Neurotherapeutics (2020) 17:392–403

https://doi.org/10.1007/s13311-020-00862-1

REVIEW

Septic-Associated Encephalopathy: a Comprehensive Review

Aurélien Mazeraud 1,2,3 & Cássia Righy 1,4 & Eleonore Bouchereau 1,2,3 & Sarah Benghanem 3,5 &
Fernando Augusto Bozza 6 & Tarek Sharshar 1,3

Published online: 6 May 2020

# The American Society for Experimental NeuroTherapeutics, Inc. 2020

Abstract
Septic-associated encephalopathy (SAE) is a key manifestation of sepsis, ranging from delirium to coma and occurring in up to 70%
of patients admitted to the ICU. SAE is associated with higher ICU and hospital mortality, and also with poorer long-term outcomes,
including cognitive and functional outcomes. The pathophysiology of SAE is complex, and it may involve neurotransmitter
dysfunction, inflammatory and ischemic lesions to the brain, microglial activation, and blood–brain barrier dysfunction. Delirium
(which is included in the SAE spectrum) is mostly diagnosed with validated scales in the ICU population. There is no established
treatment for SAE; benzodiazepines should generally be avoided in this setting. Nonpharmacological prevention and management is
key for treating SAE; it includes avoiding oversedation (mainly with benzodiazepines), early mobilization, and sleep promotion.

Key Words Sepsis . neuroinflammation . sepsis-associated encephalopathy . microglia . blood–brain barrier . neuroanatomy

Introduction about 50%. It varies from 8% to more than 70% of septic

patients, according to the sepsis severity, patients’ profile,
Since ancient times, physicians have recognized that the cen- and SAE diagnostic criteria [4–6]. It can be the revealing
tral nervous system (CNS) is one of the first organs affected in manifestation of sepsis, and alteration in mental status has
sepsis, and its clinical manifestation is the so-called sepsis- been identified as 1 of the 3 screening items for detecting
associated encephalopathy (SAE) [1]. Sepsis, defined as a sepsis [7]. Its pathophysiology comprises neuroinflammation,
life-threatening organ dysfunction caused by a dysregulated vascular changes, and metabolic failure leading to tissue le-
host response to infection [2], is a leading cause of ICU ad- sions seen in animal models and observed in humans [1].
mission and death worldwide [3]. The incidence of SAE is These mechanisms are not homogeneous in the brain and
may lead to region-specific lesions. The centers involved in
autonomic controls, arousal, awareness, and behavior are par-
Invited review for the neurocritical care special issue ticularly affected, accounting for the clinical features of SAE
varying from sickness behavior to consciousness impairment
Electronic supplementary material The online version of this article
(https://doi.org/10.1007/s13311-020-00862-1) contains supplementary (i.e., ranging from delirium to coma) [1, 8]. SAE is character-
material, which is available to authorized users. ized by changes in the electroencephalogram (EEG), with a
nonconvulsive status epilepticus that can be detected in up to
* Tarek Sharshar 20% of cases [9]. The radiologic lesions are uncommon, but
[email protected] white matter hypersignal and ischemic stroke can be observed
1
in the subgroup of patients with septic shock ischemic lesions
GHU Paris Psychiatrie et Neuroscience, Neurointensive Care and in about 30% of patients [10–12]. SAE is associated with
Neuroanesthesia Department, 1, rue Cabanis, 75014 Paris, France
2
increased mortality [4, 13] as well as long-term cognitive
Medical Intensive Care Unit, Hôpital Européen Georges Pompidou, impairment—mainly affecting memory, attention, and verbal
20 rue Leblanc, 75015 Paris, France
3
fluency [14, 15]—and psychological disorders, including de-
Université de Paris, 75006 Paris, France pression, anxiety, and posttraumatic stress disorder (PTSD)
4
Instituto Estadual do Cérebro Paul Niemeyer, Rio de Janeiro, Brazil [16, 17]. Its management relies mainly on general ICU good
5
Médecine Intensive et Réanimation, Centre Hospitalier Universitaire practices as specific treatment is still lacking.
Cochin, Paris, France The brain is an immune-privileged organ as it is less subject
6
Instituto D’Or de Pesquisa e Ensino, Rio de Janeiro, Brazil to the immune response, with consequences to make it tolerant
Septic-Associated Encephalopathy: a Comprehensive Review 393

to tumors and graft. Now the current thinking is that the CNS Specific pathways mediate this signal to the brain and trig-
can be profoundly affected by severe systemic infections. gers 3 processes which are neuroinflammation, ischemia, and
Therefore, in this chapter, we will review the main character- cellular metabolic stress [27]. It is likely that complement
istics of SAE and its pathophysiology, as well as its therapeu- system activation contributes to the BBB dysfunction [28].
tics and prognosis. Macro- and microcirculatory dysfunctions induce ischemia;
the dysfunction of the vascular complex (which includes the
endothelial cells, astrocytes, and blood–brain barrier) and the
activation of microglia lead to neuroinflammation. These 2
Epidemiology
processes add to other systemic factors (i.e., drug neurotoxic-
ity, hypoxia, dysglycemia, renal or liver failure) and induce
Because of the absence of consensual definition of SAE, there
metabolic stress (Fig. 1). Oxidative stress closely linked to
is a great variation in its incidence across studies. Thus, 20 to
dysfunction of the mitochondria alters neuronal function and
40% of septic patients admitted to the ICU will develop an
vitality [26, 29]. Specific brain regions are particularly sensi-
encephalopathy. It has been reported that delirium is observed
tive to these processes, those involved in autonomic control,
in up to 70% of elderly mechanically ventilated patients [14,
arousal, defining behavioral response to stress, and more com-
18]. The most consistent risk factors are age, previous cogni-
plex cognitive functions such as memory and attention [30].
tive impairment, kidney and liver failure, and sepsis severity
Strong neuronal activation is consistently evidenced in the
[4, 14]. Similarly, bacteremia is accompanied by changes in
hippocampus, the amygdala, the nucleus tractus solitarii, and
neurological status, ranging from lethargy to coma, in about
the locus coeruleus [31]. An increased neuronal apoptosis
70% of cases [4]. Finally, only 19% of the patients admitted to
associated with microglial activation has been specifically ev-
the ICU showed a normal EEG with alpha rhythm and 80%
idenced in these areas by the neuropathological study of the
present EEG background or epileptic anomalies [19, 20].
brain from patients who died from sepsis [32]. These findings
It has been clearly shown that occurrence of encephalopa-
provide the anatomic substrate of acute and long-term conse-
thy increases the risk of death in septic patients, although the
quences of SAE.
mechanisms underlying the relationships between SAE and
mortality are not elucidated yet. Mortality rate is mainly asso-
Brain Signaling
ciated with the clinical and electrophysiological severity of
SAE.
The inflammatory signal in the brain involves the humoral and
Septic patients are at risk also of developing long-term
neural pathways. For example, the intraperitoneal inflamma-
cognitive impairment and psychological disorders. Thus, hos-
tion signal is conveyed by vagal afferents to the medullary
pitalization for sepsis is associated with a 10% increase in the
autonomic nuclei and the vagus nerve is able to modulate local
prevalence of cognitive impairment during 8 years [14, 21].
and systemic inflammation (by its peritoneal and splenic in-
Attention, verbal fluency, executive function, verbal memory,
nervations respectively). The neural pathways could also com-
and quick mental processing are the main cognitive functions
prise other nerves and nuclei such as the trigeminal nerve and
impaired, whereas visual memory and visuoconstructive abil-
nucleus. The medullary autonomic nuclei can modulate more
ity are usually spared [15, 22]. The psychological disorders
widely the response to the sepsis by their connections to other
include anxiety, depression, and PTSD [17, 23]. Sepsis even
autonomic neuroendocrine and behavioral centers [32]. Areas
increases the risk for suicide within 2 years after its recovery
deprived of BBB form the humoral pathway involving
[24]. These psychological and cognitive disorders dramatical-
circumventricular organs and the area postrema.
ly impact quality of life and functional status [15]. At 1 year,
Inflammatory mediators are then allowed to traffic towards
up to 51% of septic patients have not returned to full-time
the local neuroendocrine and autonomic centers [33]. These
employment [25]. It is considered that the cognitive dysfunc-
2 pathways orchestrate the inflammatory stress response clin-
tion results from early sepsis-related insults of the hippocam-
ically observable as the sickness behavior.
pus and frontal lobe [21, 26] and the psychological disorders
Blood–brain barrier (BBB) dysfunction is also observed
from those involving the limbic system [1].
during sepsis. It controls the blood–brain water, molecule,
and ion balance and restrains immune cell, toxin, and patho-
gen crossing [34, 35]. Astrocytes and pericytes maintain its
Pathophysiology integrity. In septic shock, the neuropathological, electron mi-
crograph examination and MRI studies confirm the BBB im-
Systemic inflammation leads to deleterious effects on the pairment, responsible for vasogenic edema in the white matter
brain parenchyma resulting in SAE. SAE can be evidenced [11, 36]. Posterior reversible encephalopathy syndromes are
clinically, electrophysiologically, or radiologically and can be evidenced in septic patients, corroborating these findings [37].
modeled in animal studies. BBB impairment is also experimentally modeled, reproducing
394 Mazeraud et al.

Fig. 1 Schematic view of the different pathophysiological processes uncoupling, and strokes. Neuroinflammation includes microglial and as-
observed or supposed during sepsis-associated encephalopathy. trocytic activation enhancing excitotoxicity and metabolic imbalance in-
Vascular changes include blood–brain barrier dysfunction, neurovascular ducing neuronal cell death

endothelial activation, complement activation, and the de- Upon stimulation through circulating endotoxins or toll-
crease in tight junction protein expression, namely occludin, like receptor pathways, microglial cells activate and present
ZO-1, ZO-2, claudin-3, and claudin-5 [38]. Also, an upregu- morphological, immunological, and metabolic changes.
lation of aquaporine 4 has been evidenced, probably favoring Morphological changes are mostly characterized by a short-
brain edema in patients [36]. ening of their processes called deramification, up to an amoe-
boid form to the extreme. Various immunological phenotypes
of microglia activation are observed in response to such acti-
Neuroinflammation and Microglial Activation vation, ranging from proinflammatory (M1) releasing proin-
flammatory cytokines (such as gamma interferon or tumor
In the brain, microglial cells are the main macrophage cells, necrosis factor alpha) to anti-inflammatory (M2) ones releas-
representing most of the brain immune system. They also ing immune modulatory cytokines (such as IL4 or IL10). The
prune unused synapses and secrete neurotrophic factors con- neurotoxic consequences are usually considered secondary to
tributing to synaptic plasticity. They also present phagocytic, proinflammatory phenotypes whereas anti-inflammatory phe-
migration, proliferation, and various mediator release capaci- notypes could be neuroprotective [39]. The neuronal activa-
ties. Most of their various surface receptors interact with the tion and dysfunction is in part secondary to microglia activa-
peripheral immune system notably through cytokine ligation tion; thus, the neuroinflammatory process contributes to long-
but can also sense damage- and pathogen-associated molecu- term consequences of SAE. Cytokines, nitric oxide, excitatory
lar patterns (DAMPs and PAMPs) [39]. gliotransmitters, and neurotoxic metabolites (such as reactive
Septic-Associated Encephalopathy: a Comprehensive Review 395

oxygen species) mediate the microglia-mediated neurotoxici- exchange of “gliotransmitters,” ions, or small molecules be-
ty through an increase in neuronal excitability leading to hy- tween astrocytes and neurons or in the extracellular milieu
peractivation and excitotoxicity [39, 40]. (hemichannels). In a neuroinflammatory context induced by
Animal [41, 42] and human [43, 44] studies during sepsis LPS, microglial cell activation inhibits gap junction channels
consistently show microglia activation without any known and open hemichannels [59]. Whether this occurs in SAE has
cerebral infection. This concept has recently been challenged to be evidenced.
as bacterial genomic material and living bacteria have been In animal models of SAE, astrogliosis has been consistent-
found in animals and humans without them leading to infec- ly evidenced [50]; however, these findings have never been
tious encephalitis [45]. Whereas brain microbiota is consid- reported in humans. Human studies rely solely on neuropath-
ered absent in healthy subjects, this might be challenged in the ological findings and morphological analysis of glial fibrillary
critically ill [46]. Thus, targeting microbiota in the ICU could acid protein (GFAP) staining, which are insufficient to analyze
be a relevant therapy in the future [47, 48]. the functionality of neuron–glia interaction and astrocytic
As microglia activation induces most probably brain dam- network.
age during sepsis, its modulation seems a relevant approach
for treating SAE [49, 50]. Promising experimental results have
been published using various interventions including Ischemic Processes
minocycline [51], cholinergic inhibition [52, 53], and vagal
nerve stimulation [54]. Rivastigmine administration has been Ischemic processes are separated in macrocirculatory dysfunc-
an original and convincing lead in the sepsis setting. It has tion including hypotension, decreased cerebral flow, and im-
been shown to reduce microglia activation by restoring the paired autoregulation—and microcirculatory impairment,
cholinergic inhibition. However, a randomized clinical trial characterized by neurovascular uncoupling impairment, dis-
evaluating rivastigmine has been prematurely interrupted for ruption of the blood–brain barrier, and coagulation cascade
increasing mortality showing that manipulating the microglia activation. Septic shock induces ischemic damages observed
may be hazardous [55]. The absence of a microglial pheno- in all cases of patients dead of septic shock which are associ-
type biomarker limits clinical and experimental research ated in ~ 20% of cases with microhemorrhages associated to
allowing to classify easily such a dynamic phenomenon as disseminated intravascular coagulopathy [60].
microglial activation. Microglial activation has been charac- The cerebral blood flow (CBF) is maintained constant by
terized in Alzheimer disease using PET-CT, but its feasibility autoregulation as mean systemic arterial pressure ranges be-
limits its transposition to septic patients until further improve- tween 60 and 150 mmHg. Below 60 mmHg, CBF decreases
ments are made [56]. and induces olighemia whereas when MAP is above
One dementia pathophysiology hypothesis is a 150 mmHg, hyperhemia is observed. A decrease in the CBF
neuroinflammatory cascade as microglial activation is consis- has been consistently evidenced in septic shock, and an im-
tently activated in such diseases. In these pathologies, pairment of its autoregulation has been evidenced to be asso-
microglial cells are considered as “primed” and overactivate ciated with delirium [29, 61].
after being submitted to a second hit. Sepsis can be considered Brain microcirculation is finely regulated through complex
the primer or the second hit [57]. This hypothesis can there- mechanisms. These depend on the gliovascular unit composed
fore account for the relationship between the rate of sepsis of endothelial cells, astrocytes, and pericytes [62]. When neu-
occurrence and intensity of the cognitive decline but also to ronal activity increases and consumes oxygen, the
its deleterious impact on patients with pre-existing neurode- gliovascular unit couples the cerebral flow by vasodilation
generative disease. increasing locally the energy supply [62]. In experimental
models of sepsis, the microcirculatory dysfunction has been
Astrocytes and Blood–Brain Barrier Dysfunction evidenced by neurovascular uncoupling impairing cerebral
vascular supply and leading to a deleterious metabolic crisis
Astrocytes represent the brain’s most numerous cells. They [26, 63].
play a major role in brain homeostasis, and astrocyte dysfunc- In specific areas during sepsis, the metabolism is increased.
tion has been identified as a potential mechanism of SAE. For a normal mean blood pressure between 65 and 70 mmHg,
BBB permeability, brain water balance, and microcirculatory the autoregulating mechanisms of the CBF allow the blood
cerebral brain flow (via the release of NO, prostaglandins, and flow to match the energy demand. Sepsis can also impair
arachidonic) are controlled by astrocytes [58]. They also par- cerebral autoregulation, compromising a major protection of
ticipate in synaptic plasticity in the “tri-partite” synapse, re- the brain from ischemia. Thus, a decrease in mean arterial
leasing and reuptaking glutamate, GABA, and glycine. blood pressure below 70 mmHg can result in ischemic dam-
Astrocytes form extensive networks through gap junctions age. Such disturbance could induce a mismatch between ce-
of connexins (Cx), channels permitting the bidirectional rebral blood flow and metabolic demand [32].
396 Mazeraud et al.

Finally, such ischemic process could be involved in long- Functional Neuroanatomy of SAE
term cognitive decline as a similar mechanism to vascular
dementia. A neuroanatomical approach to SAE helps in understanding
not only SAE clinical features but also its associated increased
Mitochondrial Dysfunction mortality and long-term psychocognitive disorders (Fig. 2).

The metabolic and bioenergetics demands may result in oxi- Neuroanatomy of Response to Stress
dative stress and mitochondrial dysfunction during sepsis
[26]. Indeed, early mitochondrial dysfunction has been shown During sepsis, the peripheral inflammation is transmitted to
in various brain regions of septic animals, resulting in reduced structures controlling autonomic and neuroendocrine systems
ATP generation and the production of oxygen/nitrogen reac- interconnected to each other and to behavior and cognition
tive species [64]. This metabolic process is proapoptotic and centers, which in turn regulates the immune response through
involves the glial cells and neurons. the “inflammatory reflex” [77–80]. The intensity of such re-
sponse may be qualified in regard to the severity of sepsis as
Neurotransmitter Dysfunction adapted, maladapted, or pathological (overactivated or
blunted) [81, 82]. Such a distinction implies that recovery or
Aside from mitochondrial dysfunction and oxidative stress, not of sepsis might be influenced by the CNS response.
neurons are liable to excitotoxicity [65]. Neurotransmitter However, there is no clinicobiological criteria to establish
concentration in the synaptic cleft is the basis for excitotoxic such a categorization. Through the neural pathway, the vagus
processes occurring when neurotransmitter release is in- nerve activates early the nucleus tractus solitarii and the locus
creased whereas its reuptake is diminished or insufficient coeruleus [31, 83, 84], which activate compensatory mecha-
[66]. The excitotoxic process is a pathological process occur- nisms in sepsis such as the control of blood pressure, heart
ring in neuroinflammation and ischemia but is also secondary rate, and arousal. These neural centers act as entry points to the
to systemic perturbations such as hypoxemia, electrolyte dis- CNS and stimulate the other autonomic nuclei and the behav-
orders, dysglycemia, drug toxicity, and excess circulating neu- ioral and neuroendocrine centers. The precocity of the CNS
rotoxic amino acids (ammonium, tyrosine, tryptophan, and activation by the neural route might explain that sickness be-
phenylalanine) [67, 68]. Excitotoxicity through glutamate re- havior is one of the earliest features of sepsis (Fig. 3) [85].
ceptors (N-methyl D aspartate receptor) induces functional Sickness behavior is a physiological response to a systemic
impairment but also cell death, which could account for the inflammation that is clinically characterized by social with-
clinical and electrophysiological features of SAE. Various drawal, impaired cognition (psychomotor retardation, im-
synapses seem to be involved in the SAE, notably dopaminer- paired attention), altered alertness (anxiety, hypersomnia, fa-
gic, β-adrenergic [69], the GABA [70], and the cholinergic tigue, sleepiness), and eating disorders (anorexia, weight loss,
ones [52, 71]. During delirium, a dopamine/choline imbalance thirst). Thus, a severe sickness behavior can mimic a
is suspected [72] but its modulation with cholinergic drugs hypoactive delirium [86]. Use of a specific scale for these
such as rivastigmine [55] or antidopaminergic drugs, like hal- entities is therefore clinically useful [87]. Interestingly, in
operidol [73], has not shown benefits in patients. On the other brains of patients who died from septic shock, the neuropath-
side, the risk of delirium is increased by benzodiazepines ological examination of response to stress structures consis-
(GABA agonists) [74], but not reduced by noradrenergic tently showed an increased number of apoptotic cells. This
drugs (i.e., dexmedetomidine) in septic patients [75]. finding underlines their particular sensitivity to the excitotoxic
process [32, 60, 85].
Iatrogenic Factors
Brainstem and Mortality and Delirium
There are various drugs commonly administered in septic pa-
tients that can be neurotoxic, including antibiotics and seda- Four broad functions are controlled by the brainstem:
tive agents. Thus, antibiotics overdose has been shown to be brainstem reflexes, sleep–wake cycle, control of vital func-
associated with delirium. It has to be noted that antibiotic tions, and modulation of the immune response [88].
neurotoxicity is not always related to an overdose. Brainstem dysfunction might then result in swallowing disor-
Benzodiazepines have been clearly established to induce de- ders, impaired arousal, cardiovascular sympathetic activity
lirium. All pharmacological factors have to be systematically and respiratory drive dysfunction, and both cholinergic and
checked. ICU environment and sleep deprivation are also in- adrenergic immunomodulation. These changes might account
volved in the occurrence of delirium. The use of the ABCDEF for the increased mortality, development of organ failure, and
bundle can be applied for controlling potential factors of de- fluctuating wakefulness. We have found in deeply sedated
lirium in septic patients [76]. critically ill patients that a specific pattern of heterogeneous
Septic-Associated Encephalopathy: a Comprehensive Review 397

Fig. 2 Early signs observed during sepsis (upper row), relying on specific inflammation; delirium and consciousness disorders are clinical signs of
structures (middle row) and associated with worst outcome (right row). SAE. SAE = sepsis-associated encephalopathy, PTSD = posttraumatic
Sickness behavior is considered as a physiological response to systemic stress disorder

brainstem reflex abolition was associated with higher mortal- are supposed to account for the multifocal necrotizing
ity and delirium to the absence of oculocephalogyre reflex leukoencephalopathy associated with marked apoptosis with-
[89]. Also, a dysfunction of the reticular activating ascending in the brainstem autonomic nuclei reported in a patient who
substance has been suggested by the absence of EEG reactiv- died from septic shock [92].
ity, and is predictive of death in septic patients [90]. In addi-
tion, critical illness increases brainstem latencies of auditory Amygdala and Psychological Disorders
evoked potential suggesting common mechanisms of sepsis
with other pathologies [91]. Postsepsis psychological disorders, also gathered under the
Finally, in septic patients, the decrease in heart rate vari- name “postsepsis syndrome,” involve fear and anxiety cir-
ability is associated with bad outcome and reflects the sympa- cuits, notably the amygdala, the bed nucleus stria terminalis,
thovagal imbalance [81]. Sepsis is also associated with an the hypothalamus, or other brainstem nuclei [93]. Amygdala
impairment in the baroreflex, controlled by medullar nuclei. nuclei are involved in fear behavior expression (generaliza-
Circulating inflammatory mediators can diffuse at the level of tion, freezing) and traumatic memory formation as well as
the area postrema, inducing an intense neuroinflammatory anxiety and depression symptoms. Also, within the response
process leading to brainstem dysfunction. These mechanisms to stress network, amygdala mediates the response to stress
398 Mazeraud et al.

Fig. 3 Scheme for the response to stress network mobilization during the vagus nerve nuclei. Behavioral, neuroendocrine, and autonomic
sepsis. The peripheral inflammation signal is transmitted to the CNS structures are interconnected and stimulated, leading to the functional
through 3 main pathways (dark blue arrows) activating specific response (dark green arrows and rectangles). CNS = central nervous
structures (light blue arrows), the circumventricular organs (CVO), or system, CVO = circumventricular organs, BBB = blood–brain barrier

network behavioral changes and the sickness behavior fea- Finally, MRI studies in patients with septic shock showed
tures (anorexia, anxiety, avoidance). In a model of LPS injec- white matter damages as expected according to the patholog-
tions, the amygdala microcircuitry has been shown to define ical findings [11]. This sepsis-induced axonopathy is largely
anorexic behavior [94]. Within the amygdala, microglia seems unknown but might be similar to those of critical illness
to be particularly involved [84], maybe explaining the effi- polyneuropathy. It is highly conceivable that it contributes to
ciency of corticosteroids in the prevention of PTSD in septic the long-term cognitive impairment.
patients [95]. The beneficial effects of corticoids might be
explained on gene transcription but also through epigenetics
pathways involved in aversive memory formation. Diagnosis

Hippocampus, Frontal Cortex, White Matter, Clinical Features

and Cognitive Decline
SAE is defined by the combination of extracranial infection
Memory formation requires an intact hippocampus and frontal with clinical signs of neurological dysfunction. SAE clinical
cortex. In septic animals, hippocampal dysfunction has been manifestation encompasses impairment of awareness, which
modeled and excitotoxicity secondary to long-term potentia- ranges from delirium (50%) to coma (46%) [6]. Sepsis-related
tion, gliosis, neuroinflammation, and increased production of delirium is rather hypoactive than hyperactive. SAE be asso-
reactive oxygen species (ROS) hypoxemia and ischemia lead ciated with focal deficits (18%) [11], seizures (10%) [20, 96],
to late neuronal death [51]. As these processes might be nonepileptic myoclonus, tremor, or asterixis [96]. CAM-ICU
prolonged after sepsis control, it is not surprising that hippo- and ICSDC can be used for diagnosing delirium and GCS and
campal dysfunction can progress toward atrophy [21]. Sepsis FOUR for monitoring coma. In deeply sedated patients, the
alters not only memorization but also other frontal cortex- BRASS score can be helpful for detecting brainstem dysfunc-
specific cognitive functions such as attention, verbal fluency, tion [18].
and executive functions [14]. Micromacrovascular and neuro- It has to be underlined that sickness behavior is the first
inflammation within the frontal cortex can contribute widely manifestation of sepsis, which can be distinguished from
to cognitive impairment. hypoactive delirium with the help of an appropriate scale.
Septic-Associated Encephalopathy: a Comprehensive Review 399

Second, the ICU physician must look for sepsis in any patient and delirious septic patients. Moreover, EEG can be abnormal
who develops change in mood, behavior, or consciousness. even without changes in neurological status [10].
The suspicion of a brain infection must prompt the ICU phy- These findings suggest that EEG changes are rather
sician to perform brain imaging and lumbar puncture. Finally, markers of severity than pathogenic. Indeed, the prognosis
changes in the mental status of a septic patient can be related value of EEG patterns has been clearly evidenced. Mortality
to other causes besides infection (i.e., electrolyte disturbances, increased from 0% in the case of normal EEG to 50 to 67% in
drug overdose or withdrawal, vitamin deficiency, etc.). the case of triphasic waves or suppression. Generalized PED
and absence of reactivity are predictive of mortality but also
Neurophysiological Features subsequent occurrence of delirium after adjustment to severity
of sepsis and sedation [20]. Finally, it has been reported that
There are 2 prospective electrophysiological studies in septic sepsis slightly increases the risk of long-term epilepsy. There
patients applying the ACNS guidelines [20, 96, 97]. A recent is no evidence for recommending a preventive antiepileptic
systematic review indicates that EEG and evoked potentials therapy as well as no clear recommendation on how to treat
are sensitive, but not specific, to SAE [9]. electrographic seizure in septic patients. However, we think
continuous EEG should be used, whenever it is available, in
Electroencephalogram septic patients in order to detect nonconvulsive seizures, es-
pecially in case of renal or hepatic insufficiency. All potential
The electrophysiological classification developed by Young epileptogenic factors must be assessed in septic patients, in-
et al. for the classification of SAE severity takes mainly into cluding antibiotic neurotoxicity, drug overdose or withdrawal,
account the background activity and includes 4 grades (grade electrolyte disturbances, etc. Finally, we recommend brain
0, normal EEG; grade 1, predominant theta activity; grade 2, imaging in case of electrographic seizure or malignant EEG
predominant delta activity; grade 3, predominant triphasic pattern.
waves; and grade 4, suppression) [19]. In a prospective study
based on 20-min standard EEG done within the first 3 days Evoked Potentials
after ICU admission, Azabou et al. detected delta and theta
predominant rhythms in 33% and 48%, triphasic waves in 6%, Subcortical (i.e., N20–N23 interlatency) and cortical (N20–
and suppression in less than 3% among 110 septic patients N70 interlatency) pathways of somatosensory evoked poten-
[20]. Low voltage and absence of reactivity were seen in tial (SSEP) are impaired early in 34% and 84% of 68 septic
65% and 25% of cases, respectively. The predictors of mor- patients [99]. Evoked potentials are useful for assessing SAE
tality and occurrence of delirium were absence of EEG reac- in sedated critically ill patients, as, in contrast to EEG, their
tivity, a delta-predominant background, periodic discharges latencies are slightly altered by sedatives. Thus, increased
(PDs), Synek grade ≥ 3, and Young grade > 1 at day 1 to 3 P14–N20 SSEP interlatency was associated with mortality in
following admission. In a study evaluating continuous EEG in this population [91].
98 severely septic patients, Gilmore et al. detected
nonconvulsive status (NCS), PD, and lack of reactivity in Brain Imaging
11%, 25%, and 28% of cases, respectively [96]. Only the lack
of reactivity was associated with mortality at 1 year. There is retrospective or prospective imaging studies on a
The incidences of pseudoepileptic discharges (PED) and homogeneous and large cohort of septic patients. The only
epileptic seizures vary between 19 and 48% and 9 to 50% prospective MRI study was performed on 72 septic shock
[98], respectively. The discrepancies between studies in terms patients who developed an acute brain dysfunction defined
of sepsis severity, time, duration and type of EEG recording, by focal neurological sign (18%), seizure (10%), coma
and EEG criteria account for these wide ranges. It has to be (46%), and delirium (49%). The MRI was normal in half of
noted that electrographic seizures are always associated with cases in this selected cohort. New white matter hyperdensities
PED, which are predictive, when generalized and frequent, of (suggestive of vasogenic edema) and ischemic stroke were
subsequent electrographic seizure. Continuous EEG monitor- observed in half of the remaining cases [11]. Sepsis can be
ing has evidenced that 75–95% of epileptic events are detected complicated by posterior reversible encephalopathy syndrome
in the first 48 h [10]. (PRESS) [37]. Because of the risk of transporting septic pa-
There is no strong correlation between clinical manifesta- tients, brain imaging should be indicated upon relevant clini-
tion and EEG findings. Thus, clinical seizures are exception- cal deterioration, including focal neurological deficit, seizure,
ally reported in septic patients, indicating that electrographic and unexplained impaired consciousness. MRI is more specif-
seizures are mostly nonconvulsive [96]. If electrographic sei- ic and sensitive than brain CT scan, which conversely is easier
zures are more frequent in patients with delirium at time of and may complete a body scan. Transcranial Doppler imaging
EEG, EEG does not show epileptic activity in most comatose could also predict the occurrence of SAE with good diagnostic
400 Mazeraud et al.

performance; however, these results are not found consistent, showed that early goal-directed sedation based on
because of discrepancy in population, severity of sepsis, time dexmedetomidine did not show any benefit in critically ill
of the test, and TCD criteria. patients [75].
While SAE could result from a mismatch between energy
Biomarkers demand and cerebral blood flow, a personalized target of mean
arterial pressure defined upon Doppler criteria could be eval-
Biomarkers might be helpful for detecting and monitoring uated in further work. However, such an approach has never
SAE by targeting different structures involved in the patho- been evaluated.
physiological processes. Some markers of systemic inflamma- The nonpharmacological interventions recommended for
tion can be associated with the occurrence of SAE, such as C preventing or treating delirium should be applied in septic
reactive protein or procalcitonin. There are also various patients, including early mobilization [107], discontinuation
markers of brain damage. N-Terminal propeptide of CNP of sedation [106], rehydration, management of physical and
(NT-proCNP), protein S100b, and neuron-specific enolase psychological discomfort [108], and avoidance of prodelirious
(NSE) or neurofilament, are respectively biomarkers of endo- drugs.
thelial dysfunction, microglial activation, and brain injury
with axonal damage [100–102]. They can be measured in
the plasma or in the CSF. Although their increase can correlate
with the severity of SAE, their sensibility and specificity re-
Conclusion
main low, limiting their clinical relevancy. One may argue that
SAE is a major complication of sepsis. It is characterized by
their combined assessment might improve their clinical and
changes in neurological status, ranging from sickness behav-
prognosis value.
ior to delirium to coma. It is associated with changes in EEG
background activity and less frequently with electrographic
seizures. It has a major impact on outcome, as it increases
Management
mortality and causes long-term psychocognitive disorders.
Neuroinflammation, ischemia, and excitoxicity are its main
The treatment of SAE is based both on the management of
pathophysiological mechanisms. Brainstem dysfunction
sepsis and on the correction of potential neurotoxic factors and
might account for increased mortality while amygdala and
management of seizure, delirium, and coma. There is no spe-
hippocampus/frontal cortex dysfunction for psychological
cific treatment of SAE. Recent randomized clinical trials have
and cognitive disorders. There is no specific treatment for
not evidenced any benefit of statins and dexmedetomidine for
SAE, which depends on management of sepsis, delirium, sei-
preventing delirium in septic patients [103, 104], despite their
zure, and coma.
respective anti-neuroinflammatory and noradrenergic/anti-
apoptotic properties. If rivastigmine was theoretically interest-
ing because of its potential microglial effect, it has been shown
to increase mortality without reducing delirium in ICU pa-
tients, including septic ones. The SAILS and HARP2 trials
failed to further demonstrate such effects in an interventional References
trial [103, 105].
There is no specific recommendation on pharmacological 1. Mazeraud A, Pascal Q, Verdonk F, Heming N, Chrétien F,
management of sepsis-associated delirium and seizure. It has Sharshar T (2016) Neuroanatomy and physiology of brain dys-
function in sepsis. Clin Chest Med 37:333–345
to be reminded that fever can be due to neuroleptic malignant
2. Singer M, Deutschman CS, Seymour CW, et al (2016) The Third
syndrome and choice of the anti-epileptic drug depends on International Consensus Definitions for Sepsis and Septic Shock
renal and liver organ failure. To date, no specific sedative (Sepsis-3). JAMA 315:801–810
agent is particularly recommended, except that sedation has 3. Angus DC, Linde-Zwirble WT, Lidicker J, Clermont G, Carcillo J,
to be avoided or discontinued whenever possible [106]. The Pinsky MR (2001) Epidemiology of severe sepsis in the United
States: analysis of incidence, outcome, and associated costs of
effect of sevoflurane on mortality and inflammatory parame- care. Crit Care Med 29:1303–1310
ters is currently assessed (NCT03643367). Ketamine is poten- 4. Eidelman LA, Putterman D, Putterman C, Sprung CL (1996) The
tially neuroprotective in SAE but has not yet been tested in spectrum of septic encephalopathy. Definitions, etiologies, and
interventional trials. Enhanced sensitivity towards benzodiaz- mortalities. JAMA 275:470–473
epines is present in systemic inflammatory processes. 5. Young GB, Bolton CF, Austin TW, Archibald YM, Gonder J,
Wells GA (1990) The encephalopathy associated with septic ill-
Dexmedetomidine, an alpha agonist agent, has been shown ness. Clin Investig Med Med Clin Exp 13:297–304
to increase delirium-free days and reduce 28-day mortality 6. Ely EW, Shintani A, Truman B, Speroff T, Gordon SM, Harrell
rate in septic shock [104]. However, a recent large study FE, Inouye SK, Bernard GR, Dittus RS (2004) Delirium as a
Septic-Associated Encephalopathy: a Comprehensive Review 401

predictor of mortality in mechanically ventilated patients in the 24. Lund-Sørensen H, Benros ME, Madsen T, Sørensen HJ, Eaton
intensive care unit. JAMA 291:1753–1762 WW, Postolache TT, Nordentoft M, Erlangsen A (2016) A nation-
7. Seymour CW, Liu VX, Iwashyna TJ, et al (2016) Assessment of wide cohort study of the association between hospitalization with
clinical criteria for sepsis: for the Third International Consensus infection and risk of death by suicide. JAMA Psychiatry 73:912–
Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA 315: 919
762–774 25. Herridge MS, Cheung AM, Tansey CM, et al (2003) One-year
8. Cunningham C, Maclullich AMJ (2013) At the extreme end of the outcomes in survivors of the acute respiratory distress syndrome.
psychoneuroimmunological spectrum: delirium as a maladaptive N Engl J Med 348:683–693
sickness behaviour response. Brain Behav Immun 28:1–13 26. Bozza FA, D’Avila JC, Ritter C, Sonneville R, Sharshar T, Dal-
9. Hosokawa K, Gaspard N, Su F, Oddo M, Vincent J-L, Taccone FS Pizzol F (2013) Bioenergetics, mitochondrial dysfunction, and
(2014) Clinical neurophysiological assessment of sepsis- oxidative stress in the pathophysiology of septic encephalopathy.
associated brain dysfunction: a systematic review. Crit Care. Shock Augusta Ga 39 Suppl 1:10–16
https://doi.org/10.1186/s13054-014-0674-y 27. Adam N, Kandelman S, Mantz J, Chrétien F, Sharshar T (2013)
10. Oddo M, Carrera E, Claassen J, Mayer SA, Hirsch LJ (2009) Sepsis-induced brain dysfunction. Expert Rev Anti Infect Ther 11:
Continuous electroencephalography in the medical intensive care 211–221
unit. Crit Care Med 37:2051–2056 28. Flierl MA, Stahel PF, Rittirsch D, et al (2009) Inhibition of com-
11. Polito A, Eischwald F, Maho A-L, Polito A, Azabou E, Annane D, plement C5a prevents breakdown of the blood-brain barrier and
Chrétien F, Stevens RD, Carlier R, Sharshar T (2013) Pattern of pituitary dysfunction in experimental sepsis. Crit Care Lond Engl
brain injury in the acute setting of human septic shock. Crit Care 13:R12
Lond Engl 17:R204 29. Taccone FS, Su F, Pierrakos C, He X, James S, Dewitte O, Vincent
12. Finelli PF, Uphoff DF (2004) Magnetic resonance imaging abnor- J-L, De Backer D (2010) Cerebral microcirculation is impaired
malities with septic encephalopathy. J Neurol Neurosurg during sepsis: an experimental study. Crit Care 14:R140
Psychiatry 75:1189–1191 30. Hennessy E, Gormley S, Lopez-Rodriguez AB, Murray C,
13. Sprung CL, Peduzzi PN, Shatney CH, Schein RM, Wilson MF, Murray C, Cunningham C (2017) Systemic TNF-α produces
Sheagren JN, Hinshaw LB (1990) Impact of encephalopathy on acute cognitive dysfunction and exaggerated sickness behavior
mortality in the sepsis syndrome. The Veterans Administration when superimposed upon progressive neurodegeneration. Brain
Systemic Sepsis Cooperative Study Group. Crit Care Med 18: Behav Immun 59:233–244
801–806 31. Skelly DT, Hennessy E, Dansereau M-A, Cunningham C (2013)
14. Iwashyna TJ, Ely EW, Smith DM, Langa KM (2010) Long-term A systematic analysis of the peripheral and CNS effects of sys-
cognitive impairment and functional disability among survivors of temic LPS, IL-1β, [corrected] TNF-α and IL-6 challenges in
severe sepsis. JAMA 304:1787–1794 C57BL/6 mice. PloS One 8:e69123
15. Jackson JC, Hart RP, Gordon SM, Shintani A, Truman B, May L, 32. Sharshar T, Gray F, Lorin de la Grandmaison G, Hopkinson NS,
Ely EW (2003) Six-month neuropsychological outcome of medi- Ross E, Dorandeu A, Orlikowski D, Raphael J-C, Gajdos P,
cal intensive care unit patients. Crit Care Med 31:1226–1234 Annane D (2003) Apoptosis of neurons in cardiovascular auto-
16. Hatch R, Young D, Barber V, Griffiths J, Harrison DA, Watkinson nomic centres triggered by inducible nitric oxide synthase after
P (2018) Anxiety, depression and post traumatic stress disorder death from septic shock. Lancet Lond Engl 362:1799–1805
after critical illness: a UK-wide prospective cohort study. Crit Care 33. Shechter R, Miller O, Yovel G, et al (2013) Recruitment of bene-
Lond Engl 22:310 ficial M2 macrophages to injured spinal cord is orchestrated by
17. Righy C, Rosa RG, da Silva RTA, Kochhann R, Migliavaca CB, remote brain choroid plexus. Immunity 38:555–569
Robinson CC, Teche SP, Teixeira C, Bozza FA, Falavigna M 34. Cain MD, Salimi H, Diamond MS, Klein RS (2019) Mechanisms
(2019) Prevalence of post-traumatic stress disorder symptoms in of pathogen invasion into the central nervous system. Neuron 103:
adult critical care survivors: a systematic review and meta-analy- 771–783
sis. Crit Care Lond Engl 23:213 35. McCaffrey G, Davis TP (2012) Physiology and pathophysiology
18. Riker RR, Fugate JE, Participants in the International Multi- of the blood brain barrier. J Investig Med 60(8):1131–1140.
disciplinary Consensus Conference on Multimodality https://doi.org/10.2310/jim.0b013e318276de79
Monitoring (2014) Clinical monitoring scales in acute brain inju- 36. Davies D (2002) Blood–brain barrier breakdown in septic enceph-
ry: assessment of coma, pain, agitation, and delirium. Neurocrit alopathy and brain tumours. J Anat 200:639–646
Care 21 Suppl 2:S27-37 37. Racchiusa S, Mormina E, Ax A, Musumeci O, Longo M, Granata
19. Young GB, Bolton CF, Archibald YM, Austin TW, Wells GA F (2019) Posterior reversible encephalopathy syndrome (PRES)
(1992) The electroencephalogram in sepsis-associated encepha- and infection: a systematic review of the literature. Neurol Sci Off
lopathy. J Clin Neurophysiol Off Publ Am Electroencephalogr J Ital Neurol Soc Ital Soc Clin Neurophysiol 40:915–922
Soc 9:145–152 38. Danielski LG, Giustina AD, Badawy M, Barichello T, Quevedo J,
20. Azabou E, Magalhaes E, Braconnier A, et al (2015) Early standard Dal-Pizzol F, Petronilho F (2018) Brain barrier breakdown as a
electroencephalogram abnormalities predict mortality in septic in- cause and consequence of neuroinflammation in sepsis. Mol
tensive care unit patients. PloS One 10:e0139969 Neurobiol 55:1045–1053
21. Semmler A, Widmann CN, Okulla T, et al (2013) Persistent cog- 39. Wolf SA, Boddeke HWGM, Kettenmann H (2017) Microglia in
nitive impairment, hippocampal atrophy and EEG changes in sep- physiology and disease. Annu Rev Physiol 79:619–643
sis survivors. J Neurol Neurosurg Psychiatry 84:62–69 40. Becher B, Spath S, Goverman J (2017) Cytokine networks in
22. Pandharipande PP, Girard TD, Jackson JC, et al (2013) Long-term neuroinflammation. Nat Rev Immunol 17:49–59
cognitive impairment after critical illness. N Engl J Med 369: 41. Michels M, Sonai B, Dal-Pizzol F (2017) Polarization of microglia
1306–1316 and its role in bacterial sepsis. J Neuroimmunol 303:90–98
23. Nikayin S, Rabiee A, Hashem MD, Huang M, Bienvenu OJ, 42. Hoogland ICM, Houbolt C, van Westerloo DJ, van Gool WA, van
Turnbull AE, Needham DM (2016) Anxiety symptoms in survi- de Beek D (2015) Systemic inflammation and microglial activa-
vors of critical illness: a systematic review and meta-analysis. Gen tion: systematic review of animal experiments. J
Hosp Psychiatry 43:23–29 Neuroinflammation 12:114
402 Mazeraud et al.

43. Lemstra AW, Groen in’t Woud JCM, Hoozemans JJM, van 60. Sharshar T, Annane D, Gradmaison GL de la, Brouland JP,
Haastert ES, Rozemuller AJM, Eikelenboom P, van Gool WA Hopkinson NS, Gray F (2004) The neuropathology of septic
(2007) Microglia activation in sepsis: a case-control study. J shock. Brain Pathol 14:21–33
Neuroinflammation 4:4 61. Taccone FS, Castanares-Zapatero D, Peres-Bota D, Vincent J-L,
44. Zrzavy T, Höftberger R, Berger T, Rauschka H, Butovsky O, Berre’ J, Melot C (2010) Cerebral autoregulation is influenced by
Weiner H, Lassmann H (2019) Pro-inflammatory activation of carbon dioxide levels in patients with septic shock. Neurocrit Care
microglia in the brain of patients with sepsis. Neuropathol Appl 12:35–42
Neurobiol 45:278–290 62. Brown LS, Foster CG, Courtney J-M, King NE, Howells DW,
45. Singer BH, Dickson RP, Denstaedt SJ, Newstead MW, Kim K, Sutherland BA (2019) Pericytes and neurovascular function in
Falkowski NR, Erb-Downward JR, Schmidt TM, Huffnagle GB, the healthy and diseased brain. Front Cell Neurosci. https://doi.
Standiford TJ (2017) Bacterial dissemination to the brain in sepsis. org/10.3389/fncel.2019.00282
Am J Respir Crit Care Med 197:747–756 63. Rosengarten B, Krekel D, Kuhnert S, Schulz R (2012) Early
46. Mazeraud A, Bozza FA, Sharshar T (2018) Sepsis-associated en- neurovascular uncoupling in the brain during community acquired
cephalopathy is septic. Am J Respir Crit Care Med 197:698–699 pneumonia. Crit Care 16:R64
47. Branton WG, Ellestad KK, Maingat F, Wheatley BM, Rud E, 64. Azevedo LCP (2010) Mitochondrial dysfunction during sepsis.
Warren RL, Holt RA, Surette MG, Power C (2013) Brain micro- Endocr Metab Immune Disord Drug Targets 10:214–223
bial populations in HIV/AIDS: α-proteobacteria predominate in- 65. Berg RMG, Møller K, Bailey DM (2011) Neuro-oxidative-
dependent of host immune status. PloS One 8:e54673 nitrosative stress in sepsis. J Cereb Blood Flow Metab 31:1532–
48. Salter SJ, Cox MJ, Turek EM, Calus ST, Cookson WO, Moffatt 1544
MF, Turner P, Parkhill J, Loman NJ, Walker AW (2014) Reagent 66. Bezzi P, Domercq M, Brambilla L, et al (2001) CXCR4-activated
and laboratory contamination can critically impact sequence- astrocyte glutamate release via TNFα: amplification by microglia
based microbiome analyses. BMC Biol 12:87 triggers neurotoxicity. Nat Neurosci 4:702–710
49. Murray C, Sanderson DJ, Barkus C, Deacon RMJ, Rawlins JNP, 67. Darcy CJ, Davis JS, Woodberry T, McNeil YR, Stephens DP, Yeo
Bannerman DM, Cunningham C (2012) Systemic inflammation TW, Anstey NM (2011) An observational cohort study of the
induces acute working memory deficits in the primed brain: rele- kynurenine to tryptophan ratio in sepsis: association with impaired
vance for delirium. Neurobiol Aging 33:603-616.e3 immune and microvascular function. PloS One 6:e21185
50. Moraes CA, Santos G, de Sampaio e Spohr TCL, D’Avila JC, 68. Freund H, Atamian S, Holroyde J, Fischer JE (1979) Plasma ami-
Lima FRS, Benjamim CF, Bozza FA, Gomes FCA (2015) no acids as predictors of the severity and outcome of sepsis. Ann
Activated microglia-induced deficits in excitatory synapses Surg 190:571–576
through IL-1β: implications for cognitive impairment in sepsis.
69. Kadoi Y, Saito S, Kunimoto F, Imai T, Fujita T (1996) Impairment
Mol Neurobiol 52:653–663
of the brain β-adrenergic system during experimental
51. Zhu F, Zheng Y, Ding Y, Liu Y, Zhang X, Wu R, Guo X, Zhao J
endotoxemia. J Surg Res 61:496–502
(2014) Minocycline and risperidone prevent microglia activation
70. Serantes R, Arnalich F, Figueroa M, et al (2006) Interleukin-1β
and rescue behavioral deficits induced by neonatal
enhances GABAA receptor cell-surface expression by a phos-
intrahippocampal injection of lipopolysaccharide in rats. PLoS
phatidylinositol 3-kinase/Akt pathway relevance to sepsis-
ONE. https://doi.org/10.1371/journal.pone.0093966
associated encephalopathy. J Biol Chem 281:14632–14643
52. Hofer S, Eisenbach C, Lukic IK, et al (2008) Pharmacologic cho-
linesterase inhibition improves survival in experimental sepsis. 71. Zhang Q-H, Sheng Z-Y, Yao Y-M (2014) Septic encephalopathy:
Crit Care Med 36:404–408 when cytokines interact with acetylcholine in the brain. Mil Med
Res. https://doi.org/10.1186/2054-9369-1-20
53. Zaghloul N, Addorisio ME, Silverman HA, et al (2017) Forebrain
cholinergic dysfunction and systemic and brain inflammation in 72. Bleck TP (2018) Dopamine antagonists in ICU delirium. N Engl J
murine sepsis survivors. Front Immunol. https://doi.org/10.3389/ Med 379:2569–2570
fimmu.2017.01673 73. Girard TD, Exline MC, Carson SS, et al (2018) Haloperidol and
54. Schweighöfer H, Rummel C, Roth J, Rosengarten B (2016) ziprasidone for treatment of delirium in critical illness. N Engl J
Modulatory effects of vagal stimulation on neurophysiological Med 379:2506–2516
parameters and the cellular immune response in the rat brain dur- 74. Pandharipande P, Shintani A, Peterson J, Pun BT, Wilkinson GR,
ing systemic inflammation. Intensive Care Med Exp. https://doi. Dittus RS, Bernard GR, Ely EW (2006) Lorazepam is an indepen-
org/10.1186/s40635-016-0091-4 dent risk factor for transitioning to delirium in intensive care unit
55. van Eijk MMJ, Roes KCB, Honing MLH, et al (2010) Effect of patients. Anesthesiology 104:21–26
rivastigmine as an adjunct to usual care with haloperidol on dura- 75. Shehabi Y, Howe BD, Bellomo R, et al (2019) Early sedation with
tion of delirium and mortality in critically ill patients: a dexmedetomidine in critically ill patients. N Engl J Med. https://
multicentre, double-blind, placebo-controlled randomised trial. doi.org/10.1056/NEJMoa1904710
Lancet Lond Engl 376:1829–1837 76. Marra A, Ely EW, Pandharipande PP, Patel MB (2017) The
56. Horti AG, Naik R, Foss CA, et al (2019) PET imaging of microg- ABCDEF bundle in critical care. Crit Care Clin 33:225–243
lia by targeting macrophage colony-stimulating factor 1 receptor 77. Muscatell KA, Dedovic K, Slavich GM, Jarcho MR, Breen EC,
(CSF1R). Proc Natl Acad Sci 116:1686–1691 Bower JE, Irwin MR, Eisenberger NI (2015) Greater amygdala
57. Perry VH, Holmes C (2014) Microglial priming in neurodegener- activity and dorsomedial prefrontal-amygdala coupling are asso-
ative disease. Nat Rev Neurol 10:217–224 ciated with enhanced inflammatory responses to stress. Brain
58. Howarth C (2014) The contribution of astrocytes to the regulation Behav Immun 43:46–53
of cerebral blood flow. Front Neurosci. https://doi.org/10.3389/ 78. Stare J, Siami S, Trudel E, Prager-Khoutorsky M, Sharshar T,
fnins.2014.00103 Bourque CW (2015) Effects of peritoneal sepsis on rat central
59. Retamal MA, Froger N, Palacios-Prado N, Ezan P, Sáez PJ, Sáez osmoregulatory neurons mediating thirst and vasopressin release.
JC, Giaume C (2007) Cx43 hemichannels and gap junction chan- J Neurosci Off J Soc Neurosci 35:12188–12197
nels in astrocytes are regulated oppositely by proinflammatory 79. Siami S, Bailly-Salin J, Polito A, et al (2010) Osmoregulation of
cytokines released from activated microglia. J Neurosci Off J vasopressin secretion is altered in the postacute phase of septic
Soc Neurosci 27:13781–13792 shock. Crit Care Med 38:1962–1969
Septic-Associated Encephalopathy: a Comprehensive Review 403

80. Sonneville R, Guidoux C, Barrett L, et al (2010) Vasopressin sepsis: a prospective study in the medical intensive care unit uti-
synthesis by the magnocellular neurons is different in the supra- lizing continuous EEG monitoring. Intensive Care Med 41:686–
optic nucleus and in the paraventricular nucleus in human and 694
experimental septic shock. Brain Pathol Zurich Switz 20:613–622 97. Le Roux P, Menon DK, Citerio G, et al (2014) Consensus sum-
81. Annane D, Trabold F, Sharshar T, Jarrin I, Blanc AS, Raphael JC, mary statement of the International Multidisciplinary Consensus
Gajdos P (1999) Inappropriate sympathetic activation at onset of Conference on Multimodality Monitoring in Neurocritical Care: a
septic shock: a spectral analysis approach. Am J Respir Crit Care statement for healthcare professionals from the Neurocritical Care
Med 160:458–465 Society and the European Society of Intensive Care Medicine.
82. Annane D, Sébille V, Charpentier C, et al (2002) Effect of treat- Neurocrit Care 21 Suppl 2:S1-26
ment with low doses of hydrocortisone and fludrocortisone on 98. Foreman B, Mahulikar A, Tadi P, et al (2016) Generalized periodic
mortality in patients with septic shock. JAMA 288:862–871 discharges and “triphasic waves”: a blinded evaluation of inter-
83. Reyes E-P, Abarzúa S, Martin A, Rodríguez J, Cortés PP, rater agreement and clinical significance. Clin Neurophysiol Off J
Fernández R (2012) LPS-induced c-Fos activation in NTS neu- Int Fed Clin Neurophysiol 127:1073–1080
rons and plasmatic cortisol increases in septic rats are suppressed 99. Zauner C, Gendo A, Kramer L, Funk GC, Bauer E, Schenk P,
by bilateral carotid chemodenervation. Adv Exp Med Biol 758: Ratheiser K, Madl C (2002) Impaired subcortical and cortical
185–190 sensory evoked potential pathways in septic patients. Crit Care
84. Carlson DE, Chiu WC, Fiedler SM, Hoffman GE (2007) Central Med 30:1136–1139
neural distribution of immunoreactive Fos and CRH in relation to 100. Ehler J, Barrett LK, Taylor V, et al (2017) Translational evidence
plasma ACTH and corticosterone during sepsis in the rat. Exp for two distinct patterns of neuroaxonal injury in sepsis: a longi-
Neurol 205:485–500 tudinal, prospective translational study. Crit Care 21:262
85. Dantzer R, O’Connor JC, Freund GG, Johnson RW, Kelley KW
101. Ehler J, Petzold A, Wittstock M, et al (2019) The prognostic value
(2008) From inflammation to sickness and depression: when the
of neurofilament levels in patients with sepsis-associated enceph-
immune system subjugates the brain. Nat Rev Neurosci 9:46–56
alopathy – a prospective, pilot observational study. PLoS ONE.
86. Peterson JF, Pun BT, Dittus RS, Thomason JWW, Jackson JC,
https://doi.org/10.1371/journal.pone.0211184
Shintani AK, Ely EW (2006) Delirium and its motoric subtypes:
a study of 614 critically ill patients. J Am Geriatr Soc 54:479–484 102. Ehler J, Saller T, Wittstock M, et al (2019) Diagnostic value of NT-
87. Andreasson A, Wicksell RK, Lodin K, Karshikoff B, Axelsson J, proCNP compared to NSE and S100B in cerebrospinal fluid and
Lekander M (2018) A global measure of sickness behaviour: de- plasma of patients with sepsis-associated encephalopathy.
velopment of the sickness questionnaire. J Health Psychol 23: Neurosci Lett 692:167–173
1452–1463 103. Needham DM, Colantuoni E, Dinglas VD, Hough CL, Wozniak
88. Tracey KJ (2009) Reflex control of immunity. Nat Rev Immunol AW, Jackson JC, Morris PE, Mendez-Tellez PA, Ely EW, Hopkins
9:418–428 RO (2016) Rosuvastatin versus placebo for delirium in intensive
89. Rohaut B, Porcher R, Hissem T, et al (2017) Brainstem response care and subsequent cognitive impairment in patients with sepsis-
patterns in deeply-sedated critically-ill patients predict 28-day associated acute respiratory distress syndrome: an ancillary study
mortality. PloS One 12:e0176012 to a randomised controlled trial. Lancet Respir Med 4:203–212
90. Azabou E, Navarro V, Kubis N, Gavaret M, Heming N, Cariou A, 104. Pandharipande PP, Sanders RD, Girard TD, McGrane S,
Annane D, Lofaso F, Naccache L, Sharshar T (2018) Value and Thompson JL, Shintani AK, Herr DL, Maze M, Ely EW,
mechanisms of EEG reactivity in the prognosis of patients with MENDS investigators (2010) Effect of dexmedetomidine versus
impaired consciousness: a systematic review. Crit Care Lond Engl lorazepam on outcome in patients with sepsis: an a priori-
22:184 designed analysis of the MENDS randomized controlled trial.
91. Azabou E, Rohaut B, Heming N, et al (2017) Early impairment of Crit Care Lond Engl 14:R38
intracranial conduction time predicts mortality in deeply sedated 105. McAuley DF, Laffey JG, O’Kane CM, et al (2014) Simvastatin in
critically ill patients: a prospective observational pilot study. Ann the acute respiratory distress syndrome. N Engl J Med 371:1695–
Intensive Care 7:63 1703
92. Sharshar T, Gray F, Poron F, Raphael JC, Gajdos P, Annane D 106. Mehta S, Burry L, Cook D, et al (2012) Daily sedation interruption
(2002) Multifocal necrotizing leukoencephalopathy in septic in mechanically ventilated critically ill patients cared for with a
shock. Crit Care Med 30:2371–2375 sedation protocol: a randomized controlled trial. JAMA 308:
93. Tovote P, Fadok JP, Lüthi A (2015) Neuronal circuits for fear and 1985–1992
anxiety. Nat Rev Neurosci 16:317–331 107. Adler J, Malone D (2012) Early mobilization in the intensive care
94. Cai H, Haubensak W, Anthony T, Anderson DJ (2014) Central unit: a systematic review. Cardiopulm Phys Ther J 23:5–13
amygdala PKC-δ+ neurons mediate the influence of multiple an- 108. Demoule A, Carreira S, Lavault S, Pallanca O, Morawiec E,
orexigenic signals. Nat Neurosci 17:1240–1248 Mayaux J, Arnulf I, Similowski T (2017) Impact of earplugs
95. Schelling G, Stoll C, Kapfhammer HP, Rothenhäusler HB, and eye mask on sleep in critically ill patients: a prospective ran-
Krauseneck T, Durst K, Haller M, Briegel J (1999) The effect of domized study. Crit Care Lond Engl 21:284
stress doses of hydrocortisone during septic shock on posttraumat-
ic stress disorder and health-related quality of life in survivors.
Publisher’s Note Springer Nature remains neutral with regard to jurisdic-
Crit Care Med 27:2678–2683
tional claims in published maps and institutional affiliations.
96. Gilmore EJ, Gaspard N, Choi HA, Cohen E, Burkart KM, Chong
DH, Claassen J, Hirsch LJ (2015) Acute brain failure in severe