Professional Documents
Culture Documents
aDivision of Pediatric Endocrinology, bDepartment of Pathology, Divisions of cPediatric Neurology and dPediatric Critical Care Medicine, and eDepartment of Radiology,
The authors have indicated they have no financial relationships relevant to this article to disclose.
ABSTRACT
Cerebral edema is the most common neurologic complication of diabetic ketoacidosis in children. A minority of
young patients with intracerebral crises in diabetic ketoacidosis present with cerebrovascular accidents. We report 2
adolescent patients with diabetic ketoacidosis who presented with coma and diffuse white matter hemorrhages in the
absence of either cerebral edema or cerebrovascular accidents. These 2 cases illustrate a novel clinical and neuro-
pathologic description of diffuse white matter hemorrhages, possibly related to a cytotoxic process as the underlying
mechanism. These case descriptions emphasize that pediatric patients with diabetic ketoacidosis and coma can
present with pathology not related to either cerebral edema or cerebrovascular accidents.
e1540 MAHMUD et al
Downloaded from pediatrics.aappublications.org at Leeszaal Wilhelmina on March 14, 2015
TABLE 1 Comparison of Clinical, Initial Laboratory, and Diagnostic Data
Patient 1 Patient 2
Age, y 11 14
Blood glucose level (initial), mmol/L (mg/dL) 17 (306) 34.8 (685)
Sodium level (corrected), (reference range: 135–145 mmol/L) 141 140
Potassium level, (reference range: 3.0–5.5 mmol/L) 5.3 4.6
Chloride level, (reference range: 98–107 mmol/L) 110 104
Anion gap, mEq/L (reference range: 7–15 mEq/L) 29 23
Serum osmolarity (effective osmolarity), mOsm/L 359 (308) 344 (313)
Lactate, mmol/L (reference range: 0.9–2.5 mmol/L) 1.7 2.5
Arterial blood gas values
pH 6.96 6.74
PCO2 (mmHg) 31 32
PO2 (mmHg) 71 66
Bicarbonate (mmol/L) 7 4
Ketone level, mM (reference range: 0.058–0.170 mM)a 5.8 4.2
Creatine kinase level, (reference range: 26–140 U/L) 22 000 455
Myoglobin level (urine), g/L (reference range: 0–12 g/L) ⬎10 000 435
Amylase level, (reference range: 36–128 U/L) 755 800
Coagulation test results Normal Normal
Liver-function test results Normal Normal
Hemoglobin A1c, % (reference range: 4%–6%) 11.5 11.4
Initial fluids, mL/m2 per 24 h 3600 2700
Insulin therapy 0.1 U/kg bolus (transit) and 0.1 then 0.03 U/kg per h infusion
0.05 U/kg per h infusion
Cerebral imaging (CT and MRI/MRA) results Normal White matter changes
Toxicology results Negative Negative
Metabolic test results Negative Negative
Cerebrospinal fluid Normal Normal
Electroencephalography Burst suppression Diffuse slowing
Course in hospital Persistent acidosis; died at 72 h Coma for 8 d; slow recovery;
persistent neurologic deficits
The corrected sodium level was calculated as measured sodium ⫹ 0.016 ⫻ (serum glucose ⫺ 100), and effective serum osmolarity was calculated as 2(Na) ⫹ glucose (mmol/L) using values timed
with measured serum osmolarity. Arterial blood gases were measured within 2 hours after critical care admission and once stable ventilation was achieved. The insulin bolus was administered in
transit at the peripheral hospital. Coagulation studies included platelets, prothrombin time, partial thromboplastin time, fibrinogen, D-dimer, protein C, protein S, antithrombin III, and factor V Leiden
antibody. Metabolic tests included urine organic acids, serum amino acids, lactate, pyruvate, biotin, metachromatic leukodystrophy, and mitochondrial DNA point-mutation analysis. MRA indicates
magnetic resonance angiography.
a Blood  -hydroxybutyrate.
was 17 mmol/L (306 mg/dL), and results of a urine and perfusion (heart rate: 130 beats per minute; blood
dipstick test were positive for ketones. A provisional pressure: 80/30 mm Hg). Her examination was negative
diagnosis of DKA was made, and the patient was given 2 for signs of trauma, her pupils were equal at 2 mm and
L of normal saline delivered intravenously over 2 hours reactive to light; results of fundoscopy were normal, and
and a single 10-U regular intravenous insulin push. no abnormal brainstem signs were elicited. The remain-
Maintenance fluids (normal saline with 20 mEq/L KCl) der of her examination was normal except for a yellow
were continued at 100 mL/hour. Because of the limited vaginal discharge. She did not have acanthosis nigricans.
medical resources in her community, the patient was Her laboratory data are summarized in Table 1.
transferred to our tertiary care pediatric hospital by air Broad-spectrum antibiotic and antifungal coverage
ambulance. Approximately 9 hours after her initial pre- was administered, and serum osmolality was corrected
sentation, while in flight, there was a rapid deterioration slowly over 36 hours. Computed tomography (CT) and
in her level of consciousness, and the plane was diverted MRI (Fig 1A and B) of the head were unremarkable.
to the nearest medical center. There she was intubated, Serial electroencephalograms demonstrated a burst-sup-
ventilated, and given 100 g of intravenous mannitol pression pattern consistent with profound encephalopa-
followed by 500 mL of 3% saline, which resulted in 1000 thy. Imaging of the chest (radiography and CT) revealed
mL of urine output but no improvement in her level of interstitial pulmonary edema, and results of CT of her
consciousness. In addition to intravenous saline boluses, abdomen and pelvis were normal. There was no evi-
dopamine was administered to treat her hypotension dence for bacterial, fungal, or viral infection in cultures
and poor perfusion. and polymerase chain reaction testing of her blood,
On arrival at our hospital, the patient had a GCS score urine, and cerebrospinal fluid (opening pressure: 25
of 3 in the absence of sedation. She received fluid and mm H2O). Over the course of 3 days, the patient became
further inotropic support to stabilize her blood pressure euglycemic on insulin infusion but remained on inotro-
e1542 MAHMUD et al
Downloaded from pediatrics.aappublications.org at Leeszaal Wilhelmina on March 14, 2015
ensure normal intracranial pressure during correction of
her hyperosmolar state. MRI of the brain showed no CE.
There were, however, multiple well-defined hyperin-
tense T2 foci associated with petechial hemorrhages in-
volving the subcortical white matter U fibers bilaterally,
genu of corpus callosum, and the posterior limb of the
internal capsule bilaterally. There was a single focus of
enhancement associated with a lesion in the left frontal
lobe (Fig 2A and B). An angiogram showed numerous
small segmental foci of decreased vessel caliber in the
anterior and posterior circulations. On day 6 of admis-
sion, an MRI-guided left frontal craniotomy for brain
biopsy was performed on the lesion in the left frontal
lobe.
The brain biopsy showed no abnormalities of the
cortical neurons. There were extensive areas of petechial
hemorrhage (Fig 2C) surrounded by pale white matter,
with a ring-and-ball morphology (Fig 2D), with associ-
ated axonal injury but no evidence of demyelination (Fig
2E) or vasculitis (Fig 2F).
By day 10 of hospitalization, the patient demon-
strated a slow improvement in her neurologic condition
and no longer required ventilatory or inotropic support.
In the initial phase of her convalescence, she had severe
aphasia, very poor short-term memory, and motor
apraxia, which were suggestive of diffuse frontal and
temporal injury. She never exhibited focal motor or
FIGURE 2
sensory findings. She was transferred to a brain rehabil- Imaging and neurosurgical pathology findings (patient 2). A, Head MRI: single axial fluid-
itation unit and received intensive therapy over the next attenuated inversion recovery image demonstrating multiple well-defined foci of abnormal
2 months; she then was discharged from the hospital. increased T2 signal scattered throughout both cerebral hemispheres involving predomi-
nantly the subcortical white matter. B, Single coronal multiplanar gradient recalled acquisition
An MRI head scan was repeated 7 months after her image. The arrows demonstrate punctate foci of abnormal diminished T2 signal associated
initial presentation (data not shown). There was diffuse with these lesions compatible with petechial hemorrhage. C, This low-power view of a gyrus
subcortical white matter volume loss. The magnetic res- (stained with hematoxylin, eosin, and Luxol fast blue) illustrates multiple small and micro-
scopic hemorrhages (arrowheads) associated with “confluent” pallor (asterisks) of the myelin,
onance angiogram and venogram were normal. The pa- a thin layer of preserved subcortical myelin (arrows), and normal cortex. D–F, These photomi-
tient has been followed in our pediatric diabetes and crographs were taken from the same region; a vessel, labeled v, is identified for reference. In D,
neurology outpatient clinics. There is no neurologic def- stained with Luxol fast blue, hematoxylin, and eosin (which stains myelin blue), the central
hemorrhage was formed by perivascular necrosis (arrow), reminiscent of the lesions in patient
icit on formal examination, but she has moderate cog- 1, a concentric ring of red blood cells, and diffusely rarefied white matter that is speckled with
nitive deficits and short-term memory loss. She cur- eosinophilic astrocytes (arrowheads). Shown in E is a photomicrograph from a section treated
rently uses insulin glargine and aspart as part of a with neurofilament immunohistochemistry (which labels axons dusky brown). Note the
coarsening of axonal labeling in the center of the lesion, which indicates perivascular
multiple daily-injection regime and maintains her he- axonal injury (arrow), the loss (or separation) of axons in the red blood cell layer (asterisk),
moglobin A1c near 8.0%. She experiences occasional and the proportional labeling of axons compared with the myelin staining (illustrated in
episodes of moderate hypoglycemia and has not had D). Shown in F is a photomicrograph of a section treated with CD68 immunohistochem-
istry (which labels microglia [rod-shaped or spindled cells] and histiocytes [round cells]
subsequent admissions to the hospital for DKA. brown). Note that there is no predilection of the activated/proliferating microglia/histio-
cytes for the perivascular lesions. Marked accumulation of these cells only occurred ad-
DISCUSSION jacent to, and in, the layer of preserved subcortical myelin. The pathological findings in
case 2 are similar to those in case 1 with respect to the widely distributed small and
DKA is a common initial presentation of pediatric type 1 microscopic hemorrhages. The hemorrhagic lesions differ between the cases in that,
diabetes, with defined biochemical criteria including hy- compared with case 1, the lesions are more numerous, their mantle of red cells is more
perglycemia, acidosis, and ketosis and incidences among prominent, the perilesional myelin pallor is confluent, and the microgliosis and astrocy-
tosis is more prominent in case 2. The degree of gliosis in case 2 is likely to be a conse-
newly diagnosed patients ranging from 15% to 67% in quence of a longer interval between the onset of symptoms and pathological examina-
developed countries.16–18 Compared with adults, DKA in tion of the tissues. Note that in case 2 that the cortex, including the cortical neurones, was
children carries higher morbidity and mortality largely normal except for mild microgliosis.
e1544 MAHMUD et al
Downloaded from pediatrics.aappublications.org at Leeszaal Wilhelmina on March 14, 2015
children. The pathologic findings in both cases were Emergency Medicine Collaborative Research Committee of the
similar with respect to the widely distributed small and American Academy of Pediatrics. N Engl J Med. 2001;344:
264 –269
microscopic hemorrhages. The hemorrhagic lesions dif-
4. Dillon ES, Riggs HE, Wallace Dyer W. Cerebral lesions in un-
fered between the cases in that, compared with case 1, complicated fatal diabetic acidosis. Am J Med Sci. 1936;192:
the lesions were more numerous, the mantle of red cells 360 –365
was more prominent, and the perilesional myelin pallor 5. Rosenbloom AL. Intracerebral crises during treatment of dia-
was confluent in case 2. The extent of hemorrhages betic ketoacidosis. Diabetes Care. 1990;13:22–33
6. Roe TF, Crawford TO, Huff KR, et al. Brain infarction in chil-
between patients may have related also to the timing of
dren with diabetic ketoacidosis. J Diabetes Complications. 1996;
presentation, level of acidosis, ethnicity, and/or differ- 10:100 –108
ences in BMI. Although both patients met criteria for 7. Muir AB, Quisling RG, Yang MC, et al. Cerebral edema in
DKA, our first patient had multiple risk factors for type 2 childhood diabetic ketoacidosis: natural history, radiographic
diabetes, including aboriginal ancestry and elevated findings, and early identification. Diabetes Care. 2004;27:
1541–1546
BMI. Indeed, type 2 diabetes was demonstrated recently
8. Rosenbloom AL. Hyperglycemic comas in children: new in-
to cause DKA in a significant number of presenting sights into pathophysiology and management. Rev Endocr Metab
adolescent cases.47 The first case presented here also Disord. 2005;6:297–306
shares similarities with the previously reported case of a 9. Rogers B, Sills I, Cohen M, et al. Diabetic ketoacidosis: neuro-
black 16-year-old boy with a BMI of 33.8 kg/m2, hypo- logic collapse during treatment followed by severe develop-
mental morbidity. Clin Pediatr (Phila). 1990;29:451– 456
tension, and rhabdomyolysis who died as a result of
10. Atkin SL, Coady AM, Horton D, Sutaria N, Sellars L, Walton C.
renal failure and was reported to present with hypergly- Multiple cerebral haematomata and peripheral nerve palsies
cemic hyperosmolar coma.47 Our patient had profound associated with a case of juvenile diabetic ketoacidosis. Diabet
acidosis at presentation, a lower initial blood glucose Med. 1995;12:267–270
level, and significant ketonemia, which are more consis- 11. Ho J, Mah JK, Hill MD, Pacaud D. Pediatric stroke associated
with new onset type 1 diabetes mellitus: case reports and
tent with DKA than hyperglycemic hyperosmolar coma,
review of the literature. Pediatr Diabetes. 2006;7:116 –121
although mixed presentations can occur.47,48 In the sec- 12. Atluru VL. Spontaneous intracerebral hematomas in juvenile
ond patient, we were careful to optimize fluids with diabetic ketoacidosis. Pediatr Neurol. 1986;2:167–169
inotropic support and initiate insulin as a slow infusion. 13. Roberts MD, Slover RH, Chase HP. Diabetic ketoacidosis with
In addition, the second patient had intracranial pressure intracerebral complications. Pediatr Diabetes. 2001;2:109 –114
14. Keane S, Gallagher A, Ackroyd S, McShane MA, Edge JA.
measurements to monitor the effects of the slow osmolar
Cerebral venous thrombosis during diabetic ketoacidosis. Arch
correction. Dis Child. 2002;86:204 –205
15. Ishida K, Shimizu H, Hida H, Urakawa S, Ida K, Nishino H.
CONCLUSIONS Argyrophilic dark neurons represent various states of neuronal
damage in brain insults: some come to die and others survive.
We have described 2 cases of diffuse white matter focal
Neuroscience. 2004;125:633– 644
hemorrhages that may have been caused by a diabetic 16. Lévy-Marchal C, Patterson CC, Green A, et al. Geographical
small vessel injury, possibly caused by circulating cyto- variation of presentation at diagnosis of type I diabetes in
toxic factors. Physicians who care for children should be children: the EURODIAB study. European and Diabetes. Dia-
aware that hyperglycemic coma in adolescents is not betologia. 2001;44(suppl 3):B75–B80
only associated with CE. These cases provide further 17. Lindblad B, Blom L, Hanas R, et al. Metabolic acidosis at the
onset of diabetes is frequent at all pediatric ages. J Pediatr
clinical and neuropathologic insight into the cerebral Endocrinol Metab. 2002;15:1081
dysfunction present in children with severe DKA. 18. Mel JM, Werther GA. Incidence and outcome of diabetic ce-
rebral oedema in childhood: are there predictors? J Paediatr
ACKNOWLEDGMENTS Child Health. 1995;31:17–20
19. Wolfsdorf J, Glaser N, Sperling MA; American Diabetes Asso-
Dr Fraser is supported by the Children’s Health Research ciation. Diabetic ketoacidosis in infants, children, and
Institute and the Centre for Critical Illness Research and adolescents: a consensus statement from the American Diabe-
is a strategic training fellow of the Canadian Institutes of tes Association. Diabetes Care. 2006;29:1150 –1159
Health Research in the Canadian Child Health Clinician 20. Carmassi F, Morale M, Pucetti R, et al. Coagulation and fibrino-
Scientist Program. lytic system impairment in insulin dependent diabetes melli-
tus. Thromb Res. 1992;67:643– 654.
21. Dalton RR, Hoffman WH, Passmore GG, Martin SL. Plasma
REFERENCES C-reactive protein levels in severe diabetic ketoacidosis. Ann
1. Lawrence SE, Cummings EA, Gaboury I, Daneman D. Popula- Clin Lab Sci. 2003;33:435– 442
tion-based study of incidence and risk factors for cerebral 22. Stentz FB, Umpierrez GE, Cuervo R, Kitabchi AE. Proinflam-
edema in pediatric diabetic ketoacidosis. J Pediatr. 2005;146: matory cytokines, markers of cardiovascular risks, oxidative
688 – 692 stress, and lipid peroxidation in patients with hyperglycemic
2. Edge JA, Hawkins MM, Winter DL, Dunger DB. The risk and crises. Diabetes. 2004;53:2079 –2086
outcome of cerebral oedema developing during diabetic keto- 23. Hoffman WH, Burek CL, Waller JL, Fisher LE, Khichi M,
acidosis. Arch Dis Child. 2001;85:16 –22 Mellick LB. Cytokine response to diabetic ketoacidosis and its
3. Glaser N, Barnett P, McCaslin I, et al. Risk factors for cerebral treatment. Clin Immunol. 2003;108:175–181
edema in children with diabetic ketoacidosis. The Pediatric 24. Stentz FB, Kitabchi AE. Hyperglycemia-induced activation of
e1546 MAHMUD et al
Downloaded from pediatrics.aappublications.org at Leeszaal Wilhelmina on March 14, 2015
Coma With Diffuse White Matter Hemorrhages in Juvenile Diabetic
Ketoacidosis
Farid H. Mahmud, David A. Ramsay, Simon D. Levin, Ram N. Singh, Trevor Kotylak
and Douglas D. Fraser
Pediatrics 2007;120;e1540; originally published online November 26, 2007;
DOI: 10.1542/peds.2007-0366
Updated Information & including high resolution figures, can be found at:
Services http://pediatrics.aappublications.org/content/120/6/e1540.full.
html
References This article cites 48 articles, 13 of which can be accessed free
at:
http://pediatrics.aappublications.org/content/120/6/e1540.full.
html#ref-list-1
Citations This article has been cited by 1 HighWire-hosted articles:
http://pediatrics.aappublications.org/content/120/6/e1540.full.
html#related-urls
Subspecialty Collections This article, along with others on similar topics, appears in
the following collection(s):
Endocrinology
http://pediatrics.aappublications.org/cgi/collection/endocrinol
ogy_sub
Diabetes Mellitus
http://pediatrics.aappublications.org/cgi/collection/diabetes_
mellitus_sub
Permissions & Licensing Information about reproducing this article in parts (figures,
tables) or in its entirety can be found online at:
http://pediatrics.aappublications.org/site/misc/Permissions.xht
ml
Reprints Information about ordering reprints can be found online:
http://pediatrics.aappublications.org/site/misc/reprints.xhtml
The online version of this article, along with updated information and services, is
located on the World Wide Web at:
http://pediatrics.aappublications.org/content/120/6/e1540.full.html