Johannes Schulte-Pelkum, Germany
Phadia GmbH R&DPresenter of 1 Presentation
PERFORMANCE OF A NOVEL FULLY AUTOMATED IMMUNOASSAY FOR THE DIAGNOSIS OF ANTI-GAD65 AUTOANTIBODIES IN AUTOIMMMUNE TYPE 1 DIABETES
Abstract
Background and Aims
The measurement of anti Glutamic acid decarboxylase autoantibodies (anti-GAD65 AAb) is a hallmark in the serological diagnosis of type 1 diabetes (T1D). Until nowadays many of these tests have to be performed manually. A special assay architecture has to be chosen to yield highest possible assay characteristics. Here we report a fully automated assay for the measurement of anti-GAD65 AAb.
Methods
The study involved 205 serum samples (clinically-defined type1 diabetes-patient samples n=100; Clinically-defined type2 diabetes-patient samples n=55; blood donor samples n=50. All sera were measured with a well established commercial ELISA (RSR ltd.) and by EliA anti-GAD65 (research use only, Thermo Fisher Scientific, Freiburg, Germany). Data analysis was performed using Analyse-IT for MS Excel.
Results
High agreement of the results was observed between the predicate ELISA and the novel EliA GAD65. Positive and negative agreement were found at 99% (95% CI=0.931-0.999). When comparing the quantitative results, a high correlation of results could be observed (r= 0949; 95% CI=0.934-0.961). Comparative clinical ROC analysis showed an AUC of 0.941 (95% CI=0.910-0.972) for the EliA GAD 65, and an AUC of 0.932 (95% CI=0.896-0.969). We thus conclude that the novel fully automated assay is comparable to its predicate device.
Conclusions
We found a high agreement of the novel Assay and the predicate device. The novel EliA GAD 65 showed high clinical sensitivity, comparable to the state of the art ELISA assay. We successfully showed a fully automated alternative to labour intensive manual ELISA for the serological identification of autoimmune diabetes type 1.