Full text of DOI:10.1007/BF03343822

J. Endocrinol. Invest. 24: 104-106, 2001
CASE REPORT
Peritoneal dialysis in an infant with Type 1 diabetes and
hyperosmolar coma
G. Multari*, B. Werner*, M. Cervoni*, R. Lubrano*, F. Costantino*, V. Demiraj,
and P. Pozzilli**
*Departement of Pediatrics, University of Rome “La Sapienza”, **Department of Internal Medicine,
University of Rome “Tor Vergata” and University Campus Biomedico, Rome, Italy
ABSTRACT. Hyperosmolar coma which is charac-
terized by severe hyperglycemia in absence of
chetosis is very rare in pediatric age with only 11
cases reported in the literature. The outcome of
the condition is usually poor with mental retarda-
tion being the most common event. Here a case of
hyperosmolar coma is described in a female of
three months of age who was treated with peri-
toneal dialysis 11 hours after admittance to hos-
pital. This female patient has been receiving in-
sulin from three months of age and today at the
age of 10 years she leads a normal life despite be-
ing on insulin therapy. A very low level of C-pep-
tide (<0.3 ng/ml) clearly confirms she is affected
by Type 1 diabetes. To our knowledge this is the
first case report of hyperosmolar coma in a
neonate with Type 1 diabetes who survived this
condition without late neurological consequences.
(J. Endocrinol. Invest. 24: 104-106, 2001)
©2001, Editrice Kurtis
INTRODUCTION
CASE REPORT
Hyperosmolar coma is characterized by severe hy-
perglycemia (>600 mg/dl), hypernatriemia, absence
of chetosis, severe dehydratation, coma of various
degrees, convulsion and hemiparesis (1). It is un-
common in pediatric age and until today only 11
cases have been described (2). The first case was
reported in a child affected by Down syndrome, in
another four cases patients died soon after diag-
nosis, and two patients were described as severe-
ly mentally retarded (3-5). In another infant, aged
6 weeks, the adoption of “exanguinotrasfusion” for
the persistently high blood glucose level was un-
successful (6).
A case is here presented of hyperosmolar coma oc-
curring in a neonate of three months of age, resis-
tant to insulin treatment with persistent hyper-
glycemia and anuria who underwent peritoneal dial-
ysis. She survived and now at the age of ten, leads
a normal life despite being on insulin.
A 15-week-old female was brought to the hospital
with a high temperature lasting for 48 hours, diar-
rhea lasting 12 hours and polypnea. Second-born
by normal delivery at 40 weeks, she weighed 3050
g. Both Apgar score and neonatal period were nor-
mal. She was breast-fed until the second month of
age and then mix-fed with soya milk with evidence
for an increasing demand for milk over the previ-
ous 2 weeks (7 meals/die, 180 ml each or 220
ml/kg/die). There was also a noticeable decrease
in body weight (from 5,900 to 5,500 g in a week).
Maternal grandfather and both paternal grand-
mother and grandfather were affected by Type 2
diabetes.
At the time of admittance she was found to be de-
hydrated (weight decrease by 6%), with a temper-
ature of 38.5 C, heart rate of 150/min and respira-
tory rate of 60/min.
Laboratory data at the time of admittance were:
44% Ht, 13.6% Hb, 23,300 mm3 white cells (Neu-
trophils 66%), 650 mg/dl blood glucose, 6.2 mEq/l
K, 158 mEq/l Na (not corrected for glucose levels),
118mEq/l Cl, pH 7.1, 20.2 mmHg pCO2, 50.0
mmHg pO2, 5.8 mEq/l bicarbonates, 376 mOsm/l
serum osmolarity and presence of glycosuria.
Treatment began initially with 0.45% NaCl 8
ml/kg/h for 2 hours followed by 0.25 IU/h insulin in-
Key-words: Hyperosmolar coma, Type 1 diabetes, peritoneal dialysis.
Correspondence: Prof. P. Pozzilli, Università Campus Biomedico, Via
Emilio Longoni 83, 00155 Rome, Italy.
E-mail: p.pozzilli@caspur.it
Accepted July 24, 2000.
104

Peritoneal dialysis in hyperosmolar coma
fusion. After 6 hours blood glucose levels raised to
1,388 mg/dl with pH 7.34, 170 mEq/l Na, 3.6 mEq/l
K, 410 mOsm/l osmolarity, 3.3 mmol/l lactate, 23
mmHg pCO2, 74 mmHg pO2, and an arterial pres-
sure of 85/48. However 11 hours later, blood glu-
cose dropped to 650 mg/dl with 180 mEq/l Na, 4.3
mEq/l K, 135 mEq/l Cl, 9.9 mmol/l lactate, 70 mg/dl
BUN and 1.8 mg/dl creatinine. The patient was olig-
uric (60 ml x 24 h).
bodies was not performed at the time of admission
to hospital 10 years ago.
DISCUSSION
To our knowledge this is the first report of hyperos-
molar coma in a neonate undergoing peritoneal dial-
ysis who survived such a condition without showing
late neurological or other consequences than Type 1
diabetes. At the presentation therapeutic measures
were taken as expected in a case like this. However,
despite infusion therapy with saline hypotonic solu-
tion first, followed by insulin infusion after, blood glu-
cose and sodium levels remained very high after 24
hours. The clinical situation got worse with the ap-
pearance of lactic acidosis and anuria. To compen-
sate for the severe dehydration, further administra-
tion of fluids and bicarbonate were carried out to
correct lactic acidosis. In a similar case reported in
literature, exanguinotransfusion was unsuccessful (6).
Persisting lactic acidosis and the appearance of
anuria prompted us to implement peritoneal dialysis.
This was carried out to maintain liquid in the ab-
domen for a short time to prevent high blood glu-
cose from causing a massive re-absorption of fluids
from the abdomen. A progressive reduction of blood
glucose, sodium levels and lactate with a normaliza-
tion of acid-base equilibrium was obtained 72 hours
later. Neurological manifestation observed two hours
after beginning peritoneal dialysis should have been
caused by hyperosmolarity and the unlikely throm-
bosis of cerebral arteries for vascular disseminated
coagulation, even though it was not detectable at
computerized tomography.
Given that anuria with lactic acidosis developed and
that hyperglycemia and hyperosmolarity persisted,
the patient was placed on peritoneal dialysis while
simultaneously administered insulin (0.25 IU/kg/h).
Briefly, intermittent peritoneal dialysis was per-
formed using a 14 diameter gauge positioned in
the right pelvic fossa. For the in and out flow of dial-
ysis fluid we used a Y system with a separate line
for each flow. Isotonic bags (Traverol) with low glu-
cose concentration (1.36%) were employed. The to-
tal quantity of fluids administered ranged between
100-150 ml, whereas fluids removed ranged be-
tween 66 and 165 ml. The total of peritoneal dial-
ysis lasted for 45 hours. Two and a half hours after
beginning dialysis, the patient experienced con-
vulsions with each crisis lasting between 2 and 3
minutes at various intervals of 20-35 minutes. The
dialysis was stopped during convulsion. A calcium
value of 6.8 mg/dl was reported. She was then
treated with benzodiazepine first and 10% calcium
gluconate soon after, intubated and transferred to
the Neonatal Intensive Care. Pulmonary ventilation
with positive pressure followed by intermittent pul-
monary ventilation (IPPV) were implemented.
Eleven hours after beginning IPPV, her breathing
became spontaneous. Subsequently, the oxymet-
ric values gradually set off normality, improvement
of sensorium, decrease of blood glucose up to near
normal values, progressive normalization of sodi-
um, lactate and base excess were observed. Lactate
became normal in 20 hours and diuresis returned
physiological 20 hours later.
Management of the hyperosmolar hyperglicemic
coma is not easy especially when other complica-
tions arise. The key for a successful outcome is a
prompt, rapid administration of crystalloid solution
that has tonicity appropriate to the level of hyper-
osmolarity (7). In the presence of renal failure treat-
ment becomes extremely difficult.
The patient was transferred to the ward; her neu-
rological examination at this stage was normal.
Meals consisted of adapted milk every 3 hours with
small insulin boluses. She was discharged from hos-
pital 60 days later without any neurological defects,
insulin being her sole therapy.
Today (2000) this girl is 10 years old, her metabol-
ic control with subcutaneous insulin is good,
weight-height development is above average (75°
centile), neurological defects are not present and
she regularly attends school succeeding in her stud-
ies. C-peptide levels in this patient are very low
(<0.3 ng/ml) clearly indicating that she is affected
by Type 1 diabetes. Measurement of islet cell anti-
We have no evidence to state that implementation
of peritoneal dialysis was the reason for the suc-
cessful outcome of this baby girl.
However, we were forced to adopt such a procedure
because of anuria and it might be possible that its
use has accelerated the clinical improvement, there-
fore limiting the time of exposure to high blood glu-
cose which could have caused brain damage.
In conclusion, implementation of peritoneal dialysis
seemed to be the variable which might have chan-
ged the course of hyperosmolar coma in this
neonate at least compared with other cases re-
ported in the literature (2). Such an approach, there-
fore, should be considered in similar cases to re-
105

G. Multari, B. Werner, M. Cervoni, et al.
4. Rubin H.M., Kramer R., Drash A.
Hyperosmolarity complicating diabetes mellitus in
childhood.
duce the risk of severe complications associated
with hyperosmolar coma in neonates.
J. Pediatr. 1969, 74: 177-186.
REFERENCES
5. Hoffman W.H., Fernandos S.S.
Hyperglycemic hyperosmolar non-ketotic coma in a
non-diabetic child.
1. Gupta S., Prabhu M.R., Gupta M.S., Niblett D.
Severe non-ketotic hyperosmolar coma-intensive
care management.
Diabetologia 1983, 25: 531.
Eur. J. Anaesthesiol. 1998, 15: 603-606.
6. Vemon D.D., Postellon D.C.
Non-ketotic hyperosmolar diabetic coma in a child.
Management with low dose insulin, infusion and in-
tracranial pressure monitoring.
2. Eidiliz-Markus T., Nussinovic M., Varsano I.,
Kauschansky A.
Non ketotic hyperosmolar coma in children.
Isr. J. Med. Sci. 1994, 30: 585-587.
Pediatrics 1986, 77: 70-77.
7. Matz R.
3. Yamashiro Y., Yamamoto T., Mayama H.
Non ketotic hyperosmolar coma in two diabetic chil-
dren.
Management of the hyperosmolar hyperglycemic
syndrome.
Am. Fam. Physician 1999, 47: 659-665.
Acta Paediatr. Scand. 1981, 70: 337-340.
106