Type 1 Diabetes Prevention with Early Insulin a Bust

[admin]

Insulin therapy doesn't prevent type 1 diabetes in genetically high-risk infants showing early signs of the autoimmune disease, researchers found.

In children with elevated genetic risk, nasal insulin started soon after detection of diabetes-associated autoantibodies had no effect on development of type 1 diabetes (hazard ratio 0.98, P=0.91), reported Kirsti Näntö-Salonen, M.D., of the University of Turku, and colleagues online in The Lancet.

Animal and pilot studies had suggested that insulin could reduce diabetes incidence, possibly by allowing the beta cells to rest and minimizing autoantigen production, the researchers said.

The results of the current findings agree with those for a similar strategy with oral or subcutaneous insulin in the Diabetes Prevention Trial 1, noted David B. Dunger, M.D., Ph.D., and John A. Todd, Ph.D., both of the University of Cambridge, England, in an accompanying commentary.

However, it's possible that the strategy worked but wasn't started early enough to reverse autoimmune processes that destroy beta cells, they said.

"Autoantibody seroconversion in the first one to three years of life may be a common prerequisite for development of type 1 diabetes," they wrote, "which would suggest that an early window of susceptibility exists, after which seroconversion and type 1 diabetes are much less likely."

The Type 1 Diabetes Prediction and Prevention study (DIPP) aimed to exploit this window of time to prevent diabetes.

The phase III randomized, double-blind trial included genetic screening of 116,720 infants consecutively born at three Finnish university hospitals and 3,430 of their siblings. Cord blood samples were screened for the human leukocyte antigen (HLA)-DQB1 susceptibility alleles for type 1 diabetes.

Children at elevated genetic risk were followed every three to 12 months for appearance of autoantibodies.

The 224 infants and 40 siblings positive for two or more autoantibodies were randomized to double-blind treatment with short-acting human insulin (Actrapid) at a dose of 1 unit/kg or placebo given once a day intranasally.

Primary endpoint was diagnosis of diabetes. Treatment continued for an average 1.8 years until the study was stopped early for lack of efficacy.

Adherence to the study drug was difficult to verify, the researchers noted, cautioning that "the data that can be derived represent the absolute minimum of drug use."

During the intervention phase, 42 children who had received insulin and 38 on placebo developed type 1 diabetes.

Insulin therapy appeared to make no difference in the age at diagnosis or time from appearance of autoantibodies to diagnosis.

The rate of progression to diabetes was 16.8% (95% confidence interval 11.7 to 21.9) per year in the insulin group and 15.3% (95% CI 10.5 to 20.2) per year in the placebo group.

Overall, insulin had no effect on the cumulative incidence of diabetes (hazard ratio 0.98, 95% CI 0.67 to 1.43, P=0.91).

Adjustment for number of autoantibodies, young age at seroconversion, and low first-phase insulin response did not change the results.

Notably, though, insulin tended to accelerate diabetes onset in the subgroup of children with more advanced autoimmunity who presented with three or four types of autoantibodies before treatment (HR 1.50, P=0.09).

Nasal irritation from treatment -- occasional nasal congestion, discharge, nosebleed, or unpleasant feeling in the nose after administration -- occurred at similar frequency in both groups.

However, nasal passages and rates of middle-ear and sinus infection appeared unaffected by active treatment.

The researchers cautioned that screening for susceptibility to diabetes presents ethical concerns because no prevention or cure for the disease is available.

But, screening did offer early diagnosis, easier and faster control of the disease, and protection from ketoacidosis for children who progressed to overt type 1 diabetes, they said.