Tohoku J. Exp. Med., 2001, 193, 73-77
Case Report

Thrombotic Stroke in a Child with
Diarrhea-Associated Hemolytic-Uremic
Syndrome with a Good Recovery

Touru Nakauwata, Hrrosut Tanaka, TAKASHI TATEYAMA, Tomomi UEDA,
Koicut Suzuxi, Surn-tcut Osari,! Mrxio Kasar' and Suinopu Waca

Department of Pediatrics, Hirosaki University School of Medicine,
Hirosaki 036-8562 and
1 Division of Pediatrics, Hirosaki Municipal Hospital, Hirosaki 036-8182

Naxanata, T., Tanaka, H., Tatevama, T., Uspa, T., Suzuxt, K., Osarr, S.,
Kasat, M. and Waca, 8. Thrombotic Stroke in a Child with Diarrhea-Associated
Hemolytic-Uremic Syndrome with a Good Recovery. Tohoku J. Exp. Med., 2001,
193 (1), 73-77 -—— A boy aged 3.5 years with post-diarrheal hemolytic-uremic
syndrome (HUS) was referred to our hospital because of convulsion and stupor.
He had been admitted to a regional hospital with a 3-day history of bloody
diarrhea, colic abdominal pain and fever. Two days later, he had complained of
generalized seizures and oliguria. On admission, he developed anuria, and serum
blood nitrogen and creatinine increased to 56mg/100 ml and 2.8 mg/100 ml,
respectively. Platelets decreased to 42 000/y1. Under the diagnosis of HUS, a
continuous hemodiafiltration treatment had to be instituted. Computed tomogra-
phy of his head at hospital day 5 revealed abnormal low density area of infarction
with edema in both the basal ganglia involving with the posterior limb of internal
capsule. Serum titer of IgM antibody to Escherichia coli 0157 showed positive
value. Although his anuria and stupor persisted over 10 days, he recovered
without serious complications. These clinical observations may indicate that
patients with similar lesions do not necessarily have serious morbidity.
central nervous system involvement; continuous hemodiafiltration; favorable
outcome; hemolytic-uremic syndrome; stroke
© 2001 Tohoku University Medical Press

Received November 1, 2000; revision accepted for publication January 29, 2001.
Address for reprints: Hiroshi Tanaka, M.D., Department of Pediatrics, Hirosaki University School of
Medicine, 5 Zaifu-cho, Hirosaki 036-8562, Japan.
e-mail: hirotana @cc.hirosaki-u.ac.jp
73
14 T. Nakahata et al.

Although the hemolytic-uremic syndrome
(HUS) is the common cause of acute renal
failure in young children, the central nervous
system (CNS) involvement has been reported to
be a major contributor to mortality or chronic
morbidity of the HUS (Bale et al. 1980;
Upadhyaya et al. 1980; Hamano et al. 1993;
Siegler 1994; Gallo and Gianantonio 1995). It
is generally assumed that the HUS ence-
phalopathy is due to metabolic derangement
and action of Shiga-like toxin causing microvas-
cular damage (Upadhyaya et al. 1980; Hamano
et al. 1993; Siegler 1994). Of the CNS involve-
ment, about 3°%% to 5% of children with HUS
have identifiable structural brain damage in the
form of infarcts (Siegler 1994). Despite the
general assertion that severe CNS manifesta-
tions such as generalized seizures and prolonged
coma/stupor portend a poor prognosis (Bale et
al. 1980; Trevathan and Dooling 1987; Siegler
1994; Gallo and Gianantonio 1995), some chil-
dren with HUS complicated with identifiable
structural brain damage have been reported to
have a good recovery (Steele et al. 1983; Stein-
berg et al. 1986; Signorini et al. 2000).

We observed a boy with HUS complicated
with thrombotic stroke, who showed a good
recovery despite prolonged stupor.

CASE REPORT

A Japanese boy aged 3.5 years referred to
our hospital because of generalized tonic sei-
zares and lethargy. He had had bloody diar-
rhea, colic abdominal pain and fever for 3 days,
and had been treated with oral fosfomycin
(Meiji Co., Tokyo) by a physician. Then he
was admitted to a regional hospital. Labora-
tory studies revealed the following values:
white blood cells (WBC) 27 800/21, hemoglobin
14.7 g/100 ml, platelets 376 000/1, serum urea
nitrogen (UN) 14 mg/100 ml, creatinine 0.3 mg/
100 ml and C-reactive protein (CRP) 5.0 mg/
100 ml. Stool culture and a latex aggregation
test for Shiga-like toxin were negative. Two

days later, he developed generalized tonic sei-
zures and oliguria. Computed tomography
(CT) of the brain at that time was unremarka-
ble. Serum UN and creatinine increased to 62
mg/100ml and 2.4mg/100 ml, respectively.
Platelets decreased to 73 000/w1. He was trans-
ferred to us under the diagnosis of post-
diarrheal HUS.

On admission, the patient was pale and
insensitive to pain. Blood pressure was 120/70
mmHg. He developed anuria. Laboratory
studies revealed the following: WBC, 27 500/,1
with neutrophils of 74%; hemoglobin, 7.7 g/100
ml; hematocrit, 23.6%; platelets, 42 000/,1;
serum total protein, 4.8 g/100 ml; albumin, 2.1
g/100 ml; UN, 56mg/100 ml; creatinine, 2.8
mg/100 ml; sodium, 120 mEq/liter; potassium,
4.4 mEq/liter; chloride, 97 mEq/liter; calcium,
7.7 mg/100 ml and CRP, 11.8mg/100 ml. Im-
munological studies showed the following val-
ues: IgG, 1310 mg/100 ml; IgA, 128 mg/100 ml;
IgM, 148 mg/100 ml; C3, 80 mg/100 ml (normal
range, 79-152 mg/100 ml); C4, 26 mg/100 ml
(normal range, 16-38 mg/100 ml) and hemolytic
complement activity, 37.8 U/ml (normal range,
23-46 U/ml). Anti-nuclear antibody was
absent. Neither acidemia nor hyperammonemia
were identified. Peripheral blood smear
showed red blood cell (RBC) fragments.
Serum lactate dehydrogenase (LDH) increased
to 3652 U/liter (normal range, 80-200), and
haptoglobin decreased to less than 6 mg/100 ml
(normal range, 36-195 mg/100 ml). The serum
level of fibrin/fibrinogen degradation products
(FDP) showed a significant increase at 104.9
mg/100 ml (normal range, 0-5.0 mg/100 ml).
Although repeated stool cultures were negative,
serum IgM antibody to #. cols 0157 (measured
by enzyme-linked immunosorbent assay at
Institute of Infectious Diseases, National
Children’s Hospital) showed positive value.
The diagnosis was HUS associated with acute
renal failure and CNS involvement. Because
of technical problems, measurement of serum
antibodies to Shiga-like toxin (Verotoxin 1 and
Stroke in Hemolytic-Uremic Syndrome 15

2) was not done.

He was treated with continuous
hemodiafiltration (CHDF) and anticoagulants
(nafamostat mesilate and heparin), since CHDF
had been reported to be effective in treating and
preventing fluid overload in children with
multiorgan failure (Lowrie 2000). Blood pres-
sure and body weight were well controlled by
CHDF treatment. When the level of hemoglo-
bin dropped to less than 6.5 g/100 ml, transfu-
sion of packed RBC (200 ml) was performed.

Se: 1
om $50.03
In: 10

DFOV 22.0ce

Seizures recurred and stupor persisted.
Electroencephalogram showed diffuse high volt-
age slow waves without seizure discharges.
Head CT images obtained 4 days after admis-
sion disclosed abnormal widespread low density
area around both basal ganglia (Fig. 1). There
was no evidence of hemorrhage. Since the
brain images seemed to be edematous, intra-
venous dexamethasone administration (1.6 mg/
8hours) for 5days was started. Treatment
with midazolum and phenobarbital was also

DFOV 2
STND/T

Fig. 1. Axial CT scan, showing areas of decreased density in both basal ganglia in the region of the

putamen.

IMAGE 77
SER 1-4

tse2-5 180
oR
1 SAT

TR 4000.0
TE 102.0/2

sP

SL »
Fov 210*240
250 *2560s
Tracor 8&

Fig. 2. T2-weighted magnetic resonance imaging on hospital day 33, showing slit-shaped high signal

images on the both putamen.
76 T. Nakahata et al.

started and he had no further definite convul-
sions.

After 12 days of these treatment, his urine
output was gradually increased and the level of
serum UN and creatinine gradually decreased to
the normal ranges by hospital day 24. The
CHDF treatment was discontinued at hospital
day 18. A total of 11 times of packed RBC
transfusion was performed. Neither platelets
transfusion nor fresh frozen plasma infusion
were performed. Platelets count, serum LDH
and FDP returned to the normal ranges by
hospital day 10,19 and 24, respectively. The
duration of stupor was 32 days. Thereafter,
steady improvement in neurologic function was
Magnetic resonance images (MRI)
obtained hospital day 33 disclosed marked reso-
lution of the lesion except for a slit area of high
signal intensity on T2-weighted images in both
putamen (Fig. 2).

At present, 5 months after our first observa-
tion, he is well except for a residual left
hemiparesis, mild dysphemia and mild
proteinuria (approximately 60 mg/100 ml).

seen.

DISCUSSION

So far the importance of the CNS involve-
ment in HUS patients has been reported (Bale
et al. 1980; Upadhyaya et al. 1980; Cimolai et
al. 1992; Hamano et al. 1993; Siegler 1994;
Gallo and Gianantonio 1995). Although path-
ogenesis of the CNS involvement in HUS
remained to be elucidated, several factors, such
as metabolic derangement, action of Shiga-like
toxin or microvascular damage may be respon-
sible for the lesion (Upadyaya et al. 1980;
Cimolai et al. 1992; Siegler 1994; Gallo and
Gianantonio 1995). Recent report (Gallo and
Gianantonio 1995) described that extensive
microvascular damage and thrombosis may con-
tribute to the multiorgan involvement in
diarrhea-associated HUS. Further, Fujii et al.
(1996) reported that rabbits given intravenous
verotoxin 2 showed a variety of brain damage,

and that the edematous changes determined by
MRI and ultrastractual studies were caused by
the toxicity of verotoxin 2 which was conveyed
from the endothelial and ependymal cell layers.

In our patient, despite persistent anuria, no
metabolic derangement or hypertension were
seen during the clinical course. Hence, mi-
crovascular damage probably caused by Shiga-
like toxin from E. coli 0157, and subsequent
vascular thrombosis was the most likely cause
of the stroke lesion.

It has also been reported that the impor-
tance of systemic release of proinflammatory
cytokines in acute inflammatory processes in
HUS patients is emphasized (Karpman et al.
1995; Inward et al. 1997). Karpman et al.
(1995) reported that serum interleukin (IL)-6 is
higher in children with HUS who developed
extrarenal manifestations. Also, Inward et al.
(1997) reported that a patient with fatal CNS
involvement exhibited high concentration of
IL-6 in the plasma. From the current litera-
ture, circulating inflammatory cytokines as well
as thrombosis caused by microvascular damage
may be responsible for the CNS involvement.

In our patients, although measurement of
serum cytokines was not done, prompt initia-
tion of CHDF treatment may have contributed
to remove circulating inflammatory cytokines
and to prevent fluid overload, and subsequently
resulted in a good recovery.

Since severe CNS manifestations in HUS
patients are assumed to portend a poor progno-
sis (Bale et al. 1980; Trevathan and Dooling
1987; Hamano et al. 1993; Siegler 1994; Gallo
and Gianantonio 1995), CHDF treatment may
be of benefit in a proportion of HUS patients
with severe CNS involvement to change the
natural course of the disease as in the recent
report describing children with multiorgan fail-
ure (Lowrie 2000).

Several patients with severe CNS involve-
ment with a favorable outcome have been repor-
ted sporadically to date (Steele et al. 1983;
Steinberg et al. 1986; Signorini et al. 2000).

Stroke in Hemolytic-Uremic Syndrome 17

The treatment of these patients varies, ie.,
peritoneal dialysis (PD) (Steele et al. 1983), PD
with fresh frozen plasma infusion (Steinberg et
al. 1986) and hemodialysis with plasma changes
(Signorini et al. 2000). Hence, an appropriate
therapeutic intervention for HUS patients with
severe CNS involvement remains to be elucidat-
ed. Further reports to describe similar cases
are needed.

Long-term prognosis regarding quality of
life in children with episodes of HUS has been
reported to be deficit in verbal intelligence and
in the verbally based skills of reading compre-
hension, vocabulary use and in behavior
(Schlieper et al. 1992). In our patient, residual
left hemiparesis, mild dysphemia and mild
proteinuria remained. Careful follow up
should be done.

Acknowledgments

The authors thank Dr. Hirofumi Ohkawa and
Dr. Toshihito Tsubo for medical care in the emer-
gency room of Critical Care and Emergency Medi-
cine, Hirosaki University and Dr. Tae Takeda,
Institute of Infectious Diseases, National Children’s
Hospital for kindly measurement of IgM antibody
to E. coli 0157.

References

Bale, J.F., Jr., Brasher, C. & Siegler, R.L. (1980)
CNS manifestations of the hemolytic-uremic
syndrome. Relationship to metabolic altera-
tions and prognosis. Am. J. Dis. Child., 134,
869-872.

Cimolai, N., Morrison, B.J. & Carter, J.E. (1992)
Risk factors for the central nervous system
manifestations of gastroenteritis-associated
hemolytic-uremic syndrome. Pediatrics, 90,
616-621.

Fujii, J., Kinoshita, Y., Kita, T., Higuchi, A.,
Takeda, T., Tanaka, N. & Yoshida, S. (1996)
Magnetic resonance imaging and histopath-
ological study of brain lesions in rabbits given
intravenous verotoxin 2. Infect. Immun., 64,
5053-5060.

Gallo, G.E. & Gianantonio, C.A. (1995) Extrarenal
involvement in dairrhoea-associated

haemolytic-uraemic Pediatr.
Nephrol., 9, 117-119.

Hamano, S., Nakanishi, Y., Nara, T., Seki, T., Ohtani,
T., Oishi, T., Joh, K., Oikawa, T., Muramatsu,
Y., Ogawa, Y. & Akashi, 8. (1993) Neuro-
logical manifestations of hemorrhagic colitis
in the outbreak of Escherichia coli 0157: H7
infection in Japan. Acta Paediatr., 82, 454-
458.

Inward, C.D., Varagunam, M., Adu, D., Milford, D.V.
& Taylor, C.M. (1997) Cytokines in
haemolytic uraemic syndrome associated with
verocytotoxin-producing Escherichia coli
infection. Arch. Dis. Child., 77, 145-147.

Karpman, D., Andreasson, A., Thysell, H., Kaplan,
B.S. & Svanborg, C. (1995) Cytokines in
childhood hemolytic uremic syndrome and
thrombotic thrombocytopenic purpura.
Pediatr. Nephrol., 9, 694-699.

Lowrie, L.H. (2000) Renal replacement therapies
in pediatric multiorgan dysfunction syndrome.
Pediatr. Nephrol., 14, 6-12.

Schlieper, A., Rowe, P.C., Orrbine, E., Zoubek, M.,
Clark, W., Wolfish, N. & McLaine, P.N.
(1992) Sequelae of haemlytic uraemic syn-
drome. Arch. Dis. Child., 67, 930-934.

Siegler, R.L. (1994) Spectrum of extrarenal in-
volvement in postdiarrheal hemolytic-uremic
syndrome. J. Pediatr., 125, 511-518.

Signorini, E.S., Lucchi, 8., Mastrangelo, M., Rapuzzi,
S., Edefonti, A. & Fossali, E. (2000) Central
nervous system involvement in a child with
hemolytic uremic syndrome. Pediatr. Ne-
phrol., 14, 990-992.

Steele, B.T., Murphy, N., Chuang, S.H., McGreal, D.
& Arbus, G.S. (1983) Recovery from pro-
longed coma in hemolytic uremic syndrome.
J. Pediatr., 102, 402-404,

Steinberg, A., Ish-Horowitez, M., El-Peleg, O., Mor, J.
& Branski, D. (1986) Stroke in a patient
with hemolytic-uremic syndrome with a good
outcome. Brain Dev., 8, 70-72.

Trevathan, E. & Dooling, E.C. (1987) Large throm-
botic strokes in hemolytic-uremic syndrome.
J. Pedaatr., 111, 863-866.

Upadhyaya, K., Barwick, K., Fishaut, M.,
Kashgarian, M. & Siegel, N.J. (1980) The
importance of nonrenal involvement in
hemolytic-uremic syndrome. Pediatrics, 65,
115-120.

syndrome.