Biomarkers in ALI and ARDS

Mohammed Attia, MD, FRCPCH (UK)

Foretelling
using dreams
Dreaming of….
a deep well
prison
a mirror
a second wife
a shining moon
forgiveness
a large cat
a large crop
If someone sees himself ……
dead
long life
eating crocodile flesh
he will become a village official
with his face in a mirror
a new life
bringing in cattle
will evangelize the spirit of the community
If someone sees himself ……
uncovering his own backside
he will become an orphan
putting one’s face to the ground
the dead want something
plunging into cold waters
absolution of all ills
making love to his wife in daylight
god will discover his misdeeds
Introduction
ALI occurs
◦ In 1 – 4% of all PICU cases
◦ 10% of all children receiving mechanical
ventilation on PICU
ALI is associated with a mortality of 25%
to 75% depending on
◦ Diagnostic criteria used to identify cases
◦ Coexisting risk factors
◦ Presence of nonpulmonary organ failure
The Ideal Biomarker
Increases pathologically in the presence
of the disease (high sensitivity)
Does not increase in the absence of the
disease (high specificity)
Relates to the disease burden and extent
Changes in accordance with the clinical
evolution, reflecting the current status of
disease
Anticipates clinical changes before it
happens
Shows no major fluctuation in serum
level
Adds independent information about the
risk or prognosis
Reproducible
Easy and cheap determination
Markers in ARDS and ALI
The main characteristics of the syndrome
are
◦ Diffuse inflammation
◦ Increased microvascular permeability
Although a variety of insults may lead to
ARDS, a common pathway may probably
result in the lung damage
A clinically useful biomarker for ALI
might add information regarding the
development of ARDS in at-risk patients.
The clinically useful marker will help the
intensivist to monitor the disease and
evaluate or modulate treatments.
Epithelial damage

Neutrophil activation

Inflammatory
mediators release

Endothelial
activation
Epithelial damage

Neutrophil activation

Inflammatory
mediators release

Endothelial
activation
Biomarkers of ARDS and ALI
Cytokines
◦ IL-1
◦ IL-2
◦ IL-6
◦ IL-8
◦ IL-15
◦ TNF-α
◦ IL-10
◦ IL-1 ra
Epithelial damage

Neutrophil activation

Inflammatory
mediators release

Endothelial
activation
Epithelial damage

Neutrophil activation

Inflammatory
mediators release

Endothelial
activation
Biomarkers of ARDS and ALI
Markers of endothelium activation
◦ Adhesion molecules
 E, L-selectin
 I-CAM-1
 V-CAM-1
◦ VWF
Epithelial damage

Neutrophil activation

Inflammatory
mediators release

Endothelial
activation
Epithelial damage

Neutrophil activation

Inflammatory
mediators release

Endothelial
activation
Biomarkers of ARDS and ALI
Markers of neutrophil activation
◦ Matrix metalloproteinase-9 (MMP-9)
◦ LTB4
Epithelial damage

Neutrophil activation

Inflammatory
mediators release

Endothelial
activation
Epithelial damage

Neutrophil activation

Inflammatory
mediators release

Endothelial
activation
Biomarkers of ARDS and ALI
Lung epithelium specific proteins
◦ Surfactant-associated proteins
 SP-A
 SP-B
 SP-D
◦ Mucin-associated antigens
 Krebs von den Lungen-6 (KL-6/MUC1)
Cytokines
Anti-inflammatory cytokines Pro-inflammatory cytokines

IL-10 TNF-α
IL-1ra IL-1
sTNFR1 IL-6
sTNFR2 IL-8
IL-6 IL-4
Cytokines
Cytokines are produced either by
◦ local resident cells
 Alveolar macrophages
 Pneumocytes
 Endothelial cells
 Fibroblasts
◦ cells arriving to the lung in response to local
or systemic injury
 Neutrophils
 Lymphocytes
 Platelets
 TNF-α and IL-1 serum levels are related
to ARDS disease severity and
mortality(1,2).
Serum levels of IL-6, IL-8 and CD 11b
could identify patients with septic shock
at risk for organ failure, including
respiratory failure (3,4).

1- Marks JD et al., Am Rev Respir Dis 1990, 141:94-7.

2- Damas P et al., Crit Care Med 1989, 17:975-8.
3- Pinsky MR et al., Chest 1993,103:565-75.
4- Takala A et al., Clin Sci (Lond) 1999, 97:529-38.
 Some researchers found that serum
cytokines could not be used as a routine
laboratory test to predict the outcome in
septic-shock patients.

Calandra T et al., Am J Med 1991, 91:23-9.

 With multiple study limitations, it was
found that plasma TNF-α, IL-6, IL-8 fail
to associate with severity and course of
ARDS in 19 leukocytopenic patients.

Meanwhile, BALF levels appeared to

differentiate between responders and non-
responders to treatment before clinical
differences become apparent.
Kiehl MG et al., Crit Care Med 1998,26:1194-9
 Meduri et al. studied plasma levels of
various cytokines in 27 ARDS patients.
Consistent, efficient and independent
predictive value for IL-1β and IL-6 serum
concentrations over time in severe ARDS.
Superiority of IL-1β and IL-6 plasma
levels monitoring over commonly applied
clinico physiologic parameters.

Meduri GU et al., Chest 1995, 107:1062-73

 Agouridakis et al. found an excellent
predictive value of plasma IL-2 and IL-15
levels compared to those observed in
BALF.

A remarkable finding was the emergence

of the discriminative usefulness of
elevated IL-2 and IL-15 serum levels in
patients with ARDS or at risk for ARDS.
Agouridakis P, et al., Eur J Clin Invest 2002, 32:862-7.
 In contrast Lesur et al. found lower blood IL-
2 levels in patients with ARDS compared to
those that never developed ARDS.
There was a strong association of early low
serum IL-2 levels with the patients’ survival
A major criticism to this study is opposite
and disproportional fluctuations of IL-2
content in serum and BALF in patients with
or without ARDS

Lesur O et al., Crit Care Med 2000, 28:3814-22.

Anti-inflammatory cytokines
A large prospective (77 patients at risk of
ARDS) found that
◦ IL-1ra and IL-10 were elevated in patients at
risk for ARDS

◦ IL-1ra and IL-10 exhibited a remarkable

association with the disease outcome, but
could not predict the development of the
syndrome

Parsons PE et al., . Am J Respir Crit Care Med 1997, 155:1469-73.

 The inability of serum cytokine levels to
predict the development of ARDS in
patients with ALI was reproduced by
other researchers(1,2).

1- Takala A et al., Shock2002, 17:252-7.

2- Bouros D et al., BMC Pulm Med 2004, 4:6.
 There is a remarkable consistently
elevated serum levels of IL-6 and IL-8 in
ARDS and/or severe pneumonia,
differentiating these entities from
cardiogenic pulmonary oedema.
Serum IL-6 and IL-8 were unable to
differentiate ARDS from severe
pneumonia

Schutte H, et al., Eur Respir J1996, 9:1858-67.

Limitations in cytokine researches
Number of patients is small in most
studies
Including heterogeneous groups of
patients in the same study
Blood versus BALF levels of cytokines
Studying limited number of molecules
Limitations in cytokine researches
Not looking at the anti-inflammatory limb
of the cytokine cascade
Lack of serial measurements
Statistical analysis not including ROC
and not reporting cut-off values
Markers of endothelium
activation
The patho-physiologic sequence
characterizing ALI involves a cascade of
leukocyte-endothelium interactions and
adhesions.
The adhesion is followed by
transendothelial migration of neutrophils
and release of their cytotoxic products
resulting in microvascular and tissue
injury.
Adhesion of neutrophils to the
endothelium is regulated by
◦ selectins (E, L and P)
◦ Integrins
◦ immunoglobulin superfamily
 ICAM-1
 VCAM-1
Interactions of leukocytes and the
endothelium
◦ An initial “loose” contact (rolling): selectin
◦ Firm adhesion: integrin (β2) and ICAM-1
Adhesion Molecules
Soluble isoforms have been detected in the blood
under various inflammatory conditions
◦ s-E-selectin
◦ sICAM-1
◦ sVCAM-1
Mechanisms to explain an increase in circulating
adhesion molecules:
◦ overexpression by the endothelial cells induced by
cytokines (IL-1, TNF-α)
◦ increased proteolytic cleavage of endothelial-bound
adhesion molecules secondary to endothelial damage
Adhesion Molecules
The expression is almost restricted to
stimulated endothelial cells

Their presence in serum should

potentially reflect the state of endothelium
in disease and subsequently the disease
severity in ALI
von-Willebrand factor antigen
A macromolecular antigen that is
produced by
◦ endothelial cells
◦ platelets and megakaryocytes
Endothelial injury (as in ARDS) results in
the release of VWF from preformed stores
into the circulation.
 Donnelly et al. demonstrated in a large
cohort of patients at risk for ARDS that
◦ Mean circulating levels of sE-selectin were
not correlated with subsequent ARDS
development and patients' mortality.
◦ Low values of sL-selectin exhibited a
significant prognostic value

Donnelly SC et al. , Lancet 1994,23(344):215-9

 Cowley et al. showed a superiority
plasma levels of sE-selectin in predicting
organ dysfunction and death in a group of
patients with SIRS compared to sICAM-1
peripheral concentrations.

Cowley HC et al., Crit Care Med 1994,22:651-7.

 sICAM-1 sequential plasma levels, were
suggestive of a strong association between
the severity of shock
◦ the presence of hypotension
◦ the requirement of vasoactive drugs
and the circulating concentrations of the
sICAM-1.

Sessler CN et al., Am J Respir Crit Care Med 1995, 151:1420-7.

Kayal S et al., Am J Respir Crit Care Med 1998, 157:776-84.
ICAM-1 and VCAM-1
Were not found to be independent factor
for ARDS development,
Exhibited a considerable negative
predictive value for ARDS development.
ROC curve analysis showed a clear
superiority of plasma parameters in
correlating with the disease outcome
compared with BALF molecules

Agouridakis P et al., Respir Res 2002, 3:25.

VWF
Marked and independent association of
circulating VWF with the disease severity
as assessed by other commonly applied
clinical variables.

Serum VWF levels were an independent

predictor of hospital mortality and were
associated with longer duration of MV

Kayal S et al., Am J Respir Crit Care Med 1998, 157:776-84.

Ware LB et al., Crit Care Med 2001, 29:2325-31.
VWF
In a multicenter study of 559 patients with ALI and
ARDS
A significant correlation of elevated VWF plasma
levels with adverse outcomes
◦ Mortality
◦ duration of unassisted ventilation
◦ organ failure
An association between vWF and presence or
absence of sepsis, supporting the hypothesis that
ALI might be an independent cause of systemic
endothelial activation and injury.
No modulation of plasma VWF concentrations by
protective mechanical ventilation.
Ware LB et al., Crit Care Med 2001, 29:2325-31.
Markers of neutrophil
activation
There is an increasing evidence
implicating neutrophils in most cases of
ARDS.
Neutrophils have been reported by several
studies to exert an important role in the
early phase of ALI characterized by
◦ architecture remodeling
◦ Surfactant and epithelial toxicity
Neutrophils use a wide array of enzymes
during the process of transmigration through
alveolar-capillary barrier.
These enzymes include, metalloproteinases
(MMPs) such as MMP-9 also called
gelatinase B which is secreted from
preformed neutrophil granules in response to
proinflammatory cytokines (IL-8, TNF-α).
MMP-9 is secreted as a zymogen, and then
activated by other proteases such as elastase,
and plays a crucial role in digesting basement
membranes.
 Pulmonary edema fluid levels of IL-8,
MMP-2, MMP-9 could differentiate
ARDS and cardiogenic pulmonary edema
whereas plasma levels proved to be of
poor discriminative value.
These findings raise the issue that the
inflammatory response of ARDS patients
is well compartmentalized, with little
spillover into the circulation.
Pugin J et al., CritCare Med 1999, 27:304-12.
Leukotrienes
LTs (B4, C4, D4, E4) exert a synergistic role
with IL-8 in the neutrophil influx and
activation leading to a massive recruitment
of neutrophils and a massive inflammatory
response.
Their BALF levels have been found elevated
in patients with ARDS.
Contribute to alterations of microvascular
permeability and accumulation of pulmonary
edema.
 LTB4 plasma levels could serve as a
valuable predictive marker of ARDS in
terms of specificity and sensitivity.
There is a strong association of both
LTB4 and IL-8 peripheral concentrations
with the patients' survival.

Amat M et al., Crit Care Med 2000, 28:57-62.

Lung epithelium-specific
proteins
The lung epithelium produces mucus
blanket and surfactant proteins

Study of BALF is the classical means of

studying the proteins secreted by the lung
epithelium and investigating their
alterations in lung disorders.
The presence of these proteins in the
bloodstream can be explained by several
mechanisms including
◦ leakage from the lung into the bloodstream
◦ Increased production by the alveolar type II
cells
◦ diminished clearance rates from the circulation
Surfactant-associated Proteins
Pulmonary surfactant is a complex and
highly surface active material covering
the alveolar space of the lung.
Structure
◦ Mainly phospholipids
◦ Surfactant proteins
Surfactant-associated Proteins
Functions of the alveolar surfactant
system:
◦ Reduce surface tensions of alveoli
◦ Prevention of alveolar edema
◦ Pronounced influence, especially of the
collectins (SP-A and SP-D) in the innate
immune system of the lung
Serum levels reflect the epithelial damage
and turnover.
Surfactant-associated Proteins
Four surfactant-specific proteins
◦ SP-A
◦ SP-B
◦ SP-C
◦ SP-D
SP-B and SP-C are low MW hydrophobic
SP-A and SP-D are high MW hydrophilic
SP-A and SP-D belong to the collectin
subgroup of the C-type lectin superfamily.
SP-A and SP-D are produced by two
types of epithelial cells in the peripheral
airway
◦ Clara cells
◦ Alveolar type II cells.
SP-A
Acute indicator of lung function and
alveolocapillary membrane injury (1)
SP-A BALF levels are strongly related to
outcome and likelihood of disease
progression (2)
Plasma levels are associated with severity
of clinical lung injury and with disease
outcome (3)
1- Doyle IR et al., Am J Respir Crit Care Med 1995, 152:307-17.
2- Greene KE et al., Am J Respir Crit Care Med 1999, 160:1843-50.
3- Cheng IW et al., Crit Care Med 2003, 31:20-7.
SP-A
There is a potential value of SP-A plasma
levels in discriminating at risk patients
who developed ARDS from those with
sepsis and aspiration.(1)
discriminating patients with ALI of
various etiologic factors
◦ severe pneumonia
◦ cardiogenic lung edema.(2)

1- Greene KE et al., Chest 1999, 116:90S-91S.

2- Gunther A et al., Am J Respir Crit Care Med 1996, 153:176-84
SP-B
Superior to SP-A plasma levels as a
marker of
◦ lung function
◦ alveolocapillary membrane injury.(1)
SP-B cut-off plasma levels predict ARDS
development in at-risk patients suffering a
direct lung injury
An increase of blood SP-B levels was
observed before other changes of ALI
become apparent.(2)
1- Doyle IR et al., Am J Respir Crit Care Med 1997, 156:1217-29.
2- Bersten AD et al., Am J Respir Crit Care Med 2001, 164:648-52.
SP-D
SP-A BALF levels are strongly related to
outcome and likelihood of disease
progression.(1)
In a large RCT
◦ Elevated SPD levels were associated with worse
clinical outcomes
 Greater risk of death
 Fewer ventilator- free days
 Fewer organ failure-free days
◦ Attenuation of SP-D plasma levels by lower
volume ventilation strategies .(2)
1- Greene KE et al., Am J Respir Crit Care Med 1999, 160:1843-50.
2- Eisner MD et al., Thorax 2003, 58:983-8.
Mucin-associated Antigens
Mucins are major components of the
mucus layer covering the airway
epithelium.
Mucins are either
◦ associated with cell membranes or
◦ secreted at the surface of the respiratory tract
KL-6
Krebs von den Lungen-(KL)-6 is a
mucin-associated antigen mainly
associated with cellular membranes.
KL-6 is mainly detected in
◦ alveolar type II cells
◦ epithelial cells of the respiratory bronchioles.
It is also present on other somatic cells
◦ pancreatic cells
◦ eosophageal cells
◦ fundic cells of the stomach
KL-6
 KL-6 is a sensitive indicator of damage
to alveolar type II cells
Its raise would theoretically represent
◦ the destruction of the normal lung parenchyma
and architecture
◦ the increased permeability of the air-blood
barrier
KL-6
Association of optimal cut-off values of
KL-6 serum levels with patients' mortality
with high sensitivity (87%) and specificity
(100%)
This finding supports the theory that
disruption of the alveolar barrier
represents a major determinant of
prognosis of ALI

Ishizaka A et al., Am J Physiol Lung Cell Mol Physiol 2004, 286:L1088-94.

KL-6
Association of KL-6 serum levels with
variables of lung injury severity and with
mortality rates.

No statistically significant correlation

between blood levels and ventilation
strategies.

Sato H et al., Eur Respir J 2004,23:142-5.

Summary
The ideal biomarker
Cytokines
Products of endothelial activation
Products of neutrophil activation
Lung epithelium-specific proteins
Future directions
Finding and applying the ideal biomarker
in the clinical setting would be very
helpful in terms of
◦ Defining categories of patients for different
therapies or prognosis.
◦ For the purpose of counseling families and
patients regarding high/low risk of
complications.
◦ Possibly identifying novel therapeutic targets.
Future directions
Development of a prognostic index that
combines clinical and biological
determinants may be useful to ameliorate
the limitations encountered using only
biomarkers.
A combination index could be crucial for
the selection of more homogeneous
groups of patients with ALI/ARDS for
further studies.
Future directions
Crossing the boundary from research to
clinical application requires
◦ Validation in multiple settings
◦ Experimental evidence supporting a
pathophysiologic role
◦ Ideally intervention trials showing that
modification improves the outcome
Conclusion
Conclusion
Currently, the application status in routine
clinical practice for most of these biologic
markers of ALI is still in its infancy and
remains exploratory.
It is important to note that the greatest
promise is in only few (VWF, IL-1β, IL-6,
ICAM-1, VCAM-I, E-selectin) which
exhibited independent discriminatory
power.
Conclusion
There is a necessity for further
investigations in the context of large
prospective studies analyzing
homogeneous and well defined group of
ARDS or at risk patients and the
assessment of novel molecules to serve as
diagnostic and prognostic tools, as well as
markers of the disease activity and
severity.
The question remains ….
How much have we moved from
here?
Thank you