Chapter 11
Prediction of Type 1A Diabetes:
The Natural History of the Prediabetic
Period
TRIGGERING QUESTIONS
•
•
•
•
•
Is there an environmental trigger?
Does autoantibody appearance mark triggering?
Time lag between trigger and insulitis?
Time lag between insulitis and beta cell killing?
“Best Model”
?Kilham Rat Virus (Multiple Other viruses)
ACTIVATION INNATE IMMUNITY BY VIRUS
SPECIFIC MHC AND SPECIFIC TCR (Mordes et al)
ANTI-INFLAMMATORY PREVENTS (Zipris et al)
Williams et al: Pancreatic Volume is Reduced in Adult Patients with
Recently Diagnosed Type 1 Diabetes JCEM 2012, 97
“Stages” in Development of Type 1A Diabetes
(?Precipitating Event)
Beta cell mass
Genetic
Predisposition
Overt
immunologic
abnormalities
Normal insulin
release
Progressive
loss insulin
release
Glucose
normal
Overt
diabetes
C-peptide
present
Minimal
C-peptide
Age (years)
Eisenbarth 2012
nPOD 6052-02 Tail: 12 yo 1 year diabetes
-Lobular Pseudoatrophic Islets
Glucagon/anti-CD3 Staining
Insulin and Ki67 Staining
Discordant Triplets at Risk for Diabetes
350
Intravenous Glucose Tolerance Test
(IVGTT) 1+3 minute insulin
Antibody Positive
300
Antibody
Positive Initial
Test
250
200
150
100
50
DM
0
14 16 18 20 21 22 24 27 30 34 35 36 37 38 40 42 43 44 45 46 47 48 49 50
14 15 16 17 18 19 20 21 22 23
Age
Srikanta S. et al, New Engl
J Med 308:322-325, 1983
Time Course of Beta Cell Loss
Linear, Chronic Model
Eisenbarth (NEJM 1986, 314:1360)
Beta
Cell
Mass
Age
Benign:Malignant Model
Lafferty (J Aut 1997, 10:261)
Beta
Cell
Mass
Benign
Malignant
Age
Random Loss Model
Palmer (Diabetes 1999, 48:170)
Beta
Cell
Mass
Age
Time Course Beta Cell Loss
Linear: Eisenbarth
NEJM 1986, 314:1360
Prodrome> Acute
Lafferty; J Aut 1997, 10:261
Random:Palmer
Diabetes 1999, 48:170
Stages in Development of Type 1 Diabetes
GENETICALLY AT RISK
BETA CELL MASS
MULTIPLE ANTIBODY POSITIVE
LOSS OF FIRST PHASE
INSULIN RESPONSE
GENETIC
PREDISPOSITION
INSULITIS
BETA CELL INJURY
“PRE”DIABETES
DIABETES
TIME
NEWLY DIAGNOSED DIABETES
J. Skyler
Sustained beta cell apoptosis in patients
with long-standing type 1 diabetes:
indirect evidence for islet regeneration?
Meier et al, Diabetologia 2005
% Insulin/Pancreatic Area
1.2
1
0.8
0.6
0.4
0.2
0
Diabetes
Control
“Biochemical” Autoantibody Assays
• Insulin
• Glutamic Acid Decarboxylase
• ICA512 (IA-2)
• ZnT8
Non-Radioactive Electrochemiluminescent Insulin
Autoantibody Assay
Insulin
Autoantibody
Sulfo-TAG
Labeled Proinsulin
Biotin
Labeled Proinsulin
Streptavidin
Coated plate
Yu et al Diabetes Nov 2011
% NOT DIABETIC
Progression to Diabetes Among Children Positive for Anti-Islet Autoantibodies
100
90
80
70
60
50
40
30
20
10
0
0.0
1 Ab
2 Ab
3 Ab
2.5
5.0
7.5
10.0
12.5
15.0
17.5
YEARS SINCE INITIAL AB+ TEST
Steck et al Diabetes Care 2011
Age 1st Islet Ab+ vs Age DM Onset
AGE DIABETES ONSET
Steck et al Diabetes Care 2011
15
10
R2=.47 p<.0001
5
0
0.0
2.5
5.0
7.5
10.0
AGE 1st Islet Autoantibody +
12.5
Time to DM from age 1st
Ab+
MEAN LOG IAA vs Time to DM from age Islet Ab first +
15
R2=.37 P<.0001
10
5
0
-3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5
log10 Mean Insulin Abs
0.0
Steck et al Diabetes Care 2011
Quartiles of Insulin Autoantibodies Predict Faster
Progression to Overt Diabetes
Parikka ….Simell Diabetologia 2012
Predicted Onset Age=
2.6-1.3*log(mean IAA) =0.8* age first Ab+
Steck et al Diabetes Care 34:1397–1399, 2011
GAD65 Levels with Years to Diabetes ICA512 Levels with Years to Diabetes
Steck et al Diabetes Care 2011
17.5
17.5
15.0
15.0
12.5
12.5
10.0
10.0
7.5
7.5
5.0
5.0
2.5
2.5
0.0
0.0
-3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0
-4
Mean GAD65 Levels (Log10)
-3
-2
-1
Mean ICA512 Levels (Log10)
0
LONG TERM
LEVEL Autoantibodies
2.4
PRE-DM WITH MULTIPLE AUTOANTIBODIES
36336-0 DAISY Study
Diabetes Onset
1.9
1.4
0.9
0.4
-0.1 0
2
NOTE LACK IAA
4
6
GAD
8
IA-2
10
IAA
ZnT8RW
12
14
AGE (YEARS)
16
DPT-1 Ancillary Biochemical Ab
• Cytoplasmic ICA Positive (3.4%)
1/2 Negative for GAD/ICA512/Insulin Ab
0.9% = 1 Biochemical Ab
1.1% >=2 Ab
• Cytoplasmic ICA Negative (96.6%)
3.3% =1 “Biochemcial Ab
0.3% >=2 Ab
• Staging: Only 12% eligible ICA+/Bioch • Future Trials Likely without ICA
Progression to Diabetes vs Number of Autoantibodies
(GAD, ICA512, Insulin)
Percent not Diabetic
100
80
3 Abs
2 Abs
1 Ab
60
40
20
0
0
2.5
5
7.5
10
12.5
2
6
1
4
15
Years of Follow-up
3 Ab n = 41
2 Abs n = 44
1 Abs n = 93
17
27
23
8
15
14
1
4
10
Verge et al. Diabetes,
1996;45;926
New Onset Type 1 DM: Loss of Insulin Secretion (ISR
area(AUC)) related to early (peak<45 min) versus delayed
secretion Mixed Meal Steele et al Diabetes 53:26, 2004
Type 1 diabetes risk stratification models based on islet autoantibody characteristics
Model 1
Stratification based on
Model 2
Stratification based on
Model 3
Stratification based on
Model 4
Stratification based on
Number of
islet autoantibodies
(IAA, IA-2A, GADA)
High titre of
IAA (>3rd quart.)
and IA-2A (>1st quart.)
High risk characteristics:
High titre IA-2A (>1st quart.)
and IgG2 or IgG4 IA-2A
and IgG2, IgG3 or IgG4 IAA
Status of
IA-2A and IA-2βA
Category 1
One autoantibody
Category 1
Neither IAA nor IA-2A
at high titre
Category 1
No high risk
characteristic
Category 1
IA-2A negative
Category 2
Any two
autoantibodies
Category 2
One of IAA or IA-2A
at high titre
Category 2
One high risk
characteristic
Category 2
IA-2A positive and
IA-2βA negative
Category 3
All three
autoantibodies
Category 3
Both of IAA and IA-2A
at high titre
Category 3
Any two high risk
characteristics
Category 3
IA-2AβA positive
Category 4
All three high risk
characteristics
Shaded Categories: 10-year diabetes risk >50% (high-risk categories)
Achenbach et al., Diabetologia (2006) 49:2969-2976
Type 1 diabetes risk stratification considering changes in model risk category on follow-up
Model 1
Model 2
Model 3
Model 4
100
80
P = 0.02
60
P < 0.001
P < 0.001
P < 0.001
40
20
0
100
80
P = 0.02
60
40
20
0
0
2
4
6
8 10 12 0
2
4
6
8 10 12 0
2
4
6
8 10 12 0
2
4
6
8 10 12
Follow up (years)
Stable low-risk category
Changed from high-risk to low-risk category
Stable high-risk category
Changed from low-risk to high-risk category
Achenbach et al., Diabetologia (2006) 49:2969-2976
HbA1c (%)
RISING HbA1c PRECEDES DIABETES
7
7
6
6
7
5
5
6
4
4
3
3
2
4
6
8
10
12
14
HbA1c (%)
7
4
3
0
2
4
6
8
10
12
14
5
4
0
2
4
6
8
10
12
0
14
0
2
4
6
8
10
12
14
7
7
7
6
6
6
5
5
5
4
4
4
3
3
0
2
4
6
8
10
12
14
4
6
8
10
12
14
Acute?
0
2
4
6
8
10
12
14
0
2
4
6
8
10
12
14
0
2
4
6
8
10
12
14
0
2
4
6
8
10
12
14
3
0
2
4
6
8
10
12
14
7
7
7
6
6
6
5
5
4
4
5
2
7
6
5
4
3
7
6
5
4
3
6
3
HbA1c (%)
Acute?
5
0
HbA1c (%)
8
4
3
3
0
2
4
6
8
10
12
3
0
14
2
4
6
8
10
12
14
HbA1c (%)
Age (years)
7
7
7
6
6
6
5
5
5
4
4
4
3
3
0
2
4
6
8
Age (years)
10
12
14
3
0
2
4
6
8
Age (years)
10
12
14
Age (years)
Stene et al DAISY Study Pediatric Diabetes 7:247-253
NEW ONSET CHILDREN PRESENTING WITH DIABETES
FH = Family History of Diabetes; Control = DM Onset not in DAISY Study
Blood glucose, md/dL
Blood glucose values in
Control vs. Daisy children
1400
1200
1000
800
600
400
200
0
DAISY
Control-FH
Barker et al. DiabetesCare, 2004; 27:1399-1404.
Control no FH
T1DM- a slowly progressive T-cell mediated
autoimmune illness
100%
Genetic
susceptibility
Inciting
Event(s)
“Silent” 
Cell Loss
Diabetes
Onset
Islet
Strong
association
withEtiology
MHC class
II (DQ in particular)
Infectious
agent(s)?if
true?
What
is the “slope”
ofweaker,
the  cell
loss?
Cell
Other
associations
much
population
dependentEnvironmental
toxin(s)?
Is recovery
possibleCD152,
once process
begins?
Mass
e.g.
insulin
VNTR,
other
Absence
of childhood
illness?
What
underlies
the
effect
of age
on slope
We cannot
easily/accurately
measure
islet
mass
in vivo or Iex vivo
Combination
of
factors?
50%
of for
cellthe
loss?
No accepted
norm
islet number within a human pancreas
Age
of
exposure?
Why does the  cell destruction typically
occur slowly (in contrast to graft rejection)?
 cell
Mass??
0%
Time (years)
David Harlan
“Brittle”
Diabetes
II III
Is  cell
mass
Is  cell lossCan
 cell
completely
exclusively regeneration
lost?
immune mediated?
occur?
Diagnosis of Diabetes ADA
NORMAL
IMPAIRED
DIABETES
HbA1c
<6.4
5.7-6.4
>=6.5
FASTING
< 100 mg%
(5.6 mM)
100-125
>= 126 mg%
(7 mM)
ORAL GTT
<140 mg%
(7.8 mM)
140-199
>=200 mg%
(11.1 mM)
Gestational Diabetes (>=2 high)
100-g or 75-g Glucose
100-g Glucose
Fasting
1-h
2-h
3-h
75-g Glucose
Fsting
1-h
2-h
mg/dl
mmol/l
95
180
155
140
5.3
10
8.6
7.8
95
180
155
5.3
10
8.6
Greenbaum et al Diabetes June 11, 2012
Fall in C-peptide During First 2 Years From Diagnosis: Evidence of at
Least Two Distinct Phases From Composite TrialNet Data.
Progressive Loss C-peptide Post Diagnosis (SEARCH Diab Care 2009)
DCCT
Fast>=.23ng/ml
ACCELERATED LOSS OF PEAK C-PEPTIDE AFTER DIAGNOSIS OF TYPE 1A
DIABETES (“WAITING” FOR CONFIRMATORY ORAL GLUCOSE TOLERANCE TEST)
Sosenko et al, Diabetes Care August 2008
50 Year Joslin Medalist C-Peptide
Times 3 for ng/ml
Keenan et al Diabetes 2010
Gestational Diabetes: Risk at 2 years Type 1
Diabetes by Autoantibodies
ICA, GAD65, ICA512(IA-2)
90
80
70
60
50
40
30
20
10
0
0 Ab
>=1 Ab
1 Ab
2 Ab
3 Ab
Sensitivity GAD=63%; Sensitivity 3 Abs=82%
Ziegler et al. Diabetes 1997: 46:1459-67, N=437
LADA: Latent Autoimmune Diabetes Adults
in UKPDS study
% GAD +
Insulin by 6 Years
45
40
35
30
25
20
15
10
5
0
35
30
25
20
15
10
5
0
25
-34
55
-65
IC
A
+
AGE
Turner et al. Lancet 1997;350:1288-93
GA
D6
5+
0A
b
Caveats of IVGTT Testing
Caveat
Suggestion
"First" Test lack response Repeat Abnormal Tests
young children
Lack Carbohydrate
Dietary Preparation similar
to OGTT
Type 1A and insulin
Subtract 2X fasting insulin
resistance may coexist
Can be <1st Percentile in Long-term follow up
adults years prior to DM
Subset normals <1st %
In absence Abs low risk
Variation e.g. puberty
Repeat tests; caution in
interpreting changes
First-phase insulin release during the intravenous
glucose tolerance test as a risk factor for type 1
diabetes (DPT)
Chase et al. J. Peds 138,244; 2,001
1+3 Minute Insulin (uU/ml)
250
200
150
ICA Negative
ICA Positive
100
50
0
AGE
<8
8-20 21-30 31-45
BDC
log FPIR (1+3' insulin)
FPIR in pre-diabetic relatives with initial FPIR
> 50mU/L
500
50
5
-8
-7
-6
-5
-4
-3
-2
-1
Years prior to diabetes
Melbourne Pre-Diabetes Study (Colman PG & Harrison LC)
0
EARLY LOSS IVGTT ICA+ INFANTS:
Individual insulin concentrations during IVGTTs
performed to A) ICA negative children 1–5 years of age, B) ICA positive children who had not progressed
to diabetes by May 2001, and C) ICA positive children who had developed clinical Type I diabetes by May
2001. (From: Keskinen P, Korhonen S, Kupila A, Veijola R, Erkkila S, Savolainen H, Arvilommi P, Simell T, Ilonen
J, Knip M, Simell O: First-phase insulin response in young healthy children at genetic and immunological risk for
Type I diabetes. Diabetologia 45:1639-48, 2002)
18 0
C
B
A
16 0
In su lin (m U/l)
14 0
12 0
10 0
80
60
40
20
0
0
1
2
3
4
5
6
7
8
9
10
0
1
2
3
4
5
6
Time (min)
7
8
9
10
0
1
2
3
4
5
6
7
8
9
10
Insulin Secretion (IVGTT) in Obese Child
(BMI 30 to 35) Progressing to Diabetes: Type
1 + Type 2 with Elevated Fasting Insulin
Insulin (uU/ml)
250
200
150
1+3 Insulin
2X Fast
100
50
0
9
10
11
12
Age (years)
13
14
15
Lack of Progression to DM of ICA+
0602+ Relatives
Percent Not Diabetic
0602+
0602-
100
75
50
25
0
0
2
4
6
8
Years of Follow up
10
12
Number Abs: IVGTT > or <1st%
80
60
40
20
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
Percent Not Diabetic
100
0
16.0
Years of Followup
1Ab <1st
2Ab <1st
3Ab <1st
1Ab >1st
2Ab>1st
3Ab>1st
Percent Not Diabetic
Melbourne Data: Dual Parameter Prediction
Time to DM=-.12+1.35ln(IVGTT)-.59ln(IAA)
100
80
60
Predict<2.5
Predict>2.5
40
20
0
0
2
4
6
8
Years of Follow up
<2.5 N= 11 5
3
>2.5 N=70
42 32
53
1
0
24 13
6 Proc AAP:110:126-135
DM
Normal but increasing hemoglobin A1c levels predict progression from islet autoimmunity to overt
type 1 diabetes: Diabetes Autoimmunity Study in the Young (DAISY). Stene Pediatr Diabet 2005
Diabetes Autoimmunity Study in the Young
General population cohort
Sibling/offspring cohort
screened = 21,713
enrolled =
293
high risk
72
429
moderate risk
220
347
average - low risk
401
1,069
relatives
1,491
All
693
1,007
DAISY Interviews and Clinical
Interviews:
B 3m
6m
9m
diet
infections
immunizations
allergies
stress
1y
15m
2y
3y
Visits
Clinical Visits: blood sample for GAA, IAA, ICA512, ICA
DNA
throat and rectal swabs
saliva sample
Prediction of Autoantibody Positivity and Progression to Type 1
Diabetes: DAISY study
Barker et al. J Clin Endocrinol Metab 89:3896, 2004
162 Positive
Of 1,972= 8.2%
50 False +
112 Confirmed +
1/3
1/3
50 Transient
58(4) Persistent
1/3
28 X1+
24 Diabetic
22 >1+
28 Not DM
1/3 of Multiple Time+ are Transient
(22/(22+24+28)
2/3 High Risk Diabetes
DAISY AUTOANTIBODIES:Initial Test <Age 1
Percent with Persistent Autoantibodies (GAA/IAA/ICA512)
100
p<.0001
3/4SOC
3/4NEC
80
not 3/4SOC
60
not 3/4NEC
40
20
0
0
1
3/4 SOC:
15
9
3/4 NEC: 151 110
-3/4 SOC:
69 56
-3/4 NEC: 492 300
2
5
67
39
208
3
4
18
16
110
4
3
12/27/97
Relatives (SOC) vs. Population (NEC)
Persistent vs. Transient AutoAb
35
30
Percent
25
SOC 3/4
SOC # 3/4
NEC 3/4
NEC #3/4
20
15
10
5
0
PERSISTENT
TRANSIENT
Yu et al. JCEM 85: 2421, 2000
PERCENT
100
90
80
70
60
50
40
30
20
10
0
Affin>10(9)
Multipe Abs
f/u
Diabetes
Proinsulin
High Risk
"False Positive"
Mature high-affinity immune responses to (pro)insulin anticipate the autoimmune cascade
that leads to type 1 diabetes. Achenbach et al, J.Clin Invest 2004, 114:589
Candidate environmental causes of type
1 diabetes
• Definite (rare cases)
– congenital rubella 
• Putative
– enteroviruses
– rotaviruses
– components of infant diet
• gluten
• cow’s milk
Enteroviruses: Recent Studies
Study
Autoimmunity
Diabetes
Frisk 1992
CVB1-5 IgM
Dahlquist 1995
CVB2-4 IgM
Hyoty 1995
IgM,IgG CVB
CVB IgM,IgG
Clements 1995
EV RNA
Graves(DAISY) No EV RNA
1996,2000
(RNA 12%:18%)
No EV IgM
Hyoty (DIPP) EV Ab:57%:31%
2000 (<6mos.) EV RNA: 29% 6%
No EV RNA
Autoantibody development and enteroviral RNA
in a HLA-DR3/4,DQB1*0302 sibling
SD score
1000.0
ECHO 16
100.0
GAA
IAA
ICA512
TGIgA
10.0
1.0
Age [yrs]
0.1
0.8 1.1 1.4 1.8 2.4 2.7 3.0 3.3 3.6 3.8 4.2 4.9 5.5
DAISY ID 00060
EV- EV+ EV- EV+ EV+ EV- EV+ EV+ EV- EV- EV-
Beta-cell autoimmunity and presence of enteroviral
RNA in serum, saliva and stool
Graves PM, DAISY, 1999
Prevalence of EV RNA
30%
3/13 3/13
3/14 6/28
20%
10%
0%
Relatives
High risk children
from the general
population
Cases
Controls
DIPP Protocol
Main Cohort (n=38,000)
• Newborns screened for genetic risk
• High risk babies followed serially for ICA
(n=81)
• ICA-positive children randomized to nasal
insulin or placebo
Trials to Prevent Type 1 Diabetes
• Trialnet/DPT-1
• ENDIT
• TRIGR
• DIPP
Cumulative Ab frequency (%)
Development of islet autoantibodies in
1610 offspring of mothers or fathers with T1D
DR3/4-DQ8
20
DR4/4-DQ8
15
10
Moderate
DR4-DQ8
Neutral
Moderate DR3
Protective
5
0
0
2
4
6
8
Age (years)
Walter et al, Diabetologia 2003 (updated 2004)
Development of islet Abs
Cumulative Ab frequency (%)
- HLA DR-DQ and INS VNTR genotypes
30
DR3/4-DQ8 or 4/4-DQ8
+ INS VNTR I/I
25
P = 0.03
20
DR3/4-DQ8 or 4/4-DQ8
+ INS VNTR I/III or III/III
15
10
5
Other
0
0
2
4
6
Age (years)
8
Walter et al, Diabetologia 2003 (updated 2004)
Multiple autoantibodies (%)
Development of islet Abs - proband
both parents or
parent + sibling
30
25
20
P < 0.0001
15
10
father only
5
P = 0.05
mother only
0
0
2
4
6
8
Age (years)
A. Ziegler
Progression from multiple islet Abs to diabetes
- No effect of HLA DR-DQ or proband
Proband
Type 1 diabetes (%)
HLA
High
Neutral or Protective
Moderate
100
80
100
80
60
60
40
40
20
20
0
0
0
2
4
6
8
Both parents or Parent plus sibling
Mother only
Father only
0
2
4
6
8
Time from first autoantibody positive (years)
A. Ziegler
Islet autoantibody appearance in BABYDIAB
offspring
10
Any islet Abs (7.8%)
8
6
Multiple islet Abs (3.7%)
4
Single islet Abs
2
0
0
A. Ziegler
2
4
Age (years)
6
8
10
Hummel et al., Ann Intern Med, June 2004
Progression to multiple Abs is necessary for
disease
100
Diabetes (%)
80
multiple antibodies
60
40
20
Single IAA
0
0
2
4
6
8
Time from first Ab (years)
A. Ziegler
Hummel et al., Ann Intern Med, June 2004
Usual Progression
Insulin
GAD
IA-2
ZnT8
A. Ziegler
Cumulative frequency (%)
First antibody is insulin/proinsulin
8
6
IAA
4
GADA
2
IA2A
0
0
2
4
6
8
10
Age (years)
A. Ziegler
IAA affinity is high in children who develop
multiple islet Abs
1012
P<0.0001
IAA Affinity (L/mol)
1011
1010
109
108
107
106
105
104
multiple
Abs
IAA
only
Achenbach, J Clin Invest, 2004
A. Ziegler
IAA affinity is relatively stable during follow-up
1012
IAA Affinity (L/mol)
1011
1010
109
108
107
106
105
104
0
1
2
3
4
5
6
7
8
9
10
11
12
A. Ziegler
Risk for developing islet Abs in relation to
birth autoantibody status in offspring of T1D mothers
% with multiple Abs
10
P = 0.007
8
NEG GADA and IA2A at birth
n = 244
6
Father T1D
4
POS GADA or IA2A at birth
n = 476
2
0
2
4
6
Age (years)
A. Ziegler
8
10
Koczwara et al, Diabetes 2004