Sensitive detection of multiple islet autoantibodies in type 1 diabetes using small sample volumes by agglutination-PCR
Fig 2
Clinical performance of ADAP assay in an evaluation cohort of T1D and healthy control.
(A) The ROC curves of ADAP showed AUC of 0.94 (95%CI: 0.89–0.99), 0.82 (95%CI: 0.71–0.93) and 0.95 (95%CI: 0.90–1.00) for GAD, IA-2 and insulin antibodies/autoantibodies respectively. The samples were also analyzed by radioassay and showed corresponding AUC of 0.92 (95%CI: 0.85–0.99), 0.80 (95%CI: 0.68–0.92) and 0.95 (95%CI: 0.88–1.00). (B) Comparison plots of ADAP and radioassay signals. The x-axis displays radioassay signals in logarithm scales. The y-axis shows ADAP signal in ΔCt. The use of logarithm was necessary as ΔCt is a logarithmic parameter. (For instance, consider a sample of ΔCt value 2 and another sample of ΔCt of 4, their amplicon quantities differ by 4 fold (24/22) rather than 2 fold). The horizontal and the vertical dash lines denote ADAP and radioassay cutoff thresholds respectively. T1D sample data is shown in blue circle, whereas health serum signal is shown in red square. A total of 30 T1D and 39 control was analyzed without blinding.