Mechanisms of Adverse Effects of
Hyperglycemia in Acute Illness
A Review of the Basic Science
1
Potential Mechanisms of Increased
Mortality/Morbidity in Acutely Ill Patients
•
•
•
•
•
•
•
•
Hyperactive inflammatory response
Increased oxidative stress
Impaired endothelial function
Increased coagulation response
Increased free fatty acids
Poor tissue perfusion
Autonomic dysfunction
Increased risk of infections or poor response
to anti-infective agents
2
Link Between High Blood Glucose and
Poor Outcomes: Potential Mechanisms
Metabolic stress response
↑ Stress hormones and peptides
Glucose
Insulin
Immune dysfunction
Infection dissemination
FFA
Ketones
Lactate
Cellular injury/apoptosis
Inflammation
Tissue damage
Altered tissue/wound repair
Acidosis
Infarction/ischemia
Reactive O2 species
Transcription factors
Secondary mediators
Prolonged hospital stay
Disability
Death
Clement S, et al. Diabetes Care. 2004;27:553-591.
3
Mechanisms of Acute Inflammatory
Reactions
Inflammation due to acute injury/disease*
Blood monocyte, Tissue MØ
IL-1 family
Acute Phase
Proteins
TNF family
Stromal cells: IL-6, IL-8, monocyte chemoattractant protein
Neutrophil recruitment, further inflammatory cytokine release
* Trauma, burns, tissue infarction, cancer, immunologic
and crystal inflammation
4
Cytokines
• Signal cells and coordinate the immune
response (“hormones of the immune system”)
• Can be delivered to the target cell by the
systemic circulation or the local environment
• Bind to high-affinity surface receptors
• Can be produced by non-immune cells
(fibroblasts, endothelial cells)
5
NF-B as an Inflammatory Regulator
Barnes and Karin. N Engl J Med. 1997;336:1066-1071.
6
NF-B Activation Before and After
Euglycemic or Hyperglycemic Clamp
Euglycemic (90 mg/dL) and
hyperglycemic clamp (180 mg/dL)
in nondiabetic subjects (n=9)
300%
NF-κB in peripheral blood
mononuclear cells
250%
300%
p=0.08
200%
200%
150%
150%
100%
100%
50%
50%
0%
0%
0 min
Schiekofer S, et al. Diabetes. 2003;52:621-633.
250%
120 min
p=0.04
0 min
120 min
7
NF-B Activation With Oral Glucose
Dhindsa, et al. Metabolism. 2004;53:330-334.
8
What Is the Effect of Hyperglycemia and
Insulin Deficiency on Cytokine Production?
Glucose 0,33g/Kg + variable infusion
Hyperglycemic clamp
at 270 mg/dL,
octreotide to
suppress islet
hormones
Glucose (mmol/L)
20
15
10
5
N=20 control subjects, 12 with IGT
0
Glucose pulses (0,33g/Kg)
Intermittent IV
glucose pulses
every 2 h
Esposito K, et al. Circulation.
2002;106:2067-2072.
Glucose (mmol/L)
20
15
10
5
0
0
1h
2h
3h
4h
5h
9
5
4
IL-6 (pg/mL)
Cytokines
During
Hyperglycemic
Clamp Studies
3
2
1
0
Plasma cytokine
concentrations attained
during hyperglycemic
clamps in 12 IGT (blue)
and 20 control (green)
subjects
TNF-α (pg/mL)
8
0
1h
0
1h
2h
3h
4h
5h
6
4
2
0
2h
3h
4h
5h
IL-18 (pg/mL)
200
160
IGT
120
80
40
Esposito K, et al. Circulation. 2002;106:2067-2072.
control
0
0
1h
2h
3h
4h
10
5h
Cytokines
During Glucose
Pulse Studies
IL-6 (pg/mL)
8
6
4
2
0
8
TNF-α (pg/mL)
Circulating cytokine levels
after consecutive glucose
pulses in 12 IGT subjects
(blue) and 20 control
subjects (green)
0
1h
2h
3h
4h
5h
0
1h
2h
3h
4h
5h
0
1h
2h
6
4
2
0
IL-18 (pg/mL)
200
160
120
80
40
0
Esposito K, et al. Circulation. 2002;106:2067-2072.
3h
4h
5h
11
Insulin as an
Anti-inflammatory
Agent
Dandona, et al. J Clin Endocrinol Metab. 2001;86:3257-3265.
12
Anti-inflammatory Effects of Insulin in
STEMI in Nondiabetic Patients
Change in C-reactive Protein
7
∆ CRP (mg/L)
from baseline
(0–h)
6
5
*
Control
Insulin
*
4
3
2
1
0
–1
* P<0.05 between
groups
0
Chaudhuri A, et al. Circulation. 2004;109:849-854.
2
4
6
24
Time (hours)
48
13
Hyperglycemia Increases
Lipopolysaccharide-Induced
Inflammation and Insulin Suppresses It
Hagiwara S, et al. Crit Care Med. 2009;37:2223-2227.
14
Liver
-
Insulin
+
IL-1, TNF, IL-6
Acute phase proteins
+
+
glucocorticoids
Serum amyloid A, CRP, C3, C4, fibrinogen,
plasminogen, TPa, PAI-1, ferritin
15
Taken Together,
These Studies Suggest:
• Insulin and glucose co-modulate inflammation
• Hyperglycemia is proinflammatory
• Hyperinsulinemia with euglycemia is
anti-inflammatory
• Inability to control glucose in CREATE/ECLA
and DIGAMI 2 likely explains the negative
results using GIK in these two studies
16
Hyperglycemia and hyperinsulinemia
also affect coagulation mechanisms
that increase mortality and morbidity
during acute illness
17
Hyperglycemia Stimulates Coagulation
in Healthy Humans
Plasma concentrations of thrombin-antithrombin complexes (TATc) (A) and soluble tissue
factor (B) in 6 subjects studied during a low-insulin euglycemic clamp (□), low-insulin
hyperglycemic clamp (○), hyperinsulinemic euglycemic clamp (▪), and hyperinsulinemic
hyperglycemic clamp (•). * P<0.001 vs both euglycemic clamps.
Stegenga, et al. Diabetes. 2006;55:1807-12.
18
Hyperinsulinemia Impairs Fibrinolysis
in Healthy Humans
Plasma concentrations of tPA antigen (A), plasminogen activator activity (PA activity) (B),
PAI-1 antigen (C), and PAI-1 activity (D) during a low-insulin euglycemic clamp (□), lowinsulin hyperglycemic clamp (○), hyperinsulinemic euglycemic clamp (▪), and
hyperinsulinemic hyperglycemic clamp (•). * P<0.001 vs both euglycemic clamps.
Stegenga, et al. Diabetes. 2006;55:1807-12.
19
Thus, hyperglycemia stimulates
thrombus formation and, once formed,
hyperinsulinemia prevents its breakage.
Therefore, it will be ideal if one could
prevent hyperglycemia
20
And What Is the Effect of Oxidative
Stress on Inpatient …
• Infection?
• Wound healing?
• Cardiovascular and renal outcomes?
21
Measuring Oxidative Stress
• Urinary isoprostanes: best marker of oxidative
stress in total body
• Urinary 8-iso-PGF2-alpha: most common
isoprostane
• Nitrotyrosine: marker for peroxynitrite, a powerful
oxidant increased with hyperglycemia
• 8-hydroxydeoxyguanosine (8-OHdG): a sensitive
indicator of oxidative damage to DNA
• Reactive oxygen species generation by leukocytes
• “A1C of oxidative stress”
22
Oral Glucose Load Causes
Oxidative Stress
300
250
200
150
100
50
0
Reactive Oxygen Species generation in response
to 75G oral glucose
0h
1h
ROS generation by MNC
Mohanty P, et al. J Clin Endocrinol Meta. 2000;85:2970-2973.
2h
3h
ROS generation by PMNL
23
Insulin as Antioxidant
Dandona, et al. J Clin Endocrinol Metab. 2001;86:3257-3265.
24
What Is the Effect of Insulin
and Different Levels of Glycemia
on Nitric Oxide Production?
25
An Understanding of Nitric Oxide (NO)
• In endothelium, eNOS generates low levels of
NO, which has protective properties
• In virtually every cell type, iNOS (inducible)
generates high NO concentrations, which are
proinflammatory
26
Proposed Mechanisms for Production of
Endothelium-Derived NO and Vasoconstrictors
In Health
In Insulin Resistance/Diabetes
Rask-Madsen C, King GL. Nat Clin Pract Endocrinol Metab. 2007;3:46-56.
27
-inflammation
- VEGF-induced
adhesion molecules
+ vasodilation
Nitric Oxide
(low amounts)
eNOS
+
-
28
+ inflammation
++ vasodilation
+ VEGF-induced
adhesion molecules
Nitric Oxide
(high amounts)
iNOS
-
NF-kB
+
29
Intensive Insulin Therapy in Critically Ill Patients:
Blood Glucose Control, Insulin Doses, and CRP
CIT
IIT
200
200
140
175
175
150
**
**
**
**
CRP (mg/l)
100
125
Insulin dose (IU)
Blood glucose (mg/dl)
120
150
80
100
60
75
40
50
125
**
50
25
0
0
0
Day
15
Last
Day
**
75
20
Day 5 Day 7
*
100
25
Adm
**
Adm Day 5 Day 7 Day
15
Last
Day
Adm Day 5 Day 7 Day
15
Last
Day
N=405 in SICU X >7 days
Blue box = CIT (N=224); gray box = IIT (N=181)
Langouche L, et al. J Clin Invest. 2005;115:2277-2286.
30
What Data Support the Relationship
Between eNOS, iNOS, and Insulin?
• Postmortem biopsies in
CIT (white boxes, N=49)
and IIT (gray boxes,
N=27)
• NF-κB activation
evaluated by
analysis of
phosphorylated IB
Central line = the median
Box height = IQR
Whiskers = 10th-90th percentiles
* P ≤0.05; ** P ≤0.01; § P ≤0.1
31
Langouche L, et al. J Clin Invest. 2005;115:2277-2286.
What Happens to Nitric Oxide With
Intensive Insulin Therapy (IIT)?
Serum NO levels, in
CIT (white boxes) and
IIT (gray boxes)
Central line = the median
Box height = IQR
Whiskers = 10th-90th
percentiles
* P ≤0.05 vs conventional
insulin therapy
† P ≤0.05 day 7 vs admission
day
# P ≤0.05 vs healthy
volunteers
## P ≤0.01 vs healthy
volunteers
Langouche L, et al. J Clin Invest. 2005;115:2277-2286.
32
Nitric Oxide and Insulin: Conclusion
• “Maintaining normoglycemia with intensive
insulin therapy during critical illness
protects the endothelium likely in part via
inhibition of excessive iNOS-induced NO
release, and thereby contributes to
prevention of organ failure and death.”
Langouche L, et al. J Clin Invest. 2005;115:2277-2286.
33
The Implications of Glucose Variability:
Why Are Ups and Downs Bad?
34
Measuring Glycemic Variability
• Mean amplitude of glycemic excursions
(MAGE)
• Standard deviation on SMBG meter
download
35
One Possible Mechanism
• Human umbilical vein endothelial cells
(HUVECs) “exposed to intermittent
hyperglycemia results in ROS
overproduction, through a PKC-dependent
activation …”
• This suggests that glucose fluctuation may be
involved in the development of oxidative
stress and vascular injury
Quagliaro L, et al. Diabetes. 2003;52:2795-2804.
36
Oxidative Stress Marker Measurements
0.6
A
Nitrotyrosine (7 days)
Low Glucose
0.5
High Glucose
C
L
H
50
80HdG (7 days)
Low Glucose
40
H/L
High Glucose
Alternating
High and Low
Glucose
0.3
(ng/ml)
(microM)
0.4
30
Alternating
High and Low
Glucose
20
0.2
10
0.1
0
0
no inhibitor
BIMI-I
LY379196
PKCX
B
0.7
Nitrotyrosine (14 days)
mannitol
MnTBAP
no inhibitor
Low GlucoseL
D
0.3
80HdG (14 days)
mannitol
MnTBAP
Low Glucose
High Glucose
40
(ng/ml)
(microM)
0.5
Alternating
High and Low
Glucose
60
50
High GlucoseH/L
0.4
LY379196
osm con ox stress X
H
0.6
BIMI-I
Alternating
High and Low
Glucose
30
20
0.2
10
0.1
0
0
no inhibitor
BIMI-I
LY379196
mannitol
MnTBAP
no inhibitor
BIMI-I
LY379196
mannitol
MnTBAP
37
Quagliaro L, et al. Diabetes. 2003;52:2795-2804.
Urinary 8-SO-PGF2 alpha Excretion Rates
(pg/mg creatinine)
Correlation Between Urinary
8-iso-PGF2 alpha and MAGE in T2DM
1200
1000
800
R=0.86, p<0.0001
600
400
200
0
0
20
40
60
80
100
120
140
160
MAGE (mg glucose/dL)
Monnier L, et al. JAMA. 2006;295:1681-1687.
38
Why This Study Is So Important
• Type 2 diabetes
– Oxidative stress not related to A1C, fasting
glucose, fasting insulin, mean blood glucose
– Stronger correlation of oxidative stress to MAGE
than to postprandial glucose levels
– MAGE = both the ups and the downs of blood
glucose
Monnier L, et al. JAMA. 2006;295:1681-1687.
39
Glucose Fluctuations May Be Involved
in the Pathogenesis of Vascular
Diabetic Complications
% of propidium positive cells
50
5 mmol/l glucose
20 mmol/l glucose
5/20 mmol/l glucose
#$
40
&*
*
30
20
• Cell death of human
umbilical vein
endothelial cells
• “Variability in
glycemic control
may be more
deleterious than a
constant high
concentration of
glucose”
10
0
7 days
Risso A, et al. Am J Physiol Endocrinol Metab. 2001;281:E924-E930.
14 days
40
Glucose Fluctuations May Be Involved in the
Development of Vascular Injury in Diabetes
PKC activity at low glucose (5mM), high glucose (20 mM),
or alternating low and high
Quagliaro L, et al. Diabetes. 2003;52:2795-2804.
41
What About Glycemic Variability
and Inflammation?
42
Concentration (ng/ml)
E-Selectin (7 days)
8
6
6
4
4
2
2
0
0
NG
HG
H/LG
tO
NG
Concentration (ng/ml)
ICAM-1 (7 days)
1.0
0.8
0.8
0.6
0.6
0.4
0.4
0.2
0.2
0.0
H/LG
tO
0.0
HG
H/LG
tO
NG
HG
H/LG
tO
VCAM-1 (14 days)
VCAM-1 (7 days)
Concentration (ng/ml)
HG
ICAM-1 (14 days)
1.0
NG
3
3
2
2
1
1
0
0
without PJ34
NG
HG
H/LG
NG
tO
HG
H/LG
tO
with PJ34
IL6 (14 days)
IL6 (7 days)
Concentration (ng/ml)
Effect of Normal (90), High
(360) or Alternating Glucose
Levels on UVHEC Adhesion
Molecules and IL-6
E-Selectin (14 days)
8
120
120
100
80
100
80
60
40
60
Poly ADP ribose polymerase =
inhibitor of adhesion molecule
expression and inflammation
40
20
20
0
0
NG
HG
H/LG
tO
NG
HG
H/LG
tO
Conditions
43
Piconi L, et al. J Thromb Haemost. 2004;2:1453-1459.
Hypoglycemia Also Increases Cytokines
Dostson S, et al. Diabetes Care. 2008;31:1222-1223.
44
So What Is the Significance of the
Understanding of Glucose Variability?
• “It suggests that different therapeutic
strategies now in use should be evaluated for
their potential to minimize glycemic
excursion, as well as their ability to lower
A1C.”
• “Wider use of real-time continuous glucose
monitoring in clinical practice would provide
the required monitoring tool to minimize
glycemic variability and superoxide
overproduction.”
Brownlee M, et al. JAMA. 2006;295:1707-1708.
45
Summary
• Hyperglycemia is proinflammatory, pro-oxidative, and
pro-coagulant
– All these mechanisms are known to increase mortality
and morbidity in acutely ill patients
• In experimental models, correction of hyperglycemia
reverses all the above abnormalities
• In the presence of euglycemia, insulin may have
direct anti-inflammatory, anti-oxidative, and anticoagulant effects
• In the presence of hyperglycemia, role of insulin is
controversial, but it may be harmful
46
Summary
• High free fatty acid levels that are associated with
insulin resistance and hyperglycemia in acutely ill
patients can further increase inflammation and cause
endothelial dysfunction
• Fluctuating blood glucose levels are also
proinflammatory and induce endothelial dysfunction
and oxidative stress
• Therefore, the best way to suppress the harmful
cellular and molecular mechanisms during acute
illness is the maintenance of a stable glycemic
control in the physiological range
47