Askiel Bruno, Thomas A. Kent, Bruce M. Coull, Ravi R.... Becker, Brett M. Kissela and Linda S. Williams

Treatment of Hyperglycemia In Ischemic Stroke (THIS): A Randomized Pilot Trial
Askiel Bruno, Thomas A. Kent, Bruce M. Coull, Ravi R. Shankar, Chandan Saha, Kyra J.
Becker, Brett M. Kissela and Linda S. Williams
Stroke. 2008;39:384-389; originally published online December 20, 2007;
doi: 10.1161/STROKEAHA.107.493544
Stroke is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 2007 American Heart Association, Inc. All rights reserved.
Print ISSN: 0039-2499. Online ISSN: 1524-4628
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://stroke.ahajournals.org/content/39/2/384
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Stroke can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office.
Once the online version of the published article for which permission is being requested is located, click
Request Permissions in the middle column of the Web page under Services. Further information about this
process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Stroke is online at:
http://stroke.ahajournals.org//subscriptions/
Downloaded from http://stroke.ahajournals.org/ by guest on September 9, 2014
Treatment of Hyperglycemia In Ischemic Stroke (THIS)
A Randomized Pilot Trial
Askiel Bruno, MD; Thomas A. Kent, MD; Bruce M. Coull, MD; Ravi R. Shankar, MD;
Chandan Saha, PhD; Kyra J. Becker, MD; Brett M. Kissela, MD; Linda S. Williams, MD
Background and Purpose—Hyperglycemia may worsen brain injury during acute cerebral infarction. We tested the
feasibility and tolerability of aggressive hyperglycemia correction with intravenous insulin compared with usual care
during acute cerebral infarction.
Methods—We conducted a randomized, multicenter, blinded pilot trial for patients with cerebral infarction within 12 hours
after onset, a baseline glucose value ⱖ8.3 mmol/L (ⱖ150 mg/dL), and a National Institutes of Health Stroke Scale score
of 3 to 22. Patients were randomized 2:1 to aggressive treatment with continuous intravenous insulin or subcutaneous
insulin QID as needed (usual care). Target glucose levels were ⬍7.2 mmol/L (⬍130 mg/dL) in the aggressive-treatment
group and ⬍11.1 mmol/L (⬍200 mg/dL) in the usual-care group. Glucose was monitored every 1 to 2 hours, and the
protocol treatments continued for up to 72 hours. Final clinical outcomes were assessed at 3 months.
Results—We randomized 46 patients (31 to aggressive treatment and 15 to usual care). All patients in the aggressivetreatment group and 11 (73%) in the usual-care group had diabetes (P⫽0.008). Glucose levels were significantly lower
in the aggressive-treatment group throughout protocol treatment (7.4 vs 10.5 mmol/L [133 vs 190 mg/dL], P⬍0.001).
Hypoglycemia ⬍3.3 mmol/L (⬍60 mg/dL) occurred only in the aggressive-treatment group (11 patients, 35%), 4 (13%)
of whom had brief symptoms, including only 1 (3%) neurologic. Final clinical outcomes were nonsignificantly better
in the aggressive-treatment group.
Conclusions—The intravenous insulin protocol corrected hyperglycemia during acute cerebral infarction significantly
better than usual care without major adverse events and should be investigated in a clinical efficacy trial. (Stroke. 2008;
39:384-389.)
Key Words: brain infarction 䡲 diabetes mellitus 䡲 hyperglycemia 䡲 insulin
I
n animal studies, hyperglycemia during focal brain ischemia worsens outcomes, particularly in transient occlusion
with reperfusion models.1,2 In human retrospective and observational studies, hyperglycemia during acute cerebral infarction has been linked to worse outcomes as well.3–12 The
human studies were analyzed by controlling for multiple
factors, including diabetes mellitus and stroke severity, thus
suggesting that hyperglycemia worsens outcomes from acute
stroke independent of the other factors. These findings led to
the hypothesis that aggressive correction of hyperglycemia
during acute cerebral infarction will limit brain damage and
improve clinical outcomes. One recent efficacy trial of
aggressive hyperglycemia correction during acute stroke,
predominantly in patients without diabetes, has been reported
and showed no benefit.13 One additional pilot trial of aggressive hyperglycemia correction during acute stroke has been
reported,14 and another is in progress.15
Because aggressive correction of hyperglycemia with intravenous insulin during acute stroke involves substantial
effort and cost and some risk, it is essential to determine
whether such an intervention improves outcomes and to what
extent. Thus, to collect data needed to optimize the design of
a subsequent efficacy trial, we performed a pilot trial of
aggressive versus usual hyperglycemia correction, predominantly in patients with diabetes mellitus.
Patients and Methods
The Treatment of Hyperglycemia in Ischemic Stroke (THIS) study
was a National Institute of Neurological Disorders and Stroke–
sponsored randomized, blinded, multicenter trial. Patients were
enrolled between November 2002 and July 2006 at 5 US medical
centers (see Appendix). The primary objective was to collect data
about the safety, feasibility, and effectiveness of aggressive hyperglycemia correction with intravenous insulin during acute cerebral
infarction. The secondary aim was to measure clinical outcomes in
Received May 11, 2007; final revision received June 14, 2007; accepted June 27, 2007.
From the Department of Neurology (A.B.), Division of Pediatric Endocrinology (R.R.S.), and Division of Biostatistics (C.S.), Indiana University
School of Medicine, and the Roudebush Veterans Affairs Medical Center (L.S.W.), Indianapolis, Ind; the Department of Neurology (T.A.K.), Baylor
College of Medicine, Houston, Tex; the Department of Neurology (B.M.C.), University of Arizona School of Medicine, Tucson, Ariz; the Department
of Neurology (K.J.B.), University of Washington School of Medicine, Seattle, Wash; and the Department of Neurology (B.M.K.), University of Cincinnati
School of Medicine, Cincinnati, Ohio.
Correspondence to Askiel Bruno, MD, Department of Neurology, Indiana University School of Medicine, 1050 Wishard Blvd, 6th Floor, Indianapolis,
IN 46202. E-mail [email protected]
© 2008 American Heart Association, Inc.
Stroke is available at http://stroke.ahajournals.org
DOI: 10.1161/STROKEAHA.107.493544
384
Downloaded from http://stroke.ahajournals.org/
by guest on September 9, 2014
Bruno et al
this patient population to help optimize the design of an efficacy
trial.
Patients
Patients presenting with cerebral infarction were screened for enrollment. Neuroimaging was required before randomization to exclude
acute cerebral infarction mimics. Inclusion criteria were as follows:
onset of symptoms within 12 hours before randomization, baseline
blood glucose level ⱖ8.3 mmol/L (ⱖ150 mg/dL), and baseline
National Institutes of Health Stroke Scale (NIHSS) score of 3 to 22,
with at least 2 points on the motor portion. We hypothesized that
aggressive correction of hyperglycemia might be most effective
when started as soon as possible after the onset of cerebral ischemia,
and thus we enrolled patients and initiated protocol treatment as soon
as possible after stroke onset. We also hypothesized that this type of
intervention might be effective primarily when the brain cells were
still viable (cerebral penumbra) and before reperfusion occurred.3,8
Thus, we chose 12 hours as the enrollment limit because the cerebral
penumbra is often still present at 12 hours after the onset of ischemia,
based on magnetic resonance imaging and positron emission tomography criteria,16 –18 and spontaneous or tissue plasminogen activator
(tPA)–associated reperfusion in nonlacunar stroke is still occurring
in some patients.19 –21 We chose the hyperglycemia threshold for
enrollment as 8.3 mmol/L (150 mg/dL) to include predominantly
patients with diabetes mellitus, as their hyperglycemia during hospitalization is greater than in patients without diabetes,5,10,22,23 thus
enabling a greater reduction of glucose levels.
Exclusion criteria were preexisting incapacitating illness equivalent to a modified Rankin Scale score ⱖ3, indication for intravenous
insulin therapy (such as acute myocardial infarction or diabetic
ketoacidosis), or corticosteroid therapy. Hemoglobin A1C was measured in all patients on admission, and we defined diabetes mellitus
as either having it documented in the medical records or an
abnormally elevated hemoglobin A1C. Other vascular risk factors
were determined according to medical history and medical records.
Randomization and Blinding
All patients signed a valid, informed consent approved by institutional review boards at each center. To acquire a greater proportion
of data about our novel aggressive intervention, we randomized
patients to aggressive hyperglycemia correction or usual care in a 2:1
ratio, respectively. The acute intervention was single-blind (patients
and families), and the final outcome assessment was double-blind.
The Data Management Center prepared lists of random treatmentgroup allocations in the specified 2:1 ratio and distributed these lists
to the research pharmacists at each participating center. Unblinded
research pharmacists at each center randomized patients according to
the prepared lists and after randomization revealed the treatment
group to the treating physicians and nurses so that they could
administer the treatment protocols in an unblinded fashion. Before
randomization, the investigators were blinded to the next treatment
group. The patients and their families remained blinded throughout
the trial. The final clinical assessment at 3 months was done by
blinded investigators who had not participated in the acute care.
Patients in both groups received continuous intravenous infusions
from bags with concealed labels and with periodic rate adjustments,
as well as subcutaneous injections. Patients in the aggressivetreatment group received subcutaneous saline injections QID to
simulate subcutaneous insulin injections in the usual-care group.
Patients in the usual-care group received intravenous saline to
simulate the intravenous insulin infusions in the aggressivetreatment group.
Interventions
On enrollment, all premorbid antidiabetic medications were temporarily discontinued until the study protocols were completed. Patients
treated with tPA were admitted to intensive care units (medical,
surgical, or neurocritical) according to standard practice. Patients not
treated with tPA were admitted either to intensive care units or
intermediate care units (also known as step-down or progressive care
Hyperglycemia in Acute Ischemic Stroke Trial
385
units), depending on the clinical condition of the patient and the
preparedness of a specific unit to administer intravenous insulin. If
an appropriate unit bed was not immediately available, protocol
treatment began in the Emergency Department. The protocol treatments continued for 72 hours in both groups unless patients
improved rapidly and were ready for discharge or died before this
time point.
We designed a standardized usual-care protocol to resemble the
usual care given to acute stroke patients with hyperglycemia in our
communities at that time (2000). Usual care consisted of a subcutaneous regular insulin sliding scale administered QID as needed. No
insulin was given when the glucose value was ⬍11.1 mmol/L (⬍200
mg/dL). The insulin doses were 2 U for glucose values of 11.1 to
13.9 mmol/L (200 to 250 mg/dL), 3 U for 13.9 to 16.7 mmol/L (251
to 300 mg/dL), 4 U for 16.7 to 19.4 mmol/L (301 to 350 mg/dL), 6
U for 19.4 to 22.2 mmol/L (351 to 400 mg/dL), 7 U for 22.2 to
25.0 mmol/L (401 to 450 mg/dL), and 8 U for 25.0 to 27.7 mmol/L
(451 to 499 mg/dL). Capillary glucose was monitored every 2 hours
in the usual-care group. Because patients with diabetes mellitus and
acute stroke have heterogeneous insulin sensitivities, the investigators were allowed to alter the insulin doses based on the initial patient
responses to achieve glucose levels ⬍11.1 mmol/L (⬍200 mg/dL).
When patients with diabetes resumed eating, they received additional
subcutaneous regular insulin immediately after each meal at 0.12
U/kg, prorated for the portion of meal that was consumed. Investigators were also allowed to alter the meal insulin dose, based on
individual responses, to maintain glucose levels ⬍11.1 mmol/L
(⬍200 mg/dL).
Aggressive treatment consisted of continuous intravenous insulin
infusion with rate adjustments according to protocol (Table 1). This
protocol represents the current version after multiple modifications
during the preliminary study24 and the early phase of this pilot trial.
Capillary glucose was monitored hourly in the aggressive-treatment
group. When patients with diabetes resumed eating, they received
subcutaneous very rapidly acting insulin immediately after each
meal, 1 U of insulin for each 20 g of carbohydrate consumed.
Certified dieticians helped the nurses determine carbohydrate consumption. Investigators were allowed to alter the meal insulin dose
based on individual responses to maintain glucose levels of 5.0 to
7.2 mmol/L (90 to 130 mg/dL). Insulin infusions were temporarily
stopped when patients needed to leave their hospital units. Glucose
was not included in the insulin solution. Serum potassium was
measured daily, and potassium (20 mEq) was added to the insulin
solution only when the potassium level was ⬍3.5 mEq/L. For
symptomatic hypoglycemia (glucose ⬍3.3 mmol/L, or ⬍60 mg/dL,
with symptoms of hypoglycemia), the protocol called for 25 mL IV
of 50% dextrose, a glucose recheck in 20 minutes, and continuation
of the protocol.
Clinical monitoring during hospitalization included a daily NIHSS
assessment. After protocol treatment, management of hyperglycemia
and diabetes mellitus was decided by the attending physicians
according to the standard of care. All other nonglucose-related
treatments were decided by the attending physicians on the basis of
individual patient needs and the local standard of care.
Outcomes
An independent safety monitor (see Appendix) periodically reviewed
all of the data and adverse events during this trial. The prespecified
primary outcome was the mean glucose difference between the 2
groups during protocol treatment. All adverse events were monitored
and documented. The main safety concern was hypoglycemia,
defined as any glucose value ⬍3.3 mmol/L (⬍60 mg/dL) and any
associated symptoms during protocol treatment. Each episode of
hypoglycemia was documented with all associated symptoms, treatments, and duration on a designated form. For exploratory analysis,
favorable clinical outcomes were a modified Rankin Scale score ⱕ2,
a modified Barthel Index score of 19 to 20, an NIHSS score ⱕ2, and
the Stroke-Specific Quality of Life scale at 3 months. The StrokeSpecific Quality of Life scale consists of multiple domains and the
scores from each domain and the total range from 1.0 (worst) to 5.0
(best).25 This scale tends to have a normal distribution and can be
Downloaded from http://stroke.ahajournals.org/ by guest on September 9, 2014
386
Stroke
Table 1.
February 2008
IV Insulin in Acute Stroke Protocol
Start with 40 mL/h (25 U in 500 mL normal saline) and recheck glucose in 1 hour.
At midnight, decrease insulin rate to 20 mL/h (unless already lower) and then continue to follow protocol as indicated.
Glucose Level, mg/dL
Infusion Rate Adjustment
D10 Glucose Infusion
Repeat Glucose in
ⱖ160
Increase by 25% (previous rate⫻1.25⫽new rate;
minimum 40 mL/h); no increase if glucose dropped
ⱖ80 mg/dL per hour
None
1 hour
130–159
Increase by 10% (previous rate⫻1.1⫽new rate); no
increase if glucose dropped ⱖ80 mg/dL per hour
None
1 hour
90–129
No change
None
1 hour
80–89
Decrease by 50% (previous rate⫻0.5⫽new rate)
None
1 hour
STOP infusion and RESTART at 50% previous rate when
glucose ⱖ80 mg/dL (previous rate⫻0.5⫽new rate)
200 mL/h until glucose ⱖ80 mg/dL
20 minutes
⬍80
If symptomatic hypoglycemia (autonomic or neurologic symptoms), give 25 mL of 50% dextrose IV push STAT, do not give the D10 infusion, and
page the study physician.
If not eating, give normal saline (0.1 mL SC) at ⬇08:00, 13:00, 18:00, and 22:00.
If eating, give very rapidly acting insulin (Lispro, Aspart SC) right after each meal, 1 U/20 g of carbohydrate eaten.
When glucose is ⬍80 mg/dL, monitor for symptoms of hypoglycemia every 20 minutes and page study physician.
If infusion rate is ⬍10 mL/h, stop infusion and monitor every 2 hours until glucose ⬎130 mg/dL; restart infusion at the previous rate if glucose
131–159 mg/dL or at 40 mL/h if glucose ⱖ160 mg/dL.
analyzed as a continuous variable. One investigator (L.S.W.),
blinded to treatment group and outcomes, determined stroke subtype
in all patients according to Trial of ORG 10172 in Acute Stroke
Treatment criteria.26
Statistical Analysis
The Division of Biostatistics at the Indiana University School of
Medicine managed the data and performed statistical analyses. A
Wilcoxon rank-sum test compared the continuous variables, whereas
a ␹2 and Fisher’s exact tests compared the dichotomous variables
between the 2 treatment groups at baseline and at 3 months. In
addition, ANCOVA compared the glucose levels during protocol
treatment and the final clinical outcomes between the groups while
adjusting for age, baseline glucose level, baseline NIHSS score,
diabetes, tPA treatment, and stroke subtype (lacunar versus nonlacunar). All ANCOVA assumptions were satisfied.
teristics. Randomization was effective, except that all 4
patients without diabetes were randomized to the usual-care
group (1.2% chance, P⫽0.008). Table 3 shows the metabolic
outcomes. The aggressive-treatment protocol corrected hyperglycemia significantly better than did usual care
throughout the treatment period (P⬍0.001). The unadjusted
mean glucose level during protocol treatment was
10.5⫾3.6 mmol/L (190⫾64 mg/dL) with usual care and
7.4⫾0.9 mmol/L (133⫾16 mg/dL) with aggressive treatment.
After adjusting for potential confounding factors, the differTable 3.
Metabolic Outcomes in THIS Trial
Usual Care
(n⫽15)
Aggressive Treatment
(n⫽31)
190⫾64
133⫾16*
First 24 hours
185⫾70
132⫾21
25–48 hours
207⫾61
134⫾29
49–72 hours
194⫾65
132⫾22
3.7⫾0.6
3.6⫾0.4
⫺0.3⫾0.8
⫺0.6⫾0.3
All patients, n (%)
0
11 (35)*
10 (32)
Asymptomatic, n (%)
0
7 (23)
260⫾79
Autonomic symptoms, n (%)
0
4 (13)†
Neurologic signs, n (%)
0
1 (3)‡
Outcome
Results
We randomized 46 patients (31 to aggressive treatment and
15 to usual care). Table 2 shows the baseline patient characTable 2. Baseline Characteristics and Stroke Subtypes of
Patients in THIS Trial
Usual Care
(n⫽15)
Aggressive Treatment
(n⫽31)
Mean age, y (⫾SD)
53⫾15
62⫾15
Men, n (%)
9 (60)
Characteristic
17 (55)
Median NIHSS score (IQR)
10 (6–15)
Diabetes mellitus, n (%)*
11 (73)
31 (100)†
History of hypertension, n (%)
12 (80)
27 (87)
2 (13)
8 (26)
Treated with standard IV tPA, n (%)
Lacunar stroke subtype, n (%)
4 (27)
Mean blood glucose, mg/dL (⫾SD)
271⫾100
9 (5–15)
IQR indicates interquartile range.
*Diabetes mellitus defined as documented in the medical records or an
elevated hemoglobin A1C.
†P⫽0.008.
Mean glucose during protocol
treatment, mg/dL (⫾SD)
Lowest serum potassium
during protocol treatment,
mEq/L (mean⫾SD)
Change in serum potassium
(lowest⫺baseline) during
protocol treatment, mEq/L
(mean⫾SD)
Hypoglycemia (any glucose
⬍60 mg/dL)
*P⬍0.001.
†Transient mild diaphoresis, tremulousness, or both.
‡Cognitive slowing for 10 minutes.
Downloaded from http://stroke.ahajournals.org/ by guest on September 9, 2014
Mean Glucose (mg/dL)
Bruno et al
250
230
210
190
170
150
130
110
90
70
Hyperglycemia in Acute Ischemic Stroke Trial
387
Table 4.
Clinical Outcomes in THIS Pilot Trial at 3 Months
Outcome
Usual Care
(n⫽15)
Aggressive Treatment
(n⫽31)
Modified rankin scale score
ⱕ2, n (%)
7 (47)
16 (52)
Modified barthel index
19–20, n (%)
7 (47)
15 (48)
NIHSS score ⱕ2, n (%)
3/13 (23)
11/29 (38)
SSQOL score
3.63⫾0.95
3.57⫾1.08
Figure. Mean glucose levels during the first 24 hours of protocol treatment before patients resumed eating. E, Usualcare group (n⫽15); 䡺, aggressive-treatment group (n⫽31);
interrupted line, the 4 patients without diabetes randomized
to usual care.
Death, n (%)
0
ence in mean glucose levels between the 2 groups during
protocol treatment was 3.7 mmol/L (66 mg/dL; 9.7 vs
6.0 mmol/L, or 174 vs 108 mg/dL; P⬍0.001). The mean daily
glucose levels in the 4 patients without diabetes in the
usual-care group were similar to those in the aggressivetreatment group: 7.3⫾1.3, 7.5⫾1.3, and 7.0⫾1.4 mmol/L, or
132⫾24, 135⫾24, and 127⫾25 mg/dL during the initial 24
hours, 25 to 48 hours, and 49 to 72 hours, respectively (Table
3). These 4 patients did not receive any insulin according to
protocol because their glucose values did not exceed
11.1 mmol/L (200 mg/dL).
The Figure shows the glucose levels during the initial 24
hours of protocol treatment before patients resumed eating.
During the initial 4 hours, glucose levels dropped somewhat
faster and below 7.2 mmol/L (130 mg/dL) in the aggressivetreatment group. The 4 patients without diabetes had considerably lower glucose levels than did the entire usual-care
group, although they remained somewhat elevated, at 7.2 to
8.9 mmol/L (130 to 160 mg/dL) during the initial 15 hours of
treatment. After the initial 24 hours, glucose levels increased
transiently after meals in both groups, as seen in our preliminary study.24
Hypoglycemia (⬍3.3 mmol/L, or ⬍60 mg/dL) occurred
only in the aggressive-treatment group in 11 patients (35%,
Table 3), and there were a total of 12 episodes. The
hypoglycemia events were asymptomatic in a majority of
patients (7 of 11, 64%). All symptoms of hypoglycemia
resolved completely within 20 minutes. One patient (3%) had
a mild neurologic symptom (cognitive slowing) for 10 minutes associated with a glucose value of 2.8 mmol/L (50
mg/dL) and received 25 mL of dextrose 50% according to
protocol. Three patients with symptomatic hypoglycemia did
not receive dextrose 50% because they had brief autonomic
symptoms only. Two patients (6.5%) had a glucose level
⬍2.8 mmol/L (⬍50 mg/dL). The lowest glucose level was
2.3 mmol/L (42 mg/dL) and was asymptomatic. There were
no seizures. The rate of hypoglycemia decreased throughout
this trial from 44% in the initial 16 patients to 27% in the
subsequent 15 patients, likely due to 2 protocol modifications
during the early phase of this trial. These modifications were
not increasing the insulin rate when the glucose was dropping
faster than 4.4 mmol/L per hour (79 mg/dL per hour) and
lowering the insulin infusion rate to 1 U/h at midnight
regardless of glucose level. While the rate of hypoglycemia
decreased, the mean glucose level did not increase
(7.4 mmol/L, or 133 mg/dL, in the initial 16 patients and
7.3 mmol/L, or 131 mg/dL, in the subsequent 15 patients).
Functional outcomes at 3 months were somewhat better in
the aggressive-treatment group (Table 4), but none were
statistically significant either before (all P values ⱖ0.35) or
after (all P values ⱖ0.09) adjusting for potential confounding
factors. There was no indication that the hypoglycemic
episodes worsened clinical outcomes. No patients were lost to
follow-up, but 4 patients (2 in each group) could not be
examined in person and thus have missing final NIHSS
scores.
0
2
4
6
8
10
12
14
16
18
20
22
24
Hours after starting protocol treatment
2 (7)
SSQOL indicates Stroke Specific Quality-of-Life Scale. None of these
comparisons were statistically significant (all P values ⱖ0.35 for unadjusted
and ⱖ0.09 for adjusted comparisons).
Discussion
This pilot trial has demonstrated the effectiveness and tolerability of aggressive hyperglycemia correction with our
intravenous insulin protocol during acute cerebral infarction
in patients with diabetes mellitus. In the aggressive-treatment
group, the mean glucose value dropped from 14.4 mmol/L
(260 mg/dL) at baseline to 6.7 mmol/L (121 mg/dL) after 4
hours of protocol treatment. Throughout protocol treatment,
glucose levels were considerably lower in the aggressivetreatment group than in the usual-care group (Table 3, the
Figure). The mean adjusted glucose difference of 3.7 mmol/L
(66 mg/dL) between groups in this trial is greater than in any
of the previous intervention trials of aggressive hyperglycemia correction.13,27–30 A more rapid or greater hyperglycemia
correction in this patient population would likely be both
more labor-intensive and more risky.
It is important to consider the 4 patients without diabetes
who were randomized to usual care. Although they had
persistent mild hyperglycemia, their glucose levels were
considerably lower than in patients with diabetes at the start
of treatment (the Figure) and resembled the levels in the
aggressive-treatment group throughout the remainder of protocol treatment (7.0 to 7.5 mmol/L, or 127 to 135 mg/dL;
Table 3, Results section). These subjects likely had transient
reactive (stress) hyperglycemia and perhaps impaired glucose
tolerance or mild diabetes mellitus.31,32 Relatively mild and
persistent hyperglycemia in acute stroke patients without
known diabetes mellitus has been reported,22,23 but glucose
metabolism in such patients has rarely been studied in
detail.31,32 Nonetheless, it seems that the opportunity to
correct hyperglycemia with our aggressive protocol during
Downloaded from http://stroke.ahajournals.org/ by guest on September 9, 2014
388
Stroke
February 2008
acute cerebral infarction in patients without diabetes is
limited.
It is unclear at this time whether aggressive correction of
hyperglycemia during acute cerebral infarction might be
beneficial for all patients with hyperglycemia or perhaps only
for those with or without diabetes mellitus.10,33 However, an
efficacy trial of hyperglycemia correction with intravenous
insulin during acute stroke predominantly in patients without
diabetes showed no clinical benefit.13 In that trial, only 17%
of patients had diabetes mellitus and 12% had intracerebral
hemorrhage. The mean glucose level in the saline (control)
group was ⬇6.8 mmol/L (123 mg/dL) between 8 and 24
hours of treatment, and the mean glucose difference between
the 2 groups was only 0.57 mmol/L (10.3 mg/dL). We suspect
that greater reductions in glucose levels might be needed to
show a clinical benefit. However, a greater reduction in
glucose levels in patients without diabetes is likely more
challenging and risky. Although not comparable to stroke,
greater reductions in glucose levels during acute myocardial infarction improved clinical outcomes in the first
DIGAMI trial (2.2 mmol/L, or 38 mg/dL),27 but smaller
reductions in the second DIGAMI trial (0.9 mmol/L, or 16
mg/dL)29 did not.
Although glucose levels ⬍3.3 mmol/L (⬍60 mg/dL) occurred in a relatively high proportion of patients (35%) in this
trial, there were no sequelae. Our definition of hypoglycemia
was rather liberal owing to the preliminary nature of this trial.
The rate of hypoglycemia decreased in the second half of this
trial, likely the result of protocol modifications, and without
worsening glycemic control. This trial did not include glucose
or potassium in the insulin solution, whereas other trials of
aggressive hyperglycemia correction did, and it is unclear
how this approach compares between trials. It is possible that
adding glucose to the insulin solution might have reduced the
rate of hypoglycemia in this trial, although it might have also
reduced the magnitude of hyperglycemia correction. Based
on the similar potassium levels in the 2 groups, routine
addition of potassium does not seem necessary.
This pilot trial was not designed to test clinical efficacy,
and the somewhat better clinical outcomes in the aggressivetreatment group are not statistically significant (Table 4). This
patient sample is small and heterogeneous. For example,
tPA-treated patients may be more vulnerable to hyperglycemia because of their increased rate of reperfusion,8 and thus
may benefit most from tight glucose control. Conversely,
patients with lacunar infarcts caused by occlusion of small,
penetrating cerebral arterioles may benefit least from tight
glucose control because of their lack of reperfusion.3 This
trial is too small to adequately address these hypotheses.
Aggressive intravenous insulin protocols like ours are
relatively labor-intensive in an effort to maximize hyperglycemia correction while avoiding hypoglycemia. Such protocols need to be administered in care units prepared to use
intravenous insulin effectively and safely, which carries
added cost. Therefore, we believe that it is important to test
the safety and efficacy of the aggressive intervention in a
large clinical trial. In addition, computerized insulin infusion
algorithms and continuous glucose monitoring systems are
available and may offer some advantages to our protocol.
However, their widespread feasibility, safety, and cost remain
to be established in acute stroke.
Appendix
Study Personnel
Indiana University, Indianapolis, Ind: principal investigator Askiel
Bruno, MD; coordinators Alison Sears, RN, and Kelley Faber, MS;
principal statistician Chandan Saha, PhD; coinvestigators Linda S.
Williams, MD, William J. Jones, MD, and James D. Fleck, MD;
endocrinologist Ravi R. Shankar, MD.
University of Cincinnati, Cincinnati, Ohio: Brett Kissela, MD;
coordinators Kathleen Alwell and Joyce Ziegler; endocrinologist
Barbara Ramlo-Halstead.
Baylor College of Medicine, Houston, Tex: Thomas A. Kent, MD,
and Pitchaiah Mandava, MD; endocrinologist Glenn Cunningham,
MD; coordinator Jane Anderson, RN, APN.
University of Washington, Seattle, Wash: Kyra J. Becker, MD;
coordinator Michael S. Fruin, MN, RN, ARNP-CS.
University of Arizona, Tucson, Ariz: Bruce M. Coull, MD, and
Jeremy R Payne, MD, PhD; coordinator Denise Bruck.
Independent Safety Monitor
David Sherman, MD, University of Texas, San Antonio.
Source of Funding
This trial was funded by a National Institute of Neurological
Disorders and Stroke Pilot Clinical Trial grant R01 042078 (to A.B.).
Disclosures
None.
References
1. Quast MJ, Wei J, Huang NC, Brunder DG, Sell SL, Gonzalez JM,
Hillman GR, Kent TA. Perfusion deficit parallels exacerbation of cerebral
ischemia/reperfusion injury in hyperglycemic rats. J Cereb Blood Flow
Metab. 1997;17:553–559.
2. Kent TA, Soukup VM, Fabian RH. Heterogeneity affecting outcome from
acute stroke therapy: making reperfusion worse. Stroke. 2001;32:
2318 –2327.
3. Bruno A, Biller J, Adams HP Jr, Clarke WR, Woolson RF, Williams LS,
Hansen MD. Acute blood glucose level and outcome from ischemic
stroke: Trial of ORG 10172 in Acute Stroke Treatment (TOAST) Investigators. Neurology. 1999;52:280 –284.
4. Bruno A, Levine SR, Frankel MR, Brott TG, Lin Y, Tilley BC, Lyden
PD, Broderick JP, Kwiatkowski TG, Fineberg SE. Admission glucose
level and clinical outcomes in the NINDS rt-PA Stroke Trial. Neurology.
2002;59:669 – 674.
5. Williams LS, Rotich J, Qi R, Fineberg N, Espay A, Bruno A, Fineberg
SE, Tierney WR. Effects of admission hyperglycemia on mortality and
costs in acute ischemic stroke. Neurology. 2002;59:67–71.
6. Parsons MW, Barber PA, Desmond PM, Baird TA, Darby DG, Byrnes G,
Tress BM, Davis SM. Acute hyperglycemia adversely affects stroke
outcome: a magnetic resonance imaging and spectroscopy study. Ann
Neurol. 2002;52:20 –28.
7. Baird TA, Parsons MW, Phanh T, Butcher KS, Desmond PM, Tress BM,
Colman PG, Chambers BR, Davis SM. Persistent poststroke hyperglycemia is independently associated with infarct expansion and worse
clinical outcome. Stroke. 2003;34:2208 –2214.
8. Alvarez-Sabin J, Molina CA, Ribo M, Arenillas JF, Montaner J, Huertas
R, Santamarina E, Rubiera M. Impact of admission hyperglycemia on
stroke outcome after thrombolysis: risk stratification in relation to time to
reperfusion. Stroke. 2004;35:2493–2498.
9. Leigh R, Zaidat OO, Suri MF, Lynch G, Sundararajan S, Sunshine JL,
Tarr R, Selman W, Landis DM, Suarez JI. Predictors of hyperacute
clinical worsening in ischemic stroke patients receiving thrombolytic
therapy. Stroke. 2004;35:1903–1907.
10. Farrokhnia N, Bjork E, Lindback J, Terent A. Blood glucose in acute
stroke, different therapeutic targets for diabetic and non-diabetic patients?
Acta Neurol Scand. 2005;112:81– 87.
Downloaded from http://stroke.ahajournals.org/ by guest on September 9, 2014
Bruno et al
11. Gentile NT, Seftchick MW, Huynh T, Kruus LK, Gaughan J. Decreased
mortality by normalizing blood glucose after acute ischemic stroke. Acad
Emerg Med. 2006;13:174 –180.
12. Martini SR, Hill MD, Alexandrov AV, Molina CA, Kent TA. Outcome in
hyperglycemic stroke with ultrasound-augmented thrombolytic therapy.
Neurology. 2006;67:700 –702.
13. Gray CS, Hildreth AJ, Sandercock PA, O’Connell JE, Johnston DE,
Cartlidge NE, Bamford JM, James OF, Alberti KG. Glucose-potassiuminsulin infusions in the management of post-stroke hyperglycaemia: the
UK Glucose Insulin in Stroke Trial (GIST-UK). Lancet Neurol. 2007;6:
397– 406.
14. Walters MR, Weir CJ, Lees KR. A randomised, controlled pilot study to
investigate the potential benefit of intervention with insulin in hyperglycaemic acute ischaemic stroke patients. Cerebrovasc Dis. 2006;22:
116 –122.
15. https://grasptrial.org/grasp/home.aspx. Accessed 12/10/07.
16. Marchal G, Beaudouin V, Rioux P, de la Sayette V, Le Doze F, Viader
F, Derlon JM, Baron JC. Prolonged persistence of substantial volumes of
potentially viable brain tissue after stroke: a correlative PET-CT study
with voxel-based data analysis. Stroke. 1996;27:599 – 606.
17. Read SJ, Hirano T, Abbott DF, Sachinidis JI, Tochon-Danguy HJ, Chan
JG, Egan GF, Scott AM, Bladin CF, McKay WJ, Donnan GA. Identifying
hypoxic tissue after acute ischemic stroke using PET and 18 Ffluoromisonidazole. Neurology. 1998;51:1617–1621.
18. Darby DG, Barber PA, Gerraty RP, Desmond PM, Yang Q, Parsons M,
Li T, Tress BM, Davis SM. Pathophysiological topography of acute
ischemia by combined diffusion-weighted and perfusion MRI. Stroke.
1999;30:2043–2052.
19. Molina CA, Montaner J, Abilleira S, Ibarra B, Romero F, Arenillas JF,
Alvarez-Sabin J. Timing of spontaneous recanalization and risk of hemorrhagic transformation in acute cardioembolic stroke. Stroke. 2001;32:
1079 –1084.
20. Murphy BD, Fox AJ, Lee DH, Sahlas DJ, Black SE, Hogan MJ, Coutts
SB, Demchuk AM, Goyal M, Aviv RI, Symons S, Gulka IB, Beletsky V,
Pelz D, Hachinski V, Chan R, Lee TY. Identification of penumbra and
infarct in acute ischemic stroke using computed tomography perfusionderived blood flow and blood volume measurements. Stroke. 2006;37:
1771–1777.
21. Wunderlich MT, Goertler M, Postert T, Schmitt E, Seidel G, Gahn G,
Samii C, Stolz E. Recanalization after intravenous thrombolysis: does a
recanalization time window exist? Neurology. 2007;68:1364 –1368.
Hyperglycemia in Acute Ischemic Stroke Trial
389
22. Dora B, Mihci E, Eser A, Ozdemir C, Cakir M, Balci MK, Balkan S.
Prolonged hyperglycemia in the early subacute period after cerebral
infarction: effects on short term prognosis. Acta Neurol Belg. 2004;104:
64 – 67.
23. Allport L, Baird T, Butcher K, Macgregor L, Prosser J, Colman P, Davis
S. Frequency and temporal profile of poststroke hyperglycemia using
continuous glucose monitoring. Diabetes Care. 2006;29:1839 –1844.
24. Bruno A, Saha C, Williams LS, Shankar R. IV insulin during acute
cerebral infarction in diabetic patients. Neurology. 2004;62:1441–1442.
25. Williams LS, Weinberger M, Harris LE, Clark DO, Biller J. Development
of a stroke-specific quality of life scale. Stroke. 1999;30:1362–1369.
26. Adams HP Jr, Bendixen BH, Kappelle LJ, Biller J, Love BB, Gordon DL,
Marsh EE 3rd. Classification of subtype of acute ischemic stroke: definitions for use in a multicenter clinical trial. TOAST: Trial of Org 10172
in Acute Stroke Treatment. Stroke. 1993;24:35– 41.
27. Malmberg K, Ryden L, Efendic S, Herlitz J, Nicol P, Waldenstrom A,
Wedel H, Welin L, on behalf of the DIGAMI Study Group: Randomized
trial of insulin-glucose infusion followed by subcutaneous insulin
treatment in diabetic patients with acute myocardial infarction (DIGAMI
study): effects on mortality at 1 year. J Am Coll Cardiol. 1995;26:57– 65.
28. van den Berghe G, Wouters P, Weekers F, Verwaest C, Bruyninckx F,
Schetz M, Vlasselaers D, Ferdinande P, Lauwers P, Bouillon R. Intensive
insulin therapy in the critically ill patient. N Engl J Med. 2001;345:
1359 –1367.
29. Malmberg K, Ryden L, Wedel H, Birkeland K, Bootsma A, Dickstein K,
Efendic S, Fisher M, Hamsten A, Herlitz J, Hildebrandt P, MacLeod K,
Laakso M, Torp-Pedersen C, Waldenstrom A: Intense metabolic control
by means of insulin in patients with diabetes mellitus and acute myocardial infarction (DIGAMI 2): effects on mortality and morbidity. Eur
Heart J. 2005;26:650 – 661.
30. Van den Berghe G, Wilmer A, Hermans G, Meersseman W, Wouters PJ,
Milants I, Van Wijngaerden E, Bobbaers H, Bouillon R. Intensive insulin
therapy in the medical ICU. N Engl J Med. 2006;354:449 – 461.
31. Matz K, Keresztes K, Tatschl C, Nowotny M, Dachenhausenm A, Brainin
M, Tuomilehto J. Disorders of glucose metabolism in acute stroke
patients: an underrecognized problem. Diabetes Care. 2006;29:792–797.
32. Vancheri F, Curcio M, Burgio A, Salvaggio S, Gruttadauria G, Lunetta
MC, Dovico R, Alletto M. Impaired glucose metabolism in patients with
acute stroke and no previous diagnosis of diabetes mellitus. Q J Med.
2005;98:871– 878.
33. Capes SE, Hunt D, Malmberg K, Pathak P, Gerstein HC. Stress hyperglycemia and prognosis of stroke in nondiabetic and diabetic patients: a
systematic overview. Stroke. 2001;32:2426 –2432.
Downloaded from http://stroke.ahajournals.org/ by guest on September 9, 2014