1986-1995.
•As of FY2000 this represented 13% of all transfusion
fatalities.
•According to the FDA, it’s responsible of 24 fatalities
annually from 2003 to 2005.
•Is the 3
rd
cause of transfusion-related death.
•Actual Death Rate: 6 per year; Mortality rate: 6-10%.
•Actual Prevalence: 200 patients per year.

4. More statistics…
•1/5000 transfusions
•1/300 transfusions of PRBC’s

5. TRALI fatalities reported to the FDA (1999-2009).
Benjamin R J Blood 2011;117:4163-4164
©2011 by American Society of Hematology

6. TRALI in ICU
•In Patients with GI Bleeding (Benson AB
et al: Intensive Care Med 2010):
–Is common in patients with End-stage liver
disease (29% vs. 1%,p<0.01).
–FFP 86% than other products.
•TACO in ICU (Li G et al: Transfusion
2010):
–Common in left ventricular dysfunction &
FFP’s.

7. 7
•Hypoxemia due to non-
cardiogenic pulm. edema
–CXR: bilateral diffuse pulm. infiltrates (nl. ♥ & vessels)
–PaO
2
/FiO
2
< 300 mmHg
–Normal JVP; PCWP ≤ 18 mmHg; Nl./unΔ’d BNP
•Hypotension; occasionally begins with hypertension
•Fever (1-2
o
C elevation)
•Onset within 6 hrs. of transfusion, usually within
the first 1-2 hrs.
•Occ. recognized transient neutropenia (80-90%
reductions in some patients)
What Is TRALI?
“Transfusion-Related Acute Lung Injury”

8. 8
Diagnostic Features by Dr. Eder
•Onset within 1 – 6 hours of transfusion
•Acute respiratory distress
•Acute bilateral pulmonary edema (non-
cardiogenic)
•Severe hypoxemia
•Hypotension
•Fever
•Mild to severe clinical spectrum

9. 9
Prevalence & Pathogenesis
•Passive hemovigilance data (subject to
under-recognition & -reporting)
–All components: 1:55,556 (Sweden) to 1:260,000
(Germany)
–Plasma: 1:11,363 (Finland) to 1:66,667 (Denmark)
–49% Plasma, 29% RBCs, 13% Plts., 7% mixed, 2% WB
•Active surveillance
–1:1,323 (Canada) to 1:5,000 components (Mayo)
Healthy Sick
Neutrophil Priming
Strength of transfused
mediators
Threshold

10. 10
Prevalence & Pathogenesis
•28% of blood PMN pool is in the lung
–~50 capillaries per transit, ½ <PMN diameter
•Priming stimuli decrease PMN deformability and ↑ capacity
for tissue entry & toxin release; endothelial activation can
also tether and prime PMNs
–Priming w/ surgery, infection, cancer, cardiac disease, cytokine
administration (e.g., G-CSF), massive transfusion
•Infused leukocyte Abs (HNA > HLA Class II > HLA Class I)
& bioactive substances [lyso-PCs, CD40L] capable of
directly or indirectly (HLA Class II) activating PMNs to
varying degrees to reach threshold for:
–PMN toxin release & extravasation ® alveolar exudate
•Possible to have recipient antibody against donor PMNs,
but LR makes this unlikely

11. 11
ARC High Prob. Fatalities, 2003 – 2005
ARC Hemovigilance Program

12. 12
Strategies to Reduce TRALI
•Appropriate use of blood products
•Deferral of donors implicated in TRALI
•Reduce exposure to individual donor plasma
–Platelet additive solutions
–Whole blood-derived platelets instead of apheresis platelets
•Reduce exposure to alloreactive plasma
–Selective use of male plasma, apheresis platelets & WB
–Selective use of products from donors lacking a history of allo-
exposure (transfusion/pregnancy)
–Use of products from donors tested and found negative for HLA
and/or HNA antibodies
•Reduce exposure to bioactive substances in “aged”
cellular blood products in proven susceptible individuals
–Wash or plasma-reduce “aged” products
–Preferentially use fresher products

13. 13
The Plasma Problem – Nearly Solved
•Plasma was responsible for the majority of
TRALI fatalities reported to ARC (~63%)
–Female neutrophil antibody-positive donors were
implicated in the majority of deaths (71%)
•Only ~27% of recovered plasma is transfused
to patients
–Male plasma units can be diverted from fractionation
back to transfusable plasma to replace female units
no longer useable for transfusion
–Reduces available transfusable FFP; replaced by
PF24
–~99% O & A plasma is male; >95% B is male, but AB
plasma still ~1/3
rd
female

14. 14
The Intervention – and Its Results
Eder AF, et al. Transfusion 2010;50:1732-42; ARC Hemovigilance Program

15. 15
US FDA TRALI Fatalities
http://www.fda.gov/BiologicsBloodVaccines/SafetyAvailability/ReportaProblem/Transfusion
DonationFatalities/ucm204763.htm
ARC & others
implement
male-
predominant
plasma
transfusion
strategy

16. 16
Apheresis Platelets – Looking for a Solution
•Aim is to reduce TRALI risk without compromising the
availability of blood components
•SurveyMonkey responses (late 2009) from 43% of AABB-
member blood centers (producing 1.57M units; in 2006
1.86 units produced nationwide) and 24% of hospitals
Kleinman S, et al. Transfusion 2010;50:1312-21.
Implemented
Future Plans

17. 17
HLA Alloantibody Testing
In-house HLA Lab
In-house Other Lab
Send out
Who’s Being
Tested?
(n=20 blood centers)
On Which
Platform?
(n=19 blood centers)
(Only 2 very small centers
were testing transfused &
transplanted donors)
Kleinman S, et al. Transfusion 2010;50:1312-21.

18. 18
Antibody Testing
•17 of 20 do / will not use results as release
criterion
•19 of 20 planned to retest (14 after alloexposure,
4 at specified time intervals, 1 undecided)
•19 of 19 notify donors of positive test results
•19 of 20 redirect Ab-positive donors to WBD
(10 allow cryo production, but none allow transfusable
recovered plasma)
•None perform HNA Ab testing
Kleinman S, et al. Transfusion 2010;50:1312-21.

19. 19
ARC’s Automated HLA Ab Detection Study
•2432 apheresis donors from 5 regions tested in
2007 for HLA Class I or II Ab on approved GTi
automated ELISA platform (176 samples per shift;
$6 list price per test; manufacturer-fixed cutoff) and
prototype automated Luminex One Lambda
LabScreen Mixed assay (700 samples per shift; $15
per test; cutoff determined by user – 3SD from mean of
nonalloexposed males just as in REDS II LAP study)
•764 samples reactive by ELISA or
“ultrasensitive” Luminex cutoffs (NBGRs of 2.5;
used for transplant patients) tested by Luminex
single-antigen beads to identify specificity of the
reactivity

20. 20
Donor Demographics
Median Age = 51
Parity
66.9% of women were allo-exposed
(65.5% pregnancy, 1.4% tfxn. / txp. only
[overall 6.0% tfxn. / 1.0% txp. / 0.2% both])
4.6% of men were allo-exposed by
transfusion or transplantation
(3.7% tfxn. / 0.8% txp. / 0.1% both)
93.0% Caucasian
2.3% Asian / P.I.
1.6% African-American
3.2% Other
New England, PJ, GC&P, Lewis & Clark, & So. Cal. Regions

21. 21
HLA Class I / II Ab Reactivity
EF = alloexposed [pregnancy, transfusion, transplant] females, EM =
alloexposed males, UEF = unexposed females, UEM = unexposed males
No
statistically-
significant
difference
between
subgroups
37.6%
(1.4 – 5.4%)
21.0%
52.0%
31.1% - 33.5%
~6.5%
Results in
loss of 4.8%
of donor
base

22. 22
Effect of Additional Pregnancies

23. 23
HLA Ab Characteristics
ELISA
Class I
High
PRA
Med
PRA
Low
PRA
High
MFI
81.3%58.8%40.0%
Mod
MFI
61.5%41.2%14.3%
Low
MFI
30.4%12.5%8.7%
Of the 186 Class I and 177 Class II definable-antibody
specificities identified by selective single-antigen testing, ELISA &
“tuned” Luminex assays more often define higher titer (MFI) &
breadth (PRA) samples as reactive; for example, Class I ELISA:

24. 24
Study Conclusions
•Anti-HLA detection in transplant patients requires
exquisite sensitivity to identify low levels of Ab which may
increase after transplant & threaten the graft
•At lower NBG ratios, Luminex identifies more donors
with Ab specificity (progressively lower titer / breadth),
but false-positive reactivity increases as well and more
donors are deferred
–High sensitivity in TRALI mitigation which identifies fixed low-titer
/ narrower specificity Ab may not improve safety, but will erode
the donor base which results in platelet shortages
•Alloexposed females had the highest rates of reactivity
and in this group, ELISA performed similarly to Luminex
at cutoff 3SD from mean nonalloexposed males’ NBGR
–ELISA vs. Luminex platform benefits & challenges

25. 25
Study Conclusions (cont’d)
•The low, statistically-equivalent reactivity rates in
alloexposed males and nonalloexposed donors imply
that transfusion / transplantation are not important
triggers for HLA Ab in healthy apheresis donors
–Questioning for transfusion/transplantation history will not help
much in identifying donors with HLA Ab
–On average, for every 100 nonalloexposed donors screened,
only two have antibody and another one will be unnecessarily
deferred (data not shown)
•The variability of similarly-derived cutoff values between
ARC’s study and the REDS II LAP study demonstrates
the potential influence of platform, specimen, and
reagent lots, emphasizing the importance of local
validation of the Luminex to choose reactivity cutoffs

26. 26
REDS II LAP Study
•7920 donors recruited in 2007
from 6 REDS II centers (incl.
Southern & New England regions);
transfused males intentionally
oversampled; One Lambda
LabScreen Mixed beads used on
manual Luminex platform with
cutoff 3SD from mean NBGR of
nonalloexposed males; single
antigen specificities also
determined on samples reactive
using “ultrasensitive” cutoffs
(NBGR = 2.2)
Triulzi DJ, et al. Transfusion 2009;49:1825-35; Kakaiya RM, et al. Transfusion 2010;50:1328-34;
Rios JA, et al. Transfusion 2011;e-pub.
II

27. 27
REDS II LAPS Conclusions
•3SD cutoff yielded reactivity in 24% of prev. pregnant
women (30% of those both pregnant & transfused)
–Two pregnancy losses (TAb, early SAb) were required to
increase Ab risk
–Risk increased with each successive pregnancy (1 – 14.5%
to ≥4 – 32.4%)
•1.0 – 4.4% of transfused-only women & men and
nonalloexposed donors had Ab by Luminex
–Transfusion did not increase the risk of HLA Ab
•Deferring all females, all previously pregnant females or
only those testing Ab positive would reduce apheresis
availability by 37.1%, 22.5% or 5.4%, respectively
Triulzi DJ, et al. Transfusion 2009;49:1825-35; Kakaiya RM, et al. Transfusion 2010;50:1328-34;
Rios JA, et al. Transfusion 2011;e-pub.

28. 28
HNA Antibody Tests: GIF & GA
Granulocyte
Immmunofluorescence
(GIF)
Granulocyte
Agglutination
(GA)

29. 29
HNA Antibody Tests: MAIGA
*
+
Incubate, wash,
lyse
Anti-mouse IgG
Typed WBC
Patient Serum
+
Mouse anti-HNA-x
Anti-human IgG
Monoclonal Antibody Immobilization of Granulocyte Antigens
Immobilize, tag,
and read by EIA
+ +
*

30. 30
HNA Antibody Testing Issues
•REDS II study showed 0.7% prevalence of HNA Ab
amongst all donors (0.7% of nonalloexposed males had
non-specific Ab only; 0.6% of women with ≥3
pregnancies had specific Ab, only 1/5
th
of whom did not
also have HLA Ab)
•Ab ID is a low volume, time-consuming (GA requires 4-6 hr
incubation) test, not yet reliably automated
–Ab ID by serology requires fresh neutrophils from typed lab
personnel
•Concurrent presence of HLA Class I antibodies
complicates GIF / GA interpretations
–MAINA avoids these issues, but technically more challenging
•GIF generally more sensitive except for HNA-3a (the major
culprit in severe TRALI), for which GA is better
Gottschall JL, et al. Transfusion 2011;e-pub.

31. 31
Automated HNA Ab Testing
•One Lambda is introducing LabScreen Mixed
(includes beads for HLA Class I / II and HNA-1a,
-1b, -1c, -2a, -4 & -4/5)
–In a small French study, 48 of 51 sera suspected of
having alloAb were concordant by Luminex and GIF /
MAIGA (kappa coefficient = 0.85)
–ARC’s experience with LabScreen Multi was not as
positive, though significant bead & process changes
occurred after our trial
–HNA-3a polymorphism recently discovered and not yet
available
–Other companies pursuing HNA Ab automation, though
no data are as yet available
Fromont P, et al. Transfusion 2010;50:2643-8.

32. 32
So… What Are ARC’s Plans?
•Females donating apheresis platelets / plasma
asked about prior pregnancies; yes answer (!D32
added with date of last pregnancy / loss) in new donors
only → HLA Ab testing from add’l Red Top
•3d shipping & 10-14d test TAT from 1 of 5 ARC
HLA labs using LabScreen Mixed beads on
Luminex → DCSC & region notified (15-760 FTD
females to be tested per region under R1; ~6700 per year)
–!D30 for negative result, !D31 / HLA assertion for
positives (48 hours to retrieve only in-house plasma &
apheresis platelets)
–HLA assertion prevents apheresis platelets and
transfusable or injectable / frac plasma from shipping

33. 33

34. 34
Treatment & Prognosis of TRALI
•Interruption of the transfusion
•Ventilation & Hemodynamic Support
•No role for the diuretics and
corticosteroids.
•80% resolve within 96 hours.
•Best recommendation…. Prevention.

35. 35
Q & A

36. 36
Thank
You!