Patients

We studied 28 adult ambulatory patients on warfarin or acenocumarol referred to our hospital to confirm or exclude APS following objectively documented either venous thromboembolism (VTE), including thrombosis at atypical locations, or ischemic stroke. We assessed the patients in whom the attending physicians decided not to interrupt anticoagulant therapy for a diagnostic workup. All patients were tested for LA between November 2018 and October 2024. The diagnosis of APS was made as recommended.⁹ As this study was a part of a diagnostic workup, approval of the Bioethical Committee was not required.

Laboratory investigations

Citrated plasma samples were tested for coagulation parameters on the day of blood draw (from September to October 2024) or from frozen samples, stored at –80 °C if plasma was available. The international normalized ratio (INR) was measured in all patients.

LA was determined according the current recommendations.¹⁰ We used the following assays: dRVVT with LA1 Screening Reagent (LA1), LA2 Confirmation Reagent (LA2), and APTT‑LA with Actin FS for screening (LA1A) and Actin FSL for confirmation (LA2A) (Siemens Healthcare, Marburg, Germany). The APTT and dRVVT assays were performed on the Atellica COAG360, BCS‑XP coagulation system (Siemens Healthcare) and on the automated CN‑3000 coagulation analyzer (Sysmex, Kobe, Japan). LA1 and LA2 were expressed as normalized ratios by dividing patient plasma clotting time by the standard human plasma (Siemens Healthcare) clotting time. A positive LA was diagnosed, if all 3 steps showed positivity in either dRVVT and / or APTT assay.¹

We measured TSVT (Diagen Taipan Venom) and ET (Diagen Echis Clotting Time Test; both Diagnostics Reagents Ltd., Oxon, United Kingdom). In the TSVT assay, samples were defined as positive if a TSVT ratio above 1.1 and at least 10% correction by the ET ratio were observed according to the manufacturer’s instructions. The TSVT/ET testing was carried out with the automated CN‑3000 coagulation analyzer (Sysmex) with a manufacturer’s modification reducing the sample and reagent volumes by 50%. Plasma samples from 5 healthy individuals were analyzed, and all results were negative.

Antinuclear antibodies (ANA) were determined in serum using indirect immunofluorescence with human epithelial cells HEp‑20‑10 (Euroimmun Medizinische Labordiagnostika AG, Lübeck, Germany). The threshold of positivity was set at a titer equal to or above 1:160.

aPL were measured in serum using the Quanta Lite assays: anticardiolipin (ACA) immunoglobulin (Ig) G III, ACA IgM III, β2GPI IgG enzyme‑linked immunosorbent assay (ELISA) and β2GPI IgM ELISA (Inova Diagnostics Inc., San Diego, California, United States). Positive aPL results were defined as values exceeding the in‑house cutoff values, determined using the 99th percentile (>8 standard IgG units [SGU], >10 standard IgM units [SMU], >15 IgG phospholipid units [GPL], and >17 IgM phospholipid units [MPL] for aβ2GPI IgG, aβ2GPI IgM, aCL IgG, and aCL IgM, respectively). As recommended, measurements were performed twice, at least 3 months apart.^3,10

Statistical analysis

Variables were presented as numbers and percentages and mean (SD). Normality was assessed with the Shapiro–Wilk test. The Spearman or Pearson correlation coefficient was used for non‑normally and normally distributed variables, respectively. A P value below 0.05 was considered significant. Statistical analyses were performed using the STATISTICA software (Version 14.1, TIBCO Software, Palo Alto, California, United States).

Negative lupus anticoagulant

Among 12 patients with negative dRVVT results, 3 (25%) were found to be positive in the TSVT test, suggesting false‑negative results in the screening or confirmatory tests.

The first patient (No. 28; Table 1), on acenocumarol therapy, whose plasma sample was stored for 10 months before testing (INR, 4.42), was diagnosed with cerebral venous sinus thrombosis (CVST), her ANA titer was 1:160, and aPL tests were negative. Following repetition of the TSVT test to verify the initial positive result, the TSVT result was negative, similarly to the dRVVT test, which might suggest that the initial result might have been influenced by the storage time or other preanalytical factors.

The second patient (No. 26; Table 1) was a 52‑year‑old woman on warfarin therapy (INR, 2.63) due to a history of CVST with positive results for IgM anti-β2GPI (18.4 SMU) and ANA titer of 1:160. dRVVT test values were as follows: LA1 = 1.88, LA2 = 1.42, LA1/LA2 = 1.32, and LA1MIX = 1.18, while the TSVT value was 1.3.

The third case (Patient 27; Table 1) was a 43‑year‑old woman on warfarin (INR, 3.63) following VTE. Her medical history was notable for systemic lupus erythematosus (SLE), and in her past medical records APS was noted given positive IgM anti-β2GPI (31.7 SMU) and ANA speckled pattern at a 1:320 titer. The results of LA1/LA2 and LA1MIX were negative (1.14 and 1.11, respectively), while TSVT was positive at 1.32, suggesting the presence of LA, especially when considered together with her clinical history.

We confirmed the presence of LA using the TSVT assay in 15 of the 16 dRVVT‑positive cases, which is consistent with other studies,¹² and supports the view that combining multiple tests can improve diagnostic accuracy of LA testing, especially in complex cases. Among the dRVVT‑negative patients, TSVT test showed positive results (1.3 and 1.32) in 2 patients, indicating that TSVT may identify LA cases missed by dRVVT on VKAs, which is of clinical relevance. Kristoffersen et al¹³ emphasized the importance of handling and storage conditions for LA testing, showing that preanalytical factors, such as plasma preparation and storage can significantly impact test outcomes. Freezing can disrupt platelets, releasing excess phospholipids into the sample, which may reduce sensitivity for detecting aPL and lead to false‑negative results,¹⁴ and our study shows the importance of these issues for LA testing in practice.

ANA positivity, typical of autoimmune diseases, particularly SLE, correlates with the presence of other autoantibodies, including aPL. The presence of ANA alongside LA may correlate with a more complex immunological profile and elevated thrombotic risk, which requires careful interpretation of borderline or equivocal LA test results in ANA‑positive patients.¹⁵

In this study, plasma samples were mixed with standard human plasma in a 1:1 ratio, and mixing tests were performed without incubation. According to the International Society on Thrombosis and Haemostasis guidelines, the mixing tests are not considered a reliable solution, particularly in the cases where antibody levels are not high, as the dilution effect may lead to false‑negative outcomes.^10,16 However, this limitation did not apply to our 2 patients, both of whom had LA1 values of 1.88 and LA2 values of 1.42 and 1.65, respectively. Using both TSVT/ET and dRVVT/APTT assays provides a more reliable approach for diagnosing LA, as no single test can detect all forms of LA with 100% accuracy.¹⁷ Our findings show that while the dRVVT/APTT assays remain primary tests, TSVT should be considered if the former tests yield borderline values or clinical aspects, for example, coexistence of SLE, suggest in‑depth evaluation. These results support using TSVT in a combined testing strategy including dRVVT, APTT, and prothrombin time in patients on VKAs.^4,10 However, it is still advisable that all uncertain cases, where practicable, undergo evaluation after discontinuation of VKA therapy.

It would be valuable to additionally perform TSVT/ET in patients on VKA with INR below 1.2 to evaluate whether the results differ from those reported in other studies.¹⁸ Such a comparison could provide better insights into potential differences in test outcomes based on INR status and the type of anticoagulant therapy used.

Our study has several limitations. Firstly, the group is small and the results should be treated as preliminary. Moreover, some patients had 2 separate measurements, and in 1 of them the dRVVT/APTT and TSVT/ET results were usually consistent, most often when taken from a fresh blood sample. Secondly, not all samples were suitable for retesting due to extended storage times or prior thawing for use in other tests. Some additional variables, such as inflammatory markers or D‑dimer¹⁹ were not routinely assessed, and their impact on data interpretation is an open issue.

We conclude that the TSVT/ET assay, if approved in the European Union, could be useful as a complement to the standard LA testing in patients on VKA. The TSVT/ET confirmed positive results in most dRVVT‑positive cases and identified additional LA cases missed by dRVVT/APTT alone, while on VKAs. However, mixing test results that fall within the range of 1.1–1.2 remain uncertain, especially in patients with the presence of ANA, where borderline results may not provide definite answers. For such cases, we recommend performing TSVT/ET assays, as conversion from borderline to positive results may occur, which can help clarify LA status and may change patient management.