2023 ACR/EULAR antiphospholipid syndrome classification criteria.

OBJECTIVE: To develop new antiphospholipid syndrome (APS) classification criteria with high specificity for use in observational studies and trials, jointly supported by the American College of Rheumatology (ACR) and EULAR. METHODS: This international multidisciplinary initiative included four phases: (1) Phase I, criteria generation by surveys and literature review; (2) Phase II, criteria reduction by modified Delphi and nominal group technique exercises; (3) Phase III, criteria definition, further reduction with the guidance of real-world patient scenarios, and weighting via consensus-based multicriteria decision analysis, and threshold identification; and (4) Phase IV, validation using independent adjudicators' consensus as the gold standard. RESULTS: The 2023 ACR/EULAR APS classification criteria include an entry criterion of at least one positive antiphospholipid antibody (aPL) test within 3 years of identification of an aPL-associated clinical criterion, followed by additive weighted criteria (score range 1-7 points each) clustered into six clinical domains (macrovascular venous thromboembolism, macrovascular arterial thrombosis, microvascular, obstetric, cardiac valve, and hematologic) and two laboratory domains (lupus anticoagulant functional coagulation assays, and solid-phase enzyme-linked immunosorbent assays for IgG/IgM anticardiolipin and/or IgG/IgM anti-ß2-glycoprotein I antibodies). Patients accumulating at least three points each from the clinical and laboratory domains are classified as having APS. In the validation cohort, the new APS criteria vs the 2006 revised Sapporo classification criteria had a specificity of 99% vs 86%, and a sensitivity of 84% vs 99%. CONCLUSION: These new ACR/EULAR APS classification criteria were developed using rigorous methodology with multidisciplinary international input. Hierarchically clustered, weighted, and risk-stratified criteria reflect the current thinking about APS, providing high specificity and a strong foundation for future APS research.

Solid Phase Assays for Antiphospholipid Antibodies.

The diagnosis of antiphospholipid syndrome (APS) relies on the detection of circulating antiphospholipid antibodies (aPL). Currently, lupus anticoagulant (LA), anticardiolipin (aCL), and anti-ß2-glycoprotein I antibodies (aß2GPI) IgG or IgM are the laboratory criteria if persistently present over time. As aCL and aß2GPI are two out of the three laboratory criteria, the detection of aPL by solid phase assays is an essential step in the diagnosis of APS. Advancement has been made to resolve some of the methodological challenges of aCL and aß2GPI assays by providing guidelines how to measure aPL, as well as to gain a better understanding of their diagnostic role. However, solid phase assays for aCL and aß2GPI still show substantive inter-assay differences, resulting in disagreement concerning positive/negative results, but also differences in titer of antibodies. This hampers the semiquantitative classification into low-medium-high positivity. The non-criteria aPL, such as antibodies against the domain one of ß2GPI and anti-phosphatidylserine/prothrombin antibodies (aPS/PT) have roles in confirming the risk in APS, and can be useful, especially in patients with incomplete antibody profiles.

Antithrombotic prophylaxis for surgery-associated venous thromboembolism risk in patients with inherited platelet disorders. The SPATA-DVT Study.

Major surgery is associated with an increased risk of venous thromboembolism (VTE), thus the application of mechanical or pharmacologic prophylaxis is recommended. The incidence of VTE in patients with inherited platelet disorders (IPD) undergoing surgical procedures is unknown and no information on the current use and safety of thromboprophylaxis, particularly of low-molecular-weight-heparin in these patients is available. Here we explored the approach to thromboprophylaxis and thrombotic outcomes in IPD patients undergoing surgery at VTE-risk participating in the multicenter SPATA study. We evaluated 210 surgical procedures carried out in 155 patients with well-defined forms of IPD (VTE-risk: 31% high, 28.6% intermediate, 25.2% low, 15.2% very low). The use of thromboprophylaxis was low (23.3% of procedures), with higher prevalence in orthopedic and gynecological surgeries, and was related to VTE-risk. The most frequently employed thromboprophylaxis was mechanical and appeared to be effective, as no patients developed thrombosis, including patients belonging to the highest VTE-risk classes. Low-molecular-weight-heparin use was low (10.5%) and it did not influence the incidence of post-surgical bleeding or of antihemorrhagic prohemostatic interventions use. Two thromboembolic events were registered, both occurring after high VTE-risk procedures in patients who did not receive thromboprophylaxis (4.7%). Our findings suggest that VTE incidence is low in patients with IPD undergoing surgery at VTE-risk and that it is predicted by the Caprini score. Mechanical thromboprophylaxis may be of benefit in patients with IPD undergoing invasive procedures at VTE-risk and low-molecular-weight-heparin should be considered for major surgery.

How to Interpret Antiphospholipid Laboratory Tests.

PURPOSE OF THE REVIEW: This review focuses on the laboratory tests necessary for the diagnosis of antiphospholipid syndrome (APS). For the interpretation of the results of the tests for antiphospholipid antibodies (aPL), understanding of all pitfalls and interferences is necessary. RECENT FINDINGS: Progress has been made on the standardization of aPL tests and current guidelines for detection of lupus anticoagulant (LAC), anticardiolipin antibodies (aCL), and antibeta2-glycoprotein I antibodies (aß2GPI) are useful tools. LAC measurement remains a complex procedure with many pitfalls and interference by anticoagulant therapy. Solid phase assays for aCL and aß2GPI still show inter-assay differences. Measuring LAC, aCL, and aß2GPI allows making antibody profiles that help in identifying patients at risk. Other aPL, such as antibodies against domain I of beta2-glycoprotein I (aDI) and antiphosphatidylserine-prothrombin (aPS/PT) antibodies, may be useful in risk stratification of APS patients, but are not included in the current diagnostic criteria as no added value in the diagnosis of APS has been illustrated so far. The laboratory diagnosis of APS remains challenging. LAC, aCL, aß2GPI IgG, and IgM should be performed to increase diagnostic efficacy, with an integrated interpretation of all results and an interpretative comment. A close interaction between clinical pathologists and clinicians is mandatory.

Evaluation of the primary biliary cholangitis-related serologic profile in a large cohort of Belgian systemic sclerosis patients.

Background Systemic sclerosis (SSc) and primary biliary cholangitis (PBC) are autoimmune diseases that may occur concomitantly and are both strongly associated with disease-specific autoantibodies. This study investigated the prevalence and fine specificity of PBC-specific serology (PBC-Ab) and associations with the SSc-subtypes and SSc-specific antibodies as well as the association with cholestatic liver enzymes. Furthermore, three different techniques for the detection of PBC-Ab were compared. Methods Serum of 184 Belgian SSc patients with a known SSc-antibody profile, was analyzed for PBC-Ab (antimitochondrial antibodies [AMA], anti-Gp210, anti-Sp100 and anti-PML) using indirect immunofluorescence (IIF) analysis on human epithelioma-2000 (HEp-2000) cells (ANA-IIF, Immunoconcepts) and liver-kidney-stomach tissue sections (IIF-LKS) (Menarini), and a line immunoblot (LB) (EuroImmun). Alkaline phosphatase/γ-glutamyl transferase (ALP/GGT) were evaluated at time of first sampling (t0) and after 3 years of follow-up (t3). Results PBC-Ab were present in 13% of patients and significantly correlated with centromere antibodies (anti-CENP-B), but not correlated with the limited cutaneous SSc subgroup (lcSSc). The most frequent reactivities were AMA (11%, with 9% AMA-M2) and Sp-100 antibodies (5%), showing a major overlap. There was no relevant association between the presence of PBC-Ab and ALP or GGT elevation at t0 nor at t3. Detection of AMA with IIF-LKS is comparable to LB. ANA-IIF screening was less sensitive compared to LB. Conclusions A wide range of PBC-Ab is detectable in SSc in the absence of cholestatic liver enzyme elevations, even after 3 years of follow-up. However, as these antibodies may precede PBC-disease up to 10 years further prospective follow-up of our cohort will be necessary.

Comparison between manufacturing sites shows differential adhesion, activation, and GPIbα expression of cryopreserved platelets.

BACKGROUND: Transfusion of cryopreserved platelets (cryoplatelets) is not common but may replace standard liquid-preserved platelets (PLTs) in specific circumstances. To better understand cryoplatelet function, frozen concentrates from different manufacturing sites were compared. STUDY DESIGN AND METHODS: Cryoplatelets from Denver, Colorado (DEN); Sydney, Australia (SYD); and Ghent, Belgium (GHE) were compared (n = 6). A paired noncryopreserved control was included in Ghent. Microfluidic-flow chambers were used to study PLT adhesion and fibrin deposition in reconstituted blood. Receptor expression was measured by flow cytometry. Coagulation in static conditions was evaluated by rotational thromboelastometry (ROTEM). RESULTS: Regardless of the manufacturing site, adhesion of cryoplatelets under shear flow (1000/sec) was significantly (p < 0.05) reduced compared to control. Expression of GPIbα was decreased in a subpopulation of cryoplatelets comprising 45% ± 11% (DEN), 63% ± 9% (GHE), and 94% ± 6% (SYD). That subpopulation displayed increased annexin V binding and decreased integrin activation. PLT adhesion, agglutination, and aggregation were moreover decreased in proportion to that subpopulation. Fibrin deposition under shear flow was normal but initiated faster (546 ± 163 sec GHE) than control PLTs (631 ± 120 sec, p < 0.01), only in the absence of tissue factor. In static conditions, clotting time was faster, but clot firmness decreased compared to control. Coagulation was not different between manufacturing sites. CONCLUSION: Cryopreservation results in a subset of PLTs with enhanced GPIbα shedding, increased phosphatidylserine expression, reduced integrin response, and reduced adhesion to collagen in microfluidic models of hemostasis. The proportion of this phenotype is different between manufacturing sites. The clinical effects, if any, will need to be verified.

The importance of detecting anti-DFS70 in routine clinical practice: comparison of different care settings.

BACKGROUND: Screening for antinuclear antibodies by indirect immunofluorescence (ANA-IIF) is essential in the diagnostic workup of ANA-associated autoimmune rheumatic diseases (AARDs). However, also healthy individuals may test positive, making the interpretation challenging. Recent reports suggest that dense fine speckled 70 antibodies (anti-DFS70) may facilitate this challenge. Here, we investigate their clinical importance based on data from four Belgian laboratories (one primary, two secondary and one tertiary care). METHODS: At least one specific DFS70 assay (DFS70 IgG ELISA or lineblot [Euroimmun, full length antigen] and/or DFS70 IgG CLIA [Inova Diagnostics, truncated antigen]) was performed on four consecutive cohorts of homogeneous-like ANA-IIF samples (n=697). Co-occurrence with AARD-specific ANA and clinical information were documented in the anti-DFS70-positive samples. RESULTS: Using a combination of solid phase techniques, we found between 7.6% and 26% anti-DFS70 in the different cohorts. Focusing on anti-DFS70 CLIA-positive samples without co-occurrence of AARD-specific ANA, we observed a trend towards lower frequency in tertiary (8% [p=0.0786]) and secondary care (12% [p=0.1275] and 6% [p<0.001]) compared to primary care (21%). Moreover, in this specific subpopulation, AARD was less frequent (0%-50% compared to 6%-77% in the total anti-DFS70-positive group). CONCLUSIONS: Anti-DFS70 prevalence depends on the applied assay and care setting. Our data suggest that, for an ANA-IIF-positive patient, it is rather the absence of AARD-associated ANA and clinical symptoms that contribute to the exclusion of AARD than the presence of anti-DFS70. Nevertheless, isolated anti-DFS70 helps to clarify positive ANA-IIF results, especially if pretest probability for AARD is low.

Dilute Russell's viper venom time reagents in lupus anticoagulant testing: a well-considered choice.

BACKGROUND: Lupus anticoagulant (LAC) detection represents diagnostic challenges among which the multitude of available reagents and interference by anticoagulant treatment. One of the two advised tests is the dilute Russell's viper venom time (dRVVT). However, it is currently not clear whether all dRVVT reagents may be considered equivalent. The objective of the study was to evaluate the diagnostic performance of two dRVVT reagents, with special attention to the influence of anticoagulant therapy. METHODS: STA®-Staclot® dRVV Screen/Confirm (Stago, Asnières-sur-Seine, France) and dRVT-LS/dRVTL-LR (Haematex, Hornsby, Australia) were evaluated on 443 patient samples [358 consecutive patients with LAC request including six antiphospholipid syndrome (APS) patients, 18 non-consecutively selected APS patients and 37 vitamin K antagonists (VKA)-treated and 30 direct oral anticoagulants (DOAC)-treated non-APS patients]. Additionally, pooled normal plasma (PNP) was spiked with factor deficient plasma (n=33) and DOAC calibrators (n=21) to evaluate sensitivity for factor deficiencies and false-positivity rates, respectively. RESULTS: A higher number of samples were defined as LAC positive by Stago vs. Haematex [11.5% (41/358) vs. 3.63% (13/358)]. Most discordances were in the VKA and DOAC group. Haematex was less prone to VKA-related factor deficiencies, explaining the absence of false-positive LAC results in VKA-treated non-APS patients compared to 10.8% with Stago. We observed no false-positive LAC ratios with Haematex in DOAC-spiked PNP and a lower number in DOAC-treated non-APS patients. However, increased specificity seemed to be at cost of a reduced sensitivity as Haematex showed less positive APS patient samples (45.8% vs. 87.5%). CONCLUSIONS: dRVVT reagents differ in LAC sensitivity and for VKA and DOAC interference.

Differences in lupus anticoagulant final conclusion through clotting time or Rosner index for mixing test interpretation.

BACKGROUND: Lupus anticoagulant (LAC) testing includes a screening, mixing and confirmation step. Although recently published guidelines on LAC testing are a useful step towards standardization, a lack of consensus remains whether to express mixing tests in clotting time (CT) or index of circulating anticoagulant (ICA). The influence of anticoagulant therapy, e.g. vitamin K antagonists (VKA) or direct oral anticoagulants (DOAC) on both methods of interpretation remains to be investigated. The objective of this study was to contribute to a simplification and standardization of the LAC three-step interpretation on the level of the mixing test. METHODS: Samples from 148 consecutive patients with LAC request and prolonged screening step, and 77 samples from patients non-suspicious for LAC treated with VKA (n=37) or DOAC (n=30) were retrospectively evaluated. An activated partial thromboplastin time (aPTT) and dilute Russell's viper venom time (dRVVT) were used for routine LAC testing. The supplemental anticoagulant samples were tested with dRVVT only. We focused on the interpretation differences for mixing tests expressed as CT or ICA and compared the final LAC conclusion within each distinct group of concordant and discordant mixing test results. RESULTS: Mixing test interpretation by CT resulted in 10 (dRVVT) and 16 (aPTT) more LAC positive patients compared to interpretation with ICA. Isolated prolonged dRVVT screen mix ICA results were exclusively observed in samples from VKA-treated patients without suspicion for LAC. CONCLUSIONS: We recommend using CT in respect to the 99th percentile cut-off for interpretation of mixing steps in order to reach the highest sensitivity and specificity in LAC detection.

Paired analysis of plasma proteins and coagulant capacity after treatment with three methods of pathogen reduction.

BACKGROUND: The effect of photochemical pathogen reduction (PR) methods on plasma quality has been the subject of several reports but solid comparative data for the different technologies are lacking. STUDY DESIGN AND METHODS: Plasma (n = 24) photoinactivated with methylene blue (MB), riboflavin (RF), or amotosalen (AS) was compared using a pool-and-split design. Samples were taken before and after treatment with each method and tested for coagulation factors (fibrinogen, Factor [F] II, FV, FVIII, F IX, FXI), natural coagulation inhibitors (Protein C [PC], protein S [PS], antithrombin III [AT]), prothrombin time (PT), activated partial thromboplastin time (APTT), and thrombin generation (TG). The three methods were mutually compared by repeated-measures analysis of variance. RESULTS: All three PR methods cause significant reduction (p < 0.01) of activity of the procoagulant proteins fibrinogen, FII, FV, FVIII, F IX, and FXI. Coagulation is also affected, with significant changes in PT, APTT, and TG. RF treatment causes a significantly higher decrease in concentration of coagulation factors, PS, and AT than the other methods (p < 0.01). PT, APTT, and TG are also affected most by RF treatment. FII, FVIII, F IX, PC, AT, and PT are best preserved with the MB method and FV, FXI, and TG after AS treatment (p < 0.01). Coagulation factor loss due to the volume loss during PR treatment is more important for MB and AS than for RF. CONCLUSION: PR treatment of plasma affects coagulation proteins and coagulant capacity. For the RF method this effect is most pronounced, although to some extent compensated by a smaller volume loss.

Bedside monitoring of anticoagulation in chronic haemodialysis patients treated with tinzaparin.

BACKGROUND: During haemodialysis, anticoagulants are required to prevent clotting in the extracorporeal circuit. Low-molecular weight heparins (LMWH) are frequently used because of the ease of a single injection at the start of dialysis. Disadvantages of LMWH include the lack of a reliable bedside assay for measuring their anticoagulant effect. METHODS: We investigated a bedside test for LMWH activity. The relationship between anti-Xa (chromogenic assay) and Hemonox point-of-care assay was evaluated in 21 dialysis patients (12 men and 9 women) with a median age of 71 years, receiving tinzaparin at the start of a haemodiafiltration session. RESULTS: At the start, before tinzaparin administration, median (interquartile ranges) of Hemonox values were 74 (67-82) s. Thirty minutes after tinzaparin administration, Hemonox values were increased to 496 (360-736) s, followed by a decrease to 149 (135-301) s after 120 min, 102 (97-144) s after 180 min and 92 (83-100) s after 240 min. Corresponding anti-Xa activities were 0 (0-0), 1.12 (0.9-1.29), 0.74 (0.57-0.96), 0.47 (0.31-0.7) and 0.31(0.16-0.49) IU/mL. Hemonox values showed an exponential relation to anti-Xa levels. Interchangeability of tests was shown by Bland-Altman plot. CONCLUSION: Point-of-care Hemonox test is a valuable bedside method for monitoring anti-Xa activity in dialysis patients anticoagulated with tinzaparin.

Automated indirect immunofluorescence microscopy enables the implementation of a quantitative internal quality control system for anti-nuclear antibody (ANA) analysis.

BACKGROUND: Screening for anti-nuclear antibodies by indirect immunofluorescence (ANA-IIF) remains mandatory in the serological work-up of connective tissue diseases. Recently, automated approaches were introduced that may improve harmonization. Here, we investigated whether the introduction of automated ANA-IIF and more specifically the use of its quantitative measure, could improve ANA-IIF internal quality control (IQC) management. METHODS: We retrospectively reviewed results of two cohorts of routine samples and parallel IQC data collected from January 2010 to February 2013 and from February to mid October 2013. For the first cohort, data were collected using conventional microscopy. The second cohort was analyzed by an automated ANA-IIF microscope (Zenit G sight, A. Menarini). Retrospectively, we evaluated the applicability of the probability index (PI) of control material measurements and patient results for IQC management based on Westgard multirules. This approach was also compared with monthly monitoring of the %ANA-IIF positive samples. RESULTS: In our historical data set, we showed that monitoring of %ANA positives identified systematic errors that were not detected by monitoring control material results. Data resulting from automated microscopy showed that PI measurements on control material remained stable within the observed period and that Westgard multirules can be used for IQC follow-up. Parallel monitoring of the daily median patient PI and the monthly %ANA positives, showed that the daily median was a sensitive and fast tool for detecting systematic errors. CONCLUSIONS: The introduction of the automated ANA-IIF microscope could enable objective IQC procedures and should be considered an important step forward in ANA-IIF harmonization.

Noncriteria antiphospholipid antibodies in antiphospholipid syndrome.

Antiphospholipid syndrome (APS) is an autoimmune disease characterized by thrombotic manifestations and/or obstetric complications in patients with persistently positive antiphospholipid antibodies (aPL). aPL are a heterogeneous group of autoantibodies, but only lupus anticoagulant, anticardiolipin (aCL), and antibeta2-glycoprotein I antibodies (aß2GPI) IgG or IgM are included as laboratory classification criteria. Seronegative APS patients are usually defined as patients with the clinical symptoms of APS but who test negative for aPL. The negativity to classic aPL criteria does not exclude the presence of other aPL. Several noncriteria aPL have been identified. Some noncriteria aPL are well studied, such as IgA aCL and aß2GPI, the antiphosphatidylserine-prothrombin (aPS/PT) antibodies, and the antibodies against the domain I of beta2-glycoprotein I (aDI), both latter groups receiving more attention for their role in thrombotic events and pregnancy complications. Other noncriteria aPL that have been studied are antibodies against annexin V, prothrombin, phosphatidylethanolamine, phosphatidic acid, phosphatidylserine, phosphatidylinositol, vimentin-cardiolipin complex, anti-protein S/protein C. Measurement of some of these noncriteria aPL (aPS/PT, aDI) is useful in the laboratory work-out of APS in specific situations. We have to differentiate between patients who are positive for noncriteria aPL only, and patients who have both criteria and noncriteria aPL to enable us to study their role in the diagnosis or risk stratification of APS. The research on noncriteria aPL is continually developing as the clinical relevance of these antibodies is not yet fully clarified.

Toward harmonized interpretation of anticardiolipin and anti-ß2-glycoprotein I antibody detection for diagnosis of antiphospholipid syndrome using defined level intervals and likelihood ratios: communication from the ISTH SSC Subcommittee on Lupus Anticoagulant/Antiphospholipid Antibodies.

BACKGROUND: Improving harmonization of the clinical interpretation of anticardiolipin (aCL) and anti-ß2-glycoprotein I (aß2GPI) antibodies immunoglobulin G (IgG)/immunoglobulin M (IgM) in the diagnosis of antiphospholipid syndrome (APS) is desirable. Likelihood ratios (LRs) with corresponding test-result intervals can identify the power of a test to discriminate between a diseased and nondiseased patient and may be useful for the semiquantitative interpretation of aCL/aß2GPI results. OBJECTIVES: To determine moderate and high thresholds for aCL and aß2GPI IgG/IgM measured with chemiluminescent immunoassay, enzyme-linked immunosorbent assay, fluorescence enzyme immunoassay, and multiplex flow immunoassay. METHODS: aCL and aß2GPI antibodies IgG/IgM were determined with 4 solid-phase systems in a case-control study population including 381 APS patients and 727 controls. Interval-specific LRs (IS-LR) were calculated for ranges determined by prespecified specificity and sensitivity levels. Three methods were used for determining thresholds that separated low, moderate, and high positive antibody levels. Interassay agreement was checked with Cohen's kappa statistics. RESULTS: Assay- and antibody-specific thresholds demonstrated increasing IS-LR, reflecting different clinical significance for low, moderate, and high levels, especially for IgG aCL and aß2GPI and in thrombotic APS. IS-LRs per antibody and unit range were comparable across solid-phase platforms resulting in enhanced harmonization of result interpretation. Agreement between assays for identifying high levels was improved by semiquantitative interpretation compared with that by quantitative reporting. CONCLUSION: aCL and aß2GPI IgG/IgM moderate and high thresholds were determined for 4 analytical platforms. Thresholds improve harmonized interpretation of aCL/aß2GPI levels across platforms. The proposed thresholds should be verified in an independent case-control study to check interlaboratory transferability.

Intracardiac ultrasound-guided transseptal puncture in horses: Outcome, follow-up, and perioperative anticoagulant treatment.

BACKGROUND: Cardiac catheterizations in horses are mainly performed in the right heart, as access to the left heart traditionally requires an arterial approach. Transseptal puncture (TSP) has been adapted for horses but data on follow-up and closure of the iatrogenic atrial septal defect (iASD) are lacking. HYPOTHESIS/OBJECTIVES: To perform TSP and assess postoperative complications and iASD closure over a minimum of 4 weeks. ANIMALS: Eleven healthy adult horses. METHODS: Transseptal puncture was performed under general anesthesia. Serum cardiac troponin I concentrations were measured before and after puncture. Weekly, iASD closure was monitored using transthoracic and intracardiac echocardiography. Relationship between activated clotting time and anti-factor Xa activity during postoperative enoxaparin treatment was assessed in vitro and in vivo. RESULTS: Transseptal puncture was successfully achieved in all horses within a median duration of 22 (range, 10-104) minutes. Balloon dilatation of the puncture site for sheath advancement was needed in 4 horses. Atrial arrhythmias occurred in 9/11 horses, including atrial premature depolarizations (N = 1), atrial tachycardia (N = 5), and fibrillation (N = 3). Serum cardiac troponin I concentrations increased after TSP, but remained under the reference value in 10/11 horses. Median time to iASD closure was 14 (1-35) days. Activated clotting time correlated with anti-factor Xa activity in vitro but not in vivo. CONCLUSIONS AND CLINICAL IMPORTANCE: Transseptal puncture was successfully performed in all horses. The technique was safe and spontaneous iASD closure occurred in all horses. Clinical application of TSP will allow characterization and treatment of left-sided arrhythmias in horses.

A thrombin-driven neural net diagnoses the antiphospholipid syndrome without the need for interruption of anticoagulation.

ABSTRACT: Thrombosis is an important manifestation of the antiphospholipid syndrome (APS). The thrombin generation (TG) test is a global hemostasis assay, and increased TG is associated with thrombosis. APS is currently diagnosed based on clinical and laboratory criteria, the latter defined as anti-cardiolipin, anti-ß2-glycoprotein I antibodies, or lupus anticoagulant (LA). APS testing is often performed after a thrombotic episode and subsequent administration of anticoagulation, which might hamper the interpretation of clotting assays used for LA testing. We set out to develop an artificial neural network (NN) that can diagnose APS in patients who underwent vitamin K antagonist (VKA) treatment, based on TG test results. Five NNs were trained to diagnose APS in 48 VKA-treated patients with APS and 64 VKA-treated controls, using TG and thrombin dynamics parameters as inputs. The 2 best-performing NNs were selected (accuracy, 96%; sensitivity, 96%-98%; and specificity, 95%-97%) and further validated in an independent cohort of VKA-anticoagulated patients with APS (n = 33) and controls (n = 62). Independent clinical validation favored 1 of the 2 selected NNs, with a sensitivity of 88% and a specificity of 94% for the diagnosis of APS. In conclusion, the combined use of TG and NN methodology allowed for us to develop an NN that diagnoses APS with an accuracy of 92% in individuals with VKA anticoagulation (n = 95). After further clinical validation, the NN could serve as a screening and diagnostic tool for patients with thrombosis, especially because there is no need to interrupt anticoagulant therapy.