According to Johnny Ludvigsson, MD, PhD, from Linköping University in Linköping, Sweden, and colleagues, patients given the autoantigen glutamic acid decarboxylase (GAD) had similar declines in insulin secretion as measured by stimulated C-peptide levels as did patients given placebo, with an estimated treatment ratio of 1.16 (95% CI 0.97 to 1.40, P=0.10). Males, however, had a 41% greater preservation of C-peptide, with an estimated treatment ratio of 1.41 (95% CI 1.09 to 1.83, P=0.01), the researchers reported.
Dr. Ludvigsson stated that, "Instead of saying that this study shows that we should abandon GAD and autoantigen treatment, I think the result supports the concept that GAD can be used to preserve residual insulin secretion, although we need to learn more about dosing, intervals, and which groups of patients will have the best effect."
The use of autoantigens in type 1 diabetes is based on the hypothesis that induction of immune tolerance could stop the autoimmune process of pancreatic beta-cell destruction.
In a previous study that included 70 patients, Ludvigsson's group found that treatment with the 65-kD isoform of GAD preserved fasting C-peptide levels for more than two years and stimulated levels for 15 months, indicating residual insulin secretion. So they conducted a nine-country randomized trial that enrolled 327 patients ages 10 to 20 within three months of being diagnosed with type 1 diabetes.
Patients were randomized to receive subcutaneous injections of 20 µg GAD65 formulated with alum at baseline, one month, three months, and nine months; injections at one and three months plus two placebo injections; or four injections of placebo.
Patients in the two active treatment groups showed increases in levels of GAD autoantibodies by three months that were significantly higher than in the placebo group (P<0.001).
As with the primary endpoint of stimulated C-peptide levels at 15 months, there were no significant differences compared with placebo on various secondary endpoints, including change in fasting C-peptide (P=0.07), change in daily insulin dose (P=0.57), and change in glycated hemoglobin (P=0.86).
However, the investigators also performed multiple exploratory analyses to identify potential treatment effects in patient subgroups, and found differences not only according to sex, but also for insulin requirements, degree of maturation, and location.
Among patients whose baseline insulin dose was 0.398 to 0.605 IU/kg per day, the estimated treatment ratio was 1.33 (95% CI 1 to 1.78, P=0.05). For those whose baseline Tanner stage was 2 or 3, the ratio in the four-dose group was 1.52 (P=0.04), and the ratio for the more southern countries such as Italy and Spain was 1.36 (P=0.04), the researchers reported.
Adverse events were similar in the treatment and placebo groups, and there were no cases of stiff person syndrome, a neurologic disorder associated with high GAD65 levels.
In discussing why the results of this study contradicted those seen in the earlier trial, the investigators explained that there may have been differences in the study population or in participating clinicians' traditional treatment approaches.
Seasonal factors also may have contributed, because patients who began treatment in the spring seemed to fare better, and an outbreak of influenza with widespread immunization also may have influenced the results. In addition, the treatment groups included more patients ages 10 and 11, while the placebo group included greater numbers of 16-to-20 year olds.
Ludvigsson added that, "Nobody would expect to use one single drug in one single type of dosing for treatment of all types of cancer in all ages, so why would we believe that all type 1 diabetes in all ages, genders, and countries should be treated with just one design?"
"Much as treatments for childhood cancer and immunotherapy of allergy have developed in a stepwise, gradual manner through the combination of existing therapies, treatment for type 1 diabetes will most likely be based on the knowledge gained from this and other studies, as well as future studies, of single agents and combination therapies for both intervention and prevention," he and his co-investigators concluded.
- This multicenter study found no effect of the autoantigen glutamic acid decarboxylase on insulin secretion at 15 months in type 1 diabetics enrolled within three months of diagnosis.
- Insulin secretion does appear better preserved in certain subgroups, such as males, younger children, those living in more southern climates, and those taking lower baseline doses of insulin.
Ludvigsson J, et al "GAD65 antigen therapy in recently diagnosed type 1 diabetes mellitus" New Engl J Med 2012; 366: 433-442.