You are on page 1of 13

REVIEWS

Sepsis-induced immunosuppression:
from cellular dysfunctions to
immunotherapy
Richard S.Hotchkiss1, Guillaume Monneret2 and Didier Payen3
Abstract | Sepsis which is a severe life-threatening infection with organ dysfunction initiates
a complex interplay of host pro-inflammatory and anti-inflammatory processes. Sepsis can be
considered a race to the death between the pathogens and the host immune system, and it is
the proper balance between the often competing pro- and anti-inflammatory pathways that
determines the fate of the individual. Although the field of sepsis research has witnessed the
failure of many highly touted clinical trials, a better understanding of the pathophysiological basis
of the disorder and the mechanisms responsible for the associated pro- and anti-inflammatory
responses provides a novel approach for treating this highly lethal condition. Biomarker-guided
immunotherapy that is administered to patients at the proper immune phase of sepsis is
potentially a major advance in the treatment of sepsis and in the field of infectious disease.

Sepsis is defined as the host inflammatory response to in cancer and those observed in sepsis and because both
severe, life-threatening infection with the presence of agents result in improved survival in animal models
organ dysfunction1. Sepsis is the most frequent cause of sepsis7,10. In this Review, we discuss the panoply of
of mortality in most intensive care units (ICUs) and sepsis-induced defects in innate and adaptive immune
is responsible for more than 250,000 deaths in the cells and highlight several promising immunotherapies
United States annually 2. The incidence of sepsis is for the treatment ofsepsis.
increasing owing to the ageing population, who have
1
Department of
impaired immunity as a result of immunosenescence2. Controversies over host immunity in sepsis
Anesthesiology, Medicine, Despite the litany of failed clinical trials in sepsis, a The current paradigm regarding the host immune
and Surgery, Washington better understanding of different immunological phases response to sepsis is debated27,18,19. Traditionally, the
University School of Medicine, of the disorder and encouraging results from several host immune response to sepsis was considered to be
St Louis, Missouri 63110, USA.
PhaseII clinical trials of immunotherapies in sepsis characterized by an initial hyperinflammatory phase
2
Hospices Civils de Lyon,
Immunology Laboratory, have brought cautious optimism to the field37. that evolved over several days into a more protracted
Hspital E, Herriot, 5, Place Until recently, most research on sepsis was focused immunosuppressive phase79 (theory 1). However, recent
dArsonval, Lyon 69003, on blocking the initial hyperinflammatory, cytokine- studies have shown that both pro-inflammatory and anti-
France. mediated phase of the disorder. Improved treatment inflammatory responses occur early and simultaneously
3
Department of
Anesthesiology & Critical &
protocols have resulted in most patients surviving this in sepsis1820 (FIG.1a), although the net initial effect of these
Service Mobile dUrgence initial hyperinflammatory phase and entering a pro- competing processes is typically manifested by an early,
Rapide (SMUR), Hpital tracted immunosuppressive phase813. Deaths in this dominant, hyperinflammatory phase characterized by
Lariboisire, Assistance immunosuppressive phase are typically due to failure to shock, fever and hypermetabolism. The robustness of the
Publique-Hpitaux de Paris
control the primary infection or the acquisition of second- hyperinflammatory phase depends on numerous factors,
and Universit Paris 7 Denis
Diderot, 2 rue Ambroise Par ary hospital-acquired infections, often with opportunistic including pre-existing comorbidities, nutritional status,
75010, Paris, France. pathogens14,15. The recent remarkable success of cytotoxic microorganism load and virulence factors8,9.
Correspondence to R.S.H. T lymphocyte antigen 4 (CTLA4)- and programmed cell Investigators recently reported a new paradigm
e-mail: death 1 (PD1)-specific antibodies as immunotherapies (FIG. 1b) to describe the host immune response in
hotchkir@anest.wustl.edu
doi:10.1038/nri3552
to improve host immunity and to increase survival in trauma and sepsis (theory 2). Circulating leukocyte
Published online cancer patients16,17 is highly encouraging because of the gene expression data from people suffering from trauma
15 November 2013 many similarities between the immune defects observed and burns showed the rapid and sustained upregulation

862 | DECEMBER 2013 | VOLUME 13 www.nature.com/reviews/immunol

2013 Macmillan Publishers Limited. All rights reserved


REVIEWS

a Theory 1 innate immune system with resultant organ dysfunc-


Early deaths due
to overwhelming tion and failure. Although these investigators agree that
inammation the adaptive immune system is impaired, they theorize
Hyper- that patients who die of sepsis have a longer duration
inammatory Innate and a greater degree of organ injury caused by unabated
response innate immune-driven inflammation than those who
survive. They postulate that this inflammation persists
despite the downregulation of the expression of genes
Homeostasis Recovery that regulate the adaptive immune response and that
2 4 6 8 10 12 it is ultimately responsible for patient morbidity and
Time (days)
mortality 19.
Adaptive Although we agree with the provocative findings of
Immuno- Late deaths due to this group, we believe that this new model which pro-
suppression persistent poses that morbidity and mortality in sepsis are due to
immunosuppression unremitting innate immune-driven inflammation is
and recurrent
infections unlikely to reflect the actual clinical scenario in most
patients. We have shown that patients who die of sep-
sis have marked immunosuppression12. Immune cells
b Theory 2 Early deaths due that were harvested from the spleen or lungs of patients
to overwhelming with sepsis within 30 to 180minutes of death showed
inammation markedly decreased production of both pro- and anti-
Hyper- Late deaths due inflammatory cytokines, upregulated expression of
inammatory Innate
to intractable inhibitory receptors (including PD1), expansion of
response inammation- regulatory T (TReg) cell and myeloid-derived suppressor cell
induced organ injury
(MDSC) populations, and downregulation of CD28 and
HLADRmediated activation pathways12. Collectively,
Homeostasis Recovery these results show that sepsis induces numerous over-
2 4 6 8 10 12
Time (days)
lapping mechanisms of immunosuppression involving
both the innate and adaptive immune systems.
Adaptive
There are several potential explanations for the
Immuno- different findings in the two studies. First, the results
suppression
of the gene expression study involved patients with
trauma and burns 19, and although many of these
patients became septic, these patients differ from those
who developed sepsis as a primary phenomenon. The
Figure 1 | Competing theories of the host immune response in sepsis. a | Theory 1: mean age of the patients with trauma in the transcrip-
recent studies show that activation of both pro- and anti-inflammatory
Nature Reviewsimmune
| Immunology tome study was 33years, whereas the mean age of
responses occurs promptly after sepsis onset. Cells of the innate immune system, patients with sepsis in most developed countries is at
including monocytes and neutrophils, release high levels of pro-inflammatory cytokines
least twice this age (6568years). This difference in
that drive inflammation (blue line; days 13). The intensity of the initial inflammatory
response varies in individual patients depending on multiple factors, including pathogen age in the two studies is important because the older
load and virulence, patient comorbidities and host genetic factors. Early deaths in sepsis patients have underlying defects in immunity due to
(top red line; day 3) are typically due to a hyperinflammatory cytokine storm response immunosenescence21. Second, these investigators based
with fever, refractory shock, acidosis and hypercatabolism. Anexample of this scenario their conclusions on gene expression data that, because
would be a young patient dying of toxic shock syndrome or meningococcaemia. Most of post-transcriptional regulation, might not reflect
patients have a restoration of innate and adaptive immunity and survive the infection protein levels. Conversely, the post-mortem study of
(recovery; blue and green lines; day 6). If sepsis persists, the failure of crucial elements patients with sepsis quantitated actual cytokine levels.
of both the innate and the adaptive immune systems occurs such that patients enter a Other possible reasons for differences between these
marked immunosuppressive state (blue and red lines; after day 6). Deaths are due to an two studies are the duration and severity of illness of
inability of the patient to clear primary infections and the development of secondary
the patients who were analysed. Many patients with
infections. b|Theory2: a competing theory of sepsis agrees that there is an early
activation of innate immunity and a suppression of adaptive immunity; however, this sepsis had protracted disease with a high severity of
theory holds that deaths in sepsis are due to the persistent activation of innate immunity illness and increased mortality.
that results in intractable inflammation and organ injury. According to this theory, late We believe that immunosuppression rather than
deaths in sepsis are due to persistent, underlying innate immune-driven inflammation. low-grade inflammation is the predominant driving
force for morbidity and mortality in sepsis. First, our
post-mortem findings are consistent with numerous
of genes that regulate the innate immune response and studies that have examined cytokine production in
the simultaneous downregulation of genes that regu- peripheral-blood mononuclear cells and whole blood
late the adaptive immune response19. These investiga- from patients with sepsis and have shown severely
tors hypothesized that the best model to describe the decreased pro-inflammatory cytokine production2225.
host immune response in trauma and sepsis is one Second, recent post-mortem studies have reported
of protracted, unabated inflammation driven by the that a high number of patients who die of sepsis have

NATURE REVIEWS | IMMUNOLOGY VOLUME 13 | DECEMBER 2013 | 863

2013 Macmillan Publishers Limited. All rights reserved


REVIEWS

unresolved opportunistic infections14,15, which is consist- death receptor antagonists that preventing lymphocyte
Immunosenescence ent with defective host immunity as a dominant cause of apoptosis improves survival in sepsis3437. The detrimen-
The decreased function of death. Third, the failure of more than 30 clinical trials of tal effects of apoptosis are not only related to the severe
the immune system with age. different anti-inflammatory agents is inconsistent with loss of immune cells but also the impact that apoptotic
In particular, the number of
naive Tcells decreases as
the hypothesis that inflammation is a key driving mecha- cell uptake has on the surviving immune cells38. Uptake
thymic function decreases. nism. Undoubtedly, focal regions of inflammation occur of apoptotic cells by monocytes, macrophages and DCs
in infected tissue in patients with sepsis. Areas of ischae- results in immune tolerance by inducing anergy or a
Myeloid-derived suppressor mic and necrotic tissue are likely to contribute to local Thelper 2 (TH2) cell-associated immune phenotype with
cell
inflammation. It is possible that this focal inflammation increased interleukin10 (IL10) production39. The net
(MDSC). A subset of immune
cells that includes mature and contributes to organ dysfunction, morbidity and mor- result of this effect is that the surviving phagocytic cells
immature myeloid cells. They tality in a subset of patients with sepsis. Nevertheless, cannot combat the remaining pathogenic organisms.
are generated and/or activated we hypothesize that protracted sepsis is predominantly Thus, sepsis-induced apoptosis has multiple important
during an inflammatory characterized by systemic immunosuppression leading effects that impair host defences.
immune response. Through
direct interactions and
to a failure to eradicate primary infections and to the
secreted components, they acquisition of lethal secondary infections. Intriguingly, Impact of sepsis on immune cells
negatively affect Tcells, which the gene expression data from children who develop Sepsis directly or indirectly impairs the function of
leads to impaired Tcell sepsis show similarities to the data from adult patients almost all types of immune cells. The following sec-
function.
with trauma-related sepsis: for example, the expression tion provides an overview of these various immuno
Follicular dendritic cells of genes that modulate innate immunity is upregulated, suppressive effects on the different cells of the innate
(Follicular DCs). Cells with a whereas the expression of genes that modulate adap- and adaptive immune systems.
dendritic morphology that are tive immunity is downregulated26. Although these gene
present in the lymph nodes. expression data would seem to favour theory 2, post- Neutrophils and MDSCs. Neutrophils are essential for
These cells display on their
surface intact antigens that are
mortem studies of paediatric patients who die of sepsis the early control of invading pathogens40. Investigations
held in immune complexes, have shown immunosuppressive features that are char- on neutrophils obtained during the first hours of sepsis
and Bcells in the lymph acteristic of those observed in adult patients with sepsis, uncovered numerous abnormalities (FIG.3). Most neu-
nodes can interact with these who show a remarkable depletion of immune effector trophils normally undergo apoptosis within 24hours
antigens. Follicular DCs are of
cells (see below). Hopefully, future studies will deter- after release from the bone marrow 40. Surprisingly, and
non-haematopoietic origin
and are not related to DCs. mine which of these two conflicting hypotheses is cor- in contrast to lymphocytes, which undergo accelerated
rect, as the resolution of this controversy has important apoptosis, neutrophil apoptosis is delayed during sep-
Interdigitating DCs therapeutic implications. sis40. Because of an increase in the release of immature
Potent antigen-presenting cells neutrophils and the delayed apoptosis of circulating
that are rich in MHC classII
molecules. Interdigitating DCs
Apoptosis and immunosuppression neutrophils, patients with sepsis typically have mark-
take up antigens in the A fundamental discovery in the field of sepsis occurred edly increased numbers of circulating neutrophils
periphery and migrate to the when investigators showed that apoptosis causes the of various degrees of maturation41. Animal models of
paracortical region of the marked depletion of immune cells, including CD4+ sepsis and experimental studies in patients revealed
lymph nodes and the spleen,
and CD8+ Tcells, Bcells, follicular dendritic cells (folli- disrupted neutrophil functions, including impaired
where they interact with Tcells.
cular DCs) and interdigitating DCs, in various organs of clearance of bacteria, reduced production of reac-
Gut-associated lymphoid patients dying of sepsis, leading to immunosuppres- tive oxygen species (ROS) and decreased recruitment
tissues sion2731 (FIG.2). Sepsis-induced immune cell apoptosis to infected tissues42,43. Loss of chemotactic activity is
(GALTs). Lymphoid structures has now been confirmed in several post-mortem studies; probably the most frequently documented dysfunction
and aggregates that are
associated with the intestinal
it affects all age groups (neonatal, paediatric and adult of circulating neutrophils during sepsis43,44. Possible
mucosa, specifically the tonsils, populations)27,30,31 and occurs in response to various explanations for this include nitric oxide-mediated
Peyers patches, lymphoid microorganisms. Apoptosis of immune cells occurs in suppression and reduced expression of CXC-chemokine
follicles, appendix or coecal lymphoid tissues (spleen, thymus and lymph nodes) receptor 2 (CXCR2)44. A current theory is that these
patch, and mesenteric lymph
and gut-associated lymphoid tissues (GALTs)32. The loss alterations in neutrophil function are due to abnor-
nodes. They are enriched
in conventional and in intestinal intraepithelial and lamina propria lympho- malities in Toll-like receptor (TLR) signalling, which is
unconventional lymphocytes cytes might facilitate bacterial translocation into the analogous to the phenomenon that occurs in monocytes
and specialized dendritic cell systemic circulation, thereby perpetuating the systemic during endotoxin tolerance43 (see below).
and macrophage subsets. inflammatory response and predisposing to secondary Several studies have reported that impaired neutrophil
Anergy
infections. Sepsis-induced apoptosis occurs through function precedes the development of nosocomial infec-
A state of immune both death receptor- and mitochondrial-mediated tions. Patients with the most severely reduced neutrophil
unresponsiveness. Anergic pathways, suggesting that multiple cell death stimuli functions have the highest risk of acquiring nosocomial
Bcells or Tcells do not respond are activated during sepsis33. Therefore, it is unlikely infections45. Reduced neutrophil function has also been
to their cognate antigens.
that blockade of a single apoptotic trigger will prevent reported in a mouse model of polymicrobial sepsis, in
Endotoxin tolerance lymphocyte cell death during the disorder. which the mice had an increased susceptibility to second-
A transient state of A key question is whether the extensive sepsis- ary infection with Pseudomonas aeruginosa and the sub-
hyporesponsiveness of the host induced apoptosis of immune cells is an epiphenome- sequent development of pneumonia46. Furthermore, ten
or of cultured macrophages non or a major pathophysiological mechanism. Multiple patients with septic shock who underwent extracorporeal
and/or monocytes to
lipopolysaccharide (an
independent laboratories have shown through the use cell therapy with donor granulocytes showed improve-
endotoxin) following previous of various strategies, including transgenic and knockout ment in various biomarkers of sepsis, as well as decreased
exposure to lipopolysaccharide. mice, anti-apoptotic cytokines, caspase inhibitors and sepsis severity 47.

864 | DECEMBER 2013 | VOLUME 13 www.nature.com/reviews/immunol

2013 Macmillan Publishers Limited. All rights reserved


REVIEWS

Sepsis Trauma

Splenic CD4+ T cells


Splenic CD8+ T cells
Splenic HLA-DR expression

Figure 2 | Immunohistochemical staining for CD4+ Tcells, CD8+ Tcells and HLADR in spleens from patients with
sepsis or trauma. Spleens from patients with sepsis or patients who required a splenectomy for traumatic
Nature Reviews injury were
| Immunology
obtained and underwent immunostaining for CD4+ T cells, CD8+ Tcells or HLADR. Sepsis induces the apoptotic death of
splenocytes, including CD4+ and CD8+ Tcells, leading to marked depletion of these crucial immune effector cells. Note
the decrease in numbers of periarteriolar CD4+ and CD8+ Tcells (stained brown) in spleens from patients with sepsis
compared with patients with trauma (original magnification 200). The decrease in numbers of CD4+ Tcells in patients
with sepsis is frequently as severe as occurs in patients with AIDS. The number of HLADR+ cells (predominantly Bcells)
and the levels of HLADR expression (as determined by the intensity of the staining) are both decreased in patients with
sepsis compared with patients with trauma (HLADR+ cells are stained brown). Tissue samples were stained for CD4,
CD8 or HLA-DR. 3,3-diaminobenzidine 4-HCl was used as a chromogen to stain the cells of interest (brown), and a
Extracorporeal cell therapy
A therapy whereby the
hematoxylin counterstain (blue) was used for background staining. Images courtesy of P. Swanson, University of
patients blood is circulated Washington School of Medicine in Seattle, Washington, USA.
through a haemodialysis-like
device through which particular
types of cells, for example, Neutrophils show great plasticity in response to a wide Recently, a subset of mature CD16hiCD62L (also known
donor granulocytes, are added range of physiological or pathological conditions43. Sepsis as L-selectin)low neutrophils that suppressed Tcell prolif-
to the blood and subsequently can induce the formation of a subset of neutrophils with eration was identified in a human model of endotoxae-
infused back into the patient. suppressive properties. Although neutrophils are often mia in which healthy volunteers were administered with
Extracorporeal cell therapy of
autologous granulocytes has
not considered to produce large amounts of cytokines, a low dose of endotoxin49. Immunosuppressive neutro-
shown success in small clinical it has been shown that they produce large amounts of phils were also detected in five patients suffering from
trials in sepsis. the immunosuppressive cytokine IL10 during sepsis48. severe trauma49.

NATURE REVIEWS | IMMUNOLOGY VOLUME 13 | DECEMBER 2013 | 865

2013 Macmillan Publishers Limited. All rights reserved


REVIEWS

a Eects of protracted sepsis on the innate immune system

Follicular Dendritic cell Macrophage NK cell Neutrophil MDSC


dendritic cell
Apoptosis Apoptosis Anti-inammatory Apoptosis Release of immature Apoptosis
Antigen Antigen cytokine secretion Cytotoxic neutrophils Cytotoxic function
presentation presentation HLA-DR expression function IL-10 secretion Cytokine secretion
to B cells to T cells Pro-inammatory Cytokine Apoptosis
Cytokine cytokine secretion secretion Reactive oxygen
secretion Pathogen killing species release
Nitric oxide release
Expression of
adhesion markers
b Eects of protracted sepsis on the adaptive immune system

CD4+ T cell CD8+ T cell TReg cell B cell


Cell exhaustion Cell exhaustion Resistance to apoptosis Apoptosis
Apoptosis Apoptosis Suppressive activities Antigen-specic
TH2 cell polarization Cytotoxic function antibody production
Adhesion molecule Cytokine secretion
expression TCR diversity
CD28 expression
TCR diversity

Figure 3 | Impact of sepsis on innate and adaptive immune cells. a | Sepsis has diverse and important effects
on all cellular elements that comprise the innate immune system. Sepsis rapidly triggers extensive apoptosis in
follicular dendritic cells, dendritic cells, immature macrophages, natural killer (NK) cells and myeloid-derived
suppressor cells (MDSCs). Conversely, sepsis delays neutrophil apoptosis a result that is thought to be secondary
Nature Reviews | Immunology
to the mechanisms of neutrophil activation. After initial mobilization and activation of neutrophils, subsequent
neutrophils that are released from the bone marrow have lower bactericidal functions and decreased cytokine
production. Recent data show that a subset of neutrophils release large amounts of the immunosuppressive
cytokine interleukin10 (IL10). Decreased HLADR expression on antigen-presenting cells, including monocytes,
macrophages and dendritic cells, is a hallmark of sepsis, which may impair the optimal presentation of microbial
antigens to Tcells. b|Sepsis causes a marked loss of CD4+ and CD8+ Tcells, as well as Bcells. Regulatory T (TReg) cells
are more resistant to sepsis-induced apoptosis and, consequently, there is an increased percentage of TReg cells in
the circulation of patients with sepsis relative to the other lymphocyte subsets. This contributes to the formation
of a more immunosuppressive phenotype. Surviving CD4+ and CD8+ Tcells either shift from a pro-inflammatory
Thelper 1 (TH1) cell phenotype to an anti-inflammatory TH2 cell phenotype or develop an exhaustive phenotype
that is characterized by increased programmed cell death1 expression and reduced cytokine secretion. CD4+ Tcells
have decreased expression of CD28 and reduced Tcell receptor (TCR) diversity, which are both likely to contribute
to the impaired antimicrobial response to invading pathogens.

The recently investigated heterogeneity in neutro DCs. DCs are particularly vulnerable to sepsis-induced
phils, including immunosuppressive subsets of apoptosis29. In a post-mortem study of patients with
myeloid cells, offers parallels with MDSCs. In experi- trauma or with sepsis, researchers found a marked
mental models of sepsis, these cells have been shown reduction in the number of splenic DCs and the per-
to block specific Tcell functions, including prolifera- centage area of the spleen that was occupied by DCs in
tion and production of interferon (IFN) and IL2 the patients with sepsis compared with patients with
(REFS 50,51) . However, the actual effect of MDSCs trauma29. Similarly, reduced numbers of circulating DCs
might evolve during sepsis, as studies show that have been recorded in patients suffering from burns and
MDSCs can either enhance or attenuate the sepsis- subsequent sepsis and in patients with severe sepsis5557.
associated inflammatory response, depending on the Both plasmacytoid and myeloid DCs are effected by
stage of the disorder 5053. Our group recently showed sepsis-induced apoptosis55,58 (FIG.3). Interestingly, DC
that there was an increase in the number of cells that loss was more apparent in patients with sepsis who
were consistent with an MDSC phenotype in the lungs died than in survivors55, and it was also more marked
of patients who died of sepsis12. Currently, few clini- in patients who subsequently developed nosocomial
cal studies have investigated MDSCs in patients with infections than in those patients who didnot 59.
sepsis, possibly owing to the complexity of immuno Not only are DC numbers decreased in patients with
phenotyping these cells and the absence of a univer- sepsis, but the surviving DCs also have lower expres-
sally accepted phenotypic definition for these cells in sion of HLADR and produce increased levels of IL10
humans54. (REFS57,60). In addition, monocyte-derived DCs from

866 | DECEMBER 2013 | VOLUME 13 www.nature.com/reviews/immunol

2013 Macmillan Publishers Limited. All rights reserved


REVIEWS

patients with sepsis were unable to induce a robust The mechanisms responsible for endotoxin tolerance
effector Tcell response but instead induced either Tcell are not fully understood70,72. Analysis of mRNA expres-
anergy or TReg cell proliferation61. In mouse models of sion levels has shown increased expression of genes
burn injury, DCs possess immunosuppressive proper- encoding inhibitory signalling molecules and inhibitory
ties that impair defences against a subsequent bacte- cytokines and reduced expression of pro-inflammatory
rial challenge. Several investigators have shown that and chemokine receptor genes7174. Although they have
preventing sepsis-induced DC apoptosis or improving mainly been reported as useful biomarkers for pre-
DC function during the disorder results in enhanced dicting sepsis onset or prognosis, microRNAs are also
survival6264. Indeed, mice that had selective overex- thought to participate in endotoxin tolerance71. Similarly,
pression of the anti-apoptotic factor Bcell lymphoma recent work highlights a prominent role for epigenetic
2 (BCL2) in DCs were reported to have improved sur- regulation at different levels, such as histone modifica-
vival in a lethal model of endotoxic shock62. Importantly, tion and chromatin remodelling 75,76. These epigenetic
these BCL2overexpressing DCs were resistant to modifications are restored in vitro after IFN stimula-
sepsis-induced apoptosis, which suggests that DC tion and are associated with the recovery of monocyte
death is an important determinant of sepsis-induced cytokine release76,77. Thus, IFN might abrogate endo-
immunosuppression and mortality. toxin tolerance by facilitating TLR-induced chromatin
FMS-related tyrosine kinase 3 ligand (FLT3L) is a DC remodelling 78.
growth factor that induces a rapid increase in DC num- Two major consequences of endotoxin tolerance
bers. Treatment of burn-injured animals with FLT3L on monocytes and macrophages are an increase in
restored DC function and enhanced survival when mice the release of immunosuppressive mediators (mainly
were challenged with P.aeruginosa, which is a patho- IL10) and a decrease in antigen presentation as a result
gen that commonly infects patients with burns6365. of reduced expression of HLADR; both are associated
The protective effect of FLT3L was also observed fol- with a worse outcome in sepsis22,7981. Continued release
lowing the adoptive transfer of FLT3Ltreated DCs. of IL10 might contribute to or amplify sepsis-induced
Additional studies documented that treatment with immunosuppression and thus might augment suscep-
FLT3L increased the secretion of IL12, IL15 and IFN tibility to secondary microbial infections8284. Indeed,
and had broad effects on the function of CD4+ Tcells, blocking IL10 reverses endotoxin tolerance exvivo83,85
natural killer (NK) cells and neutrophils in models of and can reverse sepsis-induced immunosuppression and
burn infection6365. Intrapulmonary transfer of bone improve survival in a clinically relevant animal model
marrow-derived DCs restored DC function and pre- of sepsis86.
vented fatal Aspergillus fumigatus infection in mice that Low levels of monocyte HLADR expression function
had recovered from primary peritonitis66. TLR agonists as a surrogate marker of monocyte unresponsiveness79,87.
have also been shown to improve survival from pneu- Several studies showed an association of reduced mono-
monia occurring in mice after haemorrhagic shock67,68. cyte HLADR expression with impaired monocyte func-
Possible mechanisms for the beneficial effect of TLR tion (for example, lower levels of TNF and IL1 release
agonists include the increased expression of MHC from monocytes in response to bacterial challenges88
classII molecules and of the co-stimulatory receptors and decreased lymphocyte proliferation in response to
CD80 and CD86 on DCs67,68. Given these encourag- tetanus toxin, presumably due to impaired antigen
ing results, some authors have argued that preserving presentation89). Most importantly, decreased monocyte
and/or restoring DC function should be a primary target HLADR expression is associated with increased risk of
of investigations into sepsis6369. nosocomial infections and death90,91. It is noteworthy that
after adjustment for usual clinical confounders through
Monocytes and macrophages. A hallmark of sepsis is multivariate analysis, decreased monocyte HLADR
FMS-related tyrosine the diminished capacity of monocytes to release pro- expression remains an independent predictor of noso-
kinase 3 ligand
inflammatory cytokines in response to endotoxin comial infection occurrence and mortality after sepsis73,92.
(FLT3L). An endogenous
cytokine that stimulates the (lipopolysaccharide (LPS)), other TLR agonists and vari- This finding shows that monocyte unresponsiveness and
proliferation of stem and ous other bacterial compounds70,71 (FIG.3). This finding is immunosuppression independently contribute to the
progenitor cells through consistent with the phenomenon of endotoxin tolerance. increased risk of adverse events insepsis.
binding to the FLT3 receptor Monocytes from patients with sepsis typically show a
(which is a typeIII receptor
tyrosine kinase member of the
diminished ability to release pro-inflammatory cytokines NK cells. The preferential location of NK cells in tis-
platelet-derived growth factor such as tumour necrosis factor (TNF), IL1, IL6 and sues along with their low numbers in peripheral blood
family). FLT3L administration IL12, whereas their ability to release anti-inflammatory render the study of this cell subset difficult, especially
substantially increases the mediators, such as IL1 receptor antagonist (IL1RA) in the context of human sepsis. These facts explain why
number of dendritic cells in
and IL10, is either unimpaired or enhanced70,71. These NK cells have not been heavily investigated in sepsis93,94.
lymphoid and non-lymphoid
tissues. findings show that LPS can still activate monocytes but Studies indicate that sepsis affects both of the main
that their intracellular signalling has shifted towards human NK cell subsets (that is the CD56hi and CD56low
MicroRNAs the production of anti-inflammatory molecules, thereby NK cell subsets); in patients with sepsis the number
Single-stranded RNA supporting the concept of monocyte reprogramming72. of circulating NK cells is markedly decreased95,96, often
molecules of approximately
2123 nucleotides in length
In clinical studies, the magnitude and the persistent for weeks97, and low numbers of NK cells are associ-
that are thought to regulate nature of this refractory state is associated with increased ated with increased mortality 98. Furthermore, the
the expression of other genes. mortality and nosocomial infections73. absolute number of both NK cell subsets is decreased

NATURE REVIEWS | IMMUNOLOGY VOLUME 13 | DECEMBER 2013 | 867

2013 Macmillan Publishers Limited. All rights reserved


REVIEWS

in sepsis99. Reduced NK cell cytotoxic function and is not altered during sepsis, which supports the idea that
cytokine secretion occur during sepsis and following regulatory functions are maintained or increased dur-
burns and traumatic injuries in animal models and in ing sepsis, whereas effector Tcell responses are down-
patients96,100103. regulated (see below)95,122,124. Histone methylation and
Cytokine production by NK cells in response to TLR chromatin remodelling are thought to contribute to the
agonists was impaired after polymicrobial sepsis104, which suppression of TH1 and TH2 cell functions by acting at
suggests that NK cells become tolerant to TLR agonists, promoter regions of the IFNG and GATA3 genes125.
and have features similar to those of endotoxin tolerance Although the TH17 cell subset has been less well
in monocytes105. Impaired IFN production in response to studied than the T H1 and TH2 cell subsets, there is
TLR agonists such as LPS and CpG has also been reported now general agreement that TH17 cells protect against
in NK cells from patients with sepsis99. As NK cells have a extracellular bacterial and fungal infections by pro-
central role in antiviral defence, it is possible that impaired ducing IL17 and IL22 (REFS126129). The TH17 cell
NK cell function could lead to the reactivation of latent response is reduced in sepsis, possibly as a result of
viruses; this phenomenon has been frequently described decreased expression of retinoic acid receptor-related
in patients in ICUs106109. Decreased IFN production by orphan receptort (RORt), which is the transcription
NK cells occurs in patients with sepsis and frequently pre- factor that is specific for TH17 cells95,123. This defect in
cedes cytomegalovirus reactivation105 in patients with pro- the TH17 cell phenotype in sepsis is likely to be a con-
tracted sepsis96. In addition, defective NK cell production tributing factor to the increased susceptibility of these
of IFN could be a factor in the increased incidence of sec- patients to secondary fungal infections129. Indeed, mice
ondary infections and the decreased monocyte HLADR that survived an initial polymicrobial bacterial infection
expression that occur in septic patients110. (but not control mice) become highly susceptible to and
rapidly succumb to pulmonary A.fumigatus infection66.
Tcells. Tcells are a distinct subset of lymphocytes Recently, IL7 therapy has been shown to increase the
that reside in large numbers in the intestinal mucosa TH17 cell response and to decrease the number of deaths
and that possess qualities common to both innate and from secondary infection with Candida albicans 130,131.
adaptive immune cells. Although the breadth of anti-
gens to which Tcells respond is not completely Tcell exhaustion. Tcell exhaustion was first described in
known, it is clear that they recognize lipid antigens that mice suffering from chronic viral infections and is typi-
are present on various invading pathogens111. Tcells fied by Tcells that have severely impaired effector func-
reside in the intestines and on other mucosal surfaces, tions132. Subsequently, Tcell exhaustion has been shown
where they recognize invading pathogens and mount in patients with bacterial and parasitic infections, HIV
a prompt, innate-like immune response by releasing and cancer 133. The prolonged duration of sepsis is char-
IFN, IL17 and various chemokines. As such, Tcells acterized by high antigen load and increased levels
can be considered to represent a first line of defence of pro- and anti-inflammatory cytokines, which is an
against particular pathogens112. The number of circulat- ideal setting for the development of Tcell exhaustion.
ing Tcells is significantly decreased in patients with A recent study 12 in which spleens were obtained rap-
sepsis, and more severe depletion occurs in patients idly after the death of patients with sepsis showed evi-
who have the greatest severity of illness and mortal- dence that is highly consistent with the occurrence of
ity 112. The loss of Tcells in the intestinal mucosa Tcell exhaustion, including: profound suppression of
might be particularly detrimental to the host because it the production of IFN and TNF by stimulated Tcells;
might lead to the invasion of intestinal pathogens into increased expression of PD1 on CD4+ Tcells and of pro-
the circulation or the peritoneal cavity, thereby causing grammed cell death 1 ligand 1 (PDL1) on macrophages;
secondary infections. and decreased Tcell expression of CD127 (the IL7R
chain), which is another phenotypic feature of
CD4+ TH cell subsets. Numerous studies have reported exhausted Tcells. Furthermore, this study showed that
marked effects of sepsis on circulating and tissue Tcells, capillary endothelial cells and bronchial epithelial cells
with some of the most considerable effects occurring in patients with sepsis had increased PDL1 expres-
in CD4+ Tcells113118. Mature CD4+ TH cells have been sion, which in this setting can impair the function of
Tcell exhaustion characterized into TH1, TH2 and TH17 cell subsets on the Tcells that have migrated to the local area of infec-
The condition of functionally
impaired antigen-specific
basis of the type of cytokines that they produce on stim- tion, thereby seriously compromising the ability of the
Tcells, typified by increased ulation. Early studies suggested that both TH1 and TH2 host to eradicate the organisms. A potential causal link
surface expression of cell-associated cytokine production are decreased during between Tcell exhaustion and morbidity and mortality
programmed cell death 1, the initial immune response to sepsis and trauma119122 in sepsis was provided by studies showing that increased
which occurs in the context of
(FIG.3). Additional work showed marked reductions in expression of PD1 on circulating Tcells from patients
persistent high antigen load.
The defects in effector Tcell the expression of Tbet and GATA-binding protein 3 with sepsis correlated with decreased Tcell proliferative
function include a progressive (GATA3), which are transcription factors that modulate capacity, increased nosocomial infections, and mortal-
decrease in their ability to the TH1 and TH2 cell response, respectively, and this has ity 134. Finally, animal studies show that inhibition of the
produce cytokines, loss of reinforced the idea that both the TH1 and TH2 cell lin- PD1PDL1 interaction improves survival in several
proliferative capacity and
decreased cytotoxicity, and
eages are suppressed after trauma and sepsis123. These clinically relevant models of sepsis, which is consistent
can result in apoptotic cell studies also showed that expression of the TReg cell- with a key role for Tcell exhaustion in the pathogenesis
death. associated transcription factor forkhead box P3 (FOXP3) of sepsis (see below)135137.

868 | DECEMBER 2013 | VOLUME 13 www.nature.com/reviews/immunol

2013 Macmillan Publishers Limited. All rights reserved


REVIEWS

TReg cells. An increased percentage of circulating TReg the anti-apoptotic molecule BCL2, induces the pro-
cells has been described in patients suffering from sep- liferation of peripheral Tcells and sustains increased
tic shock. This increase was observed immediately after numbers of circulating blood CD4+ and CD8+ Tcells. In
the onset of sepsis but persisted only in those patients addition, IL7 increases the diversity of the Tcell recep-
who subsequently died138. These results were further tor (TCR) repertoire, is associated with a reduction in
extended by showing that this relative increase was the proportion of TReg cells in the circulation, rejuve-
due to a decrease in effector Tcell numbers rather than nates exhausted Tcells by decreasing PD1 expression
changes in the absolute numbers of TReg cells119. This and increases the expression of cell adhesion molecules,
finding suggests that TReg cells are more resistant to thereby facilitating the trafficking of Tcells to sites of
sepsis-induced apoptosis, possibly because of increased infection149,150 (FIG.4). Recently, a massive loss in diver-
expression of the anti-apoptotic protein BCL2. sity of the TCR repertoire was observed in patients with
Alarmins, including heat shock proteins and histones, are sepsis151. Importantly, this low TCR diversity was also
increased in sepsis, and as strong inducers of TReg cells associated with increased risk of death and nosocomial
they are also likely to contribute to increased TReg cell infections. An optimal diversity of the TCR repertoire
numbers in sepsis139. Since this initial work, many groups greatly facilitates an effective immune response against
have confirmed an increase in TReg cell numbers in the invading pathogens, and this might be one of the most
blood or the spleen in various animal models of and important benefits of IL7. Thus, beyond supporting
in patients suffering from trauma and sepsis114,140,141. the replenishment of lymphocytes, which are markedly
Initial studies investigating the role of TReg cells in depleted during sepsis, IL7 improves multiple func-
sepsis-induced immune dysfunction in mice were incon- tional aspects of Tcells that are considerably altered
clusive, possibly owing to the use of CD25specific anti- insepsis.
bodies, which lack specificity in inhibiting TReg cells142. In clinical trials, recombinant IL7 has been used
Recent studies indicate that increased TReg cells are to treat patients with idiopathic lymphopenia and with
Alarmins deleterious in sepsis and are associated with decreased lymphopenia-driven diseases, including patients who
Endogenous molecules that effector Tcell proliferation and function141. Importantly, are infected with HIV and have persistently low lympho-
are released after tissue injury these suppressive effects were totally abrogated by using cyte numbers despite effective antiretroviral therapy 152.
and that promote activation
small interfering RNA (siRNA) specific for FOXP3 to Clinical studies in over 250 patients have shown that IL7
of the innate immune system.
Excessive release of alarmins block the differentiation of TReg cells141. Other investiga- is safe and well tolerated148. The use of IL7 in clinical
might promote uncontrolled tors used glucocorticoid-induced TNF-receptor-related trials of sepsis is supported by data from animal mod-
inflammation and exacerbate protein (GITR; also known as TNFRSF18)-specific anti- els of sepsis131,150,153 (BOX 1). In mice with polymicrobial
tissue injury. Conversely, bodies to block TReg cells, which led to improved immune sepsis due to peritonitis, treatment with IL7 improved
alarmins may have beneficial
effects by activating the
function in sepsis and enhanced microbial killing 143. Tcell viability, trafficking and IFN production, and
immune system to combat In addition, sepsis-induced immunosuppression was restored the delayed-type hypersensitivity response to
potential pathogens. shown to facilitate rapid growth of solid tumours via a recall antigens150. Similar beneficial effects of IL7 were
TReg cell-mediated effect 144. observed in a murine fungal sepsis model that reproduces
Small interfering RNA
TReg cells can also suppress innate immune cells. the delayed secondary infections that frequently affect
(siRNA). Synthetic RNA
molecules of 1923 In the context of LPS stimulation, TReg cells inhibit both patients in ICUs130. Importantly, in these mouse experi-
nucleotides that are used to monocyte and neutrophil function145. Furthermore, TReg ments, there was also a more than twofold improvement
knock down (that is, to silence cells inhibit IFN production by Tcells in response to in survival. The ability of IL7 to reverse the defects in
the expression of) a specific Mycobacterium tuberculosis and precipitate an NKcell- lymphocyte function that occur in patients with sep-
gene. This is known as RNA
interference and is mediated
dependent endotoxin tolerance-like phenomenon sis has been tested in an invitro model154. Exvivo IL7
by the sequence-specific that is characterized by decreased production of IFN treatment of Tcells from patients with sepsis corrected
degradation of mRNA. and granulocytemacrophage colony-stimulating fac- sepsis-induced defects, including CD4+ and CD8+ Tcell
tor (GM-CSF)113,146,147. In summary, there is extensive proliferation abnormalities, decreased IFN production,
Central memory Tcells
evidence that patients with sepsis and trauma have impaired phosphorylation of signal transducer and acti-
Antigen-experienced Tcells
that express cell surface increased numbers of TReg cells, which, by acting both vator of transcription5 (STAT5) and decreased BCL2
receptors that are required on innate and adaptive immune cells, impair immunity levels154. These data indicate that the IL7 signalling path-
for homing to secondary and contribute to nosocomial infections and mortality. way remains fully operative during sepsis and that IL7
lymphoid organs. These cells reverses crucial sepsis-induced defects in Tcell function.
are generally thought to be
long-lived and can function
Immunotherapies in sepsis
as the precursors for effector Recombinant human IL7. Considering the severe PD1- and PDL1specific antibodies. Another approach
Tcells for recall responses. quantitative and qualitative alterations in Tcells that are that holds great potential in reversing immunosuppres-
induced by sepsis, IL7 has recently emerged as a prom- sion in sepsis involves blockade of the coinhibitory
Delayed-type
ising therapeutic agent for septic patients. IL7 is essen- molecules PD1 and PDL1 (FIG.4). As discussed previ-
hypersensitivity
A cellular immune response tial for Tcell development and function148. Its effects are ously, PD1 and PDL1 are widely expressed on immune
to antigens that develops mediated via the heterodimeric IL7R, which is com- effector cells, endothelial cells and bronchial epithelial
over 2472hours with the posed of the IL7R chain (CD127) and the common cells in patients with sepsis12. Blockade of PD1PDL1
infiltration of Tcells and cytokine receptor chain (c; CD132). This receptor is signalling improves survival in clinically relevant animal
monocytes, and that is
dependent on the production
expressed by most resting human Tcells: naive and cen- models of bacterial sepsis135137 and markedly decreases
of T helper 1 cell-associated tral memory Tcells express the highest levels and TReg cells mortality in primary and secondary fungal sepsis caused
cytokines. express the lowest levels. IL7 upregulates expression of by C.albicans155. Thus, PD1 expressed by lymphocytes or

NATURE REVIEWS | IMMUNOLOGY VOLUME 13 | DECEMBER 2013 | 869

2013 Macmillan Publishers Limited. All rights reserved


REVIEWS

HLA-DR IFN IL-7


GM-CSF
Activation TCR diversity
Macrophage Activation

PDL1 PD1- or PD1- or CD4 or LFA1


Tracking
PDL1-specic PDL1-specic CD8+ T cell
PD1 antibody PD1 antibody

IFN
CD8+ Exhausted IFN
T cell CD8+ T cell Loss of cytotoxicity Macrophage
Apoptosis Proliferation activation

Figure 4 | Immunotherapy of sepsis. There are several immunotherapeutic agents that have shown promise in
reversing the immunosuppressive phase of sepsis, including recombinant interleukin7 (IL7), Nature Reviews | cell
programmed Immunology
death1
(PD1)- or programmed cell death 1 ligand1 (PDL1)specific antibodies, recombinant interferon (IFN) and
recombinant granulocytemacrophage colony-stimulating factor (GMCSF). GMCSF and IFN primarily affect
monocytes and macrophages to increase HLADR expression and to induce activation. IL7 and PD1specific
antibodies have the advantage of targeting CD4+ and CD8+ Tcells to restore the function of the adaptive immune
system. By re-engaging CD4+ Tcells, both recombinant IL7 and PD1specific antibodies will have effects not only on
adaptive immune cells but also indirectly on monocytes and macrophages. PD1- and PDL1specific antibodies will
prevent and/or reverse Tcell exhaustion. Thus, the net effects of these antibodies will be to prevent decreased IFN
production, Tcell apoptosis and decreased CD8+ Tcell cytotoxicity. IL7 will block Tcell apoptosis, induce Tcell
proliferation, increase IFN production by Tcells, increase Tcell receptor (TCR) diversity and increase Tcell trafficking
by increasing the expression of cell integrins such as lymphocyte function-associated antigen 1 (LFA1). The application
of immunotherapeutic agents will depend on the use of various biomarkers or tests of immune function to document
that patients have entered the immunosuppressive phase of sepsis.

PDL1expressed by monocytes could function as poten- with IFN showed enhanced clearing of fungi from the
tial biomarkers for the selection of patients with sepsis cerebrospinal fluid compared to control individuals160.
as candidates for therapy. A recent important study Thus, IFN might be effective if targeted to patients with
showed that invitro blockade of PD1 improved IFN sepsis who have entered the immunosuppressive phase,
production and decreased apoptosis of Tcells from as identified by decreased monocyte HLADR expres-
patients with active M.tuberculosis infections156. A sec- sion and/or the development of fungal sepsis. A clinical
ond major finding in this study was that when patients trial of IFN is currently underway in the Netherlands in
with tuberculosis were effectively treated, the number patients with sepsis who have been determined to have
of PD1expressing Tcells decreased and inversely cor- entered the immunosuppressive phase of the disorder
related with the IFN Tcell response against M.tuber- (ClinicalTrials.gov number: NCT01649921).
culosis. We believe that this work has major implications Although IFN offers real promise as a potential
for the broader field of sepsis because of the similarities immunotherapy in sepsis because of its ability to restore
between active tuberculosis and protractedsepsis. monocyte function, it will not correct the fundamental
defect in Tcells that is a major pathological abnormality
IFN. A key immunological defect in sepsis is decreased in sepsis. In the authors opinion, recombinant IL7 and
production of IFN, which is a cytokine that is essential PD1specific antibodies are more likely to be of benefit
for the activation of innate immunity (FIG.4). Small clini- to patients with sepsis because of their favourable effects
cal trials using recombinant IFN have been carried out on the functions of CD4+ and CD8+ Tcells: they both
in sepsis. Treatment with recombinant IFN reversed restore IFN production in sepsis and have many other
monocyte dysfunction in patients with sepsis whose beneficial effects on various other immune effectorcells.
Chronic granulomatous
monocytes had decreased HLADR expression and pro-
disease duced reduced amounts of TNF in response to LPS. Out GCSF and GMCSF. Given the many defects in neutro-
An inherited disorder caused of nine patients with severe sepsis who were treated with phil function that occur in sepsis, investigators carried
by defective oxidase activity in IFN to improve monocyte function, eight survived157. out two randomized clinical trials using recombinant
the respiratory burst of
Most recently, IFN treatment in a patient with pro- granulocyte colony-stimulating factor (GCSF), which
phagocytes. It results from
mutations in any of four genes tracted Staphylococcus aureus sepsis caused an increase is a drug that increases neutrophil numbers and func-
that are necessary to generate in monocyte HLADR expression and in the numbers tion, in patients with sepsis resulting from hospital-
the superoxide radicals of IL17expressing CD4+ Tcells, and led to the eradica- acquired and community-acquired pneumonia80,81.
required for normal neutrophil tion of the bacteria158. Furthermore, IFN is effective in GCSF had already been shown to be highly beneficial
function. Affected patients
suffer from increased
treating fungal sepsis in patients with chronic granuloma- in decreasing the incidence of sepsis in patients who
susceptibility to recurrent tous disease159. Finally, in a recent trial, patients with HIV have abnormally low absolute neutrophil numbers
infections. who had cryptococcal meningitis and who were treated owing to chemotherapy or radiation therapy. Although

870 | DECEMBER 2013 | VOLUME 13 www.nature.com/reviews/immunol

2013 Macmillan Publishers Limited. All rights reserved


REVIEWS

Box 1 | Do mouse sepsis models reflect the human condition?


Key differences between the mouse and human immune systems for example, differences in Toll-like receptor (TLR)
expression and nitric oxide regulation have been recognized for years. However, much discussion arose from the recent
publication of a study indicating that mouse models of inflammation and sepsis do not reflect the complex scenario that
occurs in patients with these disorders167,168. Investigators carried out gene expression analysis on total blood leukocytes
from patients with trauma and burns (many of whom were septic), as well as from volunteers who received endotoxin, and
the results were compared with those generated from mouse models of trauma, sepsis and burn injury168. Gene expression
data from the three human conditions showed very similar genomic responses, implying a commonality of the human
immune response during inflammation induced by distinct mechanisms. By contrast, gene expression data from the three
mouse models did not reflect any of the three human states. This lack of similarity in the human and mouse genomic
response to injury led the investigators to suggest that mouse models are poor predictors of the more complex human
conditions and that results from animal models might be misleading. The investigators argued that this lack of correlation
between the mouse and human immune responses during sepsis is likely to explain why so many drugs that were
efficacious in animal models of sepsis failed in clinical trials of patients with sepsis.
Although there are major differences between the human and mouse immune responses to sepsis, there are several
important factors that should be discussed. First, the pathological anatomical changes that occur in the mouse and
human immune systems during sepsis share many common features; specifically, the marked apoptotic depletion of
immune cells and gastrointestinal epithelial cells that occurs in mouse sepsis is closely replicated in patients dying of
sepsis27,28,33,34. Second, the lack of correlation of the genomic response between mouse and human blood leukocytes
might be partly due to differences in the relative proportions of the various white blood cell components, such as
neutrophils, lymphocytes and monocytes. Perhaps a much better correlation of the genomic response between mice
and humans might have been obtained if the comparisons were restricted to a particular cell class, such as circulating
lymphocytes or monocytes. In this regard, the dramatic upregulation in mouse lymphocytes of genes encoding
pro-apoptotic Bcell lymphoma 2 (BCL2) family members and the downregulation of genes encoding anti-apoptotic
BCL2 family members is similar to the expression pattern seen in lymphocytes from patients with sepsis169,170. The
importance of this correlation for a particular class of genes in mice and humans is highlighted by the fact that apoptosis
is a key pathogenic mechanism in sepsis. A few of the other immunological abnormalities that have been reported in both
mouse and human sepsis include: a shift from a T helper 1 (TH1) to a TH2 cell phenotype; an increase in regulatory T cell
numbers; an increase in myeloid-derived suppressor cell numbers; and an upregulation of coinhibitory molecule
expression12,50,58,155. Thus, in our opinion, mouse sepsis models do provide important insights into particular mechanisms
of sepsis that seem to have a comparable role in humans.

neutrophil numbers are typically increased in sepsis, the outcome by causing an over-exuberant inflammatory
investigators postulated that administration of GCSF response if applied during the wrong phase of the disor-
to further increase neutrophil numbers might improve der. Readers are referred to numerous recent reviews that
pathogen killing. Although white blood cell numbers discuss the use of various biomarkers in sepsis91,164,165.
increased in these patients, there was no effect on over- BOX2 provides a list of clinical and laboratory findings that
all survival161,162. These two clinical studies indicate could be used to identify immunosuppressed patients
that it is unlikely that drugs that enhance neutrophil who might benefit from immunotherapies. Potential
numbers or function will be of benefit in patients with biomarkers include some currently available param-
non-neutropenia-associated sepsis. eters, such as decreased monocyte HLADR expression
Another immunotherapy that is being actively inves- and increased circulating IL10 concentrations (both of
tigated in sepsis is the administration of GMCSF, which which assess innate immune function and can be used
is a cytokine that accelerates the production of neutro- to stratify patients for GMCSF or IFN treatments), and
phils, monocytes and macrophages11 (FIG.4). Treatment decreased absolute CD4+ Tcell numbers and increased
of ventilator-dependent patients with sepsis who had percentage of TReg cells (both of which assess adaptive
entered the immunosuppressive phase of the disorder, as immunity and can be used to stratify patients for IL7
identified by persistent decreases in monocyte HLADR therapy). Ideally, exvivo functional testing of circulat-
expression, with recombinant GMCSF, resulted in the ing immune cells remains the gold standard for the
restoration of HLADR expression, fewer days on the evaluation of immunity because it directly measures
ventilator and fewer days in the ICU163. In addition, the capacity of cells to respond to pathogens.
treatment of immunosuppressed paediatric patients Two hallmarks of sepsis-induced immunosuppres-
with sepsis with recombinant GMCSF restored TNF sion have been identified: the decreased capacity of
production and reduced the acquisition of nosocomial monocytes to produce pro-inflammatory cytokines
infections11. (mainly TNF) in response to endotoxin challenge; and
decreased lymphocyte proliferation. Currently, these
Biomarker-guided therapy. A prerequisite for the two characteristics are thought to provide a reasonable
application of immunotherapy in sepsis is the proper assessment of immune effector cell status but remain
selection of patients. Various biomarkers should help barely usable in routine clinical monitoring owing to
in deciphering whether the patient is in the hyperin- methodological challenges, such as long incubation
flammatory versus the hypoinflammatory phase of the time, lengthy cell purification procedures and low stand-
disorder (FIG.1). Indeed, immunotherapy could worsen ardization. A major challenge is to develop automated

NATURE REVIEWS | IMMUNOLOGY VOLUME 13 | DECEMBER 2013 | 871

2013 Macmillan Publishers Limited. All rights reserved


REVIEWS

Box 2 | Potential biomarkers for applied immunotherapy in sepsis


of cell phenotypic assays (for example, measuring
HLADR or PD1 expression), functional assays (for
Innate immunity example, measuring TNF production in whole blood)
HLADR expression on monocytes. and genomic assays166.
Tumour necrosis factor (TNF) production by lipopolysaccharide-stimulated whole
blood cells. Conclusions
Programmed cell death ligand 1 expression on monocytes. The current understanding of the immune response
to sepsis is controversial some investigators suggest
Adaptive immunity
that sepsis-induced morbidity and mortality is the result
Persistent severe lymphopenia.
of persistent immune activation with accompanying
Reactivation of cytomegalovirus or herpes simplex virus infections. inflammation, whereas others assert that it is the result
Programmed cell death 1 expression on CD4+ or CD8+ cells. of immunosuppression. We support the idea that sepsis-
Number of circulating regulatory Tcells. induced immunosuppression with Tcell exhaustion is a
Interferon- production by Tcells. major abnormality. As discussed in this Review, sepsis
Tcell proliferation. induces numerous defects in host innate and adaptive
Innate and adaptive immunity immunity such that, if the invading organisms are not
Infections with relatively avirulent or opportunistic pathogens, such as Enteroroccus
promptly eliminated, the host becomes more susceptible
spp., Acinetobacter spp. or Candida spp. to intractable infection or to new secondary infections.
Interleukin10/TNF ratio.
Methods for identifying when patients have entered an
immunosuppressive phase of sepsis and for detecting
Delayed-type hypersensitivity response.
particular defects in immunity will enable the applica-
tion of potent new immunotherapies that have shown
or standardized biomarkers that can be used in routine great promise in animal models of sepsis and in early
daily practice. More precise cellular monitoring might clinical studies of various infectious disorders. In par-
be possible through the use of innovative therapies, ticular, recombinant IL7 and PD1specific antibodies
such as monoclonal antibody-based therapies target- reverse fundamental immune defects in sepsis, lead to
ing PD1- and CTLA4expressing cells. Quantifying the improved survival in multiple clinically relevant animal
mRNA for a panel of candidate genes in whole blood models of sepsis, are clinically well tolerated and have an
might enable clinicians to identify immunosuppressed ability to improve immunity in patients with cancer and
patients and to provide the appropriate treatment. In chronic viral infections. Immunotherapy could there-
the authors opinion, the ideal method for identifying fore represent the next major advance in the treatment
immunosuppressed patients would be a combination of infectious disease andsepsis.

1. Vincent,J.L., Opal,S.M., Marshall,J.C. & 12. Boomer,J.S. etal. Immunosuppression in patients 21. Geiger,H., de Haan,G. & Florian,M.C. The ageing
Tracey,K.J. Sepsis definitions: time for change. who die of sepsis and multiple organ failure. haematopoietic stem cell compartment. Nature Rev.
Lancet 381, 774775 (2013). JAMA 306, 25942605 (2011). Immunol. 13, 376389 (2013).
2. Martin,G.S., Mannino,D.M. & Moss,M. The effect This study is the first to document considerable 22. van Dissel,J.T., van Langevelde,P.,
of age on the development and outcome of adult defects in immunity in tissues from patients dying Westendorp,R.G., Kwappenberg,K. & Frlich,M.
sepsis. Crit. Care Med. 34, 1521 (2006). of sepsis. It determines that Tcell exhaustion Anti-inflammatory cytokine profile and mortality in
3. Cohen,J., Opal,S. & Calandra,T. Sepsis studies need is an important mechanism of immunosuppression febrile patients. Lancet 351, 950953 (1998).
new direction. Lancet Infect. Dis. 12, 503505 in this disorder. 23. Ertel,W. etal. Downregulation of proinflammatory
(2012). 13. Ward,P.A. Immunosuppression in sepsis. JAMA 306, cytokine release in whole blood from septic patients.
4. Wenzel,R.P. & Edmond,M.B. Septic shock-- 26182619 (2011). Blood 85, 13411347 (1995).
evaluating another failed treatment. N.Engl. J.Med. 14. Torgersen,C. etal. Macroscopic postmortem 24. Munoz,C. etal. Dysregulation of invitro cytokine
366, 21222124 (2012). findings in 235 surgical intensive care patients with production by monocytes during sepsis. J.Clin. Invest.
5. Ward,P.A. New approaches to the study of sepsis. sepsis. Anesth. Analg. 108, 18411847 (2009). 88, 17471754 (1991).
EMBO Mol. Med. 4, 12341243 (2012). 15. Otto,G.P. etal. The late phase of sepsis is 25. Rigato,O. & Salomao,R. Impaired production of
6. Angus,D.C. The search for effective therapy for characterized by an increased microbiological burden interferon- and tumor necrosis factor- but not of
sepsis: back to the drawing board? JAMA 306, and death rate. Crit. Care 15, R183 (2011). interleukin 10 in whole blood of patients with sepsis.
26142615 (2011). 16. Hodi,F.S. etal. Improved survival with ipilimumab in Shock 19, 113116 (2003).
7. Hotchkiss,R.S., Monneret,G. & Payen,D. patients with metastatic melanoma. N.Engl. J.Med. 26. Wong,H.R. Genome-wide expression profiling in
Immunosuppression in sepsis: a novel understanding 363, 711723 (2010). pediatric septic shock. Pediatr. Res. 73, 564569
of the disorder and a new therapeutic approach. 17. Topalian,S.L. etal. Safety, activity, and immune (2013).
Lancet Infect. Dis. 13, 260268 (2013). correlates of antiPD1 antibody in cancer. 27. Hotchkiss,R.S. etal. Apoptotic cell death in
8. Hotchkiss,R.S. & Karl,I.E. The pathophysiology and N.Engl. J.Med. 366, 24432454 (2012). patients with sepsis, shock, and multiple organ
treatment of sepsis. N.Engl. J.Med. 348, 138150 This study confirms that immunotherapy represents dysfunction. Crit. Care Med. 27, 12301251
(2003). a major breakthrough in the treatment of cancer, (1999).
This review article presents one of the first a disorder that shares many immune defects with This study was the first to demonstrate that
conceptual theories for a new understanding of sepsis. apoptosis causes massive death and depletion
the immunological basis of sepsis and points to 18. Munford,R.S. & Pugin,J. Normal responses of immune effector cells in patients with sepsis.
potential new therapeutic approaches based on to injury prevent systemic inflammation and can be 28. Hotchkiss,R.S. etal. Sepsis-induced apoptosis
the immunological phase of sepsis. immunosuppressive. Am. J.Respir. Crit. Care Med. causes progressive profound depletion of B and
9. Hotchkiss,R.S. & Opal,S. Immunotherapy for 163, 316321 (2001). CD4+ T lymphocytes in humans. J.Immunol. 166,
sepsis a new approach against an ancient foe. One of the first studies to demonstrate that both 69526963 (2001).
N.Engl. J.Med. 363, 8789 (2010). pro-inflammatory and anti-inflammatory processes 29. Hotchkiss,R.S. etal. Depletion of dendritic cells,
10. Payen,D., Monneret,G. & Hotchkiss,R. occur rapidly after sepsis. but not macrophages, in patients with sepsis.
Immunotherapy a potential new way forward in 19. Xiao,W. etal. A genomic storm in critically injured J.Immunol. 168, 24932500 (2002).
the treatment of sepsis. Crit. Care 17, 118 (2013). humans. J.Exp. Med. 208, 25812590 (2011). 30. Felmet,K.A., Hall,M.W., Clark,R.S., Jaffe,R. &
11. Hall,M.W. etal. Immunoparalysis and nosocomial 20. Stearns-Kurosawa,D.J., Osuchowski,M.F., Carcillo,J.A. Prolonged lymphopenia, lymphoid
infection in children with multiple organ dysfunction Valentine,C., Kurosawa,S. & Remick,D.G. depletion, and hypoprolactinemia in children with
syndrome. Intensive Care Med. 37, 525532 The pathogenesis of sepsis. Annu. Rev. Pathol. nosocomial sepsis and multiple organ failure.
(2011). 6, 1948 (2011). J.Immunol. 174, 37653772 (2005).

872 | DECEMBER 2013 | VOLUME 13 www.nature.com/reviews/immunol

2013 Macmillan Publishers Limited. All rights reserved


REVIEWS

31. Toti,P. etal. Spleen depletion in neonatal sepsis and 57. Poehlmann,H., Schefold,J.C., Zuckermann- 80. Gogos,C.A., Drosou,E., Bassaris,H.P. &
chorioamnionitis. Am. J.Clin. Pathol. 122, 765771 Becker,H., Volk,H.D. & Meisel,C. Phenotype Skoutelis,A. Pro- versus anti-inflammatory cytokine
(2004). changes and impaired function of dendritic cell profile in patients with severe sepsis: a marker for
32. Hotchkiss,R.S. etal. Rapid onset of intestinal subsets in patients with sepsis: a prospective prognosis and future therapeutic options. J.Infect.
epithelial and lymphocyte apoptotic cell death in observational analysis. Crit. Care 13, R119 (2009). Dis. 181, 176180 (2000).
patients with trauma and shock. Crit. Care Med. 28, 58. Dreschler,K. etal. Altered phenotype of blood 81. Hynninen,M. etal. Predictive value of monocyte
32073217 (2000). dendritic cells in patients with acute pneumonia. histocompatibility leukocyte antigenDR expression
33. Chang,K.C. etal. Multiple triggers of cell death in Respiration 83, 209217 (2012). and plasma interleukin4 and -10 levels in critically ill
sepsis: death receptor and mitochondrial-mediated 59. Grimaldi,D. etal. Profound and persistent decrease patients with sepsis. Shock 20, 14 (2003).
apoptosis. FASEB J. 21, 708719 (2007). of circulating dendritic cells is associated with ICU- 82. Oberholzer,A., Oberholzer,C. & Moldawer,L.L.
34. Hotchkiss,R.S. etal. Overexpression of Bcl2 in acquired infection in patients with septic shock. Interleukin10: a complex role in the pathogenesis of
transgenic mice decreases apoptosis and improves Intensive Care Med. 37, 14381446 (2011). sepsis syndromes and its potential as an anti-
survival in sepsis. J.Immunol. 162, 41484156 (1999). 60. Pastille,E. etal. Modulation of dendritic cell inflammatory drug. Crit. Care Med. 30, S58S63
35. Chung,C.S. etal. Inhibition of Fas/Fas ligand differentiation in the bone marrow mediates sustained (2002).
signaling improves septic survival: differential effects immunosuppression after polymicrobial sepsis. 83. Sfeir,T., Saha,D.C., Astiz,M. & Rackow,E.C.
on macrophage apoptotic and functional capacity. J.Immunol. 186, 977986 (2011). Role of interleukin10 in monocyte
J.Leukoc. Biol. 74, 344351 (2003). 61. Faivre,V. etal. Human monocytes differentiate into hyporesponsiveness associated with septic shock.
36. Oberholzer,C. etal. Targeted adenovirus-induced dendritic cells subsets that induce anergic and Crit. Care Med. 29, 129133 (2001).
expression of IL10 decreases thymic apoptosis and regulatory Tcells in sepsis. PLoS ONE 7, e47209 84. Muehlstedt,S.G., Lyte,M. & Rodriguez,J.L.
improves survival in murine sepsis. Proc. Natl Acad. (2012). Increased IL10 production and HLADR suppression
Sci. USA 98, 1150311508 (2001). 62. Gautier,E.L. etal. Enhanced dendritic cell survival in the lungs of injured patients precede the
37. Iwata,A. etal. Over-expression of Bcl2 provides attenuates lipopolysaccharide-induced immuno development of nosocomial pneumonia. Shock 17,
protection in septic mice by a trans effect. J.Immunol. suppression and increases resistance to lethal 443450 (2002).
171, 31363141 (2003). endotoxic shock. J.Immunol. 180, 69416946 85. Randow,F. etal. Mechanism of endotoxin
38. Voll,R.E. etal. Immunosuppressive effects of (2008). desensitization: involvement of interleukin 10 and
apoptotic cells. Nature 390, 350351 (1997). 63. Bohannon,J., Cui,W., Sherwood,E. & Toliver- transforming growth factor . J.Exp. Med. 181,
39. Green,D.R. & Beere,H.M. Apoptosis. Gone but not Kinsky,T. Dendritic cell modification of neutrophil 18871892 (1995).
forgotten. Nature 405, 2829 (2000). responses to infection after burn injury. J.Immunol. 86. Muenzer,J.T. etal. Characterization and modulation
40. Tamayo,E. etal. Evolution of neutrophil apoptosis in 185, 28472853 (2010). of the immunosuppressive phase of sepsis.
septic shock survivors and nonsurvivors. J.Crit. Care 64. Toliver-Kinsky,T.E., Cui,W., Murphey,E.D., Lin,C. & Infect. Immun. 78, 15821592 (2010).
27, 415.e1415.e11 (2012). Sherwood,E.R. Enhancement of dendritic cell 87. Lukaszewicz,A.C. etal. Monocytic HLADR
41. Drifte,G., Dunn-Siegrist,I., Tissieres,P. & Pugin,J. production by fms-like tyrosine kinase3 ligand expression in intensive care patients: interest for
Innate immune functions of immature neutrophils in increases the resistance of mice to a burn wound prognosis and secondary infection prediction.
patients with sepsis and severe systemic inflammatory infection. J.Immunol. 174, 404410 (2005). Crit. Care Med. 37, 27462752 (2009).
response syndrome. Crit. Care Med. 41, 820832 65. Toliver-Kinsky,T.E., Lin,C.Y., Herndon,D.N. & 88. Astiz,M., Saha,D., Lustbader,D., Lin,R. & Rackow,E.
(2013). Sherwood,E.R. Stimulation of hematopoiesis by the Monocyte response to bacterial toxins, expression
42. Alves-Filho,J.C., Spiller,F. & Cunha,F.Q. Neutrophil Fms-like tyrosine kinase 3 ligand restores bacterial of cell surface receptors, and release of anti-
paralysis in sepsis. Shock 34 (Suppl. 1), 1521 (2010). induction of Th1 cytokines in thermally injured mice. inflammatory cytokines during sepsis. J.Lab Clin.
43. Kovach,M.A. & Standiford,T.J. The function of Infect. Immun. 71, 30583067 (2003). Med. 128, 594600 (1996).
neutrophils in sepsis. Curr. Opin. Infect. Dis. 25, 66. Benjamim,C.F., Hogaboam,C.M., Lukacs,N.W. & 89. Rode,H.N. etal. Lymphocyte function in anergic
321327 (2012). Kunkel,S.L. Septic mice are susceptible to patients. Clin. Exp. Immunol. 47, 155161 (1982).
44. Cummings,C.J. etal. Expression and function of the pulmonary aspergillosis. Am. J.Pathol. 163, 90. Monneret,G., Venet,F., Pachot,A. & Lepape,A.
chemokine receptors CXCR1 and CXCR2 in sepsis. 26052617 (2003). Monitoring immune dysfunctions in the septic patient:
J.Immunol. 162, 23412346 (1999). 67. Benjamim,C.F., Lundy,S.K., Lukacs,N.W., a new skin for the old ceremony. Mol. Med. 14,
45. Stephan,F. etal. Impairment of polymorphonuclear Hogaboam,C.M. & Kunkel,S.L. Reversal of long- 6478 (2008).
neutrophil functions precedes nosocomial infections in term sepsis-induced immunosuppression by dendritic 91. Venet,A., Lukaszewica,A., Payen,D., Hotchkiss,R. &
critically ill patients. Crit. Care Med. 30, 315322 cells. Blood 105, 35883595 (2005). Monneret,G. Monitoring the immune response in
(2002). 68. Roquilly,A. etal. TLR4 agonist in post-haemorrhage sepsis: a rational approach to administration of
46. Delano,M.J. etal. Sepsis induces early alterations in pneumonia: role of dendritic and natural killer cells. immunoadjuvant therapies. Curr. Opin. Immunol. 25,
innate immunity that impact mortality to secondary Eur. Respir. J. http://dx.doi. 477483 (2013).
infection. J.Immunol. 186, 195202 (2011). org/10.1183/09031936.00152612 (2013). 92. Landelle,C. etal. Low monocyte human leukocyte
47. Altrichter,J. etal. Extracorporeal cell therapy of 69. Scumpia,P.O. etal. CD11c+ dendritic cells are antigenDR is independently associated with
septic shock patients with donor granulocytes: required for survival in murine polymicrobial sepsis. nosocomial infections after septic shock.
a pilot study. Crit. Care 15, R82 (2011). J.Immunol. 175, 32823286 (2005). Intensive Care Med. 36, 18591866 (2010).
48. Kasten,K.R., Muenzer,J.T. & Caldwell,C.C. 70. Cavaillon,J.M. & Adib-Conquy,M. Benchtobedside 93. Chiche,L. etal. The role of natural killer cells in sepsis.
Neutrophils are significant producers of IL10 during review: endotoxin tolerance as a model of leukocyte J.Biomed. Biotechnol. 2011, 986491 (2011).
sepsis. Biochem. Biophys. Res. Commun. 393, 2831 reprogramming in sepsis. Crit. Care 10, 233 (2006). 94. Souza-Fonseca-Guimaraes,F., Adib-Conquy,M. &
(2010). 71. Biswas,S.K. & Lopez-Collazo,E. Endotoxin tolerance: Cavaillon,J.M. Natural killer (NK) cells in
49. Pillay,J. etal. A subset of neutrophils in human new mechanisms, molecules and clinical significance. antibacterial innate immunity: angels or devils?
systemic inflammation inhibits Tcell responses Trends Immunol. 30, 475487 (2009). Mol. Med. 18, 270285 (2012).
through Mac1. J.Clin. Invest. 122, 327336 (2012). 72. Zhang,X. & Morrison,D.C. Lipopolysaccharide 95. Venet,F. etal. Early assessment of leukocyte
50. Delano,M.J. etal. MyD88dependent expansion structure-function relationship in activation versus alterations at diagnosis of septic shock. Shock 34,
of an immature GR1+CD11b+ population induces reprogramming of mouse peritoneal macrophages. 358363 (2010).
Tcell suppression and Th2 polarization in sepsis. J.Leukoc. Biol. 54, 444450 (1993). 96. Forel,J.M. etal. Phenotype and functions of natural
J.Exp. Med. 204, 14631474 (2007). 73. Monneret,G. etal. Persisting low monocyte human killer cells in critically-ill septic patients. PLoS ONE
This study identified MDSCs as important leukocyte antigenDR expression predicts mortality in 7, e50446 (2012).
mediators of immunosuppression in sepsis. septic shock. Intensive Care Med. 32, 11751183 97. Holub,M. etal. Lymphocyte subset numbers depend
51. Makarenkova,V.P., Bansal,V., Matta,B.M., (2006). on the bacterial origin of sepsis. Clin. Microbiol. Infect.
Perez,L.A. & Ochoa,J.B. CD11b+/Gr1+ myeloid 74. Rossato,M. etal. IL10induced microRNA187 9, 202211 (2003).
suppressor cells cause Tcell dysfunction after negatively regulates TNF-, IL6, and IL12p40 98. Giamarellos-Bourboulis,E.J. etal. Early changes of
traumatic stress. J.Immunol. 176, 20852094 production in TLR4stimulated monocytes. Proc. Natl CD4positive lymphocytes and NK cells in patients
(2006). Acad. Sci. USA 109, E3101E3110 (2012). with severe Gram-negative sepsis. Crit. Care 10, R166
52. Cuenca,A.G. etal. A paradoxical role for myeloid- 75. Ishii,M. etal. Epigenetic regulation of the (2006).
derived suppressor cells in sepsis and trauma. alternatively activated macrophage phenotype. 99. Souza-Fonseca-Guimaraes,F. etal. Toll-like receptors
Mol. Med. 17, 281292 (2011). Blood 114, 32443254 (2009). expression and interferon- production by NK cells in
53. Brudecki,L., Ferguson,D.A., McCall,C.E. & El 76. Carson,W.F., Cavassani,K.A., Dou,Y. & Kunkel,S.L. human sepsis. Crit. Care 16, R206 (2012).
Gazzar,M. Myeloid-derived suppressor cells evolve Epigenetic regulation of immune cell functions 100. Puente,J. etal. In vitro studies of natural killer cell
during sepsis and can enhance or attenuate the during post-septic immunosuppression. Epigenetics activity in septic shock patients. Response to a
systemic inflammatory response. Infect. Immun. 80, 6, 273283 (2011). challenge with -interferon and interleukin2.
20262034 (2012). 77. Turrel-Davin,F. etal. mRNA-based approach to monitor Int. J.Clin. Pharmacol. Ther. Toxicol. 31, 271275
54. Dumitru,C.A., Moses,K., Trellakis,S., Lang,S. & recombinant gamma-interferon restoration of LPS- (1993).
Brandau,S. Neutrophils and granulocytic myeloid- induced endotoxin tolerance. Crit. Care 15, R252 101. Georgeson,G.D. etal. Natural killer cell cytotoxicity
derived suppressor cells: immunophenotyping, cell (2011). is deficient in newborns with sepsis and recurrent
biology and clinical relevance in human oncology. 78. Chen,J. & Ivashkiv,L.B. IFN- abrogates endotoxin infections. Eur. J.Pediatr. 160, 478482 (2001).
Cancer Immunol. Immunother. 61, 11551167 (2012). tolerance by facilitating Toll-like receptor-induced 102. Blazar,B.A. etal. Suppression of natural killer-cell
55. Guisset,O. etal. Decrease in circulating dendritic chromatin remodeling. Proc. Natl Acad. Sci. USA 107, function in humans following thermal and traumatic
cells predicts fatal outcome in septic shock. 1943819443 (2010). injury. J.Clin. Immunol. 6, 2636 (1986).
Intensive Care Med. 33, 148152 (2007). 79. Monneret,G. etal. The anti-inflammatory response 103. Bender,B.S. etal. Depressed natural killer cell
56. Riccardi,F. etal. Flow cytometric analysis of dominates after septic shock: association of low function in thermally injured adults: successful invivo
peripheral blood dendritic cells in patients with severe monocyte HLADR expression and high interleukin10 and invitro immunomodulation and the role of
sepsis. Cytometry B Clin. Cytom. 80, 1421 (2011). concentration. Immunol. Lett. 95, 193198 (2004). endotoxin. Clin. Exp. Immunol. 71, 120125 (1988).

NATURE REVIEWS | IMMUNOLOGY VOLUME 13 | DECEMBER 2013 | 873

2013 Macmillan Publishers Limited. All rights reserved


REVIEWS

104. Chiche,L. etal. Interferon- production by natural 131. Kasten,K.R. etal. Interleukin7 (IL7) treatment 153. Pellegrini,M. etal. IL7 engages multiple mechanisms
killer cells and cytomegalovirus in critically ill patients. accelerates neutrophil recruitment through Tcell to overcome chronic viral infection and limit organ
Crit. Care Med. 40, 31623169 (2012). IL17 production in a murine model of sepsis. Infect. pathology. Cell 144, 601613 (2011).
105. Souza-Fonseca-Guimaraes,F., Parlato,M., Fitting,C., Immun. 78, 47144722 (2010). 154. Venet,F. etal. IL7 restores lymphocyte functions in
Cavaillon,J.M. & Adib-Conquy,M. NK cell tolerance 132. Zajac,A.J. etal. Viral immune evasion due to septic patients. J.Immunol. 189, 50735081 (2012).
to TLR agonists mediated by regulatory Tcells after persistence of activated Tcells without effector An important study demonstrating that IL7
polymicrobial sepsis. J.Immunol. 188, 58505858 function. J.Exp. Med. 188, 22052213 (1998). restores function in immune effector cells from
(2012). 133. Wherry,E.J. Tcell exhaustion. Nature Immunol. 12, patients with sepsis.
106. Luyt,C.E. etal. Herpes simplex virus lung infection in 492499 (2011). 155. Chang,K.C. etal. Blockade of the negative
patients undergoing prolonged mechanical ventilation. An outstanding review of the importance of Tcell costimulatory molecules PD1 and CTLA4 improves
Am. J.Respir. Crit. Care Med. 175, 935942 (2007). exhaustion in chronic infections and cancer. survival in primary and secondary fungal sepsis.
107. Limaye,A.P. etal. Cytomegalovirus reactivation in 134. Guignant,C. etal. Programmed death1 levels Crit. Care 17, R85 (2013).
critically ill immunocompetent patients. JAMA 300, correlate with increased mortality, nosocomial 156. Singh,A., Mohan,A., Dey,A.B. & Mitra,D.K.
413422 (2008). infection and immune dysfunctions in septic shock Inhibiting the programmed death 1 pathway rescues
108. Cook,C.H. & Trgovcich,J. Cytomegalovirus patients. Crit. Care 15, R99 (2011). Mycobacterium tuberculosis-specific interferon
reactivation in critically ill immunocompetent hosts: 135. Huang,X. etal. PD1 expression by macrophages -producing T cells from apoptosis in patients with
a decade of progress and remaining challenges. plays a pathologic role in altering microbial clearance pulmonary tuberculosis. J.Infect. Dis. 208, 603615
Antiviral Res. 90, 151159 (2011). and the innate inflammatory response to sepsis. (2013).
109. Chiche,L., Forel,J.M. & Papazian,L. The role of Proc. Natl Acad. Sci. USA 106, 63036308 (2009). 157. Docke,W.D. etal. Monocyte deactivation in septic
viruses in nosocomial pneumonia. Curr. Opin. Infect. This was the first study to show that blocking the patients: restoration by IFN- treatment. Nature Med.
Dis. 24, 152156 (2011). PD1 pathway can improve survival in sepsis. 3, 678681 (1997).
110. Wesselkamper,S.C. etal. NKG2D is critical for NK cell 136. Brahmamdam,P. etal. Delayed administration of One of the first studies to show that therapy with
activation in host defense against Pseudomonas antiPD1 antibody reverses immune dysfunction and agents that enhance host immunity administered
aeruginosa respiratory infection. J.Immunol. 181, improves survival during sepsis. J.Leukoc. Biol. 88, to patients in the immunosuppressive phase of
54815489 (2008). 233240 (2010). sepsis can improve survival.
111. Vantourout,P. & Hayday,A. Sixofthe-best:unique 137. Zhang,Y. etal. PDL1 blockade improves survival 158. Nalos,M. etal. Immune effects of interferon in
contributions of Tcells to immunology. Nature Rev. in experimental sepsis by inhibiting lymphocyte persistent staphylococcal sepsis. Am. J.Respir. Crit.
Immunol. 13, 88100 (2013). apoptosis and reversing monocyte dysfunction. Care Med. 185, 110112 (2012).
112. Andreu-Ballester,J.C. etal. Association of Tcells Crit. Care 14, R220 (2010). 159. Marciano,B.E. etal. Long-term interferon- therapy
with disease severity and mortality in septic patients. 138. Monneret,G. etal. Marked elevation of human for patients with chronic granulomatous disease.
Clin. Vaccine Immunol. 20, 738746 (2012). circulating CD4+CD25+ regulatory Tcells in sepsis- Clin. Infect. Dis. 39, 692699 (2004).
113. Hotchkiss,R.S. etal. Accelerated lymphocyte death induced immunoparalysis. Crit. Care Med. 31, 160. Jarvis,J.N. etal. Adjunctive interferon-gamma
in sepsis occurs by both the death receptor and 20682071 (2003). immunotherapy for the treatment of HIV-associated
mitochondrial pathways. J.Immunol. 174, 139. Zanin-Zhorov,A. etal. Heat shock protein 60 enhances cryptococcal meningitis: a randomized controlled trial.
51105118 (2005). CD4+ CD25+ regulatory Tcell function via innate TLR2 AIDS 26, 11051113 (2012).
114. Gouel-Cheron,A., Venet,F., Allaouchiche,B. & signaling. J.Clin. Invest. 116, 20222032 (2006). 161. Root,R.K. etal. Multicenter, double-blind, placebo-
Monneret,G. CD4+ Tlymphocyte alterations in 140. Huang,L.F. etal. Association between regulatory controlled study of the use of filgrastim in patients
trauma patients. Crit. Care 16, 432 (2012). Tcell activity and sepsis and outcome of severely hospitalized with pneumonia and severe sepsis.
115. Le Tulzo,Y. etal. Early circulating lymphocyte burned patients: a prospective, observational study. Crit. Care Med. 31, 367373 (2003).
apoptosis in human septic shock is associated with Crit. Care 14, R3 (2010). 162. Nelson,S. etal. A randomized controlled trial of
poor outcome. Shock 18, 487494 (2002). 141. Venet,F. etal. Increased circulating regulatory Tcells filgrastim as an adjunct to antibiotics for treatment
116. Inoue,S. etal. Reduction of immunocompetent Tcells (CD4+CD25+CD127) contribute to lymphocyte of hospitalized patients with community-acquired
followed by prolonged lymphopenia in severe sepsis anergy in septic shock patients. Intensive Care Med. pneumonia. CAP Study Group. J.Infect. Dis. 178,
in the elderly. Crit. Care Med. 41, 810819 (2013). 35, 678686 (2009). 10751080 (1998).
117. Heffernan,D.S. etal. Failure to normalize 142. Carrigan,S.O. etal. Depletion of natural CD4+CD25+ 163. Meisel,C. etal. Granulocyte-macrophage colony-
lymphopenia following trauma is associated with T regulatory cells with antiCD25 antibody does not stimulating factor to reverse sepsis-associated
increased mortality, independent of the leukocytosis change the course of Pseudomonas aeruginosa- immunosuppression: a double-blind, randomized,
pattern. Crit. Care 16, R12 (2012). induced acute lung infection in mice. Immunobiology placebo-controlled multicenter trial. Am. J.Respir.
118. Cheadle,W.G. etal. Lymphocyte subset responses 214, 211222 (2009). Crit. Care Med. 180, 640648 (2009).
to trauma and sepsis. J.Trauma 35, 844849 143. Scumpia,P.O. etal. Treatment with GITR agonistic 164. Reinhart,K., Bauer,M., Riedemann,N.C. &
(1993). antibody corrects adaptive immune dysfunction in Hartog,C.S. New approaches to sepsis: molecular
119. Venet,F. etal. Increased percentage of CD4+CD25+ sepsis. Blood 110, 36733681 (2007). diagnostics and biomarkers. Clin. Microbiol. Rev. 25,
regulatory Tcells during septic shock is due to the 144. Cavassani,K.A. etal. The post sepsis-induced 609634 (2012).
decrease of CD4+CD25 lymphocytes. Crit. Care Med. expansion and enhanced function of regulatory Tcells 165. Leentjens,J., Kox,M., van der Hoeven,J.G.,
32, 23292331 (2004). create an environment to potentiate tumor growth. Netea,M.G. &Pickkers,P. Immunotherapy for the
120. De,A.K. etal. Induction of global anergy rather than Blood 115, 44034411 (2010). adjunctive treatment of sepsis: from immuno
inhibitory Th2 lymphokines mediates posttrauma 145. Venet,F. etal. Human CD4+CD25+ regulatory T suppression to immunostimulation. Time for a
Tcell immunodepression. Clin. Immunol. 96, 5266 lymphocytes inhibit lipopolysaccharide-induced paradigm change? Am. J.Respir. Crit. Care Med. 187,
(2000). monocyte survival through a Fas/Fas ligand-dependent 12871293 (2013).
121. Heidecke,C.D. etal. Selective defects of T lymphocyte mechanism. J.Immunol. 177, 65406547 (2006). 166. Wong,H.R. Genetics and genomics in pediatric septic
function in patients with lethal intraabdominal 146. Li,L. & Wu,C.Y. CD4+ CD25+ Treg cells inhibit shock. Crit. Care Med. 40, 16181626 (2012).
infection. Am. J.Surg. 178, 288292 (1999). human memory Tcells to produce IFN- in 167. Mestas,J. & Hughes,C.C. Of mice and not men:
122. Wick,M., Kollig,E., Muhr,G. & Koller,M. The response to M tuberculosis antigen ESAT6. differences between mouse and human immunology.
potential pattern of circulating lymphocytes TH1/TH2 is Blood 111, 56295636 (2008). J.Immunol. 172, 27312738 (2004).
not altered after multiple injuries. Arch. Surg. 135, 147. Tiemessen,M.M. etal. CD4+CD25+Foxp3+ 168. Seok,J. etal. Genomic responses in mouse models
13091314 (2000). regulatory Tcells induce alternative activation of poorly mimic human inflammatory diseases.
123. Pachot,A. etal. Longitudinal study of cytokine and human monocytes/macrophages. Proc. Natl Acad. Sci. Proc. Natl Acad. Sci. USA 110, 35073512 (2013).
immune transcription factor mRNA expression in USA 104, 1944619451 (2007). 169. Brahmamdam,P. etal. Targeted delivery of siRNA to
septic shock. Clin. Immunol. 114, 6169 (2005). 148. Mackall,C.L., Fry,T.J. & Gress,R.E. Harnessing the cell death proteins in sepsis. Shock 32, 131139
124. Venet,F. etal. Regulatory Tcell populations in sepsis biology of IL7 for therapeutic application. (2009).
and trauma. J.Leukoc. Biol. 83, 523535 (2008). Nature Rev. Immunol. 11, 330342 (2011). 170. Weber,S.U. etal. Induction of Bim and Bid gene
125. Carson,W.F.t. etal. Impaired CD4+ Tcell proliferation An outstanding review article on a remarkable expression during accelerated apoptosis in severe
and effector function correlates with repressive cytokine that has great potential as an sepsis. Crit. Care 12, R128 (2008).
histone methylation events in a mouse model of immunotherapeutic agent for reversing
severe sepsis. Eur. J.Immunol. 40, 9981010 (2010). immunosuppression in various disorders, including Acknowledgements
126. Romani,L. Immunity to fungal infections. Nature Rev. sepsis and cancer. The authors would like to thank P. Swanson (Professor of
Immunol. 11, 275288 (2011). 149. Perales,M.A. etal. Recombinant human interleukin7 Pathology, University of Washington School of Medicine in
127. Gow,N.A., van de Veerdonk,F.L., Brown,A.J. & (CYT107) promotes Tcell recovery after allogeneic Seattle, Washington, USA) for providing the immunohisto-
Netea,M.G. Candida albicans morphogenesis and stem cell transplantation. Blood 120, 48824891 chemical images of patient spleens used in figure 2.R.S.H. is
host defence: discriminating invasion from colonization. (2012). supported by US National Institutes of Health grants
Nature Rev. Microbiol. 10, 112122 (2011). 150. Unsinger,J. etal. IL7 promotes Tcell viability, GM44118 and GM55194. G.M. is supported by funding
128. van de Veerdonk,F.L. etal. Deficient Candida-specific trafficking, and functionality and improves survival from the Hospices Civils de Lyon.
Thelper 17 response during sepsis. J.Infect. Dis. 206, in sepsis. J.Immunol. 184, 37683779 (2010).
17981802 (2012). 151. Venet,F. etal. Decreased Tcell repertoire diversity Competing interests statement
129. Monneret,G., Venet,F., Kullberg,B.J. & Netea,M.G. in sepsis: a preliminary study. Crit. Care Med. 41, The authors declare no competing interests.
ICU-acquired immunosuppression and the risk for 111119 (2013).
secondary fungal infections. Med. Mycol. 49 (Suppl. 152. Levy,Y. etal. Effects of recombinant human interleukin
1), S1723 (2011). 7 on Tcell recovery and thymic output in HIV-infected DATABASES
130. Unsinger,J. etal. Interleukin7 ameliorates immune patients receiving antiretroviral therapy: results of a ClinicalTrials.gov: http://clinicaltrials.gov/
dysfunction and improves survival in a 2hit model of Phase I/IIa randomized, placebo-controlled, multicenter ALL LINKS ARE ACTIVE IN THE ONLINE PDF
fungal sepsis. J.Infect. Dis. 206, 606616 (2012). study. Clin. Infect. Dis. 55, 291300 (2012).

874 | DECEMBER 2013 | VOLUME 13 www.nature.com/reviews/immunol

2013 Macmillan Publishers Limited. All rights reserved