Antiphospholipid antibody solid phase–based assays: problems and proposed solutions for the 2023 ACR/EULAR classification criteria for antiphospholipid syndrome

The 2023 ACR/EULAR antiphospholipid syndrome (APS) classification criteria have recently been published [1Barbhaiya M. Zuily S. Naden R. Hendry A. Manneville F. Amigo M.C. Amoura Z. Andrade D. Andreoli L. Artim-Esen B. Atsumi T. Avcin T. Belmont H.M. Bertolaccini M.L. Branch D.W. Carvalheiras G. Casini A. Cervera R. Cohen H. Costedoat-Chalumeau N. et al.2023 ACR/EULAR antiphospholipid syndrome classification criteria.Ann Rheum Dis. 2023; 82: 1258-1270Crossref PubMed Scopus (54) Google Scholar], replacing the prior criteria established in 2006 [2Miyakis S. Lockshin M.D. Atsumi T. Branch D.W. Brey R.L. Cervera R. Derksen R.H. DE Groot P.G. Koike T. Meroni P.L. Reber G. Shoenfeld Y. Tincani A. Vlachoyiannopoulos P.G. Krilis S.A. International consensus statement on an update of the classification criteria for definite antiphospholipid syndrome (APS).J Thromb Haemost. 2006; 4: 295-306Abstract Full Text Full Text PDF PubMed Scopus (5675) Google Scholar]. Designed primarily for application in observational clinical studies and trials, these new criteria encompass both clinical and laboratory domains. Within the laboratory domain, specific guidelines have been outlined for the "antiphospholipid antibodies (aPL) test using solid phase–based assays." The updated criteria specify the use of standardized ELISA methods for anticardiolipin IgG and IgM and anti–beta-2-glycoprotein-1 IgG and IgM, with defined thresholds for moderate positivity (40-79 units) and high positivity (≥80 units), while explicitly excluding alternative test modalities, such as emerging automated measurement platforms. From a laboratory analytical perspective, we have concerns regarding these recently updated criteria, which we substantiate with data derived from the ECAT external quality assessment (EQA) program on aPL. Firstly, the selection of fixed cut-off values is not fully supported by the existing literature. In the referenced literature [3Vandevelde A. Chayoua W. de Laat B. Gris J.C. Moore G.W. Musial J. Zuily S. Wahl D. Devreese K.M.J. Semiquantitative interpretation of anticardiolipin and antiβ2glycoprotein I antibodies measured with various analytical platforms: communication from the ISTH SSC Subcommittee on Lupus Anticoagulant/Antiphospholipid Antibodies.J Thromb Haemost. 2022; 20: 508-524Abstract Full Text Full Text PDF PubMed Scopus (24) Google Scholar], only one single ELISA assay is compared with automated platforms, where a mutual comparison between different ELISA assays is not included. Secondly, the results from our external quality control data reveal that the various available ELISA assays are noncomparable (see Table): our data demonstrate that different results can be obtained with the same sample depending on the choice of ELISA method, challenging the defensibility of a "fixed" cut-off. Additionally, the new criteria make a distinction between moderate and high positivity (with fixed cut-off values of 40-79 units and ≥80 units, respectively). This distinction is problematic given the observed variations within ELISA systems. In all ECAT surveys reported in the Table, large ranges of outcomes were obtained with a single sample among laboratories using the same ELISA assay, possibly yielding negative results in one laboratory and moderate or even high-level results in another. The absence of an internationally accepted laboratory standard for various aPL tests in solid phase assays possibly contributes to this issue. Thirdly, the preference for ELISA assays with fixed cut-off limits assumes that the results remain consistent across different ELISA assays, yielding comparable outcomes for the same sample. Our findings demonstrate that this is not the case, with large differences between suppliers. Therefore, the use of semiquantitative classification (moderate or high positivity) of solid phase aPL ranges based on fixed antibody titers is not supported by data. Similar concerns on the use of fixed cut-off limits and variability between ELISA systems were reported by others [4Favaloro E.J. Pasalic L. Lippi G. Classification criteria for the antiphospholipid syndrome: not the same as diagnostic criteria for antiphospholipid syndrome.Semin Thromb Hemost. 2023; https://doi.org/10.1055/s-0043-1776318Crossref Scopus (5) Google Scholar]. It seems more prudent to employ the 99th percentile instead, as previously recommended by the ISTH [5Devreese K.M.J. Pierangeli S.S. de Laat B. Tripodi A. Atsumi T. Ortel T.L. Subcommittee on Lupus Anticoagulant/Phospholipid/Dependent Antibodies. Testing for antiphospholipid antibodies with solid phase assays: guidance from the SSC of the ISTH.J Thromb Haemost. 2014; 12: 792-795Abstract Full Text Full Text PDF PubMed Scopus (174) Google Scholar], which may vary depending on the ELISA supplier and should be determined locally.TableMean results of anticardiolipin IgG and anti–beta-2-glycoprotein-1 IgG of ECAT EQA participants.Anticardiolipin IgGSurveySampleELISAAutomated platformsEuroimmunOrgentec (Alegria)Orgentec (ELISA)Werfen Inova Quanta liteIL ACUstar/INOVA Quanta flashThermo Scientific EliAU/mL (GPL)CU/mLU/mLMeanCVRangeMeanCVRangeMeanCVRangeMeanCVRangeMeanCVRangeMeanCVRange2022-L1LA ratio ∼ 2.046.218.526.0-58.244.2-38.2-58.441.438.48.2-91.630.816.518.4-41.0174.711.534.8-936.526.515.512.7-35.02022-L2LA ratio ∼ 1.927.0-14.0-52.714.324.99.9-57.514.733.67.4-22.620.427.111.0-27.232.311.822.2-42.35.113.23.2-6.42022-L3LA ratio ∼ 1.46.142.50.8-12.910.028.55.4-13.514.037.37.3-24.89.0-0.0-18.072.610.26.9-103.810.411.56.6-13.02022-L4LA ratio ∼ 1.79.323.06.5-13.019.735.39.0-28.725.218.912.4-45.413.223.99.8-18.5121.211.197.7-145.019.710.316.0-26.02023-L1LA ratio ∼ 1.710.324.47.0-38.217.431.411.7-24.813.420.79.6-22.010.615.07.1-14.661.211.643.9-253.86.515.44.0-9.62023-L2LA ratio ∼ 1.44.117.73.0-5.05.128.73.8-7.95.025.82.8-6.64.8-0.0-25.020.211.25.9-24.02.313.81.4-3.32023-L3LA ratio ∼ 2.214.351.01.2-27.018.721.614.8-32.614.422.29.9-32.712.519.89.0-18.970.910.456.8-85.66.814.81.4-9.9Anti–ß-2-glycoprotein-1 IgGSurveySampleELISAAutomated platformsEuroimmunOrgentec (Alegria)Orgentec (ELISA)Werfen Inova Quanta liteIL ACUstar/INOVA Quanta flashThermo Scientific EliAU/mL (GPL)CU/mLU/mLMeanCVRangeMeanCVRangeMeanCVRangeMeanCVRangeMeanCVRangeMeanCVRange2022-L1LA ratio ∼ 2.037.915.030.0-199.241.614.233.3-69.441.934.96.5-80.438.126.923.0-60.0824.316.529.7-1119.035.111.319.7-47.02022-L2LA ratio ∼ 1.95.06.64.6-8.18.019.16.6-41.49.823.94.5-24.911.9-0.0-18.5160.415.5112.0-271.35.416.93.6-7.72022-L3LA ratio ∼ 1.413.612.211.8-82.45.923.64.2-16.410.138.14.3-17.712.2-9.8-20.0376.213.5227.3-547.010.312.67.3-13.02022-L4LA ratio ∼ 1.728.024.318.0-162.110.422.07.0-13.317.634.92.8-28.917.7-12.0-24.5647.613.7312.0-899.018.711.014.0-25.92023-L1LA ratio ∼ 1.712.29.39.9-20.09.021.46.3-25.011.731.66.6-18.014.6-10.9-16.5246.815.553.0-368.110.312.53.9-13.02023-L2LA ratio ∼ 1.44.7-3.0-10.03.412.03.0-4.64.140.42.0-7.26.0-0.0-11.070.112.941.0-90.23.212.71.4-7.02023-L3LA ratio ∼ 2.213.217.39.0-58.111.521.17.3-14.610.920.07.7-21.918.722.211.9-29.5283.114.7183.6-370.312.79.33.4-17.0%CV are shown when there are ≥10 participants. IgM data are not shown because the majority of the ECAT EQA IgM results are negative, and therefore, displaying them does not provide additional information. ELISA assays: Euroimmun (Euroimmun Medizinische Labordiagnostika), Quanta Lite (Inova/Werfen), and Orgentec (ELISA/Alegria; Orgentec Diagnostika). Automated assay: Automated EliA assay on a Phadia platform (Thermo Fisher Scientific) and Werfen Acustar/Inova Quantaflash (Werfen). The ECAT EQA program for APS is part of the Lupus Anticoagulant program, which includes testing for anticardiolipin and anti–beta-2-glycoprotein-1 IgG and IgM antibodies. Lyophilized plasma samples from a patient tested positive for lupus anticoagulant were distributed to ∼625 participants in 38 countries worldwide. Participants were asked to test the samples as if it was a patient sample. Results were electronically submitted to the ECAT Foundation. Robust statistics (Algorithm A) were used to calculate the mean, SD, and between-laboratory variation for each method separately.%CV, coefficient of variation; ELISA, enzyme-linked immunosorbent assay; EQA, external quality assessment scheme; Ig, immunoglobulin. Open table in a new tab %CV are shown when there are ≥10 participants. IgM data are not shown because the majority of the ECAT EQA IgM results are negative, and therefore, displaying them does not provide additional information. ELISA assays: Euroimmun (Euroimmun Medizinische Labordiagnostika), Quanta Lite (Inova/Werfen), and Orgentec (ELISA/Alegria; Orgentec Diagnostika). Automated assay: Automated EliA assay on a Phadia platform (Thermo Fisher Scientific) and Werfen Acustar/Inova Quantaflash (Werfen). The ECAT EQA program for APS is part of the Lupus Anticoagulant program, which includes testing for anticardiolipin and anti–beta-2-glycoprotein-1 IgG and IgM antibodies. Lyophilized plasma samples from a patient tested positive for lupus anticoagulant were distributed to ∼625 participants in 38 countries worldwide. Participants were asked to test the samples as if it was a patient sample. Results were electronically submitted to the ECAT Foundation. Robust statistics (Algorithm A) were used to calculate the mean, SD, and between-laboratory variation for each method separately. %CV, coefficient of variation; ELISA, enzyme-linked immunosorbent assay; EQA, external quality assessment scheme; Ig, immunoglobulin. Finally, the new classification criteria discourage the use of automated platforms in favor of ELISA-based testing systems to ensure homogeneity in trials and studies. Nevertheless, automated platforms have become the standard in most clinical laboratories. In the ECAT EQA program, the majority (ie, >75%) of the submitted data were generated by automated platforms. This is in line with a recently published commentary that showed that a majority of clinical laboratories participating in the Royal College of Pathologists of Australasia Quality Assurance Program also used automated platforms for aPL testing [4Favaloro E.J. Pasalic L. Lippi G. Classification criteria for the antiphospholipid syndrome: not the same as diagnostic criteria for antiphospholipid syndrome.Semin Thromb Hemost. 2023; https://doi.org/10.1055/s-0043-1776318Crossref Scopus (5) Google Scholar]. Restricting APS classification to ELISA-based systems will therefore likely limit the number of patients eligible for inclusion in studies. The new APS classification criteria identify the absence of published studies comparing automated aPL detection methods with solid phase ELISA systems as the primary reason for their exclusion from the laboratory criteria. The research community is encouraged to perform these comparisons prior to using data from automated platform for patient classification. However, even if data on such comparisons were available, they would only be directly applicable to the specific method under investigation and thus not universally transferable to all ELISA methods. In addition, it is well known that solid phase assays for aPL show inconsistencies and variation in results [4Favaloro E.J. Pasalic L. Lippi G. Classification criteria for the antiphospholipid syndrome: not the same as diagnostic criteria for antiphospholipid syndrome.Semin Thromb Hemost. 2023; https://doi.org/10.1055/s-0043-1776318Crossref Scopus (5) Google Scholar]. Our data indicate that automated platforms consistently yielded results with a superior coefficient of variation (%CV) compared to ELISA assays (on average, 13%CV in comparison to 25%CV). Therefore, while we underscore the necessity for further investigation of the relationship between test results obtained with the automated platforms and clinical manifestations of disease, the presence or absence of comparisons between assays does not warrant the confinement of APS classification exclusively to the ELISA platform. In conclusion, we assert that the new ACR/EULAR APS classification criteria regarding aPL testing using ELISA solid phase–based assays and fixed cut-off limits represent a step backward compared with the previous criteria. We advocate the reintroduction of the 99th percentile as the cut-off criterion for aPL solid phase assays, with automated platforms also serving as suitable aPL classification assays. A.H. wrote and edited the manuscript. P.M. was responsible for compiling the data presented. R.U. and P.M. edited the manuscript. All authors read and approved the final version of the paper. There are no competing interests to disclose.