False negative screening tests for lupus anticoagulants--an unrecognized problem?

Practical Considerations and Testing Nuances for the Detection of Lupus Anticoagulant: Do Low Phospholipid Screen Results, Assay Type, and Test Ratio Matter?

OBJECTIVES

Lupus anticoagulant (LA) detection requires (1) prolongation of a phospholipid (PL)-dependent clot-based screening assay, (2) noncorrection upon adding normal pooled plasma, and (3) a confirmatory PL dependency test. Paired LA assays run screening and confirmatory tests simultaneously, with their test ratio (TR) or differences used to evaluate test results. We evaluated patients whose paired testing demonstrated PL dependence suggestive of LA, yet the low PL screen was not prolonged.

METHODS

Clinical and laboratory parameters are compared across (1) true positive (screen prolonged, TR positive) vs borderline (screen not prolonged, TR positive); (2) low-, moderate-, and high-TR subgroups; and (3) dilute Russell viper venom time (dRVVT) vs silica clotting time (SCT).

RESULTS

Borderline samples are not statistically different from true positives in their rate of repeat LA positivity or association with other anti-PL antibodies. Compared with true positives, borderline dRVVT is more frequent in pregnancy, women, and younger age. Elevated activated partial thromboplastin time is more frequent in true-positive dRVVT and SCT vs borderline and with an increasing dRVVT TR. LA persistence is more frequent with an increasing SCT TR. In addition, dRVVT true positivity is more frequent with thromboembolic events, while SCT is more frequent with autoimmunity and pregnancy complications.

CONCLUSIONS

Negative low-PL screens may not necessarily lack LA. A reevaluation of the laboratory criteria for LA detection may be needed.

Lupus anticoagulant assay cut-offs vary between reagents even when derived from a common set of normal donor plasmas

BACKGROUND

Multicenter studies reveal that diagnostic efficacy of lupus anticoagulant (LA) assays is enhanced if cut-offs are locally generated. However, a potential confounder is the inevitable use of separate normal donor populations.

OBJECTIVES

Generate cut-offs for multiple LA reagents with the same analyzer and normal donor plasmas.

METHODS

Cut-offs for screen ratio, confirm ratio, percent correction of screen ratio by confirm ratio, and normalized screen/confirm ratio (NSCR) were derived from the same 50 normal donor plasmas for screen and confirm pairs for two dilute Russell's viper venom time reagents, LA1/LA2 and HEMOCLOT™ LA-S/LA-C, and two APTTs, Actin FSL/FS and Cephen LS/Cephen. The cut-offs were challenged with plasmas from 20 triple-positive APS patients and 25 plasmas from LA-negative, thrombotic patients.

RESULTS

Cut-offs for screen ratio, confirm ratio, percent correction, and NSCR, respectively, were 1.12/1.08/8.3/1.09 for LA1/LA2; 1.17/1.10/13.6/1.13 for HEMOCLOT™ LA-S/LA-C; 1.12/1.13/9.7/1.10 for Actin FSL/FS; 1.09/1.13/11.0/1.11 for Cephen LS/Cephen. LA1 and LA-S screens were elevated in 19/20 and 16/20 triple-positive plasmas, respectively, while 20/20 were detected with both via integrated interpretation ie, percent correction or NSCR irrespective of screen elevation. Actin FSL and Cephen LS screens were elevated in 17/20 and 19/20 triple-positive plasmas, respectively, while one more LA was detected with Actin FSL via integrated interpretation, but not for Cephen LS. Integrated interpretation suggested 5/25 LA-negative plasmas contained weak LA (two with Actin FSL/FS, two with LA1/LA2, one with LA-S/LA-R).

CONCLUSIONS

Employing the same normal donor plasmas and analytical platform does not compensate for between-reagent differences when generating LA assay cut-offs.

Newly developed dilute Russell's viper venom reagents for lupus anticoagulant detection with improved specificity

Background Dilute Russell's viper venom time (dRVVT) is indispensible in lupus anticoagulant (LA) detection yet commercial reagents from different suppliers perform variably, no gold standard assays exist and therapeutic anticoagulation interference is problematic. Objective The objective of this study was to compare a new formulation dRVVT with two currently available dRVVTs. Materials and methods Life Diagnostics (LD) dRVVT and Stago PTT-LA were routinely used for lupus anticoagulant detection, plus Taipan snake venom time/ecarin time (TSVT/ET) for patients on warfarin or rivaroxaban. Siemens dRVVT and the new HYPHEN BioMed (HBM) dRVVT were tested with 193 patient samples. Group 1, 59 non-anticoagulated patients (NAPs) LA-positive in LD dRVVT; Group 2, 15 PTT-LA-positive/dRVVT-negative NAPs; Group 3, 24 LA-positive warfarinized patients; Group 4, 13 patients on rivaroxaban; Group 5, 62 LA-negative thrombotic NAPs; Group 6, 20 warfarinized, non-antiphospholipid syndrome patients. Results Accepting that the Life Diagnostics reagents were acting as a pseudo-gold standard, Siemens dRVVT detected 56/59, (95%) Group 1 LA and HBM dRVVT 46/59, (76%), one each from Group 2, and Siemens dRVVT detected one in Group 5. The lower HBM dRVVT detection rate mainly concerned weaker LA, where between-reagent concordance is problematic. All Group 3 patients appeared LA-positive in undiluted plasma with Siemens dRVVT, as did 16/24 (67%) with HBM dRVVT but the fewer LA-positives in mixing tests better mapped to clear LA-positives with LD dRVVT. LD and Siemens dRVVTs exhibited 87% and 95% false-positivity for Group 6 whilst HBM dRVVT had none. Increasing the cut-off improved accuracy. Applying higher cut-offs improved accuracy in Group 4 patients. Conclusion HBM dRVVT exhibited improved specificity, mainly due to less interference by anticoagulation, but reduced sensitivity, compared to the other dRVVTs employed.

Application of different lupus anticoagulant diagnostic algorithms to the same assay data leads to interpretive discrepancies in some samples

BACKGROUND

Gold standard lupus anticoagulant (LA) assays and reference plasmas do not exist and detection is based on inference in a medley of coagulation assays, creating potential for interpretive discrepancies when applying different algorithms.

OBJECTIVES

To investigate discrepancies from applying different algorithms to a common data set.

METHODS

Diagnostic data on 311 non-anticoagulated patients LA-positive by dilute Russell's viper venom time (dRVVT) and/or dilute activated partial thromboplastin time (dAPTT) assays were employed to compare algorithms. Routine testing applied interpretive criteria from guidelines endorsing classification as LA-positive despite negative mixing tests, after exclusion of other clotting abnormalities. Integrated testing without mixing tests, and the classical algorithm where negative mixing tests preclude confirm tests, were then retrospectively applied to those data.

RESULTS

Initial testing showed 92/311 (29.6%) were LA-positive by dRVVT only, 156/311 (50.1%) by dAPTT only, and 63/311 (20.3%) by both assays. All dAPTT-positive plasmas remained positive with integrated testing but eight dRVVT-positives became negative. Other data suggested they were false-negatives. The classical algorithm altered 52/155 (33.5%) dRVVT and 111/219 (50.7%) dAPTT interpretations to LA-negative because of normal mixing tests, most of which were apparently weak LA in undiluted plasma.

CONCLUSIONS

The classical algorithm improves diagnostic specificity and confidence but risks missing some genuine LA due to false-negative mixing tests. Integrated testing can be diagnostically accurate and logistically efficient but oversimplifies complex cases. Performing mix and confirm in response to an elevated screen with their interpretation based on clinical data, coagulation screens and the LA-assay design offers a potentially valuable option.

Current Controversies in Lupus Anticoagulant Detection

Antiphospholipid syndrome is an autoimmune, acquired thrombophilia diagnosed when vascular thrombosis or pregnancy morbidity are accompanied by persistent antiphospholipid antibodies. Lupus anticoagulants (LA) are one of the criteria antibodies but calibration plasmas are unavailable and they are detected by inference based on antibody behaviour in a medley of coagulation-based assays. Elevated screening tests suggest the presence of a LA, which is confirmed with mixing tests to evidence inhibition and confirmatory tests to demonstrate phospholipid-dependence. At least two screening tests of different principle must be used to account for antibody heterogeneity and controversy exists on whether assays, in addition to dilute Russell's viper venom time and activated partial thromboplastin time, should be employed. A variety of approaches to raw data manipulation and interpretation attract debate, as does inclusion or exclusion of mixing studies in circumstances where the presence of a LA is already evident from other results. Therapeutic anticoagulation compromises coagulation-based assays but careful data interpretation and use of alternative reagents can detect or exclude LA in specific circumstances, and this aspect of LA detection continues to evolve. This review focuses on the main areas of debate in LA detection.

In ECAT veritas?

The laboratory criteria of the antiphospholipid syndrome include one coagulation assay (lupus anticoagulant [LA]) and two solid phase assays (anticardiolipin [aCL] and anti-ß2glycoprotein I antibodies [aß2GPI]). External quality control (EQC) surveys show that negative and clearly positive LA samples are classified correctly by about 95% of laboratories. For ‘weak’ LA there is a wide variability in samples’ classification. Furthermore, when a weak LA sample is used in two different EQC surveys more than 50% of laboratories classify it differently. In some surveys weak LA samples were found to be positive for aCL and for aß2GPI by a majority of laboratories; the main reason for laboratories which classified these samples as LA negative was a negative result in the mixing test. It is likely that, depending on the sensitivity of the assay, a weak LA cannot be detected anymore after 1:1 dilution of the sample with normal plasma. Therefore, we recommend the use of integrated assays, such as screen/confirm ratios, for the detection of weak LA samples.

Antiphospholipid syndrome: laboratory detection, mechanisms of action and treatment

The antiphospholipid syndrome (APS) identifies a condition at increased risk of vascular occlusion and/or pregnancy complications. Patients are defined as having APS if they have at least one clinical (vascular occlusion and/or pregnancy complications) and one laboratory criterion at the same time. The laboratory criteria that define APS are repeated positivity (confirmed 12 weeks apart) for lupus anticoagulants and/or antibodies targeted against cardiolipin or β(2) -glycoprotein I immobilized on solid surfaces. Over the years, APS has attracted the interest of many medical specialties. The aim of this review is to provide an update on (i) the laboratory criteria that determine the presence of APS, (ii) how the antibodies increase the risk of vascular occlusion and foetal loss and (iii) the treatment of the related clinical events.

Further evidence of false negative screening for lupus anticoagulants

INTRODUCTION

Accurate and timely detection of lupus anticoagulants (LA) is of diagnostic and prognostic importance due to the association of persistent LA with thrombotic disease. Antibody heterogeneity and assay variability complicate LA detection and weak antibodies can go undetected.

METHODS

Screen and confirm assays on equal volume mixing studies were performed on known LA using dilute Russell's viper venom time (DRVVT), dilute activated partial thromboplastin time (DAPTT) and activated seven lupus anticoagulant (ASLA) assay. Two established calculations for phospholipid dependence were applied to ascertain whether any antibodies diluted to within screening test reference ranges maintained a significant difference between screen and confirm results sufficient to imply LA activity. We then performed confirmatory tests on neat plasma samples from patients with thrombotic disease whose screening tests were within reference ranges.

RESULTS

Forty nine of 155 DRVVT positive LA were conventionally positive in the mixing studies and 8 of the 106 negative mixing studies showed significant screen and confirm test discordance. This was the case for 21 of 56 negative DAPTT mixing studies and 2 of 39 negative ASLA mixing studies. Performance of confirm assays on the neat plasma samples with screen results within reference ranges revealed possible LA activity in 19 of 166 DRVVT results, 63 of 184 DAPTT results and 9 of 117 ASLA results.

CONCLUSIONS

LA activity can be demonstrated by assessment of screen and confirm data irrespective of screening test elevation above a reference range. Other workers have demonstrated this phenomenon in APTT using different study designs and it may be that standard interpretation criteria warrant re-assessment.

The laboratory diagnosis of lupus anticoagulant in patients on oral anticoagulation

Laboratory-specific cut-off lupus ratios (LR), above which a plasma is judged positive for lupus anticoagulant (LA), were established for both activated partial thromboplastin time-based and dilute Russell viper venom time-based methods. The validity of using these cut-off values to determine the presence of LA in patients on oral anticoagulation (OAC) was assessed. A cohort of 40 patients (23 male and 17 female), aged 22-84 years (mean 52 years) were tested for LA at the time of a thrombotic event. Repeated testing was performed after the same patients were treated with OAC (international normalized ratio 2.0-3.5). For 36 patients (90%), LA status was unchanged pre- and on-OAC. Thirteen of the 40 patients (32.5%) were positive for LA both pre- and on-OAC. Of the 27 patients negative for LA pre-OAC, 23 remained negative on-OAC. The four discordant results were interesting in that LA positivity was demonstrated only after the patient was stable on-OAC. In our cohort of 40 patients, there was a trend for LRs to decrease on-OAC, but this did not reach statistical significance. The subset (4) went against this trend and became positive after the thrombotic event.

The Ecarin time is an improved confirmatory test for the Taipan snake venom time in warfarinized patients with lupus anticoagulants

The Taipan snake venom time using dilute phospholipid as a screening test with a platelet neutralization procedure as a confirmatory test has been shown to be a sensitive and specific approach to detection of lupus anticoagulants. Taipan venom is largely insensitive to the effects of ongoing warfarin anticoagulation and this can be useful in detection of lupus anticoagulants in patients receiving this treatment. This study compared the use of the platelet neutralization procedure with the Ecarin time as confirmatory tests for the Taipan snake venom time, the Ecarin venom fraction being insensitive to both lupus anticoagulants and the effects of oral anticoagulants. Screening and confirmatory test data were assessed for phospholipid dependence by three different mathematical methods and there was no advantage in using the Ecarin time in detection of 'uncomplicated' lupus anticoagulants. In lupus anticoagulant-positive warfarinized patients, the Ecarin time achieved higher detection rates than the platelet neutralization procedure irrespective of the method used to assess correction. The Ecarin time confirmed lupus anticoagulants in all of those samples that generated elevated Taipan snake venom time ratios whereas the platelet neutralization procedure identified only 33%. Taipan snake venom time plus Ecarin time offers good diagnostic precision for lupus anticoagulant detection in a group of patients where lupus anticoagulant identification is difficult due to ongoing anticoagulation.

MIXCON-LA: a precise, sensitive and specific aPTT-based assay for detection of lupus anticoagulant

Lupus anticoagulants (LA) are associated with an increased risk of thrombosis and laboratory detection is of major importance. Multiple tests are available for screening and confirmation, but they differ in sensitivity and specificity, frequently lacking the ability to discriminate between the presence of LA, heparin, and oral anticoagulants. Based on the test-principle of the Lupus Ratio-test, an automated, sensitive APTT-based assay, using mixtures of a lupussensitive and a lupusinsensitive APTT-reagent with normal plasma for detection of lupus anticoagulants was developed. Ninety-nine healthy volunteers, ten patients treated with unfractionated heparin intravenously, 19 patients taking stable oral anticoagulation, five patients with hemophilia A, and 15 patients with antiphospholipid-antibody-syndrome (APS) were investigated. In all patients, two APTTs were performed, one with each reagent, on 1:1 mixtures of test plasma and normal plasma (MIXCON-LA assay). The ratio between the two clotting times was divided by the corresponding ratio for the normal plasma. This final lupus ratio (LR) was used for evaluation. The within-series imprecision and the between-series imprecision were excellent with coefficients of variation between 1.5% and 1.9%. The mean +/- 2 SD of the LR of the 99 healthy volunteers was used as reference range (LR: 0.95-1.07). All patients treated either with heparin or oral anticoagulants remained negative in the MLXCON-LA assay (specificity, 100%), while one of five patients with hemophilia A, in whom a factor VIII-inhibitor developed, showed a false-positive result. In 13 of 15 patients with APS, an increased ratio was observed (sensitivity, 87%). This assay system allows precise, specific, and sensitive detection of lupus anticoagulants.

Reference intervals of the dilute tissue thromboplastin inhibition and dilute Russell's viper venom tests revisited

Reference intervals for two well-recognized tests for the lupus anticoagulant were determined using 98 healthy subjects. The purpose of the study was to compare the reference intervals for the dilute tissue thromboplastin inhibition test on this group of healthy subjects calculated by parametric and nonparametric statistical methods, and to compare these results with results obtained on subsets of 20 and 40 randomly selected individuals from the group of 98. The same procedures were followed for the dilute Russell's viper venom test. Results were recorded in seconds of clotting times and in ratios (subject/mean of that set or subset). Statistical analysis revealed Gaussian distribution of the results in the large group as well as in each subset for both tests. The results showed more variation between sets of the dilute tissue thromboplastin inhibition test than of the dilute Russell's viper venom test. Nonparametrically calculated reference intervals were wider than those determined by the parametric method, especially if confidence intervals are provided for both reference limits in a group of 94 controls or in a subset of 40 subjects. The nonparametric technique uses all data for the calculation of reference interval of such sample sizes no matter how widely spread the results are. There was no significant difference between the reference intervals of subsets and the whole group (p > 0.05) calculated by both statistical techniques. Very few outliers were observed among these subjects in both tests.

Reference ranges of the dilute tissue thromboplastin inhibition and dilute Russell's viper venom tests revisited

Reference ranges for two well-recognized tests for the lupus anticoagulant were determined utilizing 98 healthy subjects. The purpose of the study was to compare the reference ranges for the dilute tissue thromboplastin inhibition test on this group of healthy subjects calculated by parametric and nonparametric statistical methods, and to compare these results with results obtained on subsets of 20 and 40 randomly selected individuals from the group of 98. The same procedures were followed for the dilute Russell's viper venom test. Results were recorded in seconds of clotting times and in ratios (subject/mean of that set or subset). Statistical analysis revealed Gaussian distribution of the results in the large group as well as in each subset for both tests. The results showed more variation between sets of the dilute tissue thromboplastin inhibition test than of the dilute Russell's viper venom test. Nonparametrically calculated reference ranges were wider than those determined by a parametric method, especially if confidence intervals are provided for both reference ranges in the group of 94 controls or in a subset of 40 subjects. The nonparametric technique utilizes all data for the calculation of reference ranges of such sample sizes no matter how wide the results are spread. There was no significant difference between the reference ranges of subsets and the whole group (p > 0.05) calculated by both statistical techniques.