Using genetic risk factors for predicting type 1 diabetes progression and prognosis

Type 1 diabetes is a multigenic disease in which the pancreatic β-cells are destroyed by an autoimmune process. At time of diagnosis, only poorly functional β-cell mass exists. Prediction of type 1 diabetes progression and prognosis using genetic markers may improve treatment strategies and increase the patient´s life quality.

by Dr Caroline A. Brorsson and Dr Joachim Størling

Type 1 diabetes
Type 1 diabetes (T1D) is a chronic disease that results from an autoimmune destruction of the insulin-producing pancreatic β-cells in the islets of Langerhans. Worldwide, T1D is affecting an increasing number of people and the strongest increase in incidence is observed among young children. The disease is complex and caused by an interplay between genetic and environmental risk factors. Genes of the human leukocyte antigen (HLA) locus are the most prominent risk-conferring genes, but dozens (>50) of other risk loci have now been established to influence the risk of T1D [1, 2]. The exact mechanisms, however, by which HLA and other associated loci affect T1D risk, islet autoimmunity and the time course of β-cell destruction, remain elusive. T1D is preceded by a pre-clinical phase characterized by the appearance of autoantibodies directed against islet antigens. Several studies have confirmed the strong predictive effect of islet autoantibodies for the risk of developing T1D in genetically susceptible individuals. Still the rate of progression from autoantibody positivity to clinical onset is highly variable between individuals, and likely influenced by a combination of genetic and environmental risk factors [3].

As any immune-modulating interventions are possible only after the first signs of autoimmunity have occurred, there is a high risk that most of the β-cells have already been destroyed. Therefore, there is a need for more precise, and preferably earlier, methods for predicting disease risk and progression in order to preserve the residual β-cell mass and to choose optimal treatment regimens. Genetic markers can be measured before the onset of autoimmunity and offers an opportunity for early screening. Also, after the onset of T1D, preservation of β-cell function, as assessed by higher C-peptide levels, has been associated with decreased risk of diabetes complications including acute hypoglycemic events and long-term microvascular complications [4–6].  
 
Prediction of T1D progression in high-risk individuals
Children with a high risk of diabetes, as characterized by either by carrying high-risk HLA genotypes or by having first-degree relatives with T1D, have been followed from birth until autoantibody development and diabetes onset in several cohort studies. A few of these studies have investigated the predictive effect of non-HLA genetic variants for islet autoimmunity and progression to T1D. Steck et al. studied the largest such prospective cohort from the U.S. population (the Diabetes Autoimmunity Study of the Young; DAISY) consisting of 861 first-degree relatives and 882 high-risk children from the general population [7]. They found that the risk alleles for PTPN22 and UBASH3A predicted both islet autoimmunity and diabetes, whereas PTPN2 predicted islet autoimmunity alone and INS predicted diabetes alone.

Studying a similar population of 1650 children of type 1 diabetic parents in the German BABYDIAB cohort, Winkler and co-workers showed that the cumulative sum of risk alleles of 12 T1D risk variants (a so-called genetic risk score; GRS) could stratify the risk of developing islet autoantibodies and diabetes, and progression from islet autoimmunity to diabetes [8]. In a subsequent study, Bonifacio et al. studied the rate of progression from the development of islet autoantibodies to diabetes in the same cohort of high-risk children [9]. They found that the genetic risk score of 12 genes could only marginally predict the risk of islet autoimmunity, but could significantly modify the risk of progressing from autoantibody positivity to diabetes. The most predictive power had a genetic risk score constructed from the five risk variants in INS, IFIH1, IL18RAP, CD25 and IL2, which could identify 80% of islet autoantibody-positive children who progressed to diabetes within 6 years and discriminate high risk (63% within 6 years) and low risk (11% within 6 years) antibody-positive children.

Achenbach and colleagues used the same cohort to investigate whether the 12 genetic variants could discriminate between slow and rapid progression to T1D in multiple autoantibody-positive children (3). Among the 1650 children, 23 developed multiple autoantibodies and progressed to diabetes within 3 years, while 24 developed multiple autoantibodies but did not progress to diabetes during more than 10 years of follow-up. The slow and rapid progressors were similar in regards to HLA risk genotypes, development of autoantibodies to insulin (IAA), glutamic acid decarboxylase (GADA) and zinc transporter 8 (ZnT8A), and progression to multiple autoantibodies. However, autoantibodies to insulinoma-associated antigen-2 (IA-2A) developed significantly later in children who progressed slowly. The GRS could clearly discriminate between the two groups of progressors. Best discriminatory power had a GRS including seven of the 12 risk variants (for the genes IL2, CD25, INS, IL18RA, IL10, IFIH1, and PTPN22). Interestingly, the risk score did particularly well in discriminating between children that carried high-risk HLA genotypes.

Prediction of T1D prognosis in new-onset patients
The Hvidoere Study Group for Childhood Diabetes (HSG) has collected a cohort of 275 newly diagnosed children with the purpose of identifying factors that control changes in β-cell function and glycemic control over time. All children underwent a standardized mixed-meal test at 1, 6 and 12 months after the diagnosis of T1D, to assess the stimulated C-peptide response at these time-points. Mortensen et al. (10) were able to demonstrate several factors that predict lower β-cell function at 12 months after diagnosis, including younger age and ketoacidosis at diagnosis, and stimulated C-peptide levels, post-meal blood glucose levels, and IAA and GADA autoantibodies at 1 month.   

Only a few studies have investigated the genetic effect on prognosis after disease onset in T1D, including the cohort collected by the HSG. Candidate gene studies of single genetic variants have shown that the INS and PTPN22 risk variants are associated with residual β-cell function, glycemic control, autoantibody titres and proinsulin in new-onset T1D [11–13]. Furthermore, in one of the first studies that used a combination of cell biology experiments and clinical observations to study the impact of a T1D risk gene, we investigated the function of CTSH. That study showed that the risk variant of CTSH was associated with β-cell function and insulin dose in the children one year after diagnosis [14]. Interestingly, it was observed that within the β-cells, CTSH is a protective gene that inhibitsβ-cell death induced by pro-inflammatory cytokines – believed to contribute to β-cell killing in T1D – thus providing a mechanistic explanation for how genetic variation in CTSH affects T1D risk. In a separate cohort studying children diagnosed with T1D before the age of 11 years, we demonstrated that the risk variant in ERBB3 was associated with better β-cell function and lower HbA1c levels, and thereby a better glycemic control, after controlling for the effects of sex, age at diagnosis and duration of diabetes [15]. In that study, we also found that ERBB3 is regulating β-cell death in response to pro-inflammatory cytokines providing a possible mechanistic link.

In our most recent study on the HSG cohort, we investigated the impact of an increasing GRS on β-cell function and glycemic control during the first year after diabetes onset (16). The GRS was constructed from 11 T1D risk genes that we found to be expressed in human pancreatic islets, and whose expression changed upon stimulation with cytokines. We chose to focus strictly on the islet-expressed risk genes because we hypothesized that these would be the best predictors of islet (β-cell) function. We found that for each additional risk variant, i.e. for each unit increase in the GRS, a decreased β-cell function and a worsened glycemic control from 6 to 12 months after onset were observed, after controlling for the effect of age at diagnosis, sex and HLA risk groups. Further, we found that several of the genes used in the GRS interacted in a network suggesting that they may cooperate to regulate important processes within the β-cells. The results from these reviewed studies are summarized in Table 1.

Benefits for patients
Use of genetics in prediction models could lead to earlier prediction useful for immune-modulatory interventions to preserve residual β-cell mass and will be beneficial both in the pre-clinical phase and after diagnosis. Better stratification for fast and slow progressors both from autoantibody positivity to diabetes and disease progression after diagnosis would be a major achievement in diabetes care. Being able to foresee which genetically-predisposed individuals progress to T1D and these patients’ remaining β-cell function at time of diagnosis and first year(s) to come would have a tremendous impact on the individual patient’s health burden and quality of life, due to lowering the risk for hypo- and hyperglycemia and long-term complications.

Conclusion
In summary, profiling of selected genetic variants may hold promise to better predict T1D progression in risk individuals and residual β-cell function in new-onset type 1 diabetics. Such knowledge may in the future be exploited to offer personalized medicine to optimize treatment regimens to increase patient care and reduce severe long-term complications.

References
1. Barrett JC, Clayton DG, Concannon P, et al. Genome-wide association study and meta-analysis find that over 40 loci affect risk of type 1 diabetes. Nat Genet. 2009; 41: 703–707.
2. Onengut-Gumuscu S, Chen WM, Burren O, et al. Fine mapping of type 1 diabetes susceptibility loci and evidence for colocalization of causal variants with lymphoid gene enhancers. Nat Genet. 2015; 47: 381–386.
3. Achenbach P, Hummel M, Thumer L, et al. Characteristics of rapid vs slow progression to type 1 diabetes in multiple islet autoantibody-positive children. Diabetologia 2013; 56: 1615–1622.
4. Steffes MW, Sibley S, Jackson M, Thomas W. Beta-cell function and the development of diabetes-related complications in the diabetes control and complications trial. Diabetes Care 2003; 26: 832–836.
5. Panero F, Novelli G, Zucco C, et al. Fasting plasma C-peptide and micro- and macrovascular complications in a large clinic-based cohort of type 1 diabetic patients. Diabetes Care 2009; 32: 301–305.
6. Pinckney A, Rigby MR, Keyes-Elstein L, et al. Correlation among hypoglycemia, glycemic variability, and C-Peptide preservation after alefacept therapy in patients with type 1 diabetes mellitus: analysis of data from the immune tolerance network T1DAL trial. Clin Ther. 2016; doi: 10.1016/j.clinthera.2016.04.032. [Epub ahead of print].
7. Steck AK, Wong R, Wagner B, et al. Effects of non-HLA gene polymorphisms on development of islet autoimmunity and type 1 diabetes in a population with high-risk HLA-DR,DQ genotypes. Diabetes 2012; 61: 753–758.
8. Winkler C, Krumsiek J, Lempainen J, et al. Ziegler AG. A strategy for combining minor genetic susceptibility genes to improve prediction of disease in type 1 diabetes. Genes Immun. 2012; 13: 549–555.
9. Bonifacio E, Krumsiek J, Winkler C, et al. Ziegler AG. A strategy to find gene combinations that identify children who progress rapidly to type 1 diabetes after islet autoantibody seroconversion. Acta Diabetol. 2014; 51: 403–411.
10. Mortensen HB, Swift PG, Holl RW, et al. Multinational study in children and adolescents with newly diagnosed type 1 diabetes: association of age, ketoacidosis, HLA status, and autoantibodies on residual beta-cell function and glycemic control 12 months after diagnosis. Pediatr Diabetes 2010; 11: 218–226.
11. Nielsen LB, Mortensen HB, Chiarelli F, et al. Impact of IDDM2 on disease pathogenesis and progression in children with newly diagnosed type 1 diabetes: reduced insulin antibody titres and preserved beta cell function. Diabetologia 2006; 49: 71–74.
12. Nielsen LB, Porksen S, Andersen ML, et al. The PTPN22 C1858T gene variant is associated with proinsulin in new-onset type 1 diabetes. BMC Med Genet. 2011; 12: 41.
13. Petrone A, Spoletini M, Zampetti S, et al. The PTPN22 1858T gene variant in type 1 diabetes is associated with reduced residual beta-cell function and worse metabolic control. Diabetes Care 2008; 31: 1214–1218.
14. Floyel T, Brorsson C, Nielsen LB, et al. CTSH regulates beta-cell function and disease progression in newly diagnosed type 1 diabetes patients. Proc Natl Acad Sci. U S A 2014; 111: 10305–10310.
15. Kaur S, Mirza AH, Brorsson CA, et al. The genetic and regulatory architecture of ERBB3-type 1 diabetes susceptibility locus. Mol Cell Endocrinol. 2016; 419: 83–91.
16. Brorsson CA, Nielsen LB, Andersen ML, et al. Hvidoere Study Group On Childhood Diabetes. Genetic risk score modelling for disease progression in new-onset type 1 diabetes patients: increased genetic load of islet-expressed and cytokine-regulated candidate genes predicts poorer glycemic control. J Diabetes Res. 2016; 2016: 9570424.

The authors
Caroline A Brorsson*¹ PhD and Joachim Størling² PhD
¹Center for Biological Sequence Analysis, Department of Systems Biology, Technical University of Denmark, Lyngby, Denmark
²Copenhagen Diabetes Research Center, Pediatric Department E, University Hospital Herlev, Herlev, Denmark

*Corresponding author
E-mail: caroline@cbs.dtu.dk