A multivariable probability score integrating routine immunoassay results for the immunological assessment of...

Accurate detection of anti-aminoacyl-tRNA synthetase autoantibodies (ASA) is central to the diagnosis of antisynthetase syndrome (ASyS). Although immunoprecipitation (IP) remains the historical reference method, its limited availability in routine clinical laboratories necessitates reliance on commercially available immunoassays. The diagnostic performance of these assays varies by specificity and cut-off interpretation. We aimed to improve the post-analytical interpretation of ASA positivity in routine practice.

We conducted a retrospective single-center study including 125 patients with at least one positive ASA detected by line-blot immunoassay (LIA) between January 2023 and September 2025. Thirty-three patients fulfilled clinician-based diagnostic criteria for ASyS, while 92 served as controls. Additional immunological data included indirect immunofluorescence on HEp-2 cells (IIF-HEp-2) and anti-Ro52 autoantibodies. Variables associated with ASyS were incorporated into multivariable logistic regression models. A diagnostic probability score was derived from the final model and evaluated using receiver operating characteristic (ROC) curve analysis.

Semi-quantitative LIA intensity alone showed limited discriminatory capacity, particularly in cases with low-intensity bands. In contrast, compatible cytoplasmic IIF-HEp-2 patterns and anti-Ro52 co-positivity significantly improved discrimination. The final multivariable model integrating semi-quantitative LIA intensity, IIF-HEp-2 cytoplasmic pattern (AC-19/20), and anti-Ro52 co-positivity achieved an AUC of 0.94 (95% CI 0.89-0.99). Stratification into low-, intermediate-, and high-probability zones provided clinically interpretable diagnostic categories.

In ASA-positive patients, probability-based integration of routinely available immunoassay results improves the interpretation of LIA positivity and enhances discrimination between ASyS and alternative diagnoses.