Immunopathogenesis of severe sepsis.

Full text

Page 1

432
Journal of the Royal College of Physicians of London Vol. 34 No. 5 September/October 2000
CME Septicaemia – I
Severe sepsis refers to the syndrome of
fever, hypotension and organ dysfunc-
tion resulting from infection, commonly
due to Gram-negative or Gram-positive
bacteria with or without bacteraemia1.
Occasionally, non-bacterial pathogens,
including fungi, rickettsiae, protozoa
and viruses, can cause a similar syn-
drome. The first line of defence against
infecting organisms comprises highly
conserved interactions of the host’s
innate immune system with microbial
products, followed by amplification of
the immune response to maximise host
defences. Whilst evidently conferring an
evolutionary survival advantage, the
humoral and cellular responses which
ensue are directly involved in the
pathogenesis of severe sepsis.
Host-microbial interactions
Microbial products which activate the
innate immune response include both
cell wall components and secreted
proteins (Table 1).
Cell wall components
Lipopolysaccharide (LPS), or bacterial
endotoxin, forms a major portion of all
Gram-negative cell walls and is the most
important microbial product implicated
in sepsis2. It has a structurally conserved
lipid component, lipid A, and an
oligosaccharide portion composed of a
core region and highly variable O-
specific chains. Lipid A activates multiple
components of the innate immune
system. Cellular activation occurs pre-
dominantly by binding to a serum
protein, LPS binding protein, which
facilitates interaction of LPS with the cell
surface receptor CD14. Transmembrane
signalling through recently described
‘Toll-like’ receptors follows3. Monocytes,
macrophages and neutrophils express
membrane CD14. LPS activation of cells
such as endothelia, which do not
express the membrane bound form, can
take place via soluble CD14. Other cell
surface receptors, such as the integrin
CD11/18 or the macrophage scavenger
receptor which can bind LPS, may
also be involved. LPS activates both
classical and alternative pathways of
complement, leading to production
of anaphylotoxins C3a and C5a and
terminal C5b-9 complexes. Direct
activation of Factor XII of the intrinsic
clotting cascade or contact system
stimulates coagulation, fibrinolytic and
plasma kallikrein-kinin systems. Gram-
positive bacterial cell wall components
such as peptidoglycans and lipoteichoic
acids similarly demonstrate CD14
dependent cellular activation4. They can
also activate complement and contact
systems, although the mechanisms are
poorly understood.
Secreted proteins
Secreted bacterial toxins which can
cause sepsis fall into several groups. So-
called ‘superantigens’ produced by
Staphylococcus aureus and Streptococcus
pyogenes can induce potent non-specific
Mahdad Noursadeghi MRCP, MRC Clinical Research Fellow
Jonathan Cohen FRCP FRCPath, Professor of Infectious Diseases and Microbiology
Department of Infectious Diseases and Microbiology, Hammersmith Hospital, London
J R Coll Physicians Lond 2000;34:432–6
Immunopathogenesis of severe sepsis
Cell wall componentsLipopolysaccharide All Gram-
negative bacteria
PeptidoglycanAll Gram-
positive bacteria
Lipoteichoic acid
Secreted toxins Superantigens Staphylococcal toxic shock
syndrome toxin 1
Staphylococcal enterotoxin B
S. aureus
Streptococcal pyrogenic
exotoxin A
S. pyogenes
Proteases Phospholipase C
Clostridia
Lipase
S. aureus
Nuclease
Neuraminidase
S. pneumoniae
Streptococcal pyrogenic
exotoxin B
S. pyogenes
Porins
a-Haemolysin
E. coli
Leukocidin
S. aureus
Catalase
Pneumolysin
S. pneumoniae
Streptolysin S & O
S. pyogenes
Table 1. Some examples of microbial products implicated in the pathogenesis of
severe sepsis.

Page 2

CME Septicaemia – I
activation of monocytes, macrophages
and T cells by cross-linking major
histocompatibility complex class II
molecules with subsets of T cell receptor
Vb chains, without intracellular process-
ing5(Fig 1). Alternatively, bacteria can
provoke a host response by causing
cellular injury by production of pore
forming toxins and a variety of
proteolytic enzymes. Some bacterial
protease enzymes
immune system directly by cleavage of
inactive substrates to active pro-
inflammatory mediators,
release of bradykinin from high
molecular weight kininogen or the con-
version of the pro-form of the cytokine
interleukin (IL)-1 to its active form by
streptococcal pyrogenic exotoxin B4.
Although there
differences in the way in which Gram-
negative and Gram-positive bacteria
interact with the host6, many similarities
are evident in the events that follow
activation of the innate immune system,
irrespective of the infecting organism.
stimulate the
such as
may be subtle
Amplification of the host response
Following the initial host-microbial inter-
action, widespread activation of other
components of the innate immune
response takes place, schematically out-
lined in Fig 2. This pro-inflammatory
cascade is mediated by either direct cell-
cell interactions or soluble factors
derived from serum proteins and cells,
which act in autocrine, paracrine or
endocrine fashion.
Journal of the Royal College of Physicians of London Vol. 34 No. 5 September/October 2000433
LPS
LPS
LPS
LPS
LPS
LPS
LPS
LBP
Intracellular
signalling
Neutrophils,monocytes,macrophages
endothelia
membrane
CD14
TLR
soluble CD14
Vb-restricted
TCR
Antigen-
specific
TCR
Superantigens
directly cross
link MHC II
and subsets
of T cell
receptors
Conventional antigens
undergo intracellular
processing before
presentation by MHC II
molecules to specific
T cell receptors
LPS
superantigen
MHC II
MHC II
APC
APC
T cell
T cell
?
Fig 1. Schematic outline of lipopolysaccharide (LPS) and superantigen (in contrast to
conventional antigen) mediated cellular activation (APC = antigen presenting cell;
LBP = LPS binding protein; MHC = major histocompatibility complex; TCR = T cell receptor;
TLR = Toll-like receptors).
Fig 2. Schematic outline of the interactions between cellular and humoral components which amplify the innate immune response to
infection (IFN = interferon; IL = interleukin; PAF = platelet activating factor; TNF = tumour necrosis factor).
Host-microbial interactions
Plasma proteins
Hepatic acute-
phase protein
synthesis
Cell adhesion molecules
IL-8
PAF
Histamine eicosanoids
Factor XII
IL-6
IL-12
Endothelia NeutrophilsT cellsMonocytes
Endothelia NeutrophilsT cellsMonocytes
Cell adhesion
molecules
IFN-g
Cellular activation
IL-1 TNF-a
Thromboplastin
Plasma kallikrein
system
Increased vascular
permeability
neutrophil recruitment
Increased inflammatory
exudate
Clotting/fibrinolytic
cascades
Complement
activation

Page 3

CME Septicaemia – I
434
Journal of the Royal College of Physicians of London Vol. 34 No. 5 September/October 2000
Cell-cell signalling
Direct
through constitutive or inducible expres-
sion of cell adhesion molecules (CAMs).
This occurs between monocytes or
macrophages and T cells during super-
antigenic activation, and between
monocytes or neutrophils and endo-
thelia during cellular extravasation and
infiltration into tissues. Several families
of CAMs involved in these cell-cell
signals are well characterised, in
particular the interaction of selectins
with glycoprotein ligands such as the
sialylated Lewis X antigen and that of b2
integrins, such as CD11/18 with
members of the immunoglobulin super-
gene family, intercellular CAM (I CAM)-
1/2 and vascular CAM (V CAM)7.
cellular interactions occur
Plasma proteins
Factor XII activation simultaneously
induces the fibrinolytic and intrinsic
clotting cascades. Widespread activa-
tion of these pathways during sepsis
leads to disseminated intravascular
coagulation (DIC)8. Bradykinin, also
generated from contact system activa-
tion, causes vasodilatation
increased capillary permeability, as do
C3a and C5a fragments from the
complement cascade
neutrophil chemoattractants, thereby
increasing inflammatory exudates9.
and
which are
Cytokines
Activated monocytes, macrophages,
neutrophils and endothelia produce
small soluble proteins,
together as cytokines. These provide
important amplification signals in the
immune response to infection and have
been directly implicated in the patho-
genesis of sepsis10. In particular, tumour
necrosis factor (TNF)-a and IL-1 are
produced very soon after cell activation
and are potent stimuli for activation of
similar cells locally and distally. IL-6 is
the major acute-phase stimulant induc-
ing de novo hepatic acute phase protein
synthesis. As part of the acute phase
response, cytokine induced synthesis of
complement and coagulation factors,
grouped
for example, may contribute to the
amplification of these plasma protein
cascades. Amongst other important
cytokines, Interleukin-8 (IL-8) is a potent
neutrophil chemoattractant; IL-12 from
monocytes and interferon g produced
by T cells form a paracrine amplification
loop; platelet activating factor (PAF) acts
on endothelia, increasing vascular per-
meability, augmenting the clotting cas-
cade and the production of an inflam-
matory exudate. PAF is also a neutrophil
chemoattractant and induces superox-
ide generation and degranulation11.
Many non-cytokine
derived from activated cells are also
evident. Nitric oxide (NO), originally
identified as endothelium-derived
relaxing factor, is produced by the enzy-
matic action of nitric oxide synthase
(NOS) on l-arginine. Different isoforms
of NOS are classified by virtue of consti-
tutive (cNOS) or inducible (iNOS) expres-
sion. cNOS is produced by endothelia
and regulates vascular tone, whilst
iNOS, produced primarily by monocytes
and macrophages in response to
microbial products and activation by
cytokines, causes NO-mediated vaso-
dilatation and increases endothelial
permeability, thereby enhancing inflam-
matory exudates12. NO also reacts with
superoxide anions generated by respira-
tory bursts in phagocytic cells to form
peroxynitrite which has direct anti-
microbial activity. Other mediators
released by activated cells include tissue
thromboplastin, which activates the
extrinsic clotting cascade, histamine
and eicosanoids such as prostaglandins
mediators
(PGs) which similarly enhance inflam-
matory exudates.
Numerous negative
mechanisms are employed by the host
to regulate amplification of the innate
immune response (Fig 3). Severe sepsis
is thought to occur when the pro-inflam-
matory cascade escalates beyond the
control of anti-inflammatory regulation.
feedback
Pathogenesis of clinical features in
severe sepsis
Fever
Thermoregulation is achieved
balancing heat production and heat loss
strategies under control of the auto-
nomic nervous system to maintain a
hypothalamic temperature set-point. In
sepsis, the cytokines IL-1, IL-6 and TNF-
a stimulate endothelia and astrocytic
terminals in the perivascular spaces of
the hypothalamic organum vasculosum
laminae terminalis to produce PGE213.
This, in turn, inhibits warm-sensitive
neurons and raises the hypothalamic
temperature set-point. The autonomic
nervous system then increases heat
production and reduces heat loss until
the new set-point has been reached. In
severe sepsis, hypothermia can super-
vene for reasons that are not clear, but
may reflect decompensation
thermoregulatory mechanisms rather
than a low hypothalamic temperature
set-point. Heat loss increases with
peripheral vasodilatation, and heat
production decreases as organ failure
ensues.
by
of
©
Sepsis is a systemic inflammatory response syndrome resulting from an
infection which is usually bacterial
©
Highly conserved, non-specific host-microbial interactions initiate cellular and
humoral innate immune responses
©
In severe sepsis, pro-inflammatory amplification mediated by cell-cell, cytokine
and plasma protein cascades exceeds the control of anti-inflammatory
regulation
©
Cellular, cytokine and plasma protein components of innate immune responses
are directly implicated in the pathogenesis of fever, hypotension and organ
dysfunction
Key Points

Page 4

CME Septicaemia – I
Hypotension
Many of the products of cellular and
plasma protein activation, in particular
NO, bradykinin, complement anaphylo-
toxins, prostacyclin, histamine and PAF
induce vasodilatation in the micro-
vascular circulation by a direct effect on
vascular smooth
Inflammatory exudates also cause local
tissue hypoxia and hypoperfusion.
Homeostatic neurohormonal mecha-
nisms attempt to compensate, produc-
ing a hyperdynamic circulation typical
of sepsis, characterised by decreased
systemic vascular
increased cardiac output. In severe
sepsis, cardiac output falls, in part due
to hypovolaemia
increased capacitance of dilated blood
vessels and fluid extravasation due to
increased vascular
Cardiogenic shock is also evident, due
to poorly characterised myocardial
depressant factors, one component of
which is NO-mediated inhibition of
positive inotropic and chronotropic
responses to b-adrenergic stimulation14.
Coronary hypoperfusion and inflam-
matory infiltration of the myocardium
also contribute to the immuno-
pathogenesis.
muscle cells.
resistance and
resulting from
permeability.
Organ dysfunction
The typical manifestations of organ
dysfunction are listed in Table 2. Their
pathogenesis in sepsis is complex,
multifactorial and incompletely under-
stood15. Tissue hypoperfusion and
hypoxia are dominant factors. The
mechanisms for this include widespread
fibrin deposition in DIC causing
microvascular occlusion, inflammatory
exudates, and vasoactive substances
resulting in abnormalities of micro-
vascular homeostasis and systemic
hypotension. The metabolic acidosis
that supervenes further disables normal
biochemical processes.
infiltrates, particularly
damage tissues directly by releasing
lysosomal enzymes and superoxide-
derived free radicals which peroxidate
cell membrane lipids. As tissue injury
progresses and organ function is com-
Cellular
neutrophils,
Journal of the Royal College of Physicians of London Vol. 34 No. 5 September/October 2000435
TissueClinico-pathological features
BrainCerebral infarction
Diffuse micro-abscesses
Encephalopathy
Myocardial inflammatory exudate
Myocardial depression
Arrhythmias
Diffuse alveolar inflammatory exudate
Acute lung injury
Acute respiratory distress syndrome
Acute tubular necrosis
Hepatocellular necrosis
Ischaemic necrosis
Perforation
Purpura
Ischaemic necrosis
Heart
Lung
Kidney
Liver
Gastrointestinal tract
Skin
Table 2. Common clinico-pathological manifestations in severe sepsis.
Fig 3. Anti-inflammatory mediators and their actions (IL = interleukin;
LPS = lipopolysaccharide; MIF = macrophage migration inhibitory factor;
TAF = transforming growth factor; TNF = tumour necrosis factor).
a
b

Page 5

CME Septicaemia – I
promised, for example by exacerbation
of hypoxia due to lung injury, sequential
multiple organ failure follows. Organ
dysfunction in sepsis is covered in more
detail in the accompanying article by
Singer.
Conclusion
Whilst an infecting organism may
produce toxins which injure tissues
directly, this is often inadequate to
explain the clinico-pathological
sequelae in severe sepsis. Instead, the
dominant role in pathogenesis may lie
with components of the host immune
response to infection. The highly
conserved responses of the innate
immune system comprise sequential
activation and amplification of humoral
and cellular antimicrobial defence
mechanisms which can escape the
control of anti-inflammatory regulation,
inadvertently causing injury to the host.
Further understanding of the immuno-
pathogenesis of severe sepsis may
unveil new opportunities for therapeutic
intervention.
References
1 Rangel-Frausto MS. The epidemiology of
bacterial sepsis. Review. Infect Dis Clin
North Am 1999;13:299–312, vii.
2 Morrison DC, Ryan JL. Endotoxins and
disease mechanisms. Annu Rev Med
1987;38:417–32.
3 Medshitov R, Janeway C. Innate immun-
ity. N Engl J Med 2000;343:338–44.
4 Sriskandan S, Cohen J. Gram-positive
sepsis. Mechanisms and differences from
gram-negative sepsis. Infect Dis Clin North
Am 1999;13:397–412.
5 Bernal A, Proft T, Fraser JD, Posnett DN.
Superantigens in human disease. J Clin
Immunol 1999;19:149–57.
6 Takeuchi O, Hoshino K, Kawai T, Sanjo H,
et al. Differential roles of TLR2 and TLR4
in recognition of gram-negative and
gram-positive bacterial cell wall compo-
nents. Immunity 1999;11:443–51.
7 Brown EJ, Lindberg FP. Leucocyte
adhesion molecules in host defence
against infection. Review. Ann Med
1996;28:201–8.
8 Mammen EF. The
manifestations of sepsis. J Antimicrob
Chemother 1998;41(Suppl A):17–24.
9 Haeney MR. The role of the complement
cascade in sepsis. J Antimicrob Chemother
1998;41(Suppl A):41–6.
10 van der Poll T, van Deventer SJ.
Cytokines and anticytokines in the
haematological
436
Journal of the Royal College of Physicians of London Vol. 34 No. 5 September/October 2000
pathogenesis of sepsis. Review. Infect Dis
Clin North Am 1999;13:413–26, ix.
11 Mathiak G, Szewczyk D, Abdullah F,
Ovadia P, Rabinovici R. Platelet-activat-
ing factor (PAF) in experimental and
clinical sepsis. Shock 1997;7:391–404.
12Parratt JR. Nitric oxide in sepsis and
endotoxaemia. J Antimicrob Chemother
1998;41(Suppl A):31–9.
13Dinarello CA. Cytokines as endogenous
pyrogens. J Infect Dis 1999;179(Suppl
2):S294–304.
14Brady AJ. Nitric oxide, myocardial failure
and septic shock. Int J Cardiol 1995;
50:269–72.
Thijs LG, Groeneveld AB, Hack CE.
Multiple organ failure in septic shock.
Curr Top Microbiol Immunol
216:209–37.
15
1996;
Address for correspondence: Professor J
Cohen, Hammersmith Hospital, Du Cane
Road, London W12 0NN.
E-mail: j.cohen@ic.ac.uk
Pathophysiology and management of
meningococcal septicaemia
Neisseria meningitidis (meningococcus) is
a major infection risk globally. In the
UK, it is the leading cause of death from
infection in childhood1, with a mortality
around 10%2. Most deaths from
meningococcal infection are due to the
development of
shock2,3. Yet N. meningitidis is a frequent
commensal of the human upper
respiratory tract. Carriage rates increase
from less than 1% in infancy to a
maximum of 25% in adolescence,
declining to around 10% in adulthood.
The meningococcus is a Gram-
negative diplococcus.
meningococci possess a polysaccharide
capsule, differences in the structure of
which form the basis of separation into
subgroups. The lack of suitable vaccines
for all the meningococcal serogroups is
because of a high level of genetic
diversity caused by intraspecies re-
combination and transformation. A
single mutation or genetic exchange
fulminant septic
Pathogenic
may lead to an outbreak of clinical
disease if associated with a change in
an immunologically important surface
antigen.
Epidemiology
Meningococcal disease is endemic
worldwide. Serogroups B and C pre-
dominate in the UK with an incidence of
5–6 per 100,0002,4. In sub-Saharan
Africa, serogroup A predominates in
cyclical epidemics every eight years and
can affect up to 1,000 per 100,000 of
the population5. The reasons for
regional variation in disease-causing
serogroups are not well defined.
Immunopathology
Transmission
Transmission is by close contact or
respiratory droplet spread.
Colonisation and invasion of
nasopharyngeal epithelium
The risk of colonisation may be
enhanced by disruption
respiratory epithelial cell layer by
irritants (such as cigarette smoke) or by
a preceding viral illness, for example
influenza A6. Binding to epithelial cells is
established by pili7,8
membrane proteins9. Certain outer
of the
and outer
Nazima Pathan BSc MRCP MRCPCH,
Clinical Research Fellow
Simon Nadel MRCP MRCPCH,
Consultant, Paediatric Intensive Care Unit
Michael Levin PhD FRCP FRCPCH,
Professor of Paediatrics
Imperial College School of Medicine at
St Mary’s Hospital, London
J R Coll Physicians Lond 2000;34:436–44