1. Genetic Testing and the Prevention of Type 1 Diabetes
Janice S. Dorman, Ph.D.
September 4, 2001
mnb

2. Type 1 Diabetes One of most frequent chronic diseases of children
- Prevalence ~ 2 / 1000 in Allegheny County, PA
Epidemiology of type 1 diabetes has been studied at the University of Pittsburgh since 1979
- Dr. Allan Drash and Dr. Lewis Kuller

3. Type 1 Diabetes Incidence Allegheny County, PA

4. Type 1 Diabetes Incidence Allegheny County, PA

5. Type 1 Diabetes Incidence Allegheny County, PA

6. FIN

7. Type 1 Diabetes Incidence Worldwide

8. Specific Environmental Risk Factors
Case-control studies - conflicting
Possible risk factors
- Infant diet or lack of breast feeding
- Childhood diet
- Viruses (exposure as early as in utero)
- Hormones
- Stress
May act as initiators or precipitators

9. Evidence for Genetic Risk Factors
Increased risk for 1st degree relatives of affected individuals
Concordance in MZ twins %
Recent genome wide screens have revealed 15+ possible susceptibility genes
Associations with HLA class II alleles in all populations

12. Genome Screens for Type 1 Diabetes
IDDM1 6p21.3
IDDM2 11p15.5
IDDM3 15q26
IDDM4 11q13.3
IDDM5 6q15
IDDM6 18q12-q21
IDDM7 2q31-33
IDDM8 6q25-27
IDDM9 3q21-25
IDDM10 10p11-q11
IDDM11 14q24-q31
IDDM12 2q33
IDDM13 2q34
IDDM14 ND
IDDM15 6q21
* Candidate Gene *Possible Candidate *No Candidate Gene

13. Interpreting Linkage Analysis for Type 1Diabetes
Need to control for effect of HLA
Some genes confer susceptibility in absence of high risk HLA haplotypes
Need model- free statistical methods
Account for gender, parent-of-origin effects and environmental risk factors
May not be appropriate phenotype

14. Genome Screens for Type 1 Diabetes
Chromosome 6
IDDM8 6q25-27
IDDM15 6q21
Chromosome 2
IDDM7 2q HOX8, IL-1 family IDDM12 2q33 CTLA4, CD IDDM13 2q34 IGFBP2, IGFBP5
* Candidate Gene *Possible Candidate *No Candidate Gene

15. Genome Screens for Autoimmune Diseases
Candidate Genes - Type 1 Diabetes
IDDM1 6p21.3 DR-DQ, 2nd loci - TNF?
IDDM2 11p15.5 INS-VNTR
IDDM12 2q33 CTLA4, CD28
Candidate Genes - Other Disorders
IDDM1 ATD, CD, RA, MS, SLE
IDDM2 SLE, ankylosing spondylitis
IDDM12 ATD

16. WHO DiaMond Molecular Epidemiology Study
Have evaluated HLA DQ
Best single genetic marker
Evaluate other candidate genes
IDDM1 HLA DR, DP
IDDM2 INS-VNTR
IDDM12 CTLA4
Others VDR, HLA class I

17. WHO Multinational Project for Childhood Diabetes (DiaMond)
What is Causing the Tremendous Geographic Variation in Incidence of Type 1 Diabetes?
Monitored Incidence Worldwide

18. University of Pittsburgh
WHO Collaborating Center for Diabetes Registries, Research and Training
Ron LaPorte, Ph.D. Disease Monitoring & Telecommunications
Jan Dorman,Ph.D. Molecular Epidemiology
University of Pittsburgh

19. WHO DiaMond Molecular Epidemiology Study
Hypothesis
Geographic differences in type 1 diabetes incidence reflect population variation in the frequencies of disease susceptibility genes
20+ countries participating
Focus on 2, 1, or 0 high risk HLA-DQ haplotypes (SS, SP, PP)

21. Relative Increase In Risk
Population SS SP PP
Caucasian† *
Af Americans† *
Asian‡ *
* p < 0.05, test for trend
†Allegheny Co, PA and Jefferson Co, AL
‡Hokkaido, Japan and Seoul, Korea

22. Cumulative Risk Through Age 30 Years
Population SS SP PP
Caucasian† 2.6% 0.7% 0.2%
Af Americans† 3.1% 0.5% 0.1%
Asian‡ % 0.1% 0.02%
†Allegheny Co, PA and Jefferson Co, AL
‡Hokkaido, Japan and Seoul, Korea

23. Population Attributable Fraction
Population SS SS or SP
Caucasian† 36.2% 66.6%
Af Americans† % 74.9%
Asian‡ % 53.3%
†Allegheny Co, PA and Jefferson Co, AL
‡Hokkaido, Japan and Seoul, Korea

24. What do these data tell us?
Increased risk for individuals with SS and SP genotypes, relative to PP, with a significant dose response
Cumulative risk for SS individuals in high-moderate incidence countries approaches rates for first degree relatives; 3 - 6%

25. What do these data tell us?
Contribution of the highest risk HLA-DQ genotypes to type 1 diabetes incidence varied from 19% - 43% across populations
More than 50% of the incidence of type 1 diabetes is NOT explained by the highest risk HLA-DQ genotypes

26. Gene - Environment Interactions
Finland
Exposure increased risk by 1/100,000 / year among susceptibles
Overall population risk would increase by 0.8%

27. Gene - Environment Interactions
China
Exposure increased risk by 1/100,000 / year among susceptibles
Overall population risk would increase by 10%

28. Molecular Epidemiology of Type 1 Diabetes in China
What is contributing to the low overall incidence and large variation in risk within China?
- Etiological heterogeneity
- Susceptibility genes
- Environmental risk factors
Project based on DiaMond registry network
Model study for molecular epidemiology

29. 1.8
Rate (per 100,000)

30. * * * * * * * * * * * * * * * * * *

31. Molecular Epidemiology of Type 1 Diabetes in China
Data collection completed in Dr. Yang Ze
296 cases, 528 controls; 18 centers
Molecular analyses - Beijing
- HLA DRB1, DQB1 typing
Serological analyses - Pittsburgh
- GAD, IA-2, TPOAb, TGAb, C-pep
Environmental data - Pittsburgh
- Nutrition, infections, pollution
Dissertation for Dr. Elsa Strotmeyer

32. Jan Alice Lew Yang Ze