We thank Dr. Peters (1) for his thoughtful comments regarding our study (2). Dr. Peters suggests that etanercept may effect the partial remission period of type 1 diabetes by mediating declines in autoantibodies, GAD-65 in particular. Clinical data supports the concept that titer and number of autoantibodies correlates with accelerated β-cell demise and a lower likelihood of clinical remission compared with individuals who tested negative for diabetes-associated autoantibodies (3). Changes in GAD-65 antibody levels were not an outcome measure for our study. However, reported herein are the data in subjects whose GAD-65 antibody levels were measured at baseline, week 24 (end of treatment period), and week +12 (washout visit). Based on reported interassay coefficients of variation (3), GAD-65 levels were defined as changed if the value varied 10% or more from baseline. A trend toward declining GAD-65 levels in the etanercept versus placebo group (66% vs. 33%, n = 6 for each group) was observed at week 24, but not at week +12. While this observation supports Dr. Peters' hypothesis, the small sample size limits statistical analysis.
We would argue that potential changes in GAD-65 autoantibody levels do not preclude the hypothesis that the primary mechanism of action of etanercept in improving β-cell function is through tumor necrosis factor-α (TNF-α) blockade. In fact, TNF-α blockade is proposed as the rationale for improved diabetes control in patients receiving adalimumab for the treatment of rheumatoid arthritis (4). The role of TNF-α in the development and progression of type 1 diabetes is multifactorial (5). TNF-α can regulate tolerance of T-cells to β-cell autoantigens, including GAD-65. Therefore, any decline in GAD-65 levels seen in etanercept-treated subjects may be the result of decreased GAD-65 autoantibody secreting T-cells mediated by TNF-α blockade. While this is speculative and cannot be addressed by our study, future immunomodulatory trials may examine the effects of these therapies on immune cells that contribute to insulin-producing β-cell demise.
Acknowledgments
The funding and study drug were provided by Immunex/Amgen.
No other potential conflicts of interest relevant to this article were reported.
