Monitoring of anticoagulation in thrombotic antiphospholipid syndrome

3.1 Prothrombin time-international normalized ratio (PT-INR)

The INR is derived from the PT screening coagulation test (sensitive to reduced activity of factors (F)I (fibrinogen), II (prothrombin), V, VII, and X, according to the formula: INR=(patient's PT/MNPT)^ISI, where the MNPT is the mean normal PT and ISI is the international sensitivity index, a correction factor applied to adjust for differences in instrument and reagent sensitivity. Accurate INR determination is critically dependent upon accurate definition of the MNPT and ISI for the reagent/instrument combination. Guidance is available for appropriate ISI verification/validation, to help reduce inter-laboratory variability.²⁰ A European Committee for External Quality Assurance Programmes in Laboratory Medicine study demonstrated that between-laboratory INR variation was lower after local INR calibration (coefficient of variation [CV] 6.7% vs. 8.6), which is strongly recommended.²¹

3.2 Point-of-care PT-INR testing

Several point-of-care (POC) devices have been shown to give reliable and accurate INR values when compared to traditional laboratory methods.²² The World Health Organisation (WHO) recommended inclusion of plasmas from patients with INRs 1.5-4.5 for calibration of a thromboplastin's ISI.²³ Thus, the INR system is strictly valid up to an INR of 4.5. However, higher INRs are measured to assess bleeding risk and guide appropriate management.²⁴ A study of 312 patients on warfarin showed that INRs > 4.5 with the CoaguChek XS Plus were comparable to laboratory INRs.²⁵ Patient self-management (PSM) reduces the risk of mortality and VTE, but probably has no influence on the risk of major bleeding, and patient self-testing (PST) possibly reduces the risk of VTE and major bleeding.²⁶ Accordingly, the American Society of Hematology (ASH) recommends PSM over any other management approach, and suggests PST over any other INR testing approach except for PSM, for suitable patients on maintenance VKA for VTE.²⁶ British Society for Haematology (BSH) guidance for health-care professionals exists on PST/PSM.²⁷

3.3 Activated partial thromboplastin time (aPTT)

The aPTT is sensitive to the activity of coagulation FI, II, V, VIII, IX, X, XI, and XII. It can thus be used to monitor the anticoagulant effect of agents that reduce the activity of these factors, if a therapeutic range can be established. Baseline aPTT prolongation in an acute situation should prompt assessment of fibrinogen to investigate acute phase, and together with assessment of D-dimer, to check for disseminated intravascular coagulation. C-reactive protein (CRP) interferes with the aPTT, prolonging clotting times proportional to CRP concentration and depending on the type of the aPTT reagent.²⁸ Raised levels of FVIII, an acute phase reactant, lead to aPTT shortening.²⁹

3.4 Thrombin time (TT) modifications

The TT measures the conversion of fibrinogen to fibrin following addition of thrombin. The protamine sulphate neutralization assay, a TT-based assay, can be used to calculate the plasma concentration of UFH. Protamine titration produces reliable and reproducible results, but is not generally undertaken in modern hemostasis laboratories, as it is not readily automated³⁰ and has been superseded by anti-Xa assays. In the dilute TT (dTT), a simple, rapid, sensitive, quantitative method, patient sample is diluted with normal plasma prior to the addition of thrombin. The dilution eliminates most interferences on the assay including factor deficiencies, D-dimer, and LA.

3.5 Chromogenic factor X (CFX)

The CFX activity is an alternative assay to the PT-INR. There is an inverse linear relationship between CFX and INR^{18, 19}; a CFX range of 23.5% to 35.5% is suggested to be therapeutic (INR range 2.0-3.0), with the lower CFX limit corresponding to an INR of 3.0.³¹ CFX assays showed poor utility for INR values ≥ 3.5 in non-APS patients, possibly due to interference by acarboxy-FX when the level of fully functional FX becomes very low, ≤12.0%.¹⁸ Therapeutic ranges for CFX are not established, and the assay is not well standardized, nor widely available or practicable for routine use.

3.6 Clotting factor II and X (FII and FX)

PT-based clotting assays for FII and X have been proposed as alternatives to the INR. Plasma concentrations of FII and X were proportional to increases in INR in patients on warfarin, for INRs 2.0 to 3.5. However, for INRs > 3.5, a non-proportional relationship was found between INR and both factors. The authors suggested that below critical clotting factor concentrations (20.6% and 15.6% for factors II and X activity, respectively), the time required for clot formation becomes non-proportional and hemostasis will be jeopardized.³²

3.7 Chromogenic anti-Xa or anti-IIa activity

Measurement of anti-Xa or anti-IIa activity is the preferred method when quantitative information is useful. Reference ranges for anti-Xa levels depend on the anticoagulant in use, the type, dose, schedule, and indication. Monitoring is based on peak levels, with the timing of these depending on the half-life of the anticoagulant, and/or trough levels.³³ The anti-Xa assay has important advantages, including that it is unaffected by raised CRP or FVIII related to an acute phase reaction; however, it has limitations, including considerable inter-laboratory variation³⁴ and inter-assay variability.^{35, 36}

3.8 Ecarin based assays

Ecarin is a metalloprotease from the saw-scaled viper, Echis carinatus, that converts prothrombin to meizothrombin, which can be inhibited by direct thrombin inhibitors, but not heparin.³⁷ The accuracy of the ecarin clotting time (ECT) may be impacted by fibrinogen and prothrombin deficiencies. The ecarin chromogenic assay (ECA) pre-dilutes the patient sample with a buffer containing prothrombin to minimize the prothrombin factor limitation; and, as it not a clot-based assay, it is not influenced by fibrinogen levels.³⁷

3.9 Thrombin generation (TG)

TG assesses overall functional coagulation status of plasma and the dynamic processes of thrombin generation. The TG curve is quantified in terms of the lag time, time to peak TG, peak TG, and endogenous thrombin potential (ETP), which is the area under the curve.³⁸ Increased ex vivo TG is a marker of thrombogenic potential and has predictive value for recurrent VTE.³⁹ TG assays remain limited to specialized laboratories, and standardized protocols and data normalization should lead to better reproducibility and ability to compare data from different laboratories.⁴⁰ A more standardized automated TG method has become available: the ST Genesia (Diagnostica Stago, Asnières sur Seine Cedex, France).⁴¹ TG is not currently suitable for anticoagulant monitoring, as it is not standardized and no therapeutic ranges have been established.

4.1 Anticoagulant monitoring of VKAs in APS

The PT-INR is traditionally used to monitor VKA anticoagulation intensity. LA-induced PT prolongation, typically unassociated with bleeding, is unusual, likely due to the high phospholipid concentration in PT reagents (Table 1). Use of an instrument-specific ISI reduced the differences observed between reagents in earlier studies, for samples with INRs within the therapeutic range, with similar results in APS and non-APS patients.⁴⁹ A multicenter study demonstrated that differences in PT-INR when measured with the majority of commercial thromboplastins are not enough to cause concern, if thromboplastins insensitive to LA and instrument-specific ISI calibration are used¹⁹. Recombinant thromboplastins produce consistently higher INRs,¹⁹ which could potentially alter clinical management.¹⁸ New thromboplastins should be checked for sensitivity to LA.¹⁷ LA can prolong the PT when associated with the rare LA-hypoprothrombinemia syndrome with acquired factor II deficiency, caused by anti-prothrombin antibodies.⁵⁰ Despite its limitations for monitoring VKA-treated APS patients, the INR remains in general use.

POC INR testing, however, shows variable results in APS patients. A study of 29 APS versus 31 non-APS patients using the ProTime InRhythm^TM System (finger stick or venous non-citrated blood) showed a difference of < 0.4 INR units compared to venous citrated blood laboratory INR (both measured with human recombinant thromboplastin) in 97% of patients, suggesting that this device is accurate in APS- and non-APS patients.⁵¹ However, in a comparative study, 5/59 (8%) patients with APS and both elevated aβ₂GPI and LA had non-measurable ProTime^® INR results and generally higher Hemochron^TM Signature INR results than the plasma-based INR, indicating elevated POC INRs in a subset of APS patients. The divergence between POC and laboratory INRs increased as INR values rose.⁵² Taylor et al found that the mean INR difference between the CoaguChek XS and laboratory INR was 0.68, with a difference significantly higher in APS patients than in controls with a target INR of 2 to 3.⁵³ Isert et al did not observe a higher disagreement between INR results from CoaguChek versus laboratory INR, using two thromboplastins, in APS patients.⁵⁴ However, INR variations above 0.5 were more frequent in APS patients compared to controls (55.6% vs. 67.8%, P = .05). No in-depth analysis has been performed to identify the reason for variability in INR by POC devices compared to plasma-based laboratory INR in APS patients, or the reason for variability between POC devices. More investigation to unravel potential differences in PT-INR measured in APS patients by POC devices is needed.

Pending further data to establish the place of POC INR testing in LA-positive patients, results should be interpreted with caution.¹⁷ A pragmatic approach is to restrict POC testing to APS patients in whom concordance of POC and laboratory INRs (INR difference < 0.5 units) is demonstrated at therapeutic intensity initially and every 6-12 months; ensure regular internal quality control (IQC), and ongoing review of external quality assessment (EQA) results to check that concordance is maintained.^{27, 55} It is prudent to apply this to all APS patients rather than only those who are LA-positive, as LA testing could yield false negative results, but might influence INR monitoring.

CFX provides an LA-independent assessment of VKA anticoagulant intensity. This is because the CFX assay requires minimal phospholipid, and the initial dilution is large, which would make the assay less sensitive to LA.³¹ A literature review of CFX use that selected 9/55 articles for their relevance, concluded that in a subgroup of APS patients, INRs may be falsely elevated and thus may not accurately inform the dose of warfarin.⁵⁶ Notwithstanding its limitations, CFX could assist in monitoring LA-positive APS patients, particularly in those who have a prolonged PT prior to commencement of VKA or who experience recurrent VTE while on apparently therapeutic VKA anticoagulation.

A study of concurrent INR, CFX, and PT-based clotting FII levels in 36 patients (26 aPL positive), showed limited discordance between CFX and FII. The authors concluded that FII and CFX testing are well correlated in patients in whom anticoagulation with warfarin requires alternative monitoring to INR and thus that either test can be used in this population.⁵⁷ In contrast to the study of Rosborough et al that showed, in 14 of 21 LA-positive patients, a significantly lower FII/CFX ratio than in LA-negative patients,⁵⁸ Baumann Kreuziger et al did not find a difference in FII/CFX ratios between patients with or without LA.⁵⁷ FII (clotting or chromogenic) assays should be avoided for anticoagulant monitoring in APS patients with LA-hypoprothrombinemia syndrome.⁵⁰

ETP and peak TG parameters showed significant inverse correlations with INR in APS patients on warfarin, including those with ≥ 3.5 INR, suggesting that TG might inform anticoagulant intensity in patients on high-intensity VKA.¹⁸ Additionally, increased peak TG in a subgroup of APS patients, despite therapeutic INR and CFX, suggests that TG might be able to identify an ongoing prothrombotic state in VKA-treated patients.¹⁸ The normalized TG-derived peak height/lag time ratio identifies LA in plasma with high sensitivity, irrespective of the patient’s treatment with oral anticoagulants, if TG is performed on mixtures of patient plasma and pooled normal plasma.⁵⁹

5.1 Anticoagulant monitoring of DOACs in APS

As in non-APS patients, DOAC anticoagulant effect is not routinely monitored, but there is increasing demand on the laboratory to have the capacity to adequately assess DOAC anticoagulant effect (pharmacodynamics) or levels (pharmacokinetics) in emergent or routine situations (Table 2).⁶⁸ DOAC levels should also be measured if testing for LA (as they may cause false positives and negatives), after the DOAC has been stopped for at least 48 hours and longer if there is renal impairment.⁶⁹ Alternatively, LA can be measured after preanalytical neutralizing or adsorbing DOAC.⁷⁰ As with VKAs, DOACs should be used with caution in thrombotic APS patients with thrombocytopenia, particularly as based on their half-lives,³⁷ DOAC anticoagulant effects persist for > 48 hours; LMWH should be considered instead.

Drug-calibrated chromogenic anti-Xa assays are suitable to provide quantitation of direct anti-Xa inhibitors^{33, 37, 68} in APS patients. Drug-calibrated dTT and anti-FIIa chromogenic methods are suitable ways to provide quantitation of dabigatran.^{33, 37, 68} The ECT, also generally suitable to monitor dabigatran, should not be used in patients with LA-hypoprothrombinemia syndrome.⁵⁰ The ECA has prothrombin added and therefore could be used in this context.

TG has been found to be sensitive to all kinds of anticoagulants and may best represent inter-individual response more so than exploring merely plasma drug concentrations.⁶⁸ In non-APS patients with VTE, rivaroxaban was reported to provide effective anticoagulation, as assessed by inhibition of TG and in vivo markers of coagulation activation.⁷¹ Notably, reduced TG parameters have been reported to persist at 12 hours after DOAC intake, despite low residual DOAC plasma levels < 50 ng/mL.⁷² In the RAPS (Rivaroxaban in Antiphospholipid Syndrome) RCT in APS patients with VTE requiring standard-intensity anticoagulation, the primary outcome, percentage change in ETP for rivaroxaban, did not reach the non-inferiority threshold. However, peak thrombin was significantly lower on rivaroxaban and the authors concluded that the overall TG curve, in which the higher ETP reflects the altered reaction kinetics with rivaroxaban, was not indicative of increased thrombotic risk.⁷³

A more standardized method for TG measurement has become available recently: the ST Genesia (Diagnostica Stago, Asnières sur Seine Cedex, France), an automated analyzer for TG. This may be a “game changer” for the assessment of DOAC anticoagulant effects in the future.⁶⁸ TG parameters measured with ST Genesia correlate with drug levels of anti-Xa DOACs. Peak thrombin and velocity index are of special interest for the determination of residual anticoagulant effect at low drug levels. For dabigatran-treated patients, only lag time shows a correlation with the dabigatran plasma levels.⁷⁴ As with VKAs, TG is not currently suitable to monitor anticoagulant intensity in DOAC-treated APS patients.

6.1 Anticoagulant monitoring of LMWH in APS

The aPTT is not usable for LMWH monitoring (Table 3). The aPTT will be variably prolonged at peak levels, and often only shows a mild prolongation over that due to the LA, depending on the sensitivity of the individual reagents used for LMWH and LA.⁸² The anti-Xa chromogenic assay, unaffected by LA, is the most appropriate assay for monitoring LMWH in APS patients, as in non-APS patients. Evidence on LMWH dosing and monitoring in APS is lacking and may be extrapolated from use in general population patients. Dose reduction of LMWH, following the manufacturer's summary of product characteristics, should be undertaken in accordance with the SmPC in APS patients with creatinine clearance (CrCl; Cockcroft-Gault [C&G]) <30mL/min, with anti-Xa monitoring considered to check for accumulation. The most appropriate LMWH dosing in patients with obesity remains undefined.⁸³

LMWH dosing should be adjusted for APS-associated thrombocytopenia. Therapeutic dosing in thrombocytopenic APS patients may be extrapolated from use in other thrombotic situations, such as in cancer patients. ISTH guidance recommends full-dose anticoagulation following acute VTE in cancer patients with a high risk of thrombus propagation and a platelet count > 50 x 10⁹/L and advise platelet transfusion support in those with thrombocytopenia to maintain platelet counts of 40-50 x 10⁹/L.⁸⁴ In APS patients, immunomodulatory treatment may increase platelet counts sufficiently for therapeutic anticoagulation. Platelet counts should be monitored while on LMWH, during the first 14 days in patients at intermediate risk for HIT,⁷⁷ and beyond, to guide consideration of dose reduction, if required, for secondary thromboprophylaxis.

The optimal dosage regimen of LMWH for treatment of VTE in pregnancy (once daily versus split-dosing, ie, two divided doses) and the value and role of anti-Xa monitoring merits further investigation.⁸⁵

7.1 Anticoagulant monitoring of UFH in APS

Monitoring of therapeutic doses of intravenous UFH is recommended and can be achieved using an anti-Xa assay or an aPTT with a therapeutic range corresponding to 0.3 to 0.7 IU/mL anti-Xa activity (Table 3).^{33, 45} The aPTT is typically prolonged in APS, particularly if reagents sensitive to LA are used. Thus, for UFH monitoring by aPTT, a relatively LA-insensitive aPTT reagent should be used.^{33, 69} If the baseline aPTT is prolonged, the aPTT should not be used for UFH monitoring as this may overestimate heparin effect, resulting in subtherapeutic anticoagulation and a potentially increased risk of recurrent thromboembolism. BSH guidelines state that the anti-Xa assay could be the test of choice to monitor UFH.³³ There are great advantages to use of the chromogenic anti-Xa assay instead of the aPTT in all cases in which a prolonged baseline aPTT is suspected; this would include LA positive patients. Particularly as aPTTs may be unreliable due to LA-induced effects and the potential for patient safety issues as a result, anti-Xa rather than the aPTT is more appropriate to monitor UFH in APS patients.

Unusually high doses of UFH may be required to achieve a therapeutic aPTT, ie, “heparin resistance,”⁴⁵ in the acute situation such as in CAPS, with potential mechanisms including raised levels of FVIII and fibrinogen, and other heparin-binding acute phase reactants, which lead to aPTT shortening⁹¹ Decreased antithrombin levels are also associated with “heparin resistance,”⁹² reported to be reversed by supplementation with antithrombin concentrate in patients undergoing cardiac surgery.⁹³ Anti-Xa levels, interpreted in the context of concomitant aPTTs, are advised.⁸⁶ Platelet counts should be monitored during UFH treatment and HIT considered should the platelet count fall.⁷⁷

The activated clotting time (ACT), a phospholipid-dependent test widely used to monitor UFH during cardio-pulmonary bypass (CPB), may be prolonged by LA. There is no consensus in the literature regarding optimal anticoagulation for APS patients in this challenging situation. Reported approaches in CPB include targeting twice the baseline ACT, measuring heparin concentrations by an automated protamine titration device, using preoperative in vitro heparin-ACT titration curves, and also measuring heparin anti-Xa levels;⁹⁴ and ACT coupled with thromboelastography.⁹⁵

8.1 Fondaparinux

The specific anti-Xa activity of fondaparinux is seven times higher than that of LMWH (~700 units/mg and 100 units/mg, respectively), and its half-life after subcutaneous injection is longer than that of LMWH (17 and about 4 hours, respectively).⁴⁵ The successful use of fondaparinux was reported in two patients with APS in combination with mycophenolate mofetil,⁹⁶ and in three APS patients with recurrent thrombosis.⁹⁷

The risk of fondaparinux-related HIT is low (<0.1%).⁷⁷ In vitro studies show no significant inhibitory effect of fondaparinux on osteoblast proliferation or activity,⁹⁸ although clinical data in this context are lacking. Fondaparinux has been reported to cross the placenta, resulting in low but measurable anti-Xa activity in umbilical-cord blood.^{85, 99} A review of maternal and pregnancy outcomes of 65 pregnancies in women using fondaparinux (11 on therapeutic dose) reported that it was well tolerated and the rate of pregnancy complications was similar to that observed in the general population.¹⁰⁰ Fondaparinux is assigned to category B by the U.S. Food and Drug Administration (FDA), ie, that it should only be given during pregnancy when need has been clearly established (https://www.accessdata.fda.gov/drugsatfdadocs/label/2008/020883s014lbl.pdf).

8.2 Anticoagulant monitoring of fondaparinux in APS

Fondaparinux dosing is weight based. It is nearly completely eliminated by renal clearance and therefore contraindicated in patients with severe renal impairment (CrCl [C&G] <30mL/min; Table 4). Routine anticoagulation monitoring is not recommended; however, it may be useful in some circumstances, such as after re-thrombosis while on fondaparinux to assess patient adherence to inform consideration of an escalated dose. Fondaparinux-specific anti-Xa assays are available, although a therapeutic anti-Xa range has not been established.

8.3 Direct thrombin inhibitors: argatroban and bivalirudin

Argatroban, half-life 45 minutes after intravenous (IV) injection,¹⁰¹ is not renally excreted and is therefore an option in patients with HIT and severe renal impairment,⁷⁷ and with CAPS.⁸⁶ It is metabolized in the liver via the cytochrome P450 3A4/5 enzyme system and should therefore be used with caution in patients with hepatic dysfunction.¹⁰¹ Argatroban has been used successfully in case reports during pregnancy;^{102, 103} however, there are no controlled data in human pregnancy. Bivalirudin has a half-life of 25 minutes after IV injection.¹⁰⁴ Approximately 20% is renally excreted and dose reduction should be considered in patients with moderate to severe renal impairment(45) or hepatic dysfunction.⁷⁷ Bivalirudin is suggested as an option in patients with acute HIT or subacute HIT (ie, normal platelets, positive functional test for HIT) who require cardiovascular surgery that cannot be delayed.⁷⁷ Both argatroban and bivalirudin are assigned to category B by the FDA.

8.4 Anticoagulant monitoring of argatroban and bivalirudin in APS

Argatroban is routinely monitored by the aPTT.¹⁰⁶ However, the aPTT could give unreliable results due to LA-induced effects, as well as during the acute phase (Table 4).^{28, 29} In such situations, the dTT, which has shown comparable results using different analyzers, should be considered for the measurement of blood concentrations of the direct thrombin inhibitors argatroban^{107, 108} and bivalirudin.¹⁰⁸ A therapeutic range for either remains to be established. ECT or ecarin chromogenic assays have been considered as alternatives, although the ECT has been demonstrated to have a slightly sigmoid response curve to concentration with argatroban¹⁰⁹ and is falsely elevated by prothrombin deficiency,¹⁰⁶ present in APS patients with LA-hypoprothrombinemia.⁵⁰ The ECA, although unaffected by the presence of factor deficiencies (due to the addition of prothrombin) or LA, has been shown to have a low sensitivity to argatroban.¹⁰⁸ In critically ill patients, it has been suggested that TT and rotational thromboelastography (ROTEM) parameters may provide better correlation to argatroban and lepirudin (unavailable since 2012) plasma concentrations than APTT.¹¹⁰ Drug-calibrated chromogenic anti-II assays are suitable to monitor argatroban and bivalrudin,³³ and in APS patients, would not be affected by any LA effects.

8.5 Danaparoid sodium and anticoagulation monitoring in APS

Danaparoid sodium is a heparinoid that catalyzes the antithrombin inhibition of factor Xa (Table  4). It has a long half-life of approximately 25 hours, which is a disadvantage for patients requiring urgent surgery or invasive procedures.⁴⁵ It has low cross-reactivity in vitro with sera containing HIT (heparin-platelet factor 4) antibodies and is suggested as an option for HIT.⁷⁷ In APS, a case report describes thrombosis of a cardiopulmonary bypass circuit in an APS patient with acute HIT, despite therapeutic danaparoid.¹¹¹ Danaparoid was reported to be effective for improving the live birth rate and safe for patients with obstetric APS.¹¹² Monitoring is with a danaparoid calibrated anti-Xa assay.³³ Danaparoid has also been assigned to category B by the FDA.