Authors: Emmanuel J. Favaloro, Leonardo Pasalic
Categories: Review Articles, DOACs, apixaban, clinical laboratory techniques, dabigatran, direct oral anticoagulants, lupus anticoagulant, rivaroxaban, Review Article
Source: Research and Practice in Thrombosis and Haemostasis
Doi: 10.1002/rth2.12676
Lupus anticoagulants (LA) are one laboratory criterion for classification of antiphospholipid syndrome, with presence of vascular thrombosis and/or pregnancy/fetal morbidity being clinical criteria. The presence of LA is detected (or excluded) by laboratory testing, with the activated partial thromboplastin time and dilute Russell's viper venom time the most commonly used tests. Given the association of thrombosis with LA, it is no surprise that anticoagulants are used to treat or manage such patients.
To review and discuss interferences from anticoagulants on LA testing, and strategies to mitigate these.
This narrative review assessed interference from commonly used anticoagulants, focusing on LA testing while on direct oral anticoagulants (DOACs), including use of DOAC neutralizers.
The classical anticoagulants comprise vitamin K antagonists such as warfarin, and heparins, predominantly unfractionated heparin and low molecular weight heparin (LMWH). DOACs have emerged with favorable efficacy and safety. These comprise two direct anti‐thrombin (anti‐IIa; dabigatran) or direct anti‐Xa (rivaroxaban, apixaban, edoxaban) agents. All anticoagulants affect clotting assays, although there are differences in effects according to anticoagulant and assay. Nevertheless, because of such interferences, anticoagulants can lead to false‐negative or false‐positive LA findings. Several strategies can mitigate such interferences, including avoidance of testing while patients are on such anticoagulants, temporarily switching to an anticoagulant (i.e., LMWH) with less assay interference, testing for LA at nadir levels of anticoagulants, and/or use of anticoagulant neutralizers.
Whilst the best approach is to avoid LA testing on patients taking anticoagulants; if unavoidable, testing may be facilitated by various mitigating strategies.
Keywords: apixaban, clinical laboratory techniques, dabigatran, direct oral anticoagulants, DOACs, lupus anticoagulant, rivaroxaban
Lupus anticoagulants (LA) represent one of the laboratory criteria for patient classification as “definite” antiphospholipid (antibody) syndrome (APS), ^1^ with presence of antibodies against cardiolipin (aCL) or beta‐2‐glycoprotein I (aB2GPI) representing alternate (or additional) laboratory criteria. Clinical criteria for APS comprise vascular thrombosis and/or pregnancy/fetal morbidity. ^1^ In turn, LA, aCL, and aB2GPI represent autoantibodies directed against phospholipids, generally in complex with a cofactor, which may be B2GPI or prothrombin. The term “lupus anticoagulant” is actually a (double) misnomer because these antibodies are associated with thrombosis, and the “anticoagulant” effect is solely expressed in in vitro assays, generally observed as a prolongation of clotting times; second, LA are not a significant feature of most cases of lupus, and the association with lupus evolved from initial case descriptions. ^2^ , ^3^ , ^4^ The presence of LA is detected (or excluded) by laboratory testing. ^5^ , ^6^ Although expanded on later in this review, the most common tests used for assessing LA are the activated partial thromboplastin time (aPTT) and the dilute Russell's viper venom time (dRVVT). However, there may be a number of other tests that are used in addition or in place of these common tests. ^7^ For example, the silica clotting time (SCT) represents a form of LA‐sensitive aPTT that may be used in place of a standard LA aPTT. ^8^ In addition, assays such as Taipan snake venom time and Textarin time are insensitive to some of the anticoagulants that compromise dRVVT and aPTT, and assays such as dilute prothrombin time (dPT) can detect LA unreactive in dRVVT and aPTT. ^7^
Anticoagulants represent a class of drugs that are predominantly used for treatment and/or prevention of thrombosis. ^9^ , ^10^ Accordingly, it should come as no surprise that they may be used to treat or prevent thrombosis in patients with symptomatic APS, or otherwise clinically symptomatic and found to be positive for LA. The classical anticoagulants comprise the vitamin K antagonists (VKAs) such as warfarin, and the heparins, predominantly unfractionated heparin (UFH) and low molecular weight heparin (LMWH). Because UFH and LMWH represent parenteral agents that need to be administered by injection (either intravenously [UFH] or subcutaneously [LMWH and sometimes UFH]), VKAs (administered orally) have for long represented the anticoagulant of choice for extended or long‐term treatment. More recently, a separate class of anticoagulants, namely the direct oral anticoagulants (DOACs), have been developed that have favorable efficacy and safety compared with the classical anticoagulant agents. These comprise two separate classes, being direct anti‐thrombin (anti‐IIa; dabigatran) or direct anti‐Xa (e.g., rivaroxaban, apixaban, edoxaban) agents. ^9^ , ^10^ , ^11^ , ^12^ Although DOACs are not the preferred anticoagulant for thrombotic APS, especially for patients with a high‐risk APL profile (i.e., so‐called triple positive), it is not uncommon to perform testing for LA as part of thrombophilia screens in patients treated with DOACs, or indeed other anticoagulants.
Of relevance to this review, it should be recognized that all anticoagulants, both classical and DOACs, can have affect clot‐based assays, inclusive of aPTT, dRVVT, and SCT; this is summarized in Table 1. The anticoagulants may have some differential effects on these tests, and indeed also on other common coagulation tests such as prothrombin time (PT) and thrombin time (TT) (Table 1). Such differential effects should be noted by clinicians requesting LA testing and laboratories performing such tests. Thus, anticoagulants can adversely impact tests used for LA detection/exclusion, and thus lead to potential false‐positive and false‐negative LA findings. This may then affect future choice of anticoagulant and duration of treatment, with risk of adverse outcome if based on an incorrect premise. Alternatively, the observed test patterns in patients where the anticoagulant may not be known may be useful to help identify the anticoagulant in use. For example, only the heparins and anti‐Xa agents (including the DOACs apixaban, rivaroxaban, edoxaban) will yield activity in anti‐Xa assays, whereas only anti‐IIa agents (such as dabigatran and UFH) will affect the TT. But moreover, no two anticoagulants have exactly the same profile regarding effects on coagulation assays (Table 1).
In regard to LA assays, such interferences can lead to both false‐positive and false‐negative LA results. ^5^ In recognition of such assay interference, manufacturers have produced reagents for dRVVT (and some aPTT and SCT reagents) that are resistant to heparin within their therapeutic level (generally up to around 1 U/ml heparin) by using heparin “neutralizers” (e.g., heparinase, protamine, polybrene). However, most aPTT reagents do not contain such neutralizers, and indeed, many aPTT reagents purposely exclude such additives because they may alternately be used as surrogate markers of UFH level for patients treated therapeutically. ^13^ , ^14^ No test reagent manufacturer currently includes any neutralizers to VKAs or to DOACs in any commercial assay. However, some DOAC neutralizers exist, and this is expanded on later.
There are now a plethora of recent guidelines advising on laboratory test procedures to aid in the detection (or exclusion) of LA. ^5^ , ^6^ , ^15^ , ^16^ , ^17^ The most widely used guidelines have been developed by the LA Scientific Standardisation Committee (SSC) of the International Society on Thrombosis and Haemostasis (ISTH), of which the latest were published in 2020. ^5^ , ^17^ Perhaps less well‐known is that these guidelines build on previous iterations, ^18^ , ^19^ , ^20^ in particular each previous version in sequence. Additional recent guidelines on LA testing are available from the Clinical and Laboratory Standards Institute (CLSI ^6^ ) and the British Committee for Standards in Haematology ^15^ , ^16^ . There are both similarities and differences in the recommended testing approaches between the guidelines. ^21^ , ^22^ , ^23^ Of some relevance to the current review, at the time of the earlier 2009 ISTH guidelines, ^18^ VKAs and heparin represented the predominant available anticoagulants, with the modern DOACs only emerging in the early 2010s. ^9^ , ^10^ , ^24^ , ^25^ Thus, the earlier 2009 and 2012 LA guidelines only provided guidance on LA testing in the absence or presence of VKAs and heparin ^15^ , ^18^ (Table 2; Table S1). Although the 2014 CLSI guidelines did to some extent cover testing in the presence of DOACs, ^6^ the main recommendation was to avoid such testing on patients, given the known test interferences, and the difficulty in discriminating true LA from false (“DOAC‐induced”) LA. Of additional relevance, an update to the British Society for Haematology guidelines published in 2020 do provide some guidance on LA testing in the presence of DOACs ^16^ (Table 2; Table S1). Also, the recent 2020 ISTH guidelines ^5^ , ^17^ do cover testing in anticoagulated patients, with the latter ^17^ providing the most guidance on LA testing in the presence of DOACs (Table 2; Table S1).
All LA guidelines recognize the use of heparin neutralizers (e.g., heparinase, protamine sulphate, polybrene) in use in LA reagents, predominantly dRVVT reagents,
^5^
,
^6^
,
^15^
,
^16^
,
^17^
able to quench therapeutic levels of heparin (up to 1 U/ml), and thus enable some LA testing without heparin influence on clotting tests in most clinical situations. However, the guidelines correctly caveat that should heparin levels exceed the reagent's neutralizing ability, some residual effects may be observed, potentially leading to false‐positive LA findings. Such heparin neutralizers are not present in most aPTT reagents because, in general, most aPTT reagents are used to assist in the monitoring of heparin therapy,
^13^
,
^14^
and thereby are made purposely sensitive to heparin. As an alternative to a heparin neutralizer in the aPTT reagent, at least one manufacturer has produced a CaCl2~ reagent with added heparin neutralizers (http://haematex.com/hrrs.html#title_bar). This then permits use of standard aPTT reagents (without added neutralizers) for both heparin monitoring (use of standard CaCl2) and for LA investigation (CaCl2 with heparin neutralizer), assuming that such aPTT reagents are otherwise suitable for said purposes.
There is no means to specifically “neutralize” the effect of VKAs because such anticoagulants act in vivo to alter the vitamin K–dependent coagulation factors (II, VII, IX, X) and thereby reduce their activity, thus ultimately affecting all clotting assays in which such factors are represented. In vivo, patient overexposure to VKAs, thus yielding very high International Normalized Ratio (INR) values can be mitigated by use of vitamin K and/or factor replacement therapy. However, there is no specific agent available to alter a VKA effect in vitro, although, to some extent, neutralization of VKA effects can be achieved by performing tests as mixtures with normal plasma. This acts to correct the loss of factor II, VII, IX, and X “deficiency” caused by VKA use, and thus provides a means of assessing LA somewhat free of the VKA effect. Indeed, this was a recommended strategy in the 2009 ISTH LA guidelines ^18^ (Table 2; Table S1). However, this strategy has lost favor in the revised 2020 ISTH LA guidelines ^5^ , ^17^ (Table 2; Table S1) because some experts believe this may lead to false‐negative or false‐positive LA findings. Nevertheless, it may remain the only option available for laboratories faced with assessing LA in a VKA‐treated patient.
Given experience with heparin neutralizers in dRVVT assays, to negate the effect of therapeutic heparin and permit more accurate detection/exclusion of LA, it should therefore come as no surprise that manufacturers have now produced “DOAC neutralizers” for similar in vitro application. However, as stated earlier, such neutralizers have not yet been included in any LA assay by manufacturers of aPTT, dRVVT, or SCT, but rather represent a separate laboratory step before LA testing by such assays. There are four main commercial products available. The first reported ^26^ is called DOAC‐Stop and was produced locally in Australia by Thomas Exner at his research and manufacturing facility of Haematex in Sydney. As a historical link to LA guidelines, readers may be interested to know that Exner was lead author of the 1991 ISTH LA guidelines, ^21^ as well as authoring dozens of other papers on LA. The product and its use have now been reported in several studies ^26^ , ^27^ , ^28^ , ^29^ , ^30^ , ^31^ , ^32^ , ^33^ , ^34^ , ^35^ , ^36^ , ^37^ , ^38^ , ^39^ , ^40^ , ^41^ , ^42^ , ^43^ (Table 3; Table S2). The product represents a form of activated charcoal, and one pellet of the commercial product can remove a therapeutic level of all the DOACs (dabigatran, rivaroxaban, and apixaban being those most well studied) from 1 ml of citrate anticoagulated plasma. In brief, after adding one pellet to 1 ml of plasma and mixing, the treated plasma is then centrifuged to pellet out the black charcoal (complexed to the DOAC), and laboratory testing then progressed on the “DOAC‐free” supernatant plasma. Another activated charcoal‐based product called DOAC‐Remove has subsequently been produced by another commercial manufacturer (5‐Diagnostics, Switzerland), and seems to have similar features to DOAC‐Stop ^38^ , ^39^ , ^44^ , ^45^ , ^46^ , ^47^ , ^48^ , ^49^ , ^50^ (Table 3; Table S2). A third product is available from a third manufacturer (Stago Diagnostics) and is called DOAC‐filter ^51^ , ^52^ (Table 3; Table S2). A different type of filter has recently been released for sale by 5‐Diagnostics, called DP‐filter; studies on the device appear to only have been published in abstract form. ^53^ , ^54^ An additional publication discusses the use of another activated charcoal product ^55^ in this setting, but is not clearly reflective of the use of any of the four commercial products mentioned previously. A series of reviews on this topic have also been published. ^56^ , ^57^ , ^58^ , ^59^ , ^60^
Some of the concerns raised with the use of such products include a fear that they may not remove all the DOAC present (especially if super‐therapeutic levels are present), similar to the situation with heparin “exceeding” the capability of heparin neutralizers. This may thus give a false sense of assurance regarding LA testing, and still potentially lead to false‐positive or false‐negative results. A second concern is that these products may have an unknown effect on other components of the test systems, which may in itself adversely affect test results and conclusions around LA presence or absence. For example, historical experience with use of filters in LA testing to help filter out platelets ahead of plasma freezing showed that although such filters successfully removed platelets, some also removed large plasma proteins such as von Willebrand factor, and accordingly also factor VIII. ^61^ In theory, loss of FVIII could lead to effects on aPTT‐based tests for LA. However, the major concerns related to potential for false diagnosis of von Willebrand disease or hemophilia, should such additional tests be performed on the filtered plasma, for example if LA testing was just one test of a panel performed for investigation of a raised aPTT. Such findings led to withdrawal of recommendations to use filtration devices to remove platelets before sample freezing, and instead to initiate a process of double centrifugation. ^18^ It is not known if use of the DOAC filter products or other neutralizers will lead to similar or other unwanted effects. Thus, the general recommendation on their use (expanded on later) is to only use such agents in test samples from patients known to be on a DOAC, and to perform specific DOAC testing before and after the use of such agents (to verify DOAC levels before, and absence of DOAC after, use). Of course, all this adds to the complexity, cost, and time taken to investigate patients on DOAC therapy. Use of these products also leads to loss of plasma sample volume, which is often already in short supply given the sample requirements for full investigation of APS or associated conditions. This may be compounded if additional tests are used to help identify the anticoagulant in question. Thus, the best strategy remains to avoid testing LA on patients under DOAC therapy, or if unavoidable, to undertake such testing at trough levels (i.e., collect blood sample just prior to next dose of DOAC), and then potentially use a DOAC neutralizer. However, even these strategies do not guarantee a successful outcome. Additional unknowns include a lack of information about repeat use of neutralizers in case a single use has not removed all the DOAC and whether a mixed approach of DOAC neutralizer/filter provides additional value.
As noted previously, given emergence of DOACs in the early 2010s, ^9^ , ^10^ only the most recent published guidelines from ISTH, ^5^ , ^17^ British Society for Haematology, ^16^ and CLSI ^6^ provide recommendations on LA testing in the presence DOACs (Table 2; Table S1). Some of these later guidelines may also provide guidance on LA testing in the presence of classical anticoagulants (VKA, heparins), but greater acknowledgment of these classical anticoagulants is the purview of the earlier guidelines. ^15^ , ^18^ Importantly, three of the latest guidelines (ISTH, ^5^ , ^17^ BCS ^16^ ) comment on the use of DOAC‐removal agents. These recommendations are summarized in Table 2, alongside various comments made within the guidelines to help caveat some of the recommendations (Table S1).
One of the authors (E.J.F.) participated in the development of both the CLSI ^6^ and one of the recent ISTH ^5^ guidelines, and through his activities in the ISTH also had input into the final version of the 2009 ISTH guidelines. ^18^ Such guidelines are both evidence‐based (where evidence exists) and eminence‐based (where evidence base is weak or does not exist). Thus, there is a smattering of expert opinion in all LA guidelines. ^62^ Moreover, the guidelines tend to be consensus‐based (i.e., essentially requiring “support” of the participants), and here, sometimes a majority view may arise that is not reflective of all‐inclusive agreement. ^62^ All authors have personal experiences and biases; for example, for us having experience around use of DOAC‐Stop ^33^ , ^36^ , ^37^ , ^56^ but not the other DOAC‐neutralizers/filters. Also, because we run a laboratory that is required to provide a broad diagnostic service, there may be pressure exerted on us by colleagues and other requesting clinicians to perform tests while patients may be on anticoagulant therapy, despite our personal protestations and misgivings. Thus, although we would agree with the guidelines that it is best practice to perform LA testing when patients are not on anticoagulant therapy, this may not always be possible. Examples of reasons where LA testing on anticoagulants may be unavoidable
In such situations, our recommended approach would entail a different approach based on the anticoagulant in question, as summarized
As a brief overview, based on our experience, we would also proffer the following. As already noted, different DOACs have variable effects on aPTT, dRVVT, and other routine assays such as PT and TT (Table 1). Particularly, although all DOACs can prolong the aPTT, the extent of prolongation is both DOAC and reagent dependent (Figure 1). Dabigatran affects the aPTT more than rivaroxaban, and apixaban affects the aPTT the least of the three. In terms of reagent dependence and LA testing, the effect can perhaps be highlighted using a common reagent pair used for such testing, Siemens FSL (LA sensitive) and FS (LA insensitive). FS tends to be more affected than FSL with all the DOACs (Figure 1); however, their relative sensitivity compared with other aPTT reagents differs according to the DOAC. The three DOACs also differ in regard to dRVVT sensitivity (Figure 2). Here, rivaroxaban affects dRVVT more than dabigatran, with apixaban showing least affect. However, the effects on screen and confirm reagent testing also differ, such that rivaroxaban, and to a lesser extent dabigatran, affects the screen more than the confirm, thus potentially yielding an abnormal LA ratio (or a false LA result; Figure 2). In contrast, apixaban affects the confirm more than the screen, thus potentially yielding a reduced LA ratio at “within therapy” levels, that in a patient with a weak LA can lead to a false‐negative result. That apixaban can lead to a false‐negative LA finding has also been inferred from studies using ex vivo samples ^38^ , ^63^ ; however, such false‐negative phenomena are harder to prove than false positives because they are reliant on finding studies using rare potentially weak LA patients on apixaban therapy. Additional local information, looking at comparative assay ratios for PT, aPTT, and dRVVT (Figure 3), provides additional context. In general, the dRVVT is affected more by the DOACs than the aPTT or PT, but there is variability in extent and relative prolongations among the assays.
FIGURE 1 Summarizing the effect of direct oral anticoagulants (DOACs) on the activated partial thromboplastin time (aPTT). An original figure highlighting historical data in which the lead author performed in collaboration with the Royal College of Pathologists of Australasia Quality Assurance Program (RCPQAP) ^64^ , ^65^ and showing differential effects on various commercial aPTT reagents according to type of DOAC. The aPTT data are shown as APTT ratios
FIGURE 2 Summarizing the effect of direct oral anticoagulants (DOACs) on the dilute Russell's viper venom time (dRVVT). An original figure highlighting historical data in which the lead author performed in collaboration with the Royal College of Pathologists of Australasia Quality Assurance Program (RCPQAP) ^64^ , ^65^ and showing differential effects according to type of DOAC. Data shown as dRVVT screen and confirm ratios (left y‐axis) and arising dRVVT screen/confirm ratio (right y‐axis), using box and whiskers showing 10th–90th percentiles
FIGURE 3 Summarizing the effect of direct oral anticoagulants (DOACs) on the various coagulation assays. An original figure highlighting historical data in which the lead author performed in collaboration with the Royal College of Pathologists of Australasia Quality Assurance Program (RCPQAP) ^64^ , ^65^ and showing differential effects on the three assays according to type of DOAC. Data shown as comparative activated partial thromboplastin time (aPTT), dilute Russell's viper venom time (dRVVT), and prothrombin time (PT) ratios. In general, a ratio above 1.2 can be considered as “abnormal”
In regard to DOAC neutralizers, our experience with DOAC‐Stop has shown several noteworthy findings related to LA testing, ^33^ as was also highlighted within the ISTH SSC guidance on LA detection in anticoagulated patients. ^17^ First, when rivaroxaban was added to pooled normal plasma, this (as expected) caused clotting time prolongation for most LA tests performed by participants of an external quality assessment program and generated falsely elevated dRVVT screen/confirm ratio results that mimicked the presence of LA. Second, when the rivaroxaban plasma sample was treated with DOAC‐Stop, results showed correction of the prolongation of the clotting time and the screen/confirm ratio for most LA tests. Notably, all study participants correctly identified the rivaroxaban plasma treated with DOAC‐Stop as LA‐negative. Third, andexanet‐alfa, an in vivo antidote for rivaroxaban, when added to the rivaroxaban plasma in vitro was able to correct the prolonged clotting time induced by rivaroxaban. It also corrected the screen/confirm ratio, but to such an extent (i.e., overcorrection) that such reduction in LA ratio could potentially lead to a false‐negative LA in those patients with weak positive LA while on rivaroxaban, should andexanet‐alfa be used as an in vitro DOAC neutralizer. Thus, in summary, andexanet‐alfa is not recommended as an in vitro DOAC neutralizer ahead of LA testing. The effect of in vivo use of andexanet‐alfa on LA test patterns from treated patients is to our knowledge unknown.
The investigation of LA represents a common activity for hemostasis laboratories. The presence of LA is detected (or excluded) by laboratory testing, with the aPTT and the dRVVT being most commonly used. Anticoagulants are commonly used to treat or manage thrombosis, which may include many patients being investigated for LA. All anticoagulants will affect the assays used to investigate LA, but to variable extent. Ideally, investigation of LA will occur at a time when patients are not on an anticoagulant. However, should this be unavoidable, there are several strategies available to mitigate anticoagulant interferences, including the use of various anticoagulant neutralizers. As an alternative to LA testing while on anticoagulant therapy, some authors instead advocate for performance of anti–phosphatidyl‐serine/prothrombin (aPS/PT) antibodies, which can be used as a surrogate test for LA and is not affected by anticoagulants. ^66^ , ^67^ , ^68^ The premise for such use is that most patients with APS and triple positivity are also positive in aPS/PT (tetra‐positive aPL), ^67^ and that aPS/PT more than aβ2GPI is responsible for LA activity in these patients. ^68^ Figure 4 provides an algorithm that summarises the sentiments expressed in this review, representing a potential approach to the investigation of LA when a patient is on anticoagulant therapy.
FIGURE 4 One potential algorithm to support the identification/exclusion of lupus anticoagulants from patients on anticoagulant therapy and applying some of the recommendations from the current review. The algorithm is based on the authors' personal preferences, but also considers options used by other workers in the field
The authors have no competing interests.
Emmanuel J. Favaloro wrote the original draft of this manuscript. Both authors contributed content, helped revise the manuscript, and approved its submission. The views expressed herein are those of the authors and are not necessarily those of NSW Health Pathology or the Royal College of Pathologists of Australasia Quality Assurance Program.
The authors thank various current and past employees of NSW Health Pathology and the Royal College of Pathologists of Australasia Quality Assurance Program for past contributions permitting reuse of some data for educational purposes in the current manuscript. There was no specific funding for this work. NSW Health Pathology, as the employer of the authors, is acknowledged for in‐kind support that permitted some allocation of time for its completion.
Favaloro EJ, Pasalic L. Lupus anticoagulant testing during anticoagulation, including direct oral anticoagulants. Res Pract Thromb Haemost. 2022;6:e12676. 10.1002/rth2.12676
Please contact the lead author.
Please contact the lead author.