Interlaboratory Evaluation of Multiplex Autoantibody Assay Performance in the Islet Autoantibody Standardization Program 2024 Workshop.

OBJECTIVE

The Islet Autoantibody Standardization Program (IASP) evaluates type 1 diabetes autoantibody assays through international interlaboratory comparison studies. This report describes 2024 IASP workshop results for multiplex assays that simultaneously measure multiple islet autoantibodies, increasingly adopted for population screening programs.

RESEARCH DESIGN AND

METHODS

Participating laboratories analyzed coded sera from individuals with new-onset type 1 diabetes (n = 50), blood donors with no diabetes (n = 98), and replicate samples (n = 3). Performance was assessed using sensitivity, specificity, adjusted sensitivity at 95%, 99%, and 100% specificity, area under the receiver operating characteristic curve analysis (ROC-AUC), and partial ROC-AUC at 95% specificity (pAUC95).

RESULTS

Fifteen laboratories contributed results from 19 multiplex assays using five formats: antibody-dependent agglutination PCR (ADAP) (n = 6), bridge-ELISA (n = 8), electrochemiluminescence (n = 2), luciferase immunoprecipitation system (LIPS) (n = 1), and protein A luciferase/solid-phase capture LIPS (n = 2). Median ROC-AUC was 0.98 across all platforms, with a pAUC95 of 0.048 against a theoretical maximum of 0.05. Laboratory-reported sensitivity ranged from 86% to 96% with specificity 81% to 100%. Standardized adjusted sensitivity at 99% specificity thresholds eliminated most false positives without compromising sensitivity, achieving 93-97% adjusted sensitivity. Bridge-ELISA platforms demonstrated potential for universal cutoff standardization, while ADAP showed substantial interlaboratory threshold variation.

CONCLUSIONS

Multiplex assays demonstrated good overall performance for detecting islet autoantibodies and discriminating case subjects from control subjects. However, interlaboratory variability and platform-specific recognition patterns necessitate further standardization efforts. While not directly tested in this workshop, the results suggest that sequential testing strategies using complementary platforms may be necessary to achieve the stringent specificity required when these assays are used in a screening context.