INTENSIVE CARE

Brain stem death Learning objectives

Richard Cowan After reading this article, you should be able to:
Barbara Miles C describe the anatomy of the brain stem and the physiological
changes associated with brain stem death
C identify the preconditions required and the clinical tests then
Abstract used to confirm brain stem death
The concept of brain and brain stem death developed from the obser- C discuss when ancillary testing may be required and describe
vation of apnoeic comatose patients. In the UK, the diagnosis of brain such tests
stem death is made by clinically testing brain stem function once spe-
cific preconditions have been met. The exact definition of brain death
and some details regarding the tests required to make this diagnosis hypotension. These patients proceeded to circulatory death if
vary across the globe. However, the majority of tests carried out are artificial ventilation or vasopressors were discontinued. These
similar to those in the UK. In this review we define brain stem death findings were characteristically described in 1959 as ‘coma
and the clinical tests used to confirm it. The use of ancillary testing depasse’, i.e. ‘beyond coma’. The autolytic neuropathological
can have a role in patients where clinical tests are not possible and findings in such patients was also discussed and described in the
this is also discussed. literature as ‘respirator brain’.
Keywords Apnoeic coma; brain death; brain stem death; brain stem The formalization of criteria to diagnose brain death occurred
death testing in 1968 with the publication in JAMA of ‘A definition of irre-
versible coma, the Report of the Ad Hoc Committee of the Harvard
Royal College of Anaesthetists CPD Matrix: 1AO1, 2CO1, 2CO6
Medical School to Examine the Definition of Brain Death’. On the
same day, the World Medical Assembly issued a statement dis-
cussing some key philosophical issues surrounding the definition
Anatomy of death. This included the concept that a definition of death
should regard the fate of a person rather than the preservation of
The brain stem is the inferior part of the brain, adjoining and
isolated cells. The development of these standards for the diag-
structurally continuous with the spinal cord. It is divided in to
nosis of brain stem death allowed appropriate treatment limita-
three distinct areas: the medulla, the pons and the midbrain
tions to be put in place for patients who were irreversibly
(Figure 1). The pons and midbrain contain the nuclei of the
comatose. Although the concept of brain death evolved relatively
reticular activating system which are vital for cortical arousal and
contemporaneously with the first allogenic organ transplants, it
conscious awareness, whereas the medulla contains the control
developed independently.
centres of cardio-respiratory homoeostasis. The nuclei and ori-
The statements have been refined over time and different
gins of cranial nerves III to XII are also contained within the
countries have produced their own guidelines. The current UK
brainstem.
definition of death is that described by The Academy of Medical
Royal Colleges (UK):
History of diagnosing death
Prior to the development of mechanical ventilation, the diagnosis ‘Death entails the irreversible loss of those essential charac-
of death was relatively straightforward: death occurred at the teristics which are necessary to the existence of a living
cessation of respiration, which inevitably led to the cessation of human person and, thus, the definition of death should be
circulation. The advent of long-term ventilation techniques in the regarded as the irreversible loss of the capacity for con-
1950s meant inadequate ventilation no longer immediately led to sciousness, combined with irreversible loss of the capacity to
circulatory death. With the regular use of these techniques and breathe.’
the emergence of intensive care units, case series of patients with
profound irreversible apnoeic coma began to be described. Geographical differences in brain death definitions
These patients had no angiographic evidence of blood flow to There remain global differences in nomenclature regarding brain
the brain, although no mechanical obstruction to flow was death. In some countries (e.g. Australia, New Zealand, USA),
demonstrated at autopsy. They had no spontaneous respiratory demonstration of whole brain death is required to determine
effort, an absence of all electroencephalogram (EEG) activity and death. In the majority of cases of brain stem death, acute whole
were areflexic and polyuric with vasopressor dependent brain injury is evident radiologically. However, in cases of iso-
lated brain stem injury, whole brain injury is not evident and
blood flow to the rest of the brain remains. Such patients
Richard Cowan MBChB (Hons) MRCP(UK) FRCA FFICM is a Specialist therefore cannot be defined as brain dead in some countries.
Trainee Registrar in Anaesthesia and Intensive Care at Glasgow Royal Confirming brain stem death is the basis of determining death
Infirmary, Glasgow, UK. Conflict of Interest: none declared. in patients in apnoeic coma in the UK. The cessation of brain
Barbara Miles MBChB FRCA FFICM is a Consultant in Anaesthesia and stem function does not necessarily entail cessation of neurolog-
Intensive Care at Glasgow Royal Infirmary, Glasgow, UK. Conflict of ical activity in the whole brain. Any potentially remaining
Interest: none declared. function in brain other than the brainstem is not seen to

ANAESTHESIA AND INTENSIVE CARE MEDICINE --:- 1 Ó 2018 Published by Elsevier Ltd.

Please cite this article in press as: Cowan R, Miles B, Brain stem death, Anaesthesia and intensive care medicine (2018), https://doi.org/10.1016/
j.mpaic.2018.08.010
 INTENSIVE CARE

Cranial nerves assessed

with brainstem function tests:

II III
Midbrain

III IV V

V VII
Pons

VI VII V IX VIII
III IV VI VIII

Medulla
V VII
X VII V IX XI XII Response to
supraorbital stimulus

Location IX X
of cranial
Gag, cough and
nerve
nuclei

Adapted from an illustration by Patrick J Lynch

Anatomy of the cranial nerves assessed clinically during UK brain stem death testing

Figure 1

constitute the essential characteristics necessary for a living Endocrine

human existence and so in the UK these patients can be defined Pituitary failure causes a reduction in anti-diuretic hormone
as dead. (ADH) production. This results in an inappropriate diuresis
(neurogenic diabetes insipidus) with resulting hypovolaemia and
Physiological changes associated with brain stem death hypernatraemia. There is reduced thyroid hormone synthesis
and secretion as well as reduced cortisol production. These
Cardiovascular
hormonal changes can adversely affect the cardiovascular state.
With initial brain stem compromise there is a massive sympa-
Hypothalamic failure leads to loss of thermoregulation. The pa-
thetic discharge, resulting in tachycardia and hypertension. With
tient may need active control of body temperature.
further brain swelling, in an attempt to maintain cerebral
perfusion, Cushing’s reflex (hypertension with bradycardia) oc-
Diagnosis of brain stem death
curs in one-third of patients. The sympathetic surge subsequently
dissipates and in the majority of patients is followed by vasodi- Clinical assessment of brain stem function alone can determine
latation induced hypotension. Arrhythmias can also occur. Any brain stem death in the UK. However, clinical testing can only be
hypotension may be worsened by concomitant hypovolaemia. undertaken when specific preconditions are met.

ANAESTHESIA AND INTENSIVE CARE MEDICINE --:- 2 Ó 2018 Published by Elsevier Ltd.

Please cite this article in press as: Cowan R, Miles B, Brain stem death, Anaesthesia and intensive care medicine (2018), https://doi.org/10.1016/
j.mpaic.2018.08.010
 INTENSIVE CARE

UK tests must be carried out twice by two doctors registered

for >5 years (at least one must be a consultant). They should be Conditions precluding brain stem death testing (UK)
competent to conduct and interpret the tests and have no clinical Temperature <34 C
conflict of interest. There is no set time limit to repeat tests but Circulatory Inability to maintain MAP consistently
both physicians take part in both sets of tests. If both sets confirm >60 mmHg
brain stem death, then the time of death is the time of completion Metabolic Na <115 or >160 mmol/L
of the first set of tests. K < 2 mmol/L
The UK testing process for children is the same as the adult Mg <0.5 or >3 mmol/L
process. Diagnosis of brain death in premature newborns and full Phosphate <0.5 or >3 mmol/L
term infants <2 months old is not possible. Glucose < 3.0 or >20 mmol/L
PaCO2> 6.0 kPa
Preconditions
pH <7.35 or >7.45
 Evidence of irreversible brain damage of a known
PaO2 < 10 kPa
aetiology.
Endocrine Clinical suspicion of significant endocrine
 The exclusion of potentially reversible causes of coma:
abnormality
 depressant drugs
 primary hypothermia Table 1
 significant circulatory, metabolic or endocrine
disturbance. Clinical assessment
 The exclusion of potentially reversible causes of apnoea. Cranial nerve reflexes from all levels of the brain stem are tested
In most cases of apnoeic coma the cause of brain damage is by clinical examination (Figure 1) and the respiratory response to
known. The most common causes are trauma, subarachnoid hypercarbia (apnoea test) is assessed.
haemorrhage, intracranial haemorrhage and hypoxic brain  Pupillary light reflex (cranial nerves (CN) II and III): pupils
injury. Occasionally, the cause of coma cannot be established. are fixed and do not respond to bright light.
Without a known cause of irreversible brain injury, patients  Corneal reflex (CN V and VII): the cornea is touched with
cannot be determined brain dead and testing cannot be cotton wool. There should be no response. Care should be
performed. taken to avoid damage to the cornea
The action of sedative or depressant medication should be  Oculo-vestibular reflex (CN III, IV, VI and VIII): establish
excluded prior to testing. The drugs in question, co-existing renal that the eardrum is visible and slowly inject at least 50 mls
or hepatic failure and the patient’s pharmacokinetics will affect of ice-cold water. The head should be at 30 to the hori-
what this entails. Specific drug levels can be tested (e.g. thio- zontal, unless contraindicated. No eye movement should
pentone or midazolam). Antagonists may be useful if opioid or be seen whilst the eyes are held open and observed for 60
benzodiazepines could be contributing to the coma. Neuromus- seconds.
cular blocking drugs can be excluded as the cause of apnoea by These three tests may be prevented by local injurye only one
using a peripheral nerve stimulator. side may be available to test. This does not invalidate the test
Prior to testing, patients should be warmed to a core tem- process but if neither side can be tested, ancillary testing may be
perature of >34 C to exclude primary hypothermia as a cause of necessary.
coma. If therapeutic hypothermia post cardiac arrest has been  Trigeminal and facial (CN V and VII): no motor response
instituted, the patient must be allowed to rewarm before brain within cranial nerve or somatic distribution after supra-
function is formally assessed. UK recommendations are that orbital pressure.
brain stem death testing is deferred for at least 24 hours once  Gag, cough and tracheal reflexes (CN IX and X): no
normothermia is achieved, to allow any residual sedative drugs response to stimulation of the posterior pharyngeal wall
to be metabolized (sedative drug clearance possibly being with a tongue depressor. No cough response on stimu-
reduced by therapeutic hypothermia). lating the trachea and carina with a suction catheter.
Potentially reversible circulatory, metabolic and endocrine  Apnoea testing: this is performed only when all preceding
causes of coma should be excluded (Table 1). Many of these tests have shown absent response. Inspired oxygen is
changes may occur during the brain stem death process. The increased to 1.0 for 5 minutes. An arterial blood gas is
majority of patients will need vasopressor support to maintain a taken, correlating peripheral saturation with PaO2 and
blood pressure sufficient to allow testing to proceed. Electrolyte ETCO2 with PaCO2. Minute ventilation is decreased to
correction may be difficult and rapid correction purely for the allowing PaCO2 to rise >6.0 kPa and pH to fall < 7.4.
purposes of brain death testing may be detrimental e too rapid Higher PaCO2 levels may be required in chronic CO2
correction of hyponatraemia may itself result in the development retention to achieve pH < 7.4. The ventilator is then
of central pontine myelinolysis and coma. Although rare, severe disconnected and oxygen at 5L/min is provided down an
hypothyroidism and Addisonian crisis may result in coma and endotracheal catheter. If saturations fall, CPAP or recruit-
should be excluded by hormonal assays if any clinical suspicion ment manoeuvres may be required. Cardiovascular sta-
is raised. bility should be maintained throughout.
Co-existing high cervical spinal injuries can limit clinical The patient is observed for evidence of respiration for 5 mi-
testing and ancillary tests may be required. nutes. Repeat blood gases are obtained after 5 minutes e a PaCO2

ANAESTHESIA AND INTENSIVE CARE MEDICINE --:- 3 Ó 2018 Published by Elsevier Ltd.

Please cite this article in press as: Cowan R, Miles B, Brain stem death, Anaesthesia and intensive care medicine (2018), https://doi.org/10.1016/
j.mpaic.2018.08.010
 INTENSIVE CARE

improves with follow up scans. The scan looks for the ‘hollow
Ancillary tests for the diagnosis of brain stem death skull phenomenon’ indicating the absence of brain perfusion.
Blood flow in the Four-vessel angiography The use of EEG is recommended in US guidelines for the
larger cerebral arteries Transcranial doppler diagnosis of brain stem death in adults and children when clin-
Magnetic resonance angiography ical testing is inconclusive or impossible. Adherence to these
Spiral CT angiography guidelines is variable and a recent survey revealed that 6.5% of
Brain tissue perfusion Cerebral scintigraphy 99mTc-labelled American hospitals mandate such ancillary testing. The EEG re-
hexamethylpropylene-amine oxime cords summated synaptic potentials. A flat or isoelectric trace
(HMPAO) confirms the absence of brain activity. This should be interpreted
Xenon CT with caution however as there are numerous causes of flat/iso-
Positron emission tomography electric traces (including anaesthetic drugs). Evoked potentials
Neurophysiology EEG (somatosensory and brainstem auditory) have a limited role in
Evoked potentials ancillary testing.
The reliability, accuracy, availability and ease of interpreta-
Table 2 tion of these tests tend to limit their usage unless specifically
indicated.
gain  0.5 kPa demonstrates an adequate respiratory stimulus.
The patient’s mechanical ventilation is then recommenced,
Cautions
returning blood gases to the pre-testing level. Recruitment ma-
noeuvres may be undertaken after the apnoea test to maintain Profound neuromuscular weakness resembling absent brain stem
baseline oxygenation. reflexes may occur as a consequence of some neurological dis-
Spinal reflexes may remain after brain stem death. These can orders. These include the locked-in syndrome (LIS) and Guillain-
produce reflex movements of any limb, plantar reflexes and even Barre syndrome (GBS) affecting the brainstem. Organophosphate
flexion at the waist resulting in a sitting position. The presence of poisoning, lidocaine toxicity, baclofen overdose, high cervical
these reflexes does not preclude or refute the diagnosis of brain spine injuries and prolonged neuromuscular blocker clearance
death. They can be distressing for relatives and staff unaware of have also been reported as causing absent brain stem reflexes.
their origin. These cases highlight the importance of ensuring all pre-
conditions are met prior to commencing brain stem testing.
Ancillary tests Preservation of eye movements in LIS should alert the physician
and preclude a diagnosis of brain stem death. Rapidly progressive
In some countries ancillary testing is mandatory for the diagnosis
GBS can mimic brain death but such patients have no diagnosis
of brain death. In the UK the diagnosis of brain stem death is a
compatible with irreversible brain damage and therefore do not
clinical diagnosis. However, in certain circumstances (e.g. major
satisfy the necessary preconditions and cannot be tested. A
facial trauma) ancillary testing may be needed to confirm the
diagnosis. Tests can be generally divided into assessment of
blood flow in cerebral arteries, assessment of cerebral perfusion FURTHER READING
and neurophysiological tests (Table 2). Academy of the Medical Royal Colleges. A code of practice for the diag-
Tests capable of demonstrating absent blood flow are accepted nosis and confirmation of death. London, 2008. Available at: http://
as establishing whole brain death. The sequence of events relates aomrc.org.uk/wp-content/uploads/2016/04/Code_Practice_
to the elevation in intracranial pressure exceeding systemic arte- Confirmation_Diagnosis_Death_1008-4.pdf (accessed 26 Sept 2017).
rial pressure, resulting in cessation of cerebral blood flow. Greer DM, Wang HH, Robinson JD, Varelas PN, Henderson GV,
Four-vessel cerebral angiographies were considered the ‘gold Wijdicks EFM. Variability of brain death policies in the United
standard’ modality for assessing cerebral blood flow. In brain States. JAMA Neurol 2016; 73: 213e8.
death, angiography will normally demonstrate absent blood flow Lewis A, Greer D. Current controversies in brain death determination.
at or beyond the carotid bifurcation or the circle of Willis. False Nat Rev Neurol 2017; https://doi.org/10.1038/nrneurol.2017.72.
negatives can occur when intracranial pressure is lowered (e.g. Advance online publication. (accessed 26 Sept 2017) at, http://
by surgery or trauma). www.nature.com/nrneurol/journal/vaop/ncurrent/abs/nrneurol.
Transcranial Doppler has the advantage of being non-invasive 2017.72.html#abstract.
and safe. It requires expertise but can be done at the bedside. The Machado C, Korein J, Ferrer Y, Portela L, de la C Garcia M,
Doppler aims to find small systolic peaks with retrograde dia- Manero JM. The concept of brain death did not evolve to benefit
stolic flow or a reverberating flow pattern. This pattern suggests organ transplants. J Med Ethics 2007; 33: 197e200.
high vascular resistance and supports the diagnosis or brain stem Vargas F, Hilbert G, Gruson D, Valentino R, Gbikpi-Benissan G,
death. Patients with external ventricular drains or craniotomies Cardinaud JP. Fulminant Guillain-Barre syndrome mimicking ce-
may have false negative testing. rebral death: case report and literature review. Intens Care Med
Techniques that look at tissue perfusion tend to be more 2000; 26: 623e7.
accurate. CT angiography/perfusion and nuclear medicine Wijdicks EFM, Varelas PN, Gronseth GS, Greer DM. Evidence-based
(cerebral scintigraphy) are used for this purpose. 99mTc-labelled guideline update: determining brain death in adults: report of the
hexamethylpropylene-amine oxime (HMPAO) scintigraphy quality standards subcommittee of the American Academy of
testing has been shown to be very useful in brain stem death Neurology. Neurology 2010; 74: 1911e8.
testing and can be utilized in the paediatric population. Sensitivity

ANAESTHESIA AND INTENSIVE CARE MEDICINE --:- 4 Ó 2018 Published by Elsevier Ltd.

Please cite this article in press as: Cowan R, Miles B, Brain stem death, Anaesthesia and intensive care medicine (2018), https://doi.org/10.1016/
j.mpaic.2018.08.010