Accurate detection of islet autoantibodies is central to identifying individuals at risk for type 1 diabetes mellitus (T1DM), especially in population screening settings. Multiplex assays that measure multiple autoantibodies in a single test are increasingly used, but their performance across platforms requires standard evaluation.

An international interlaboratory comparison conducted under the Islet Autoantibody Standardization Program (IASP) assessed multiplex assay performance in 2024. The analysis included 15 laboratories that tested coded serum samples from 50 individuals with new-onset T1DM, 98 blood donors without diabetes, and three replicate samples. A total of 19 multiplex assays were evaluated using five formats, including antibody-dependent agglutination PCR (ADAP), bridge enzyme-linked immunosorbent assay (ELISA), electrochemiluminescence, luciferase immunoprecipitation system (LIPS), and protein A luciferase-based LIPS. Performance metrics included sensitivity, specificity, receiver operating characteristic area under the curve (ROC-AUC), and partial ROC-AUC at 95% specificity (pAUC95).

The analysis showed a median ROC-AUC of 0.98 across platforms, with a pAUC95 of 0.048 compared with a theoretical maximum of 0.05. Reported sensitivity ranged from 86% to 96%, while specificity ranged from 81% to 100%. Adjusting thresholds to 99% specificity reduced false-positive results while maintaining sensitivity between 93% and 97%. Bridge-ELISA platforms showed consistent cutoff performance, while antibody-dependent agglutination PCR assays showed variation across laboratories.

These findings indicate that multiplex assays demonstrate high accuracy in detecting islet autoantibodies, with differences across platforms and thresholds observed in interlaboratory comparisons.