Minimum effective low dose of antithymocyte globulin in people aged 5–25 years with recent-onset stage 3 type 1 diabetes (MELD-ATG): a phase 2, multicentre, double-blind, randomised, placebo-controlled, adaptive dose-ranging trial

Background: Type 1 diabetes remains an important health-care problem, with no disease-modifying therapies available in people with recent-onset, clinical type 1 diabetes. Adaptive trial designs, allowing faster evaluation of treatment modalities, remain underexplored in this stage of the disease. We aimed to identify the minimum effective dose of antithymocyte globulin (ATG) in people aged 5–25 years with recent-onset, clinical type 1 diabetes. Methods: MELD-ATG was a phase 2, double-blind, randomised, placebo-controlled, multi-arm, adaptive dose-ranging, parallel-cohort trial done in 14 accredited trial centres in eight countries (the UK, Denmark, Germany, Finland, Italy, Belgium, Austria, and Slovenia). Participants aged 5–25 years, diagnosed with clinical, stage 3 type 1 diabetes 3–9 weeks before treatment, with random C-peptide concentrations 0·2 nmol/L or more and at least one diabetes-related autoantibody (GADA, IA-2A, or ZnT8) were randomly assigned by a web-based randomisation system into seven consecutive cohorts receiving placebo, 2·5 mg/kg ATG, 1·5 mg/kg ATG, 0·5 mg/kg ATG, or 0·1 mg/kg ATG. Participants in cohort 1 were randomly assigned 1:1:1:1:1, participants in cohorts 2 and 3 were randomly assigned 1:1:1:1, and participants in cohorts 4–7 were randomly assigned 1:1:1. All cohorts included one placebo group and one 2·5 mg/kg ATG group. The other groups were assigned to ATG doses that were determined based on accruing data and the decision of the dose determining committee. The trial cohorts were stratified by age group (5–9 years, 10–17 years, and 18–25 years) with block sizes varying by cohort. Concealment lists, outlining the treatment allocation, were only available for the pharmacists; participants and study teams were masked to treatment allocation. ATG was administered by an intravenous infusion over 2 consecutive days. The primary outcome was the area under the curve (AUC) of the stimulated C-peptide concentration during a 2-h mixed-meal tolerance test at 12 months measured as ln(AUC C-peptide + 1). Conditional on finding a statistically significant difference at p<0·05 for 2·5 mg/kg ATG versus placebo, the minimum effective dose of ATG was determined. All randomly assigned participants were included in the primary analysis. All participants who received the study drug were included in the safety analysis. The trial was registered at ClinicalTrials.gov (NCT04509791) and is completed. Findings: Between Nov 24, 2020, and Dec 13, 2023, 152 people were recruited and screened, 117 of whom were randomly assigned (placebo n=31, 0·1 mg/kg ATG n=6, 0·5 mg/kg ATG n=35, 1·5 mg/kg ATG n=12, and 2·5 mg/kg n=33). 54 (46%) of 117 participants were male and 63 (54%) were female. Participants were mainly European. The 0·1 mg/kg dose and the 1·5 mg/kg dose were progressively dropped from the study. At 12 months, the mean ln(AUC C-peptide + 1) was 0·411 nmol/L per min (SD 0·032) in the placebo group and 0·535 nmol/L per min (0·032) in the 2·5 mg/kg ATG group. The mean difference in the ln(AUC C-peptide + 1) between 2·5 mg/kg ATG and placebo was 0·124 nmol/L per min (95% CI 0·043–0·205; p=0·0028). At 12 months, the mean ln(AUC C-peptide + 1) in the 0·5 mg/kg ATG group, the remaining middle dose, was 0·513 nmol/L per min (SD 0·032), with a mean baseline-adjusted difference from placebo of 0·102 nmol/L per min (95% CI 0·021–0·183; p=0·014). Cytokine release syndrome occurred in 11 (33%) of 33 participants in the 2·5 mg/kg ATG group, eight (24%) of 34 in the 0·5mg/kg ATG group, and no participants in the placebo group. Serum sickness occurred in 27 (82%) participants in the 2·5 mg/kg ATG group, 11 (32%) in the 0·5 mg/kg ATG group, and no participants in the placebo group. Ther