A model-based approach to sample size estimation in recent onset type 1 diabetes.

Area of Science:

• Endocrinology
• Immunology
• Metabolic Diseases

Background:

• Type 1 diabetes is an autoimmune disease characterized by the destruction of pancreatic beta cells.
• C-peptide levels are a marker of endogenous insulin production and beta cell function.
• TrialNet studies have provided valuable data on recent-onset type 1 diabetes.

Purpose of the Study:

• To model the rate of C-peptide loss in individuals with recent-onset type 1 diabetes.
• To identify predictors of C-peptide decline and reduce variance in estimates.
• To optimize sample size calculations for future clinical trials in type 1 diabetes.

Main Methods:

• Area under the curve (AUC) C-peptide levels from a 2-hour mixed meal tolerance test were analyzed.
• Data from 498 individuals across five prior TrialNet studies were used.
• Analysis included modeling C-peptide loss from baseline to 12 months post-enrollment.

Main Results:

• Age at diagnosis and baseline C-peptide levels were significant predictors of C-peptide loss.
• Adjusting for these factors using ANCOVA (Analysis of Covariance) reduced the variance of the estimates.
• The developed model allows for nearly a 50% reduction in target sample size for new studies.

Conclusions:

• The study provides a robust method for estimating C-peptide decline in type 1 diabetes.
• The findings enable more efficient clinical trial design by reducing required sample sizes.
• The model's expected C-peptide values can aid in assessing treatment efficacy in ongoing trials.