# Antiphospholipid Antibody Testing in a Maximum Care Hospital: Method-Dependent Differences

## TL;DR

This study compares ELISA and CLIA methods for antiphospholipid antibody testing and finds significant differences in results that could affect antiphospholipid syndrome classification.

## Contribution

The study reveals method-dependent discrepancies in aPL testing outcomes in a real-world hospital setting, impacting APS classification.

## Key findings

- ELISA and CLIA assays showed poor agreement for aCL IgG/IgM and aß2GPI IgM in APS patients.
- Antibody levels in APS patients were significantly higher using CLIA compared to ELISA.
- Applying the new classification criteria threshold reduced the number of aPL-positive specimens.

## Abstract

Background: Antiphospholipid antibody (aPL) testing is critical for the classification of antiphospholipid syndrome. The 2023 ACR/EULAR classification criteria recommend the use of enzyme-linked immunosorbent assays (ELISAs) and specific thresholds for aPL positivity. Since non-ELISA methods are increasingly used, we compared and evaluated ELISA and non-ELISA aPL assays in a real-world maximum care hospital setting. Methods: Between January 2021 and June 2024, anticardiolipin (aCL; IgG and IgM) and anti-beta2 glycoprotein I (aß2GPI; IgG and IgM) antibodies were measured using ELISA (n = 5115) and a chemiluminescence-based automated immunoassay (CLIA) (n = 3820). Results of parallel testing were compared, and associations with clinical and laboratory characteristics were evaluated. Results: A total of 946 samples were tested using ELISA and CLIA in parallel. A total of 136 (14%) specimens were positive for at least one aPL, and 55 (6%) specimens were from patients diagnosed with APS. Among the latter, 47 (85%) and 41 (75%) patients were positive when ELISA- or CLIA-based aPL assays were used, respectively. After applying the >40 units threshold of the new classification criteria, the number of aPL-positive specimens was significantly lower. In the entire cohort, the agreement between ELISA and CLIA aPL assays was acceptable only for aß2GPI IgG; the results from the two methods did not agree for aCL IgG/IgM and aß2GPI IgM. In APS patients, the agreement between ELISA and CLIA aPL assays was acceptable for aß2GPI IgG and IgM but poor for aCL IgG and IgM. Antibody levels in APS patients were significantly higher using CLIA compared to ELISA. Conclusions: The method-dependent discrepancies between ELISA- and CLIA-based aPL assays regarding the quantitative and qualitative results are substantial. Both methods are suitable for APS classification, but the choice of aPL assay may influence the classification, and therefore, aPL results should be interpreted carefully in the clinical context.