Just Say “NO!” To Ammonia
Kelley E. Capocelli, MD
Clinical Pathology Conference
April 29, 2005
Overview
Review of hepatic
encephalopathy
 Neurotoxins involved with
pathogenesis of hepatic
encephalopathy
 Discussion of ammonia
 Alternatives to ammonia?

Hepatic Encephalopathy
Mechanisms
Accumulation of nitrous
metabolites
 Accumulation of toxic
substances
 Hypoglycemia
 Brain edema
 False neurotransmitters
 Electrolyte changes

EEG Abnormalities

Nonspecific



Slow waves of large amplitude
Bursts of triphasic wave patterns
Generally read as “consistent with
metabolic encephalopathy” because
similar patterns are observed in
uremia, pulmonary or heart failure, and
acid-base disorders

Subject to variability in interpretation
PET and Proton MR Studies

PET



Shows significantly decreased glucose
utilization in the cerebral cortex and
concomitant increased utilization in the
thalamus, caudate lobe, and cerebellum
These findings suggest that
hypometabolism in the brains of patients
with chronic liver disease could explain the
neuropsychiatric abnormalities
characteristic of hepatic encephalopathy
Proton MR spectroscopy of brains of
cirrhotic patients showed depletion of
myoinositol (a sign of increased
osmolality) and increased glutamine
CT and MRI


Important in ruling our intracranial
lesions when the diagnosis of hepatic
encephalopathy is in question
MRI is able to demonstrates
hyperintensity of the globus pallidus on
T1-weighted images, a finding that is
commonly described in hepatic
encephalopathy

May be due to manganese accumulation
Neurotoxins in Liver Disease

Ammonia

Normal NH3 metabolism
By-product of protein catabolism by gut
flora
 Absorbed by gut and enters portal vein
 Metabolized in the liver to urea and
glutamine (Krebs cycle)


NH3 in liver disease
Not cleared by the liver
 Delivered to the brain by systemic
circulation
 Associated with hepatic encephalopathy

Not necessarily the cause!
Neurotoxins in Liver Disease

False neurotransmitters – amino
acids

Patients with cirrhosis have a
decreased ratio of branched-chain
amino acids (BCAA) to aromatic
acids (AAA), from 3.5:1 to 1:1.
BCAA include valine, leucine, and
isoleucine
 AAA include phenylalanine, tyrosine, and
tryptophan


The decrease in BCAA is caused
predominantly by their extensive use
by skeletal muscle
Neurotoxins in Liver Disease

False neurotransmitters (cont.)

It has been postulated that the
increase in AAA in the CNS may
interfere with physiologic
neurotransmission by competitively
inhibiting “normal” neurotransmitters
(i.e., DA, NE) and favoring formation
of weak, “false” neurotransmitters
(i.e., octopamine)
Neurotoxins in Liver Disease
 This
attractive hypothesis
raises the possibility that
correction of the AAA:BCAA
ratio may lead to amelioration of
hepatic encephalopathy

A multitude of clinical trials have failed
to prove that changes in the ratio
through IV or oral administration of
BCAA result in significant improvement
of clinical signs or symptoms of this
condition
Neurotoxins in Liver Disease

Accumulation of manganese





Observation that more than 80% of
patients with cirrhosis in hepatic coma
have increased concentrations of
manganese
Prolonged exposure to manganese results
in extrapyramidal symptoms and abnormal
MR images
Long-term exposure to manganese also
leads to Alzheimer type II astrocytosis
Direct measurements of basal ganglia
manganese levels in autopsy have shown a
twofold to sevenfold increase
Proposed that ammonia and manganese
act synergistically
Neurotoxins in Liver Disease

Monoamines

Many of the early neuropsychiatric
symptoms of hepatic encephalopathy (e.g.
altered sleep patterns) have been
attributed to modification of the
monoamine neurotransmitter serotonin


A two to fourfold increase in cerebral
concentration of the serotonin metabolite 5hydroxyindoleacetic acid is the most consistent
neurochemical finding in HE
In addition, HE is associated with alterations in
the number of 5HT1A and 5HT2 receptors and
increased activity of both MAOA and MAOB
(enzymes catabolizing 5-HT)

The findings suggest an increased serotonin
turnover rate in HE
Neurotoxins in Liver Disease

Histamine
Autopsied brain tissue from cirrhotic
patients with HE displayed a higher
density and a lower affinity of
histamine H1 receptors compared
with control human frontal cortex
 Binding was highest in the parietal
and temporal cortices and lowest in
the caudate-putamen

Neurotoxins in Liver Disease

Histamine (cont.)
A selective increase in H1 receptor
density was also observed in parietal
and insular cortices of patients with
HE
 A selective up-regulation of brain H1
could contribute to the
neuropsychiatric symptoms
characteristic of human HE


May be amenable to treatment with
selective histamine H1 receptor
antagonists
Neurotoxins in Liver Disease

Oxindole



A tryptophan metabolite formed by gut
bacteria (via indol) that can cause
sedation, muscle weakness, hypotension,
and coma
Cerebral concentrations of oxindole are
increased 200-fold in rats with acute liver
failure
More than a 50-fold increase in the plasma
concentration of oxindole has been found
in patients with HE

The mechanism is currently under investigation
History of Ammonia

The first experiment implicating a
nitrogenous substance as a cause
of hepatic encephalopathy was
performed by Eck

Created portal-systemic shunts in
healthy dogs and observed that
these dogs promptly became
comatose after eating meat
Ammonia is the culprit because…

The role of ammonia has been
postulated on the basis of the
following:



A reproducible increase in blood ammonia
levels of patients with cirrhosis
The development of hepatic coma in
patients with advanced liver disease and in
experimental animals after ingestion of
ammonia
Elevated serum levels in children with
genetic abnormalities of urea cycle
synthesis, which are associated with
neuropsychiatric changes similar to those
of patients with hepatic encephalopathy
Ammonia is the culprit because…

Role of ammonia (cont.)
Increased cerebral metabolism of
ammonia as detected by PET and
ammonia 13 isotope
 Increased permeability of the bloodbrain barrier to ammonia
 Chronic elevations of blood ammonia
levels causing characteristic
changes in astrocytes

Ammonia is the culprit because…

Role of ammonia (cont.)
The exposure of the brain to
millimolar concentrations of
ammonia may impair neuronastrocyte trafficking and lead to
Alzheimer type II astrocytosis
 Ammonia inhibits excitatory
postsynaptic potentials, thereby
depressing overall CNS function

Are all of these data true?

Not all data are consistent with
the ammonia toxicity theory

There is poor correlation of ammonia
with hepatic encephalopathy

This condition can be present in the
absence of elevated ammonia levels


Approximately 10% of patients with
significant encephalopathy have normal
serum ammonia levels
Many patients with cirrhosis have
elevated ammonia levels without
evidence of encephalopathy
Are all of these data true?
Low ammonia concentrations are
associated with neuroexcitatory
effects
 Ammonia does not induce that
classic EEG changes associated with
hepatic encephalopathy when it is
administered to patients with
cirrhosis

Ammonia as a diagnostic test

Serum ammonia levels are often
elevated in patients with cirrhosis

The serum ammonium test measures both
ionized ammonia (ammonium) and unionized ammonia


However, only un-ionized ammonia crosses the
membranes
In patients with normal serum pH, the ammonia
represents only a small fraction of total
ammonium concentrations

This factor may be in part responsible for the poor
predictability of total ammonium measurement as
a diagnostic or prognostic test
Ammonia and Chemistry

In 1859, Berthelot described a
reaction between ammonia
and an alkaline solution of
phenol hypochlorite suitable
for the determination of
ammonia
 Assay
was subject to
interferences
Ammonia and Chemistry

In 1963, Kirsten et al introduced
an enzymatic method for ammonia
determination based on the action
of glutamate dehydrogenase

Although the enzymatic reaction was
highly specific and utilized direct
evaluation based on the molar
absorption of NADH, there were
difficulties in stabilizing the end
reaction
Ammonia and Chemistry

Current method is based on Da
Fonseca-Wollheim’s modification of the
Kirsten reaction
 Improvements to the original
reaction
Addition of ADP to the reaction mixture
 The use of NADPH in place of NADH to
eliminate interference from the reaction
of endogenous LDH with endogenous
pyruvate
 The substitution of plasma for
deproteinized supernatant

Ammonia and Chemistry

In the reaction catalyzed by glutamate
dehydrogenase (GLDH), ammonia
reacts with α-ketoglutarate and NADPH
to form glutamate and NADP+
GLDH
α-ketoglutarate + NH4+ + NADPH -----> L-Glutamate + NADP+ + H2O

The inclusion of ADP in the reaction
mixture causes an acceleration of the
rate of conversion and stabilizes the
GLDH in the indicated pH range
Ammonia and Chemistry


The amount of NADPH oxidized
during the reaction is equivalent
to the amount of ammonia in the
specimen and can be measured
photometrically by the resulting
decrease in absorbance
The decrease in absorbance due
to consumption of NADPH is
measured kinetically.
Ammonia and Chemistry

Reagent Preparation
R1 Working Solution = Buffered
Substrate, ready to use
R2 Working Solution =
NADPH/GLDH/Buffered Substrate
Step 1: Reconstitute Bottle 2b (GLDH)
with 0.5 mL of Bottle 2a
(Buffered Substrate)

Let stand 10 minutes at room
temperature, occasionally swirling
gently
Ammonia and Chemistry

Reagent Preparation (cont.)
Step 2: Reconstitute one bottle of Bottle 2
(NADPH) with 4.5 mL of Bottle 2a
(Buffered Substrate). Mix by gentle
inversion.
Step 3: Add 150 µL of reconstituted Bottle
2b
(GLDH) to Bottle 2 from Step 2. Mix
by gentle inversion. Transfer into the
R2 empty bottle supplied. Remove
bubbles.
Ammonia and Chemistry

Specimen Collection and Preparation




The patient should fast for 6 hours and
should not clench fist during phlebotomy.
Plasma: Use nonhemolyzed EDTA (lavender
tube) plasma. Draw the specimen from a
stasis-free vein and centrifuge in a
stoppered tube as soon as possible. Collect
1.0 mL of blood.
Sample: Place the sample on ice and
assay immediately.
Keep all tubes, cuvettes, and vials
covered to prevent uptake of ammonia
from the air
Ammonia and Chemistry

Specimen Handling



The specimen must be collected on ice and
rushed to the lab.
Centrifuge the specimen and immediately
separate plasma from cells, stopper tube,
and return plasma to cup of ice, or freeze
immediately.
NOTE: Do not leave sample,
callibrator, or quality control cups open
for longer than 15 minutes. Results
may be affected. Repeat assays must
be performed on freshly poured cups,
due to evaporation of ammonia.
Ammonia and Chemistry


Ammonia is stable in plasma for 3
hours at 4-8ºC in a stoppered tube. If
the specimen is not immediately
tested, freeze the specimen.
Reportable Range: 10-450µg/dL

Expected Values = Adults 17-80 µg/dL (for
venous plasma)


Venous ammonia concentrations may be as
much as twice that of arterial
If the serum result is greater than 450, dilute
with water to bring the result into the analytical
measuring range and multiply the result by the
dilution factor.
Ammonia and Chemistry

NOTE: Automatic rerun should
not be used for ammonia samples.
If a sample result exceeds 1000
µg/dL, a fresh sample must be
poured.
Problems with ammonia as a
diagnostic test for HE

Increased levels of ammonia may also
be seen with:





GI bleeding – blood cells are hemolyzed in
the intestines, releasing protein
Muscular exertion – muscles produce
ammonia when active and absorb it when
resting
Tourniquet use – ammonia levels can be
increased in the blood sample collected
Drugs that can increase ammonia –
alcohol, barbiturates, diuretics, and
narcotics
Smoking
Problems with ammonia as a
diagnostic test for HE

Venous ammonia levels are
inconsistent
Levels are influenced by fist
clenching, use of a tourniquet, and
whether the sample was placed on
ice
 HE is only directly related to
ammonia levels up to about a twofold
increase above normal


Any further increase of ammonia
concentration does not contribute to the
further evolution of HE
Problems with ammonia as a
diagnostic test for HE

Arterial ammonia levels tend to be more
accurate than venous but are not more useful



Serial ammonia measurements are inferior to
clinical assessment in gauging improvement or
deterioration in a patient under therapy for hepatic
encephalopathy
The grade of HE is more closely related to
the partial pressure of gaseous ammonia
since gaseous ammonia readily enters the
brain
Postprandial ammonia levels may be more
closely related to the grade of HE than
fasting levels

In clinical trials, ammonia is often measured after a
standard meal (or glutamine load)
Why then is ammonia considered an
appropriate diagnostic test?


Ong et al (2003) prospectively
evaluated the correlation between
ammonia levels and the severity
of HE in 121 patients with
cirrhosis
Methods
Based the diagnosis of HE on clinical
criteria
 Severity of HE based on the West
Haven Criteria for grading of mental
status

Why then is ammonia considered an
appropriate diagnostic test?

Methods (cont.)
 Obtained
arterial and venous
blood samples from each patient
 Four
types of ammonia
measurements were analyzed
Arterial and venous total ammonia
 Arterial and venous partial pressure
of ammonia

 Spearman
rank correlations (r s)
were calculated
Why then is ammonia considered an
appropriate diagnostic test?
 Results
 30
(25%)
 27 (22%)
 23 (19%)
 13 (11%)
had
had
had
had
grade
grade
grade
grade
0
1
2
4
HE
HE
HE
HE
Why then is ammonia considered an
appropriate diagnostic test?

Results (cont.)

Each of the four measures of
ammonia increased with the severity
of HE
Arterial total ammonia (r s = 0.61, p ≤
.001)
 Venous total ammonia (r s = 0.56, p ≤
.001)
 Arterial partial pressure of ammonia (r s =
0.55, p ≤ .001)
 Venous partial pressure of ammonia (r s =
0.52, p ≤ .001)

Why then is ammonia considered an
appropriate diagnostic test?

Conclusions
Ammonia levels correlate with the
severity of HE.
 Venous sampling is adequate for
ammonia measurement
 There appears to be no additional
advantage of measuring the partial
pressure of ammonia compared with
total ammonia levels

Things aren’t always as they seem

The authors evaluated this
clinical controversy in 121
patients
41% had documented cirrhosis by
liver biopsy
 59% had evidence of portal
hypertension
 Alcohol was the cause of cirrhosis in
64% of patients
 53% of patients were receiving
lactulose before hospital admission
 69% were a Child-Pugh Class C

Things aren’t always as they seem

Not stated in their abstract

There was substantial overlap in the total
ammonia concentrations by grade of HE



At least 1 patient with grade 4 HE had a total
ammonia (arterial and venous) <25 mmol/L
Several patients with grade 0 to 2 had ammonia
concentrations >100 mmol/L
69% of their patients without signs or symptoms
of HE had total arterial ammonia concentrations
>47 mmol/L, which was the upper limit of normal
for their institution’s clinical laboratory
Things aren’t always as they seem

Phenomenon due to?




The variability in ammonia concentrations
observed throughout the day
The delay between elevation of ammonia
concentrations and degree of
encephalopathy seen in some patients
Other compounds may be involved in the
development of HE
The most common precipitating
factors for grade 1 or higher HE in their
patients were azotemia, infection, GI
bleeding, lactulose noncompliance,
and constipation
Things aren’t always as they seem

Clinical implications of this article
A single ammonia concentration,
whether venous or arterial, has little
utility in the diagnosis of HE
 The diagnosis of HE should be based
on the mental status assessment and
clinical course of the patient during
their hospitalization

Are there alternative tests?

Glutamine

CSF levels
Not available at DHMC or Specialty Labs
 Available at Biochemical Genetics
Laboratory, Dr. Goodman




1 or 2 day TAT
$100
Manganese

Not an available test at either DHMC
or Specialty Labs
Are there alternative tests?

Serotonin

Procedure


Patient should avoid avocado, banana, tomato,
plum, walnut, pineapple, eggplant, tobacco, tea,
and coffee for 3 days before the test
Requires 4 mL of whole blood EDTA




Transfer EDTA whole blood to plastic tube, add 4
mg ascorbic acid, mix and freeze within 2 hours of
collection
Test discontinued at DHMC due to difficulty
of preparing specimen
Available as send out to Specialty Labs
with a 30-day TAT
No test available to test levels in CSF
Are there alternative tests?

Histamine
Test available to test levels in urine
 No test available at DHMC or
Specialty Labs to test levels in
serum or CSF


Oxindole

No test available at either DHMC or
Specialty Labs for either serum or
CSF
How our medicine
colleagues should respond
A 56 year old male is brought to the
hospital by his daughter. She found
him unconscious that morning on the
sidewalk outside his home. At the
hospital he is still unconscious and
does not respond to painful stimuli.

Laboratory results



Ammonia level 80 (URL = 65)
AST 90 (URL = 44)
Albumin 2.2 (LRL = 3.5)
How our medicine
colleagues should respond
The intern says
“Well, I thought this might be
hepatic encephalopathy, but
given how sick this guy is, I
expected a MUCH higher
ammonia level!”
How our medicine
colleagues should respond
The attending responds
“I am not surprised. While elevated ammonia
is associated with hepatic encephalopathy,
there is not a correlation between this
level and disease severity. The patient has
other lab findings suggestive of end-stage
liver disease. In the future, please do not
waste my time, or that of the laboratory
techs and the pathology resident, by
ordering such a useless test in an adult.
Let’s try lactulose and see how the patient
responds!”
References




Abou-Assi S, Vlahcevic ZR. Hepatic
encephalopathy: metabolic consequence of
cirrhosis is often reversible. Postgrad Med
2001;109(2): 52-70.
Ammonia Technical Procedure, DHMC,
Reviewed 1/10/2005
Blei AT, Cordoba J. Hepatic Encephalopathy.
The American Journal of Gastroenterology 2001;96(7):
1968-1975.
Ferenci, P. Clinical manifestations and diagnosis
of hepatic encephalopathy. 2005 UpToDate.
References



Ferenci, P. Pathogenesis of hepatic
encephalopathy. 2005UpToDate.
Ong JP, Aggarwal A, Krieger D, Easley KA, et
al. Correlation between ammonia levels and the
severity of hepatic encephalopathy. Am J Med
2003;114:188-193.
Wolf, DC. Hepatic Encephalopathy.
eMedicine.com