Can we refine type 1 diabetes risk with non-hla genetic markers in children Pre-screened for high-risk HLA genotypes

Abstract

Class II HLA genotypes in combination with islet autoantibodies can predict type 1 diabetes (T1D) risk in first-degree relatives (FDR); however, accurate prediction in the general population is still lacking.The Diabetes Autoimmunity Study in the Young follows prospectively for development of T1D and persistent islet autoimmunity (IA) children at increased genetic risk. A total of 1709 non-Hispanic White participants were genotyped for 30 non-HLA single nucleotide polymorphisms.In multivariate analyses adjusting for presence of HLA-DR3/4,DQ8 and family history of T1D, PTPN22 (rs2476601) and 2 UBASH3A (rs11203203 and rs9976767) SNPs were associated with development of both IA (HR=1.99, 1.61and 1.65 respectively, all p<0.001) and T1D (HR=1.80, 1.84, 2.01 respectively, all p<0.02), while INS was associated with development of T1D only (HR=1.72, p=0.02). We performed survival analysis stratified by high and low genetic risk groups (high risk: either PTPN22 rs2476601 TT or HLA-DR3/4 and UBASH3A rs11203203 AA; low risk: all others). In the general population (N=843), the risk of T1D by age 15 reached 45% for high risk compared to 3% for low risk group (p<0.0001) (Figure). Addition of non-HLA markers to HLA-DR3/4,DQ8 did not improve T1D prediction in FDR. The other non-HLA markers tested did not predict development of IA or T1D.A T1D genetic risk score including HLA class II, rs2476601 and rs11203203 might be applicable to general newborn screening. Follow up of these children at high genetic risk with autoantibody testing may prove to be valuable screening for prevention trials. (Figure presented) .