Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Authors
See Wai Chan, MD, MPH
Ann R Stark, MD Section Editors
Steven A Abrams, MD
Joseph I Wolfsdorf, MB, BCh Deputy Editor
Melanie S Kim, MD
Last literature review version 17.3: September 2009 | This topic last updated: February
2, 2009 (More)
PARENTERAL GLUCOSE — Most infants who are preterm or ill require parenteral
administration of glucose because adequate enteral feeding is delayed. Neonatal
hyperglycemia often occurs in this setting.
The net effect is that glucosuria alone is not a good marker for hyperglycemia since it
can occur at normal blood glucose concentrations. In one study of sick preterm infants
25 to 33 weeks gestation, for example, fractional glucose excretion varied widely and
glucosuria was often seen at normal blood glucose concentrations [2] . These variations
presumably are related to immaturity of the proximal tubule.
On the other hand, mild hyperglycemia may be associated with little or no glucosuria in
infants with mature proximal tubules. This was illustrated in a study of newborns who
were given glucose infusions; at a mean blood glucose concentration of 197 mg/dL (11
mmol/L), there was little glucosuria and no significant osmotic diuresis [3] .
What remains unclear is why the same degree of glucose infusion produces variable
elevations in blood glucose in newborns. A possible contributing factor is increased
secretion of counterregulatory hormones associated with stress (particularly epinephrine
and cortisol). (See "Physiologic response to hypoglycemia in normal subjects and
patients with diabetes mellitus", section on Response to hypoglycemia in normal
subjects).
Extremely low birth weight — Hyperglycemia during glucose infusion is most likely
to occur in ELBW infants with very high fluid requirements, often >200 mL/kg per day.
As noted above, some of these infants have an impaired insulin response to glucose [5] .
However, ELBW infants frequently develop hyperglycemia in the absence of high rates
of glucose infusion [9] . The mechanisms include reduced insulin secretion, incomplete
suppression of hepatic glucose production, and a stress response.
These studies demonstrate the lack of consensus in defining both hyperglycemia and the
outcome measures to demonstrate its effect in ELBW infants [12] . Without a better
understanding of the consequences of hyperglycemia and which infants require therapy,
it is difficult to assess the long-term effect of treatment of hyperglycemia. (See
"Definition" above and see "Insulin therapy" below).
In one report of 93 ELBW infants cared for at a single tertiary unit, persistent
hyperglycemia during the first week of life, defined as plasma glucose >150 mg/dL (8.3
mmol/L), was associated with an increased risk of early death and severe (grade III and
IV) intraventricular hemorrhage by 10 days of life [13] .
Stress — The stress response to critical illness (epinephrine and cortisol) may result in
hyperglycemia, especially in preterm infants who require mechanical ventilation. The
stress response also may be responsible for hyperglycemia occurring after surgery. In
this setting, increased rates of fluid administration may also contribute.
Most patients with permanent neonatal diabetes mellitus have mutations that affect the
ATP-sensitive potassium channel (KATP channel), which regulates the release of insulin
from pancreatic beta cells. Activating mutations increase the number of open KATP
channels at the plasma membrane, hyperpolarizing the beta cells, and preventing the
release of insulin.
The KATP channel is composed of a small subunit Kir6.2 that surrounds a central pore
and four regulatory SUR1 subunits. Activating gene mutations that affect these subunits
can prevent insulin release, resulting in hyperglycemia. KCNJ11 gene encoding Kir6.2
— The most common cause of permanent neonatal diabetes is due to activating
mutations in the KCNJ11 gene, which encodes Kir6.2 [26-28] . The diagnosis is made
within the first two months of life [26] . Infants are small for gestational age but exhibit
postnatal catch-up growth with insulin therapy [29] . Affected patients can also have
neurologic abnormalities including severe developmental delay, epilepsy, muscle
weakness, and dysmorphic features [26] . These findings are also known as the DEND
syndrome (developmental delay, epilepsy, neonatal diabetes) [30] .
Subcutaneous insulin was routinely used in the past to treat patients with this disorder.
However, oral sulfonylurea therapy appears to be more effective in controlling
hyperglycemia [31,32] . In a study of 49 patients with neonatal diabetes due to
activating mutations of KCNJ11 gene, 44 were able to discontinue insulin therapy after
starting oral sulfonylurea therapy [31] . In patients switched to sulfonylurea therapy,
insulin secretion and glycated hemoglobin (8.1 to 6.4 percent) improved. SUR1 —
Activating mutations of the ABCC8 gene, which encodes SUR1 (the type 1 subunit of
the sulfonylurea receptor), can cause both transient and permanent forms of neonatal
diabetes. In a series of 73 patients with neonatal diabetes, nine had activating mutations
of the ABCC8 gene [33] . Two had permanent diabetes and the others had transient
diabetes. The patients were diagnosed at a median of 32 days (range 3 to 125 days).
Oral sulfonylurea therapy normalized glycemic control in patients with genetic
mutations of SUR1.
Neonatal diabetes mellitus has also been associated with mutations in other genes
including IPF-1, EIF2AK3, GCK, FOXP3, PTF1A, GLIS3, and the Ins2 genes [34-49] .
In some cases, these mutations result in pancreatic agenesis or hypoplasia, or absent
beta cells. As an example, patients with permanent neonatal diabetes mellitus due to
Wolcott-Rallison syndrome (diabetes mellitus, exocrine pancreatic insufficiency, and
multiple epiphyseal dysplasia) have been shown to have hypoplastic pancreatic islets
and a mutation in the EIF2AK3 gene that encodes translation initiation factor 2-alpha
kinase 3 [38,39] . FOXP3 mutations cause a rare, x-linked disorder that presents in
infancy with autoimmune endocrinopathy, enteropathy, and eczema. (See "IPEX:
Immune dysregulation, polyendocrinopathy, enteropathy, X-linked").
However, reducing the glucose infusion rate is a short-term solution because it results in
decreased caloric intake and compromises growth. Glucose tolerance typically improves
when enteral feedings are established. (See "Nutritional composition of human milk and
preterm formula for the premature infant").
The efficacy of insulin therapy was illustrated by a study of 24 ELBW infants with
glucose intolerance who were randomly assigned to receive standard intravenous
therapy with glucose and total parenteral nutrition with or without a continuous insulin
infusion [51] . Over a mean duration of 15 days, insulin therapy resulted in significantly
higher glucose infusion rates (20.1 versus 13.2 mg/kg per min), greater nonprotein
energy intake (124 versus 86 kcal/kg per day), and greater weight gain (20.1 versus 7.8
g/kg per day) compared to control. There were no differences between groups in
hypoglycemia, electrolyte abnormalities, chronic lung disease, or mortality.
Benefit from insulin therapy was also demonstrated in a trial of 23 ELBW infants who
became hyperglycemic while receiving glucose at a rate up to 12 mg/kg per min [5] .
The infants were randomly assigned to reduced glucose intake or to insulin infusion.
The duration of caloric intake less than 60 kcal/kg per day was significantly shorter with
the insulin infusion (4.1 versus 8.6 days). There were no differences in morbidity or
mortality between the two groups.
This was illustrated in a multicenter, open-label trial of 389 very low birth weight
(VLBW) infants (birth weights <1500 g) who were randomly selected to receive either
standard care for glycemic control or a parenteral infusion of 20 percent dextrose with
early insulin therapy (0.05 U/Kg per hour) starting within 24 hours of birth until seven
days of age [53] . The following findings were noted: The early insulin group had lower
mean glucose levels compared to the standard care group (112 vs 121 mg/dL [6.2 vs 6.7
mmol/L]), were less likely to be hyperglycemic (defined as serum glucose greater than
180 mg/dL [10 mmol/L]) for more than 10 percent of the first week of life (21 versus 33
percent), were able to receive greater amounts of glucose infusion (51 versus 43 kcal/kg
per day), and had less weight loss during the first week of life. More patients who
received early insulin had episodes of hypoglycemia (29 versus 17 percent), which was
defined as serum glucose levels less than 47 mg/dL (2.6 mmol/L) for more than one
hour. There were no differences between the groups in the primary end point of
mortality at the expected date of delivery or in the secondary end points of sepsis,
necrotizing enterocolitis, retinopathy of prematurity, and growth parameters (ie, weight,
length, and head circumference) at 28 days of age. However, the early insulin group had
a higher mortality rate at 28 days of life.
This trial was ended early because of concerns of futility with regard to outcomes and
concern for potential harm from insulin therapy. Follow-up is ongoing to determine
whether the increased incidence of hypoglycemia in the early insulin group had a
detrimental effect on neurodevelopmental outcomes. (See "Neonatal hypoglycemia",
section on Prognosis).
Based upon these data, routine insulin therapy should NOT be used in VLBW infants.
Insulin should, however, be used to treat hyperglycemia when reducing the glucose
infusion rate to approximately 6 mg/kg per minute is ineffective or not possible.
Amino acid infusion — Insulin infusion during euglycemia reduces proteolysis and
protein synthesis in preterm infants who are not also given amino acids. In a report of
four ELBW infants at two to five days of age, whole body proteolysis and protein
synthesis decreased by 20 percent during a continuous infusion of insulin (0.05 units/kg
per hour) and glucose (without amino acids) [54] . Glucose utilization doubled (8 to
16.7 mg/kg per min) but there was no net anabolic effect. In addition, plasma lactate
concentration tripled (2.1 to 5.7 mmol/L), possibly because the high rate of glucose
infusion exceeded the maximal capacity for glucose oxidation. Whether administration
of amino acids during insulin infusion would further reduce proteolysis or increase
protein synthesis and improve protein balance is uncertain.
Based upon the limited current data, we suggest amino acid solution and lipid emulsion
also should be administered to infants receiving glucose infusion to provide substrate
for gluconeogenesis, spare glucose utilization, and stimulate insulin release.
Risk of hypoglycemia — The blood glucose concentration should be monitored
frequently during insulin infusion, although the risk of hypoglycemia appears to be
small. This was documented in a retrospective review of 34 ELBW infants who
developed hyperglycemia and glucosuria while receiving parenteral nutrition and were
treated with insulin [55] . Before therapy, mean blood glucose concentration was 195
mg/dL (11.1 mmol/L) while receiving glucose at a mean rate of 7.9 mg/kg per min.
During insulin infusion, given for 1 to 58 days, blood glucose values of 25 to 40 mg/dL
(1.4 to 2.2 mmol/L) were detected in fewer than 0.5 percent of samples (26 episodes of
hypoglycemia in 7368 samples) and no values <25 mg/dL were seen.
Similar findings were noted in another study of 10 ELBW infants treated with insulin
[56] . Glucose measurements were normal (46 to 130 mg/dL [2.6 to 7.2 mmol/L]) in 78
percent of samples and less than 24 mg/dL (1.3 mmol/L) in less than 1 percent [56] .
(See "Neonatal hypoglycemia").
An alternative approach is to use an initial insulin dose of 0.05 to 0.1 units per gram of
glucose. This dose can be increased in increments of 0.05 to 0.1 units per gram of
glucose until the target blood glucose concentration is achieved.
As glucose tolerance improves, the insulin infusion should be tapered and discontinued
to avoid hypoglycemia. In general, reductions in the insulin infusion rate can be made
more rapidly than can increases.
Adherence of insulin to plastic tubing — Plastic tubing used for infusion should be
primed with insulin for at least 20 minutes before treatment because insulin
nonspecifically binds to the tubing, resulting in decreased availability to the patient. In
one report, recovery of insulin from effluent of primed polyvinyl chloride tubing at a
flow rate of 0.2 mL/h was greater at one, two, four, and eight hours (42, 85, 91, and 95
versus 22, 38, 67, and 75 percent, respectively) compared to unprimed tubing [57] .
© 2009 UpToDate, Inc. All rights reserved. | Subscription and License Agreement
Licensed to: DR GONZALEZ