Antigen-based therapy with glutamic acid decarboxylase (GAD) vaccine in patients with recent-onset type 1 diabetes: a randomised double-blind trial

Abstract

Background: Glutamic acid decarboxylase (GAD) is a major target of the autoimmune response that occurs in type 1 diabetes mellitus. In animal models of autoimmunity, treatment with a target antigen can modulate aggressive autoimmunity. We aimed to assess whether immunisation with GAD formulated with aluminum hydroxide (GAD-alum) would preserve insulin production in recent-onset type 1 diabetes.

Methods: Patients aged 3-45 years who had been diagnosed with type 1 diabetes for less than 100 days were enrolled from 15 sites in the USA and Canada, and randomly assigned to receive one of three treatments: three injections of 20 μg GAD-alum, two injections of 20 μg GAD-alum and one of alum, or 3 injections of alum. Injections were given subcutaneously at baseline, 4 weeks later, and 8 weeks after the second injection. The randomisation sequence was computer generated at the TrialNet coordinating centre. Patients and study personnel were masked to treatment assignment. The primary outcome was the baseline-adjusted geometric mean area under the curve (AUC) of serum C-peptide during the first 2 h of a 4-h mixed meal tolerance test at 1 year. Secondary outcomes included changes in glycated haemoglobin A(1c) (HbA(1c)) and insulin dose, and safety. Analysis included all randomised patients with known measurements. This trial is registered with ClinicalTrials.gov, number NCT00529399.

Findings: 145 patients were enrolled and treated with GAD-alum (n=48), GAD-alum plus alum (n=49), or alum (n=48). At 1 year, the 2-h AUC of C-peptide, adjusted for age, sex, and baseline C-peptide value, was 0·412 nmol/L (95% CI 0·349-0·478) in the GAD-alum group, 0·382 nmol/L (0·322-0·446) in the GAD-alum plus alum group, and 0·413 nmol/L (0·351-0·477) in the alum group. The ratio of the population mean of the adjusted geometric mean 2-h AUC of C-peptide was 0·998 (95% CI 0·779-1·22; p=0·98) for GAD-alum versus alum, and 0·926 (0·720-1·13; p=0·50) for GAD-alum plus alum versus alum. HbA(1c), insulin use, and the occurrence and severity of adverse events did not differ between groups.

Interpretation: Antigen-based immunotherapy therapy with two or three doses of subcutaneous GAD-alum across 4-12 weeks does not alter the course of loss of insulin secretion during 1 year in patients with recently diagnosed type 1 diabetes. Although antigen-based therapy is a highly desirable treatment and is effective in animal models, translation to human autoimmune disease remains a challenge.

Funding: US National Institutes of Health.

Figure 1

(A): Enrollment, Randomization, and Follow-up of Study Participants. (B): Rate of enrollment over time. Indicated is the point at which approval was granted to change the lower age for eligibility of enrollment from 16 years to 3 years.

Figure 1

(A): Enrollment, Randomization, and Follow-up of Study Participants. (B): Rate of enrollment over time. Indicated is the point at which approval was granted to change the lower age for eligibility of enrollment from 16 years to 3 years.

Figure 2

(A): The population mean (and 95% confidence intervals) of stimulated C-peptide 2-hour AUC mean (expressed as nmol/L) over time for each treatment group. The estimates are from the analysis of covariance model adjusting for age, gender, baseline value of C-peptide, and treatment assignment. Y-axis is on a log(y + 1) scale. (B): The fitted line representing the population mean of stimulated C-peptide 2-hour AUC mean over time for each treatment group. The estimates are from the analysis of mixed effects model adjusting for age, gender, baseline value of C-peptide, and treatment assignment, and including a fixed effect for time as a linear line on the log(y + 1) scale. (C): The proportion of subjects with 2 hour peak C-peptide remaining at or above 0.2 nmol/L over time for each treatment group.

Figure 2

(A): The population mean (and 95% confidence intervals) of stimulated C-peptide 2-hour AUC mean (expressed as nmol/L) over time for each treatment group. The estimates are from the analysis of covariance model adjusting for age, gender, baseline value of C-peptide, and treatment assignment. Y-axis is on a log(y + 1) scale. (B): The fitted line representing the population mean of stimulated C-peptide 2-hour AUC mean over time for each treatment group. The estimates are from the analysis of mixed effects model adjusting for age, gender, baseline value of C-peptide, and treatment assignment, and including a fixed effect for time as a linear line on the log(y + 1) scale. (C): The proportion of subjects with 2 hour peak C-peptide remaining at or above 0.2 nmol/L over time for each treatment group.

Figure 3

(A): The population mean of HbA1c over time for each treatment group. The estimates are from the analysis of covariance model adjusting for age, gender, baseline value of HbA1c, and treatment assignment. (B): The population mean of insulin use per kilogram of body weight over time (in 3 month interval) for each treatment group. The estimates are from the analysis of covariance model adjusting for age, gender, baseline value of HbA1c, and treatment assignment.

Figure 4

(A): The ratio (GAD-alum ×3 to alum ×3) of treatment effect on 1 year stimulated C-peptide AUC mean within categories of pre-specified baseline factors. The estimates are from the analysis of covariance modeling log of C-peptide adjusting for age, gender, baseline value of C-peptide, the indicated categorized factor, treatment assignment, and treatment interaction terms. The homogeneity test of treatment effect was significant for DR3 allele status (p = 0.04). (B): The ratio (GAD-alum ×2 to alum ×3) of treatment effect on 1 year stimulated C-peptide AUC mean within categories of pre-specified baseline factors. The estimates are from the analysis of covariance modeling log of C-peptide adjusting for age, gender, baseline value of C-peptide, the indicated categorized factor, treatment assignment, and treatment interaction terms.