HEPATIC ENCEPHALOPATHY MIMICKING STROKE

A Case Report’

James W. Atchison,” DO, Mark Pellegrino, MD, Patricia Herbers, CC, RN, Brent Tipton, MD

and Velimir Matkovic, MD, PhD

ABSTRACT Atchison JW, Pellegrino M, Herbers P, Tipton B, Matkovic V: Hepatic
encephalopathy mimicking stroke: a case report. Am J Phys Med Rehabil 1992;
71:114-118, :

This case demonstrates focal neurologic deficit mimicking stroke with under-
lying hepatic encephalopathy. Unilateral weakness in patients with hepatic en-
cephalopathy has not been previously described in the English language literature.
A 46-yr-old white woman was admitted to an acute care hospital for left shoulder
manipulation, underwent general anesthesia and appeared to have had a right
cerebrovascular accident. At transfer to the rehabilitation hospital, in addition to
the left hemiparesis, there were inconsistencies in the neurologic examination
and signs of cognitive impairment and liver failure. The patient’s response to an
intensive, multidisciplinary inpatient rehabilitation program along with treatment
of the liver dysfunction led to resolution of left-sided weakness and flapping
tremor with independence in ambulation and activities of daily living. Relevant
literature is reviewed. A thorough history and physical examination with liver
function assessment should always be performed in patients with cerebrovascular

accident and unusual recovery.

Key Worps: Encephalopathy, Ammonia, Stroke, Rehabilitation

Hepatic encephalopathy is a metabolic disorder
of the nervous system in conjunction with hepato-
cellular disease or portal-systemic venous shunting.
The disease is characterized by fluctuating neuro-
logic signs, asterixis and electroencephalographic
(EEG) abnormalities. Disturbances in consciousness
can progress to disorientation and coma.’

The fluctuating neurologic signs reported include
rigidity, hyperreflexia, asymmetric deep tendon re-
flexes, extensor plantar signs, suck and grasp re-
flexes,” impaired motor coordination, sleep disturb-
ances,? seizure activity, diminished mental status
and paraparesis.* Asterixis or “flapping tremor,” a
nonrhythmic, asymmetric lapse in voluntary sus-
tained posture of the head, neck and extremities, is
commonly noted in the precoma stage. There are no
previous English language reports of unilateral
weakness or hemiparesis as the presenting picture
of hepatic encephalopathy.

0894-9115/92/7102-0114$03.00/0
“AMERICAN JOURNAL OF PHYSICAL MEDICINE & REHABILITATION
Copyright © 1992 by Williams & Wilkins

Presented in poster form at the 51st Annual Assembly of The
American Academy of Physical Medicine and Rehabilitation, San
Antonio, Texas, November 1989.

From the Department of Rehabilitation Medicine, University
of Kentucky College of Medicine (J.W.A.), Lexington, Kentucky
and Department of Physical Medicine and Rehabilitation, The
Ohio State University (M.P., P.H., B.T., V.M.}, Columbus, Ohio.

? All correspondence and requests for reprints should be ad-
dressed to: Department of Rehabilitation Medicine, Cardinal Hill
Hospital, 2050 Versailles Rd., Lexington, KY 40504.

114

CASE REPORT

A 46-yr-old white woman was referred for left
hemiparesis believed to be secondary to right cere-
brovascular accident. She had been admitted to an
acute care hospital as the result of a painful left
shoulder with severe restriction of motion and
underwent manipulation under anesthesia as well
as Groshong catheter placement for long-term intra-
venous meperidine therapy. Her previous medical
history included adult onset diabetes mellitus for 9
yr with some complaint of left lower limb sensory
neuropathy, hepatitis of unknown etiology, a left
Bell palsy, which had resolved, and esophageal var-
ices for which she had recently undergone sclerosis
treatment. Past surgeries included hysterectomy,
splenectomy and cholecystectomy. She had no his-
tory of left-sided weakness and was entirely inde-
pendent in ambulation and activities of daily living
before admission.

After general anesthesia, the patient experienced
an episode of loss of consciousness and on arousal
was noted to have left-sided weakness and episodes
of “shaking.” EEG showed generalized slowing with
questionable background epileptiform activity, and
the patient was put on phenytoin. Initial computed
tomography (CT) of the head without contrast and
a repeated CT 7 days later with contrast were both
normal, but T2-weighted magnetic resonance imag-
ing (MRI) of the brain showed a high intensity lesion
of 5-6 cm in the right frontal lobe and multiple
lesions in both parietal and occipital lobes. These
Vol. 71, No. 2, April 1992

findings were interpreted as being consistent with
multiple infarcts or plaques.

On presentation for rehabilitation, the patient was
lethargic with a flat affect, and pupils were equal
but sluggish. Spider nevi were present throughout,
and some right upper quadrant abdominal tender-
ness was noted with palpable liver. The neuromus-
culoskeletal examination revealed some mild left-
sided facial weakness and ptosis, which were not
noted on admission to acute hospital. All deep ten-
don reflexes were 2+ except the left biceps brachii,
which was 3+. There was an equivocal Babinski on
the left, but no Hoffmann’s reflex, clonus or abnor-
mal tone was present. Motor testing showed only
trace contraction at the left shoulder. The rest of the
left arm and the entire right side had greater than
antigravity strength, but this response was incon-
sistent and could be demonstrated only with func-
tional use, not on command. Passive range of motion
of the left shoulder showed only 45° of flexion and
80° of abduction because of severe complaints of
pain. The left lower limb showed extremely dys-
functional contractions with both agonist and antag-
onist muscle groups firing simultaneously, but very
slowly. The sensory examination was normal except
for some hyperesthesia of the left leg below the
knee. Rapid alternating movements were normal
except for slowing with the right hand. Gait was
markedly abnormal with a narrow base, slow
rhythm and continual falling to the left.

250 Begin treatment

200

HEPATIC ENCEPHALOPATHY MIMICKING STROKE = 115

Treatment on admission included glucose checks
every 6 h with sliding scale insulin coverage, phen-
ytoin (100 mg twice/day), spironolactone (100 mg
twice/day), trimethobenzamide (200 mg every 6 h
as needed for nausea or vomiting) and meperidine
HCI (25 mg intravenously every 2 h as needed for
pain). The meperidine was changed to 50 mg intra-
muscularly three times per day, trimethobenzamide
was discontinued, and 5000 units of heparin sub-
cutaneously was initiated two times per day.

Initial testing showed some abnormal liver en-
zymes (Fig. 1) with an aspartate aminotransferase
of 108 units/liter (normal, 0-60 units/liter) and al-
kaline phosphatase of 164 units/liter (normal, 0-
100 units/liter), plus a markedly abnormal serum
albumin of 2.1 g/dl (normal, 3.8-5.1 g/dl). Partial
thromboplastin time was slightly elevated at 46 s
(control, 23-35 s) as was prothrombin time at 12.7
s (control, 9.5-12.5 s). Target cells were noted in the
peripheral smear. Repeat EEG showed diffuse slow-
ing of background activity maximal in the right
hemisphere with no epileptiform activity. Phenytoin
level was therapeutic at 14.9 mg/liter (normal, 10-
20 mg/liter).

The patient began physical and occupational ther-
apy with emphasis on transfer training, balance
activities and gait training, as well as progressive
resistance exercises of all limbs, concentrating on
strengthening, increasing the range of motion and
decreasing the pain of the left shoulder. Even though

150

100

$0 +

Figure 1. Comparison of
serum liver enzymes with
level of encephalopathy.
Alk phos, .alkaline phos-
phatase; AST, aspartate
aminotransferase; ALT, al-

0
04/18

416 Q4/10 4/12 am 4/12 pm 04/13
Admit

— Alk phos
-- Ammonia p mol/l

---AST pil

I—- ALT pil

~I- Stage of Enceph

04/21

04/25 4/28
Disch.

anine aminotransferase.
116 ATCHISON ET AL.

she was oriented on admission, a neuropsychologist
was also asked to evaluate the patient because of
the inconsistencies of her examination.

On the 1st day after admission, the patient suf-
fered an unwitnessed fall while trying to transfer
herself in the bathroom. A hematoma developed at
the site of previous subcutaneous heparin injection
and at the same time, her gums were bleeding
without evidence of trauma to the area. Stat partial
thromboplastin time was >180 s (control, 23-35 s).
The heparin was discontinued, and protamine sul-
fate was administered because of her previous his-
tory of bleeding esophageal varices. Partial throm-
boplastin time decreased to 47 s by afternoon and
33 s by the next day.

By the 6th day of admission, the patient was
incontinent of urine and stool, confused, oriented to
name only, thought she was “at home” and was
performing very inconsistently in therapies as the
result of lethargy and fluctuating awareness. She
once again fell while trying to transfer herself im-
pulsively, and marked tremors (asterixis) were
noted. Hepatic encephalopathy was suspected and
her serum ammonia level was 243 ymol/liter (nor-
mal, 14-55 umol/liter).

Treatment included 30 g of oral lactulose, a Fleet’s
enema, a no protein diet and intravenous hydration.
The spironolactone was decreased, and phenytoin
was replaced with carbamezepine for continued sei-
zure prophylaxis. Following five doses of lactulose
(a total of 120 g) over a 12-h period, extensive
diarrhea occurred and a repeat ammonia level was
56 umol/liter. After 24 h, the patient was more alert
and participating in therapies. A urinary tract infec-
tion was isolated and erythromycin was instituted.
A low protein diet was restarted and the lactulose
was decreased to 20 g twice per day.

Over the succeeding 5 days, the patient’s range of
motion of the left shoulder increased markedly to
120° of abduction and 160° of flexion; she required
no pain medication. The asterixis, ptosis and left
facial weakness resolved completely, and her en-
durance and concentration in therapy improved dra-
matically. She began to ambulate with supervision
only, and became totally independent in transfers
and activities of daily living on the unit. The am-
monia level continued to fluctuate (Fig. 1); however,
the neuropsychologist noted improved memory
skills and less evidence of any perceptual difficulties
or visual-spatial neglect. The results of a repeat MRI
of the brain (Fig. 2) showed T2-weighted images
with high signal changes in the cortical convexities
and deep symmetric areas bilaterally. These changes
were felt to be consistent with anoxia or hypoper-
fusion, or to be of metabolic origin, even though the
large right frontal hyperintensity was still present
(Fig. 2).

The patient experienced no further medical prob-
lems and the ammonia level stabilized in the high-
normal range. After a family conference, she was
discharged to home entirely independent in ambu-

Am. J. Phys. Med. Rehabil.

P1646

Figure 2. MRI scan showing abnormalities consistent
with a metabolic origin.

lation and activities of daily living, with full, painless
range of motion and strength of the left shoulder.
She was alert and oriented with no verbal memory
problems, although she demonstrated some “de-
pendent behaviors” around her family. She was lost
to formal follow-up, but was seen by chance at
another hospital by one of the authors (M. P.) 3
months later. She had experienced recurrent medical
problems, but had no evidence of left hemiparesis,
shoulder pain or other neurologic abnormalities.

DISCUSSION

The diagnosis of hepatic encephalopathy is based
on four criteria: (1) the presence of advanced hepa-
tocellular disease, extensive portal-systemic collat-
eral shunts, or both; (2) disturbances in awareness
and mentation such as confusion, disorientation or
coma; (3) changes in the peripheral neurologic ex-
amination, which often involve fluctuating findings
in the sensory, motor and reflex examinations with
an inconsistent pattern of pathologic reflexes; and
(4) an EEG pattern of symmetrical, high voltage,
slow waves, usually 2~5 per s (Table 1).

The pathogenesis of hepatic encephalopathy in-
volves toxic substances absorbed from the intestine
that the liver cannot metabolize and which thus
accumulate in the brain. There are four commonly
mentioned potential pathogens: (1) increased mer-
captams in blood and cerebrospinal fluid; (2) the
synthesis of false neurotransmitters, such as octo-
Vol. 71, No. 2, April 1992

HEPATIC ENCEPHALOPATHY MIMICKING STROKE 117

TABLE 1
Levels of encephalopathy
Stage Motor/Reflexes Asterixis Mental Status EEG
J Slowed dyscoordination +/- Mild confusion, euphoria, Normal
depression, irritability

I Ataxia, dysarthria, + Lethargic, drowsy, moderate Abnormal
suck/grasp confusion

I Myoclonus, hyperre- + Somnolent but arousable; pro- Abnormal
flexia found confusion/delirium

IV Decerebrate + Coma, may be unresponsive Abnormal

pamine; (3) increased plasma levels of short and
medium chain fatty acids; and (4) hyperammonie-
mia.

Ammonia is thought to be the major toxic agent
in this condition,> although any condition that
causes elevated brain or blood ammonia levels re-
sults in the formation of Alzheimer type II astro-
cytes.° These astrocytes are the distinctive morpho-
logic alteration in the brain of humans who have
died as the result of hepatic encephalopathy,® and
the extent of change correlates well with the severity
of the encephalopathy and the blood levels of am-
monia.’ The formation of Alzheimer type II astro-
cytes preferentially effects the basal ganglia, the
dentate nucleus of the cerebellum and the prefrontal
and temporal regions of the cerebral cortex. This
neurotoxic action is not well understood: earlier
hypotheses included vasoparalysis, energy failure,
neurotransmitter abnormalities or electrophysiologic
derangements.°

A number of conditions can mimic hepatic en-
cephalopathy, particularly in the alcoholic patient.
Wernicke’s encephalopathy, Korsakoff’s psychosis,
delirium tremens and acute intoxication are accom-
panied by confusion and can be confused with
hepatic encephalopathy.’ Previous reports have
shown that subdural hematomas, meningitis, drug
overdose and hypoglycemia can also have the same
clinical presentation.

Although the distinctive abnormal laboratory
finding in hepatic encephalopathy is an elevated
serum ammonia,! Weissenborn et al.2 have shown
that P300 auditory-evoked potentials are sensitive
to early neuropsychologic changes because of he-
patic encephalopathy. Serial neurologic examina-
tions, CT or MRI of the brain, EEG and occasionally
lumbar puncture are often necessary to exclude
these other disease states.

In the case presented, the initial CT scans were
normal, but the MRI scan seemed to indicate a right
frontal infarct with many other areas of abnormality.
After treatment for hepatic encephalopathy, a re-
peated MRI (Fig. 2) continued to show multiple areas
of hyperintensity on the T,-weighted images, which
were thought to be the result of hypoperfusion or
metabolic changes. These findings may be consistent

with earlier hypotheses regarding the mechanism of
neurotoxicity in hepatic encephalopathy, which in-
cluded vasoparalysis, neurotransmitter abnormali-
ties and electrophysiologic derangements to specific
areas of the brain.® The clinical presentation of left
hemiparesis, left facial weakness and other incon-
sistent neurologic findings correlates well with the
areas of involvement on the MRI scan. Previous
reports indicate that patients with subclinical stable
lesions may develop focal neurologic signs when
hepatic encephalopathy develops.*

The patient’s rapid improvement after treatment
is most likely the result of improved neurotransmit-
ter function’ and increased perfusion to the areas
involved. Newer neuroradiologic procedures such
as positron emission tomography and single-photon
emission computed tomography have the ability to
measure these parameters, and might have been
helpful in evaluating this patient.’

In the patient with stable inactive cirrhosis, there
is usually a precipitating factor leading to hepatic
encephalopathy. These factors may include gas-
trointestinal bleeding, increased dietary protein con-
sumption, electrolyte disturbances because of exces-
sive diuresis or vomiting, acute infection, heart fail-
ure and general anesthesia for surgery. In effect,
anything that causes further deterioration of liver
function can precipitate hepatic encephalopathy. As
in this case, there may be more than one contributing
factor: the patient had a known history of chronic
hepatitis and bleeding esophageal varices, was on a
diuretic, underwent general anesthesia and had a
documented urinary tract infection. The simultane-
ous presence of these conditions most likely contrib-
uted to the onset of hepatic encephalopathy.

Treatment is centered on decreasing the serum
ammonia level. Specific treatment should eliminate
any blood from the bowel lumen, and then concen-
trate on decreasing ammonia production by excre-
tion through defecation. Within 24 h neomycin or
lactulose will reduce serum ammonia, and therefore
an improvement in the stages of encephalopathy
should be noted (Table 1).’ Other supportive ther-
apy should be instituted, including elimination of
all dietary proteins, correction of fluid and electro-
lyte abnormalities, discontinuation of all medica-
118 ATCHISON ET AL.

tions metabolized by the liver, if possible, and treat-
ment of any infection with antibiotics.

Our patient showed rapid improvement over the
first 24 h (Fig. 1), but fluctuated for several days
after the initial treatment. Once clinical improve-
ment has been noted, some dietary protein should
be reinstituted to maintain positive nitrogen balance,
especially when intensive physical and occupational
therapy are being performed. Greenberger et al.’
demonstrated that not only the quantity but the
quality of the dietary protein may be important.
Vegetable-derived protein has been shown to be
superior to animal protein in the management of
chronic encephalopathy.” Indeed, our patient had
lower and more stable ammonia levels on a vege-
table protein-only diet. This allowed her to stabilize
medically and to demonstrate marked improve-
ments in her functignal status while being monitored
by the multidisciplinary team.

CONCLUSION

This case shows'the difficulty in diagnosing he-
patic encephalopathy and the need for continued
investigation in a patient with fluctuating neurologic
status. The left-sided weakness and other neurologic
signs changed contiriually, and our diagnosis was
not confirmed until arterial serum ammonia was
shown to be elevated. *~"

This patient responded: quickly to the administra-
tion of oral lactulose and protein restriction, and her
acute and chronic neurologic.abnormalities resolved.
Not all patients with hepatic encephalopathy enjoy
such a complete and rapid recovery after a decrease
in the ammonia level. Neurologic. and functional
deficits may persist, and multidisciplinary rehabili-
tation may be indicated.

Am. J. Phys. Med. Rehabil.

Therefore, in any patient with elevated liver en-
zymes and evidence of neurologic abnormality, a
serum ammonia should be checked to rule out he-
patic encephalopathy as the source of neurologic
compromise. As always, maximal medical and func-
tional improvements are dependent on the correct
diagnosis, treatment and rehabilitation.

REFERENCES

1. Lamont JT, Isselbacher K: Disorders of the hepatobiliary
system, in Petersdorf R, Adams R, Braunwald E, et al. (eds):
Harrison’s Principles of Internal Medicine, ed 10. New York,
McGraw-Hill, 1983, pp 1814-1816,

2. Victor M, Adams RD: Metabolic diseases of the nervous
system, in Petersdorf R, Adams R, Braunwald E, et al. (eds):
Harrison's Principles of Internal Medicine, ed 10. New York,
McGraw-Hill, 1983, pp 2107-2108.

3, Butterworth RF, Gugere J, Michand J, et al: Ammonia: key
factor in the pathogenesis of hepatic encephalopathy. Neurochem
Pathol 1987;6:1-12.

4. Rothstein J, Herlong HF: Neurologic manifestations of he-
patic disease. Neurol Clin 1989;7:563-578.

5. Conn HO, Lieberthal ML: The Hepatic Coma Syndromes and
Lactulose. Baltimore, Williams & Wilkins, 1979.

6. Norenberg M: The role of astrocytes in hepatic encephalop-
athy. Neurochem Pathol 1987;6:13-33.

7. Neary JT, Norenberg LB, Gutierrez M, Norenberg MD:
Hyperammonemia causes altered protein phosphorylation in as-
trocytes. Brain Res 1987;437:161-164.

8. Weissenborn K, Scholz M, Hinrichs H, Wiltfang J, Schmidt
F, Kunkel H: Neurophysiological assessment of early hepatic
encephalopathy. Electroencephalogr Clin Neurophysiol 1990;75:
289-295.

9. Corenblum B, Shaffer E: Hyperprolactinemia in hepatic
encephalopathy may result from impaired central dopaminergic
neurotransmission. Horm Metab Res 1989;21:675-677.

10. Maurer A: Nuclear medicine: SPECT comparisons to PET.
Radiol Clin North Am 1988;26:1059-1074.

11. Greenberger NJ, Carley J, Schenker S, et al.: Effect of
vegetable and animal protein diets in chronic hepatic encepha-
lopathy. Am J Dig Dis 1977;22:845-855.

12. Uribe M, Marques M, Ramos G, et al.: Treatment of chronic
portal-systemic encephalopathy with vegetable and animal pro-
tein diets. Dig Dis Sci 1982;27:1109-1116.