Stratifying the rate of disease progression by Progression Likelihood
Scores in children and adolescents with stage 1 and stage 2 type 1
diabetes in Germany.
OBJECTIVE: To stratify the progression rate to clinical stage 3 type 1 diabetes in children with early-stage disease. RESEARCH DESIGN AND METHODS: The Fr1da study tested 211,464 children aged 1.75-10 years for islet autoantibodies. Children with early-stage type 1 diabetes were classified as stage 1 or stage 2 by oral glucose tolerance test (OGTT) and hemoglobin A1c (HbA1c) using current American Diabetes Association criteria and were followed 3-6 months. We applied our previously developed progression likelihood score (PLS), a composite of HbA1c, 90-min OGTT glucose, and islet antigen 2 antibodies (IA-2A) titer, and developed a non-OGTT-based score using multivariable Cox proportional hazards models to stratify progression rates to stage 3. RESULTS: Of 485 children who participated in staging, 360 (74.2%) were diagnosed with stage 1. Of these, stage 3 developed in 105 (median follow-up 3.3 years). PLS stratified the 2-year risk for stage 3 from 43.7% (95% CI 24.3-58.1) in children with high PLS to 4.7% (1.7-7.7) and 0% in those with intermediate or low PLS. Adding the variable obesity improved the existing model. In children with stage 2 with a single dysglycemic abnormality, PLS could stratify 2-year risk for stage 3 from 42.4% (95% CI 22.8-57.0) to 5.6% (0.0-15.6). A non-OGTT-based score based on IA-2A titer categories, HbA1c, obesity, and autoantibody positivity for IA-2 juxtamembrane epitopes could identify individuals with low (1.7%) and moderate (24.6%) 2-year risk. CONCLUSIONS: The PLS and a novel non-OGTT-based score can stratify the short- to medium-term progression rates to stage 3 and should be considered for guiding monitoring practices and clinical trial eligibility.