A Type 1 diabetes genetic risk score can discriminate monogenic autoimmunity with diabetes from early onset clustering of polygenic autoimmunity with diabetes

Abstract

Aims/hypothesis Identifying individuals suitable for monogenic autoimmunity testing and gene discovery studies is challenging: early-onset type 1 diabetes mellitus (T1D) can cluster with additional autoimmune diseases due to shared polygenic risk and islet and other organ specific autoantibodies are present in both monogenic and polygenic aetiologies. We aimed to assess if a type 1 diabetes genetic risk score (T1D-GRS) could identify monogenic autoimmune diabetes and be useful to prioritise individuals for gene discovery studies. Methods We studied 79 individuals with diabetes and at least 1 additional autoimmune disease diagnosed before 5 years. We screened all participants for the 7 genes known to cause monogenic autoimmunity that can include diabetes (AIRE, IL2RA, FOXP3, LRBA, STAT1, STAT3, STAT5b). We genotyped the top 10 risk alleles for type 1 diabetes, including HLA and non-HLA loci, to generate a T1D-GRS. Results 47% (37/79) of individuals had mutations in the monogenic autoimmunity genes. The T1D-GRS was lower in these individuals compared to those without mutations in these genes (median 9th vs. 49th centile of type 1 diabetes controls, p<0.0001). Age of diabetes diagnosis and T1D-GRS combined to be highly discriminatory of monogenic autoimmunity (ROC-AUC: 0.88). Most individuals without a mutation in a known gene had a high T1D-GRS, suggesting they have polygenic clustering of type 1 diabetes and additional autoimmunity and should not be included in gene discovery studies. Conclusions We have shown that the T1D-GRS can identify individuals likely to have monogenic autoimmunity helping both diagnostic testing and novel monogenic autoimmunity gene discovery. Individuals with monogenic autoimmunity have a different clinical course to those with polygenic type 1 diabetes and can respond well to therapies targeting the underlying genetic defect.