Development of a *two hit* in vivo autologous transfusion model in healthy volunteers

Transfusion-related acute lung injury (TRALI) is the leading cause of
transfusion-related morbidity and mortality. The incidence is high and reported
up to 8% in specific patient groups. TRALI is thought to be a *two hit* event.
The first event is the underlying condition, often sepsis or an infection, of
the patient resulting in priming of neutrophils. The second event is the
transfusion of a blood product, after which either antibodies present in the
blood product or pro-inflammatory mediators present in stored cell-containing
blood products (e.g. Red Blood Cells (RBCs)) or aged erythrocytes and/or
platelets themselves activate the primed neutrophils, resulting in pulmonary
edema. Opposed to the traditional view that TRALI has a good prognosis,
evidence is accumulating that TRALI has a significant impact on morbidity and
outcome, at least in specific patient groups such as critically ill patients.
The association of transfusion with adverse outcome resulted in blood product
and donor management strategies aimed at decreasing the risk of acquiring
TRALI. Plasma products originating from female donors and specifically
multiparous donors have up to 40% presence of antibodies directed against
leukocyte antigens and are associated with the onset of TRALI. From this point
of view the Netherlands and the UK have started from 2006 and 2003 respectively
a male only plasma donation policy. Excluding female donors for plasma donation
seems to have reduced, but not prevented the occurrence of TRALI. Additional
research is needed to determine whether the use of fresh (at present RBCs are
stored up to 35 days in The Netherlands) cell-containing blood products may be
an additional measure to reduce TRALI.
We recently demonstrated in an in vivo animal model that the supernatant of
stored RBCs induce mild lung injury in the presence of a *first hit* of
lipopolysaccharide (LPS). Other pre-clinical studies showed that transfused
aged erythrocytes themselves induce mild lung injury by loss of the Duffy
antigen and by induction of tissue inflammation by acute tissue iron
deposition. However, clinical studies on the impact of transfusion of aged red
blood cells on respiratory complications have yielded conflicting results. In
cardiothoracic surgery patients, respiratory insufficiency and mortality was
lower in patients that had received blood stored for less then 14 days compared
with patients that had received blood stored for more then 14 days (7.4% vs.
11.0%, P<0.001). However, similar studies did not confirm these findings.
The discrepancy between results from pre-clinical and clinical studies on the
effect of storage time of cell-containing blood products and the onset of mild
lung injury calls for a randomized trial. Although recent studies showed a
relative high incidence of TRALI, still the numbers are low and the presence or
absence of a *first hit* is hard to measure in the clinical setting which in
total makes it difficult to perform a clinical randomized trial. For this
reason we propose to develop a *two hit* in vivo autologous RBCs transfusion
model, i.e. a mild TRALI model in healthy volunteers. When the model is
developed it has to confirm the hypothesis that stored RBCs products induce
mild lung injury in the presence of a *first hit* in the human setting (i.e. a
mild form of TRALI). The advantages of such a model are the following; 1) an
autologous transfusion model makes it possible to investigate the effect of
storage time on the onset of lung injury as the effect of anti-body mediated
TRALI is excluded; 2) the *first hit* is standardized; 3) this model will help
us to investigate pathways involved in onset of stored blood induced lung
injury and may enable us to test preventive or therapeutic measurements aimed
at improving storage conditions. The latter aspect of the model will become
very important in the near future and results from these studies may prevent
impeding a continuous reliable blood supply when a policy of fresh blood only
is proposed.

We propose the development of a *two hit* in vivo autologous RBCs transfusion
model, i.e. a mild TRALI model in healthy volunteers to confirm that stored
RBCs are associated with the onset of mild lung injury in the presence of a
*first hit* of LPS in the human situation.

to develop a *two hit* in vivo autologous RBCs transfusion model in healthy
male volunteers.

Subjects:
Healthy male volunteers.

Study groups:
Group 1 (n=6): *First hit* Lipopolysaccharide (LPS) 2ng/kg + *Second hit*
Saline
Group 2 (n=6): *First hit* LPS + *Second hit* Fresh Red Blood Cells (RBCs) (2
day storage)
Group 3 (n=6): *First hit* LPS + *Second hit* Stored RBCs (35 day storage)

Methods:
All subjects will be screened (medical history, physical examination, ECG,
blood examination, spirometry, DLCO, chest x-ray) by the research physician of
our hospital and of Sanquin Blood Bank prior to involvement in the experiment.
All included healthy volunteers (n=18) will donate 1 unit of whole blood at
Sanquin Blood Bank which will be processed into 1 unit of RBCs (approximately
300ml). Processing and storage will be according to Sanquin Blood Bank
protocol. Prior to transfusion stored RBCs will be biotinylated (Vitamin B8) to
allow their identification with flow cytometry. In short stored RBCs will be
labelled with Sulfo-NHS biotine of Pierce (6-20µg/ml).
Subsequently on the study day healthy volunteers receive a *first hit* of
either E. coli lipopolysaccharide (LPS) 2 ng/kg i.v. (n=18). Two hours after
the *first hit* they receive an autologous transfusion of 1 unit of fresh (2
day storage) biotinylated RBCs or an autologous transfusion of 1 unit aged
biotinylated (35 days of storage) RBCs or an equivalent volume of saline 0.9%
infusion. The transfusion itself will be performed in one hour. During the
experiment subjects will be monitored for blood pressure and arterial
oxygenation using an indwelling arterial line. Blood samples will be drawn from
an indwelling artery line prior to the *first hit*, prior to the transfusion,
directly after transfusion, 0,5, 1, 2, 4 and 6 hours after transfusion. At the
same time points exhaled air will be collected for E-Nose measurement.
Furthermore, 6 hours after transfusion spirometry and DLCO measurement will be
repeated. A chest x-ray and a directed broncho-alveolar lavage (BAL) will be
performed 6 hours after transfusion. The BAL will be performed by an
experienced pulmonologist according to the Dutch pulmonologist guidelines
(NVALT Guidelines, 2004). In the BAL-fluid and plasma samples markers of
inflammation, neutrophil activation and coagulation activation are measured to
confirm whether we have developed a model of TRALI. Three months after the
study day a venous sample of 4 ml will be collected to measure prevalence of
biotin antibodies.

Group 1: *First hit* Lipopolysaccharide (LPS) 2ng/kg + *Second hit* Saline
Group 2: *First hit* LPS + *Second hit* Fresh Red Blood Cells (RBCs) (2 day
storage)
Group 3: *First hit* LPS + *Second hit* Stored RBCs (35 day storage)

Nature and extent of the burden and risks associated with participation.

De blood donation:
1. The donation of the blood transfusion will take place at the Sanquin Blood
Bank. The blood donation can be accompanied by some pain and the chance of a
bruise. The blood transfusions are processed by Sanquin according to their
standard protocol for clinical blood products.

The blood products:
1. Blood transfusions originate from the test person himself and will not bare
risk of for example virus transmission. The possibility exists that the
transfusion will induce mild transient lung injury. The test person might
perceive this as as mild stuffiness.
2. During storage of the blood product samples from the blood product will be
taken to measure the storage related changes in the blood product. The sampling
finds place under sterile conditions.
3. Prior to transfusion stored RBCs will be biotinylated (Vitamin B8) to allow
their identification according to previously published protocols.6,28 In short
stored RBCs will be labelled with Sulfo-NHS biotine of Pierce (6-20µg/ml).
Preparation will be done under sterile conditions. Cultures will be taken of
transfused products to confirm sterile conditions. Three months after the study
day a venous blood sample will be collected to detect development of biotine
antibodies. This data will be used to investigate antibody prevalence after
exposure to biotin. The presence of absence of antibodies has no clinical
relevance and repeated intravenous exposure to biotin does not produce adverse
effects.5,7

The experiment:
1. The test person will undergo two times a chest X-ray. The radiation exposure
is considered to be minor.

2. During the experiment an artery line will be inserted in one of arteries of
the arms of the test person. The placing of the artery line is performed by an
experienced anesthesiologist. There is a small chance this procedure results in
bruises or blood clot formation in the blood vessel. During the last years no
major complications occurred with this procedure in our center.

3. The administration of endotoxin can lead to mild flu-like symptoms, which
can include a slight increase in body temperature, muscle pain and/or fatigue.

4. The bronchial lavage (lung rinse) is performed by an experienced
pulmonologist. The main source of discomfort during this procedure is a dry
cough or mouth. However, these complaints are suppressed by the lidocaine
spray. In addition, there may be onset of fever within 24 hours.

5. Blood sampling during the experiment will be from the artery line and will
not result in additional punctures. The total amount of blood that is sampled
during the experiment is 135 ml. The 135 ml sampling is next to the 500 ml of
blood donation. However, depending on the experimental group the test person is
in, the majority of the 500ml donation will be returned during the experiment
by the blood transfusion. The human body can easily handle these changes in
blood volume. It is not allowed to donate blood or participate in another study
within the three months prior to this study or during this study.

6. Participation in this study may result in exclusion of future blood donation
at Sanquin since the test person after this study is registered as a person who
has received a blood transfusion.

7. The pulmonary function measurement and the collection of exhaled air is
estimated as no burden or risk for the healthy volunteers.

Risks assessment:
1. Infusion of E. coli LPS with a dose of 4 ng/kg has previously been proven to
be safe in healthy adult volunteers in our institution. We will use a lower
dose of 2 ng/kg which will induce neutrophil priming but will not result in a
SIRS reaction. 2. The use of an autologous transfusion human volunteer model
has also been proven to be safe. 3. Furthermore transfusions will be prepared
and transfused using the standard clinical protocols by Sanquin and our
hospital. The combination of these two models is expected only to cause mild
temporary side effects because of the following reasons: 1. Pre-clinical
studies show only mild lung injury after a *first hit* of LPS 2mg/kg and a
*second hit* of transfusion of stored blood products. 2. In the present model a
1000 fold lower dose of LPS (2ng/kg) will be used compared to the pre-clinical
model which we assume will make the model less severe.
2. Studies in healthy human volunteers showed that biotinylated (Vitamin B8)
RBCs can be safely administered without any side-effects. Although in a healthy
volunteer study 1 out of 8 subjects developed a transient positive test for
antibody to biotinylated RBCs, at 11 months post transfusion antibodies to
biotinylated RBCs had disappeared. Biotine labeling has no effect on RBCs
survival. Thus, biotinylation of RBCs is considered safe.