The effects of immunostimulation with GM-CSF or IFN-γ on immunoparalysis following
human endotoxemia.
A parallel randomised double-blind placebo-controlled study

Sepsis is a major medical challenge associated with a high mortality rate.
Release of pro-inflammatory mediators can result in hemodynamic instability,
coagulation abnormalities and end-organ dysfunction. Previous strategies have
aimed to treat sepsis by inhibition of pro-inflammatory mediators, however,
most of these approaches have failed. This might be due to the fact that the
majority of septic patients do not succumb to the initial pro-inflammatory
*hit*, but die at a later time-point in a pronounced immunosuppressive state.
This so-called *immunoparalysis*, which renders patients extremely vulnerable
to secondary infections, results from the triggering of counter-regulatory
anti-inflammatory pathways in response to initial pro-inflammation response.
Interferon-gamma (IFN-γ) and granulocyte macrophage colony-stimulating factor
(GM-CSF) are known for their immunostimulatory effects. GM-CSF has been
demonstrated to increase monocyte HLA-DR expression and endotoxin-induced
pro-inflammatory cytokine production ex vivo in whole blood of patients with
severe sepsis. Moreover, recent pilot trials in septic patients indicate that
long-lasting monocyte deactivation in sepsis can be reversed by treatment of
sepsis patients with immunostimulants, such as GM-CSF and IFN-y. Controlled,
biomarker-guided studies focused on the mechanism of action of
immunostimulatory therapies in a standardized setting have not been performed
yet.
We have previously shown that endotoxin administration in healthy volunteers
leads to pronounced immunosupression. Consequently, human endotoxemia can serve
as a model for sepsis-induced immunoparalysis. Furthermore, to date, all
studies on the subject have analyzed the *immune status* of the patients using
ex vivo stimulation of leukocytes or flow cytometry analysis of circulating
leukocytes obtained from the immunocompromised host. We have recently
demonstrated that, following endotoxin administration, ex vivo leukocyte
hyporesponsiveness returns to normal within one day while the in vivo response
to endotoxin is impaired for 2 weeks. These data indicate that ex vivo
measurements do not accurately reflect the in vivo situation. In the present
project, we wish to study the effects of IFN-γ and GM-CSF in a parallel
double-blind placebo-controlled randomized manner on the immunoparalysis
following human endotoxemia, both in vivo and ex vivo.

Primary objective: The primary objective of the study is to determine the
effects of GM-CSF/IFN-γ on the in vivo immunoparalysis induced by human
endotoxemia. This will be determined by measuring plasma levels of various pro
and anti-inflammatory cytokines and assessing the difference in the LPS-induced
cytokine response between day 1 and 7. The primary outcome measure is the
difference in the LPS-induced increase in plasma TNF-α concentration between
day 1 and day 7 during experimental endotoxemia.

Secondary Objective(s): There are 8 secondary objectives:

1. To determine the effects of GM-CSF/IFN-γ on ex vivo responsiveness of
leukocytes to various inflammatory stimuli.

2. To determine the effects of GM-CSF/IFN-γ on monocyte HLA-DR expression.

3. To determine the effects of GM-CSF/IFN-γ on activation of nuclear
factor-*B (NF-*B) by reactive oxygen and nitrogen species (ROS and RNS).

4. To determine the effects of GM-CSF/IFN-γ on inflammatory transcriptional
pathways (by use of microarrays).

5. To determine the effects of GM-CSF/IFN-γ on changes in phenotype and gene
expression caused by mechanisms other than changes in the underlying DNA
sequence (by use of epigenetics).

6. To determine the effects of GM-CSF/IFN-γ on (subclinical) tubular injury
known to occur during human endotoxemia.

7. To determine the effects of GM-CSF/IFN-γ on LPS-induced clinical symptoms
(illness score) and hemodynamic/temperature changes.

8. To determine the effects of endotoxin on (subclinical) diaphragm function
and the effects of GM-CSF/IFN-γ on endotoxin-induced diaphragm dysfunction.

Parallel, randomized double-blind placebo-controlled intervention pilot study
in healthy human volunteers during repeated experimental endotoxemia. Both the
mechanism and the extent of restoration of the immunoparalysis may be different
between GM-CSF and IFN-y. Use of a parallel study design enables us to evaluate
the effects of GM-CSF and IFN-y against a placebo group and head-to-head
against each other.

All subjects (n=18) will receive an intravenous bolus of endotoxin (LPS derived
from E coli O:113, 2 ng/kg) twice, with an interval of 6 days (LPS
administration on day 1 and 7). Subjects will receive either GM-CSF (4 µg/
kg/day subcutaneously, n=6), IFN-γ (100 µg/day subcutaneously, n=6) or placebo
(NaCl 0.9% subcutaneously, n=6) in a randomized, double-blind manner on day 2,
4 and 6.

A medical interview and physical examination are part of this study. During
endotoxemia, volunteers will be monitored on the research unit of our intensive
care and receive an arterial line to facilitate blood pressure monitoring and
blood sampling. The arterial line will be placed under local anaesthesia using
2% lidocaine. Furthermore, a venous cannula will be placed for the
administration of fluids and LPS and an esophageal catheter will be placed for
measurement of twitch transdiaphragmatic pressure. To elicit twitch
transdiaphragmatic pressure, cervical magnetic stimulation will be performed.
The administration of LPS induces flu-like symptoms for approximately 4-6 hrs.
This model of systemic inflammation has been applied for many years in various
research centres in the world. Endotoxin administration is considered safe and
no long-term effects have ever been documented.
At the Radboud University Medical Centre, over a 150 volunteers have received
more than 250 injections of lipopolysaccharide. Therefore, there is sufficient
experience with this model at this centre. Administration of GM-CSF and IFN-γ
is well-tolerated in healthy volunteers. In total, approximately 600 ml blood
will be drawn during the 8 days of the study and urine will be collected.
Subjects will not benefit directly from participation to the study. A subject
fee is provided.