Advertisement

In vitro validation of chromogenic substrate assay for evaluation of surrogate FVIII-activity of emicizumab

Open AccessPublished:January 12, 2023DOI:https://doi.org/10.1016/j.thromres.2023.01.007

      Abstract

      [Introduction] Emicizumab, a bispecific antibody mimicking activated factor VIII (FVIII), is increasingly used in prophylaxis against bleeding in hemophilia A. Human factor-based chromogenic substrate assay (hCSA) shows concentration-dependency between emicizumab and reported FVIII activity. However, the assay measurement settings have not been optimized for emicizumab, and the reported FVIII activity cannot be directly referred as surrogate FVIII activity.
      [Materials and Methods] For in vitro validation of hCSA-reported surrogate FVIII activity, we compared the equation curves for emicizumab concentration with surrogate FVIII activity using spiked plasma in the thrombin generation assay (TGA), hCSA, and clot waveform analysis (CWA). Then, we generated conversion equations for hCSA-reported surrogate FVIII value to that of TGA. We also assessed the additive effect of rFVIII onto 340 nM (i.e., 50 μg/mL) emicizumab using the same assays.
      [Results] With 1:20 diluted plasma, halving hCSA-reported surrogate FVIII activity can be approximated to that in TGA triggered by the extrinsic pathway reagent (27.3 IU/dL vs. 13.9 IU/dL) under therapeutic emicizumab concentration. Both in TGA and hCSA, the additive effect of added FVIII on therapeutic emicizumab concentration (340 nM) was maintained at low levels of FVIII but gradually decreased at higher levels.
      [Conclusions] Surrogate FVIII activity can be estimated simply by halving hCSA-reported FVIII value, and the additive effect of FVIII on emicizumab diminishes at high concentrations. Based on our in vitro study, a clinical study is currently being conducted to compare individual variation of surrogate FVIII activity in hCSA and TGA.

      Keywords

      1. Introduction

      Hemophilia A (HA) is an inherited bleeding disorder associated with factor VIII (FVIII) deficiency or abnormalities. Emicizumab (Roche, Basel, Switzerland) mimics activated FVIII function of activating factor X (FX) by binding to activated factor IX (FIXa) and FX, and provides excellent prophylaxis against bleeding in people living with HA (PwHA) [
      • Shima M.
      • Hanabusa H.
      • Taki M.
      • et al.
      Factor VIII-mimetic function of humanized bispecific antibody in hemophilia A.
      ,
      • Callaghan M.U.
      • Negrier C.
      • Paz-Priel I.
      • et al.
      Long-term outcomes with emicizumab prophylaxis for hemophilia A with or without FVIII inhibitors from the HAVEN 1–4 studies.
      ]. The use of emicizumab has been increasing rapidly worldwide based on its effectiveness in PwHA (with or without anti-FVIII inhibitors), ease of administration through subcutaneous injection, and its significantly long half-life [
      • Oldenburg J.
      • Mahlangu J.N.
      • Kim B.
      • et al.
      Emicizumab prophylaxis in hemophilia A with inhibitors.
      ,
      • Young G.
      • Liesner R.
      • Chang T.
      • et al.
      A multicenter, open-label phase 3 study of emicizumab prophylaxis in children with hemophilia A with inhibitors.
      ,
      • Mahlangu J.
      • Oldenburg J.
      • Paz-Priel I.
      • et al.
      Emicizumab prophylaxis in patients who have hemophilia A without inhibitors.
      ,
      • Pipe S.W.
      • Shima M.
      • Lehle M.
      • et al.
      Efficacy, safety, and pharmacokinetics of emicizumab prophylaxis given every 4 weeks in people with haemophilia A (HAVEN 4): a multicentre, open-label, non-randomised phase 3 study.
      ].
      At present, there is no easy-to-access laboratory monitoring assay for PwHA treated with emicizumab. The conventional one-stage clotting assay (OSA) is assumed to be useless since the activated partial thromboplastin time (aPTT) is excessively shortened and the measured FVIII activity by OSA is falsely high [
      • Schmitt C.
      • Adamkewicz J.I.
      • Xu J.
      • et al.
      Pharmacokinetics and pharmacodynamics of emicizumab in persons with hemophilia a with factor VIII inhibitors: HAVEN 1 study.
      ]. Although the modified one-stage FVIII assay with emicizumab calibrators is promising [
      • Shinohara S.
      • Saito T.
      • Noguchi-Sasaki M.
      • Ishiwata T.
      • Morris M.
      Evaluation of emicizumab calibrator and controls with a modified one-stage FVIII assay on an automated coagulation analyzer. Int Soc Thromb Haemost.
      ], its clinical availability is limited. In the clot waveform analysis (CWA), although clotting time (i.e., aPTT) is extremely shortened by emicizumab, clotting velocity or acceleration, especially adjusted acceleration using PT/APTT mixed agents, provides better reflection of emicizumab-induced coagulation [
      • Nogami K.
      • Matsumoto T.
      • Tabuchi Y.
      • et al.
      Modified clot waveform analysis to measure plasma coagulation potential in the presence of the anti-factor IXa/factor X bispecific antibody emicizumab.
      ].
      Although the thrombin generation assay (TGA) is not approved as a routine clinical coagulation assay due to the combination of difficulty in interpretation, low reproducibility, and low accessibility, its usefulness has been repeatedly reported with regard of emicizumab [
      • Schmitt C.
      • Adamkewicz J.I.
      • Xu J.
      • et al.
      Pharmacokinetics and pharmacodynamics of emicizumab in persons with hemophilia a with factor VIII inhibitors: HAVEN 1 study.
      ,
      • Müller J.
      • Pekrul I.
      • Pötzsch B.
      • Berning B.
      • Oldenburg J.
      • Spannagl M.
      Laboratory monitoring in emicizumab-treated persons with hemophilia a.
      ,
      • Kitazawa T.
      • Esaki K.
      • Tachibana T.
      • et al.
      Factor VIIIa-mimetic cofactor activity of a bispecific antibody to factors IX/IXa and X/Xa, emicizumab, depends on its ability to bridge the antigens.
      ,
      • Leksa N.C.
      • Aleman M.M.
      • Goodman A.G.
      • Rabinovich D.
      • Peters R.
      • Salas J.
      Intrinsic differences between FVIIIa mimetic bispecific antibodies and FVIII prevent assignment of FVIII-equivalence.
      ]. Especially, the peak thrombin in TGA is considered the best coagulation marker because it shows bell-shaped concentration-dependency between emicizumab concentration and surrogate FVIII activity, which strictly correlates with the formation of FIX-emicizumab-FX ternary complex [
      • Kitazawa T.
      • Esaki K.
      • Tachibana T.
      • et al.
      Factor VIIIa-mimetic cofactor activity of a bispecific antibody to factors IX/IXa and X/Xa, emicizumab, depends on its ability to bridge the antigens.
      ,
      • Leksa N.C.
      • Aleman M.M.
      • Goodman A.G.
      • Rabinovich D.
      • Peters R.
      • Salas J.
      Intrinsic differences between FVIIIa mimetic bispecific antibodies and FVIII prevent assignment of FVIII-equivalence.
      ]. In TGA, tissue factor (TF) is widely used in clinical practice, [
      • Kizilocak H.
      • Marquez-Casas E.
      • Malvar J.
      • Carmona R.
      • Young G.
      Determining the approximate factor VIII level of patients with severe haemophilia A on emicizumab using in vivo global haemostasis assays.
      ] compared with the limited use of activated factor XI (FXIa) for evaluation of emicizumab or other non-FVIII therapies [
      • Muto A.
      • Yoshihashi K.
      • Takeda M.
      • et al.
      Anti-factor IXa/X bispecific antibody (ACE910): hemostatic potency against ongoing bleeds in a hemophilia A model and the possibility of routine supplementation.
      ,
      • Sampei Z.
      • Igawa T.
      • Soeda T.
      • et al.
      Identification and multidimensional optimization of an asymmetric bispecific IgG antibody mimicking the function of factor VIII cofactor activity.
      ,
      • Kitazawa T.
      • Igawa T.
      • Sampei Z.
      • et al.
      A bispecific antibody to factors IXa and X restores factor VIII hemostatic activity in a hemophilia A model.
      ,
      • Uchida N.
      • Sambe T.
      • Yoneyama K.
      • et al.
      A first-in-human phase 1 study of ACE910, a novel factor VIII-mimetic bispecific antibody, in healthy subjects.
      ,
      • Waters E.K.
      • Hilden I.
      • Sorensen B.B.
      • Ezban M.
      • Holm P.K.
      Thrombin generation assay using factor XIa to measure factors VIII and IX and their glycoPEGylated derivatives is robust and sensitive.
      ]. However, there is little information about the best reagent that can provide better evaluation of the emicizumab-evoked coagulation, or whether TGA can report surrogate FVIII activity consistent with the well-established value in animal studies [
      • Muto A.
      • Yoshihashi K.
      • Takeda M.
      • et al.
      Anti-factor IXa/X bispecific antibody ACE910 prevents joint bleeds in a long-term primate model of acquired hemophilia A.
      ].
      The chromogenic substrate assay (CSA) seems to be the best alternative to TGA as an easy-to-access monitoring assay for coagulation activity under emicizumab, due to the following favorable properties; 1) high accessibility following the recommendation of the World Federation of Hemophilia (WFH) [
      • Srivastava A.
      • Santagostino E.
      • Dougall A.
      • et al.
      WFH guidelines for the management of hemophilia, 3rd edition.
      ], 2) it measures the amount of produced substance as a result of the reaction, rather than coagulation time, which is affected to a large extent by emicizumab, 3) it shows concentration-dependency correlation between surrogate FVIII activity and emicizumab concentration, when it employs the human factor-based reagents instead of bovine factor [
      • Schmitt C.
      • Adamkewicz J.I.
      • Xu J.
      • et al.
      Pharmacokinetics and pharmacodynamics of emicizumab in persons with hemophilia a with factor VIII inhibitors: HAVEN 1 study.
      ,
      • Leksa N.C.
      • Aleman M.M.
      • Goodman A.G.
      • Rabinovich D.
      • Peters R.
      • Salas J.
      Intrinsic differences between FVIIIa mimetic bispecific antibodies and FVIII prevent assignment of FVIII-equivalence.
      ,
      • Hartmann R.
      • Feenstra T.
      • Knappe S.
      • Schrenk G.
      • Scheiflinger F.
      • Dockal M.
      A bispecific antibody demonstrates limited measurability in routine coagulation assays.
      ,
      • Lowe A.
      • Kitchen S.
      • Jennings I.
      • Kitchen D.P.
      • Woods T.A.L.
      • Walker I.D.
      Effects of emicizumab on APTT, FVIII assays and FVIII inhibitor assays using different reagents: results of a UK NEQAS proficiency testing exercise.
      ,
      • Adamkewicz J.I.
      • Chen D.C.
      • Paz-Priel I.
      Effects and interferences of emicizumab, a humanised bispecific antibody mimicking activated factor VIII cofactor function, on coagulation assays.
      ]. However, while hCSA employs different dilution rates based on FVIII concentration [
      • Novembrino C.
      • Boscolo Anzoletti M.
      • Mancuso M.E.
      • Shinohara S.
      • Peyvandi F.
      Evaluation of an automated chromogenic assay for factor VIII clotting activity measurement in patients affected by haemophilia A.
      ], the dilution rate or other settings of hCSA have not yet been optimized for emicizumab. More importantly, hCSA-reported FVIII activity cannot be directly referred as surrogate FVIII activity of emicizumab [
      • Muto A.
      • Yoshihashi K.
      • Takeda M.
      • et al.
      Anti-factor IXa/X bispecific antibody ACE910 prevents joint bleeds in a long-term primate model of acquired hemophilia A.
      ], and no correction method for CSA is available yet.
      Evaluation of the additive effect of FVIII or bypassing agents on emicizumab is also an unsolved clinical issue. Although idiotype antibodies to emicizumab can be used to evaluate residual FVIII activity and inhibitors, they are not useful for real-time monitoring in acute clinical settings, due to their limited accessibility. While some investigators reported the usefulness of TF-triggered TGA in evaluating the additive effects of FVIII on emicizumab [
      • Kizilocak H.
      • Marquez-Casas E.
      • Malvar J.
      • Carmona R.
      • Young G.
      Determining the approximate factor VIII level of patients with severe haemophilia A on emicizumab using in vivo global haemostasis assays.
      ,
      • Bravo M.I.
      • Raventos A.
      • Perez A.
      • Costa M.
      • Willis T.
      Non-additive effect on thrombin generation when a plasma-derived factor VIII/von Willebrand factor (FVIII/VWF) is combined with emicizumab in vitro.
      ], there are only a few studies on the usefulness of hCSA in real-time monitoring under the coexisting state of emicizumab and FVIII. In this regard, although hCSA can measure the coagulation activity of both FVIII and emicizumab, it is not clear whether the measured coagulation activity under the combination of FVIII and emicizumab by hCSA is comparable to that of TGA.
      The aim of our in vitro validation study was to establish a protocol that allows the conversion of hCSA-reported surrogate FVIII activity by approximation to TGA-estimated one, with the following steps; 1) comparison of equation curves between emicizumab concentration and surrogate FVIII activity using spike plasma in TGA, hCSA, and CWAs, and 2) establishing a method to convert hCSA to TGA either experimentally and/or mathematically. In addition, we compared the additive effects of FVIII on emicizumab among hCSA, TGA, and CWA.

      2. Material and methods

      2.1 Sample preparations and materials

      In sample series 1, the final target concentration of emicizumab (HEMLIBRA®; Chugai Pharmaceutical Co., Tokyo, Japan) ranged from 0.01 to 10 μM (= 10 and 10,000 nM). In sample series 2, rFVIII (Rurioctocog alfa; Advate®; Takeda Pharmaceutical Co., Tokyo) concentration was set between 0.01 and 6 IU/mL (= 1 to 600 IU/dL). The target final concentrations of emicizumab and advate in spiked sample series 1 and 2 were achieved by pre-dilution with saline. The samples were prepared by mixing of 150 μL of the above solutions with 2850 μL of FVIII-deficient plasma samples (see below). For the spiking of advate, we referred to the values measured accurately by hCSA at our laboratory. For spiking of emicizumab, the concentration of the commercial formation, 60 mg/0.4 mL, was referred. To evaluate the additive effects of emicizumab, we prepared sample series 3 in which a fixed dose of 340 nM emicizumab was further spiked into sample series 2, by adding emicizumab at 1/200 volume of the total volume. The FVIII-deficient plasma sample (drawn from patients with severe hemophilia A, with FVIII activity of <0.01 IU/mL) was purchased from George King Bio-Medical Inc. (Overland Park, KS). Emicizumab was kindly provided by Chugai Pharmaceutical. All human coagulation factors used in this study (Factor IX, Factor IXa, Factor X, Factor Xa, and Factor Xia) were purchased from Enzyme Research Labs Inc. (South Bend, IN).

      2.2 Thrombin generation assay

      The TGA was performed with both extrinsic (TF) and intrinsic (FXIa) triggers. The TF trigger reagent (PPP-reagent-low), FluCa reagent, and human thrombin calibrator were purchased from Stago (Asnières-sur-Seine, France). FXIa trigger reagent was prepared at our laboratory for concentrations of FXIa and liposomal phospholipid of 0.16 nM and 20 μM, respectively. The liposomalized phospholipid, phosphatidylserine (PS): phosphatidylcholine (PC): phosphatidylethanolamine (PE) (10:60:30), was purchased from Sysmex (Kobe, Japan) [
      • Okuda M.K.N.
      • Uemura Y.
      Development of new APTT reagent using synthetic phospholipids.
      ]. TGA was measured using a calibrated automated thrombogram (CAT) in a Fluoroskan Ascent reader (Thermo Fisher Scientific., Waltham, MA) equipped with a dispenser [
      • Hemker H.C.
      • Giesen P.
      • Al Dieri R.
      • et al.
      Calibrated automated thrombin generation measurement in clotting plasma.
      ]. In this assay, 80 μL of the sample plasma was dispensed into the wells of 96-well microtiter plates, followed by the addition of 20 μL of the trigger reagent before incubation for 10 min at 37 °C. Thrombin generation was initiated by the addition of 20 μL of FluCa solution (Stago). All CAT parameters, lag time (lag time), endogenous thrombin peak (ETP), peak thrombin, and time to peak (ttPeak) were calculated using the Thrombinoscope™ software (version 5.0). All samples were measured in triplicate.

      2.3 Chromogenic substrate assay

      CSA was performed using two different kits; 1) the Revohem™ FVIII Chromogenic kit (Sysmex Corporation, Kobe, Japan), containing human coagulation factors similar to that used in the BIOPHEN™ FVIII:C (Hyphen Biomed, Neuville-sur-Oise, France), and 2) the Chromogenix Coatest SP Factor VIII kit (DiaPharma, West Chester Township, OH) containing bovine coagulation factors.
      All samples analyzed with the Revohem™ FVIII Chromogenic were processed using the CS-2400 analyzer [
      • Novembrino C.
      • Boscolo Anzoletti M.
      • Mancuso M.E.
      • Shinohara S.
      • Peyvandi F.
      Evaluation of an automated chromogenic assay for factor VIII clotting activity measurement in patients affected by haemophilia A.
      ]. Briefly, 50 μL of the sample plasma was diluted 1-, 5-, 10-, 20- (as recommended by the manufacturer), and 40-fold with the kit's dilution buffer. After incubation at 37 °C for 40 s, 50 μL of human FX solution was added before incubation at 37 °C for 20 s. This was followed by the addition of 50 μL of the human FIXa/thrombin/phospholipid mixture and incubation at 37 °C for 310 s. Finally, 50 μL of the chromogenic substrate was added and the first-minute change in optical density (OD) was considered to represent an index of FXa generation. All samples were measured in triplicate.
      All measurements with Chromogenix Coatest SP Factor VIII were performed manually by Multiskan FC mass spectrometry (Thermo Fisher Scientific) with a microtiter plate. Human FIXa was spiked into the reagent at 1 and 90 nM, while human FX was spiked at 140 nM. The protocol included initial dilution of the plasma sample at 1-, 2-, 5-, and 121-fold using the kit's buffer. In the next step, 25 μL of the diluted plasma was mixed with 50 μL of the kit reagent, spiked into human FIXa/FX, and incubated for 4 min at 37 °C. After adding 25 μL of pre-warmed CaCl2 reagent enclosed with the kit, the reaction mixture was incubated for 5 min at 37 °C. Finally, 50 μL of pre-warmed chromogenic substrate provided with the kit was added. The ODs were measured every 15 s up to 10 min. All samples were measured in quadruplicate.

      2.4 Clot waveform analysis

      A modified CWA was performed using the CS-2400 analyzer (Sysmex) with Thrombocheck APTT-SLA (Sysmex) as the aPTT reagent. In the standard CWA, the baseline light transmittance is influenced by interfering substances, including chyle, hemoglobin, and bilirubin. Furthermore, the light transmittance immediately after coagulation is influenced by fibrinogen concentration and fibrin clot density [
      • Nogami K.
      • Matsumoto T.
      • Tabuchi Y.
      • et al.
      Modified clot waveform analysis to measure plasma coagulation potential in the presence of the anti-factor IXa/factor X bispecific antibody emicizumab.
      ]. To avoid the effects of these factors, we modified the standard CWA as follows. Briefly, the state of no light intensity was defined as 0 % in the conventional CWA, whereas in the modified CWA, the minimum light intensity in the postcoagulation state was defined as 0 %. The adjusted CWA parameters; coagulation velocity (dT/dt) and maximum coagulation velocity (ad|min1|), were generated automatically by the CS-2400 program.

      2.5 Surface plasmon resonance

      Surface plasmon resonance (SPR) was employed for kinetic analysis of emicizumab using the Biacore T200 instrument (Cytiva, MA) with FIX, FIXa, FX, and FXa. Briefly, emicizumab was immobilized on the Protein A Biacore sensor chip (Cytiva) at around 1000 resonance units. After the immobilization, the analytes (FIX, FIXa, FX, and FXa) were injected into the sensor chip at a concentration ranging from 0.031 to 2.0 μM under a flow rate of 30 μL/min. The association and dissociation times of the analytes were 120 and 120 s, respectively. The binding assay was carried out in 10 mM HEPES (pH 7.5), 150 mM NaCl, and 2.5 mM CaCl2 containing 0.05 % (v/v) Tween-20 at 25 °C. At the end of each cycle, a regeneration procedure was performed with 10 mM glycine-HCl at pH 1.5. Data were analyzed and the kinetic parameters were computed by the BIAevaluation software (Cytiva).

      2.6 Estimation of FVIII activity surrogate to emicizumab concentration

      Surrogate FVIII activity for each emicizumab concentration was calculated in peak thrombin of TGA, dOD of CSA, and ad|min1| of the CWA. To derive the equations of surrogate FVIII activity, we employed commonly used units (IU/dL for FVIII and nM for emicizumab, respectively). For example, the surrogate FVIII activity equation by TGA was estimated as follows. First, from the measured peak values using FVIII-deficient plasma spiked with rFVIII or emicizumab, two regression equations were derived; with the peak thrombin set as the dependent variable and FVIII or emicizumab as the independent variable.
      Log10peak thrombinnM=fa(Log10rFVIIIIU/dL


      Log10peak thrombinnM=fb(Log10emicizumabnM


      Second, these two equations were integrated into one “surrogate FVIII activity” equation, where FVIII was set as the dependent variable and emicizumab as the independent variable.
      Log10rFVIIIIU/dL=fc(Log10emicizumabnM


      Similarly, we obtained two sets of equations for CSA and CWA, where dOD (FXa generation) was set as the dependent variable in CSA and ad|min1| in CWA. Finally, each set of two equations for CSA and CWA was integrated into one surrogate FVIII activity equation. The best fit curve was derived by comparing coefficient of correlation (r) and Pearson's correlation test. When the result of emicizumab and FVIII showed a linear pattern, a single regression analysis was performed.

      2.7 Reproducibility

      The precision of the method was assessed by analyzing the intra-day precisions of three to four replicates at concentrations ranging from 0.01 and 10 μM per sample of emicizumab and 0.01 to 6 IU/mL per sample of rFVIII using spiked plasma. Precision was represented by the coefficient of variation (CV).

      2.8 Statistical analysis

      Data analysis was performed with Microsoft Excel software (Redmond, WA) and EZR (Saitama Medical Center, Jichi Medical University, Saitama, Japan). The latter is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria) [
      • Kanda Y.
      Investigation of the freely available easy-to-use software 'EZR' for medical statistics.
      ]. The mean values and standard deviations/standard errors were calculated for each group.

      3. Results

      3.1 Thrombin generation assay

      Fig. 1A-B shows the thrombograms of rFVIII and emicizumab in TF-triggered TGA, respectively. The lag time was almost constant irrespective of FVIII concentration, whereas it was unexpectedly prolonged in emicizumab concentration-dependent manner (Fig. 1C). The endogenous thrombin potential (ETP) was unmeasurable at low rFVIII concentrations, and showed poor reproducibility at all emicizumab concentrations (Fig. 1D). With regard to peak thrombin (Fig. 1E), rFVIII showed dose-dependent correlation between peak thrombin and rFVIII concentration, whereas emicizumab showed a bell-shaped relationship, with a peak around 2 μM (Fig. 1E), as reported previously [
      • Kitazawa T.
      • Esaki K.
      • Tachibana T.
      • et al.
      Factor VIIIa-mimetic cofactor activity of a bispecific antibody to factors IX/IXa and X/Xa, emicizumab, depends on its ability to bridge the antigens.
      ]. However, a linear correlation was noted between peak thrombin and emicizumab at concentration range of 0.01–2 μM. The time to peak (ttPeak) was concentration-dependent for rFVIII but not for emicizumab (Fig. 1F). The CVs ranged from 3.98 to 15.83.
      Fig. 1
      Fig. 1Thrombin generation assay (TGA) of rFVIII (0.01–6 IU/ mL) and emicizumab (0.01–10 μM) spiked hemophilia A (HA) plasma triggered by tissue factor (TF) as the extrinsic trigger reagent. (A-B): Representative thrombograms of rFVIII (A) and emicizumab (B). (C-F): Dose-response by lag time (C), endogenous thrombin potential (ETP) (D), peak thrombin (E), and time to peak (ttPeak) (F). Data are mean ± SEM. rFVIII, recombinant factor VIII; Emi, emicizumab.
      The thrombograms of rFVIII and emicizumab by FXIa-triggered TGA were different from those of TF-triggered TGA (Fig. 2A-B ). At low concentrations, emicizumab showed markedly shorter lag time than rFVIII (Fig. 2C). The ETP could not be measured under low concentrations of rFVIII and emicizumab (Fig. 2D). The peak thrombin also showed a bell-shaped curve for emicizumab, similar to that seen with the TF-triggered TGA, with a linear correlation with emicizumab at concentration range of 0.01–2 μM. However, the value of peak thrombin was significantly higher in rFVIII than emicizumab (Fig. 2E). Unlike TF-triggered TGA, the ttPeak also showed a bell-shaped curve for emicizumab. Moreover, both rFVIII and emicizumab showed almost parallel linear correlation at mid-range concentrations (FVIII: 0.1–1 IU/mL, emicizumab: 0.1–1 μM) (Figs. 2F). The CV ranged from 5.34 to 18.88.
      Fig. 2
      Fig. 2Thrombin generation assay (TGA) of rFVIII (0.01–6 IU/ mL) and emicizumab (0.01–10 μM) spiked hemophilia A (HA) plasma triggered by activated factor XI (FXIa) as the intrinsic trigger reagent. (A-B): Representative thrombograms of rFVIII (A) and emicizumab (B). (C-F): Dose-response by lag time (C), endogenous thrombin potential (ETP) (D), peak thrombin (E), and time to peak (ttPeak) (F). Data are mean ± SEM. rFVIII, recombinant factor VIII; Emi, emicizumab.

      3.2 Chromogenic substrate assay

      In CSA, FIXa, FX, thrombin, and phospholipid are artificially replaced irrespective of FVIII or emicizumab concentration in the plasma samples. For this reason, a lower plasma dilution (higher proportion of plasma) can be a reaction-limiting factor. To investigate the effect of plasma dilution on FXa generation of emicizumab, we evaluated FXa generation at various dilutions of spiked plasma samples. For emicizumab-spiked plasma samples, FXa generation did not show bell-shaped curve but rather linear correlations with both FVIII and emicizumab, although the generated amount was lower at low dilutions (Fig. 3A-B ). The average CV was 1.60.
      Fig. 3
      Fig. 3Human-factor based chromogenic substrate assay (hCSA) for FXa generation (dOD min-1) of rFVIII and emicizumab using BIOPHEN FVIII:C kit. The spiked plasma samples were diluted at 1-, 5-, 10-, 20- and 40-fold. Various concentrations of either rFVIII (0.01–6 IU/ mL (A), or emicizumab (0.01–10 μM (B), were spiked into hemophilia A (HA) plasma samples. (C) Chromogenic substrate assay for FXa generation (dOD min-1) of emicizumab by Chromogenix Coatest® SP Factor VIII containing bovine FIXa and FX as reagents spiked with 140 nM human FX and 1 or 90 nM human FIXa. The plasma samples were diluted 1-, 2-, 5- and 121-fold. Data are mean ± SEM. rFVIII, recombinant factor VIII; Emi, emicizumab; OD, optical density; FXa, activated factor X.
      As another approach for FIXa restriction, we used the Chromogenix Coatest SP Factor VIII kit, which does not react with emicizumab due to the bovine nature of the kit. The analysis was conducted in the presence of extremely small concentrations of human FIXa through the use of pre-selected sequential dilutions of the plasma samples (Fig. 3C). At low FIXa concentrations (FIXa = 1 nM), FXa generation was lower than that at normal FIX levels (90 nM), however, the relation between FXa generation and emicizumab did not show bell-shaped curve.

      3.3 Clot waveform analysis

      The results of CWA for rFVIII are shown in Supplementary Fig. 1A (transmittance) and 1B (first derivative; dT/dt). Ad|min1| correlated linearly with rFVIII (IU/mL) (Supplementary Fig. 1C). The results of CWA for emicizumab are shown in Supplementary Fig. 1D (transmittance) and 1E (first derivative; dT/dt)). The relation between Ad|min1| and emicizumab exhibited bell-shaped curve, with plateau at 1 μM of emicizumab (Supplementary Fig. 1F). The CV was 0.81–1.02.

      3.4 Surface plasmon resonance

      We also conducted kinetic analysis of emicizumab with FIX, FIXa, FX, and FXa using SPR. Supplementary Fig. 2A-E summarizes the results of SPR and shows the data of various kinetic parameters. All the dissociation constants (KD) for these interactions were within the μM range. The values of the association rate constant (kon) were in the order of 104, and those of the dissociation rate constant (koff) were in the order of 10−2. These parameters were almost identical to the values reported previously by other investigators [
      • Kitazawa T.
      • Esaki K.
      • Tachibana T.
      • et al.
      Factor VIIIa-mimetic cofactor activity of a bispecific antibody to factors IX/IXa and X/Xa, emicizumab, depends on its ability to bridge the antigens.
      ].

      3.5 Equation curves between surrogate FVIII activity and emicizumab

      In the log-log plot within the range of 0.01 to 2 μM of emicizumab concentrations, both FVIII and emicizumab showed linear correlation with the peak thrombin in TF-triggered TGA (r = 0.906 and 0.931 for FVIII and emicizumab, respectively) and FXIa-triggered TGA (r = 0.877 and 0.889, respectively), FXa generation in hCSA (r = 0.989 and 0.991, respectively), and ad|min1| by CWA (r = 0.966 and 0.961, respectively) (Fig. 4). For example, in the peak of TF-triggered TGA, two equations for log10 [Peak (nM)] were obtained: 0.4976 x log10 [FVIII (IU/dL)] + 0.7212, and 0.3404 x log10 [emicizumab (nM)] + 0.4286. These two linear equations were integrated with the linear equation for FVIII activity and emicizumab; log10 [FVIII (IU/dL)] = 0.6841× log10 [emicizumab (nM)] - 0.5880. Similarly, linear equations between surrogate FVIII activity and emicizumab concentration were obtained from FXIa-triggered TGA, hCSA, and CWA (Fig. 5A , Table 1).
      Fig. 4
      Fig. 4Relationships between each parameter and rFVIII (0.01–6 IU mL-1) or emicizumab (0.01–10 μM) in log-log plots by various clinical assays. (A-B): Peak thrombin by thrombin generation assay (TGA) using (A) tissue factor (TF) and (B) FXIa as the trigger reagent. (C) ad|min1| of the clot waveform analysis (CWA). (D-G): FXa generation (dOD min-1)) by human-factor based chromogenic substrate assay (hCSA). (D) 40-, (C) 20-, (D) 10-, (E) 5-fold dilution. rFVIII, recombinant factor VIII; Emi, emicizumab; FXIa, activated factor XI.
      Fig. 5
      Fig. 5Integrated equation curves between emicizumab and surrogate FVIII activity estimated from log-log plots in the thrombin generation assay (TGA), human factor-based chromogenic substrate assay (hCSA), and clot waveform analysis (CWA). X = Log10 (emicizumab (nM)), Y = Log10 (surrogate FVIII activity (IU/dL)). (A) Integrated equation curves from peak thrombin by TGA using the tissue factor (TF) trigger reagent, hCSA with 20-fold dilution, and CWA. The hCSA-derived equation curve has a slightly steeper coefficient than that of TGA-derived one, but hCSA-reported surrogate FVIII activity is almost double of the TGA-estimated one around 340 nM of emicizumab. (B) Integrated equation curves by hCSA with 40-, 20-, 10-, 5-fold dilution. Higher dilution was associated with higher value of surrogate FVIII activity for the same emicizumab concentration.
      Table 1Relational expression of FVIII and emicizumab concentration expressed by the first-order linear equation, measured by various methods.
      Measuring methodEquivalent expressionSurrogate FVIII activity (IU/dL)
      Emicizumab concentration
      70 nM (10 μg/mL)340 nM (50 μg/mL)700 nM (100 μg/mL)
      TGATrigger reagentParameter
      TFPeak thrombinY = 0.6841×-0.58804.7213.9222.81
      FXIa 0.16 nMPeak thrombinY = 2.0294×-4.85350.081.928.33
      CSADilution rate
      40Y = 0.8836×-0.72498.0432.561.51
      20Y = 0.8245×-0.65067.4327.3349.57
      10Y = 0.8856×-0.88175.6522.9243.45
      5Y = 0.8889×-1.02274.1416.8932.09
      CWAY = 0.7477× + 0.032525.984.2144.5
      X = Log10 Emicizumab (nM), Y = Log10 FVIII activity (IU/dL) and calculated surrogate FVIII activities (IU/dL) by each method to emicizumab at the concentration of 70, 340, 700 nM, representing the minimum, mean, and maximum of clinical steady state. The values of TF-triggred TGA and CSA with 20-fold dilution were highlighted with bold.
      TGA, thrombin generation assay; TF, tissue factor; FXIa, activated factor XI; CSA, chromogenic substrate assay; CWA, clot waveform analysis.
      Fig. 5A shows the integrated equations between surrogate FVIII and emicizumab in TGA, hCSA, and CWA. The hCSA-derived integrated equation curve had a slightly steeper coefficient than that of TGA-derived curve. Strictly, hCSA-based surrogate FVIII activity can be converted to the TGA-based value by the equation; log10 (FVIII-TGA) = 0.7890 x log10 (FVIII-hCSA) - 0.0483. For easier correction, Table 1 shows the surrogate FVIII activity (IU/dL) to emicizumab at 70, 340, 700 nM concentrations, in TF-TGA, FXIa-TGA, hCSA, and CWA. These concentrations corresponded to the minimum, mean, and maximum of therapeutic emicizumab concentrations, respectively. The values calculated from the peak thrombin of TF-triggered TGA (13.9 IU/dL) were the nearest to those reported previously [
      • Shima M.
      • Hanabusa H.
      • Taki M.
      • et al.
      Factor VIII-mimetic function of humanized bispecific antibody in hemophilia A.
      ,
      • Schmitt C.
      • Adamkewicz J.I.
      • Xu J.
      • et al.
      Pharmacokinetics and pharmacodynamics of emicizumab in persons with hemophilia a with factor VIII inhibitors: HAVEN 1 study.
      ]. The surrogate FVIII activity calculated by hCSA measured at 20-fold dilution was approximately twice that of TGA-triggered TF.
      In addition, the plasma dilution rate significantly affected the integrated equations in hCSA [Fig. 5B]. High dilutions resulted in upper shift of the integrated equations, and were associated with higher value of surrogate FVIII activity for the same emicizumab concentration (Table 1). Since hCSA generally reports FVIII activity value based on FXa generation (dOD), we also compared the hCSA-reported value of FVIII activity and dOD-based estimation value. The values showed good agreement (Fig. 6).
      Fig. 6
      Fig. 6Comparison of measured FVIII activity and estimated value by hCSA. (A) Relationship between spiked FVIII concentration and measured FVIII activity. (B) Relationship between spiked emicizumab concentration and measured FVIII activity and estimated value by hCSA. Data are mean ± SEM. rFVIII, recombinant factor VIII.

      3.6 Additive effects of FVIII on emicizumab

      The coagulation activity of the combination of emicizumab and FVIII was compared among TGA, hCSA, CWA and OSA using FVIII-spiked plasma samples containing therapeutic concentration of emicizumab (340 nM). The aPTT remained unchanged irrespective of FVIII dose (Fig. 7A ). In aPTT-based CWA, although ad|min1| slightly increased in proportion to the amount of additional FVIII, the additive effect of FVIII on emicizumab decreased markedly, and the values of ad|min1| for FVIII+emicizumab were paradoxically lower than those of FVIII monotreatment when additional FVIII was ≥1 IU/mL (Fig. 7B). On the other hand, both hCSA and TF-triggered TGA consistently showed stable additive effect at low concentrations of additional FVIII. However, both assays showed that the additive effect gradually decreased from 0.5 IU/mL of additional FVIII, and both FXa generation (hCSA) and the Peak (TF-triggered TGA) showed paradoxically lower values at about ≥1 IU/mL of additional FVIII (Figs. 7C, D). The peak thrombin of FXIa-triggered TGA was higher under rFVIII alone from the low concentration range of rFVIII (Fig. 7E).
      Fig. 7
      Fig. 7Additive effects of rFVIII (0.01–6 IU mL-1) with 340 nM of emicizumab.
      (A) APTT of one-stage assay by the thrombocheck APTT-SLA reagent. (B) ad|min1| of clot waveform analysis (CWA). (C) FXa generation (dOD min-1) of human-factor based chromogenic substrate assay (hCSA). (D-E): Peak thrombin of thrombin generation assay (TGA) triggered by (D) tissue factor (TF), (E) activated factor XI (FXIa). Data are mean ± SEM.; rFVIII, recombinant factor VIII; Emi, emicizumab; FXa, activated factor X; OD, optical density.

      4. Discussion

      In this in vitro validation study, we found that the reported FVIII values in hCSA with 1:20 diluted plasma can be simply approximated to that in TF-triggered TGA. Since hCSA-reported FVIII activity strictly correlates to the emicizumab concentration under a wide range of concentrations, hCSA with correction can directly evaluate the coagulation activity of emicizumab without measuring emicizumab concentration. Our validation study suggested that hCSA can be the best feasible and alternative assay to TGA. Meanwhile, in clinical settings, large fluctuations in emicizumab concentrations were observed in the four HAVEN 1–4 studies [
      • Yoneyama K.
      • Schmitt C.
      • Kotani N.
      • et al.
      A pharmacometric approach to substitute for a conventional dose-finding study in rare diseases: example of phase III dose selection for emicizumab in hemophilia A.
      ], and the HAVEN 4 study showed considerable variation in hCSA-reported FVIII activity in HA patients [
      • Pipe S.W.
      • Shima M.
      • Lehle M.
      • et al.
      Efficacy, safety, and pharmacokinetics of emicizumab prophylaxis given every 4 weeks in people with haemophilia A (HAVEN 4): a multicentre, open-label, non-randomised phase 3 study.
      ]. These studies suggested that even under identical blood emicizumab concentrations, surrogate FVIII activity can vary widely among individuals. Moreover, it is not clear whether the hCSA-reported surrogate FVIII activity can be uniformly approximated to that by TGA by halving the hCSA-based value. To evaluate these unsolved questions, a clinical study is currently being conducted to compare individual variation of surrogate FVIII activity in hCSA and TGA [
      • Masato B.
      Clinical validation of chromogenic substrate assay as a novel coagulation test for evaluating coagulation function in hemophilia a patients without inhibitor during treatment with emicizumab. Japan Registry of Clinical Trials.
      ].
      The peak thrombin measured by TF-triggered TGA showed surrogate FVIII activity of 11.2–16.1 IU/dL at the therapeutic range of emicizumab (36–61 μg/mL) and 13.9 IU/dL at emicizumab 50 μg/mL (≒340 nM), whereas the peak thrombin by FXIa-TGA showed significantly lower value of surrogate FVIII activity. In FXIa-TGA, low FVIII concentration presented higher value of peak thrombin than that of TF-TGA, which caused underestimation of surrogate FVIII activity of emicizumab. Other than peak thrombin, the lag time or time to Peak (ttPeak), which reflects the speed of coagulation, seemed useful according to our results, especially in FXIa-TGA. Both the lag time and ttPeak showed bell-shaped curves to emicizumab, however, they were not detected at low concentration range of both FVIII and emicizumab. In TF-TGA, the lag time was paradoxically prolonged at higher emicizumab concentrations. The ttPeak did not show emicizumab dose-dependency. ETP could not be obtained at the low concentration ranges of FVIII and emicizumab.
      Although hCSA exhibited concentration dependency between emicizumab concentration and surrogate FVIII activity, it did not show bell-shape curve at higher emicizumab concentrations, while TGA did. To obtain bell-shape curve between FXa generation and emicizumab concentration in hCSA, we applied several experimental settings. First, when the amount of FIXa or FX is limited for the reaction, the dilution ratio is considered a key factor in the reaction [
      • Lowe A.
      • Kitchen S.
      • Jennings I.
      • Kitchen D.P.
      • Woods T.A.L.
      • Walker I.D.
      Effects of emicizumab on APTT, FVIII assays and FVIII inhibitor assays using different reagents: results of a UK NEQAS proficiency testing exercise.
      ]. However, no bell-shape curve was noted at any dilution ratio tested in our study, and higher dilution ratio was associated with higher value of surrogate FVIII activities, similar to the finding of a previous study [
      • Lowe A.
      • Kitchen S.
      • Jennings I.
      • Kitchen D.P.
      • Woods T.A.L.
      • Walker I.D.
      Effects of emicizumab on APTT, FVIII assays and FVIII inhibitor assays using different reagents: results of a UK NEQAS proficiency testing exercise.
      ]. The BIOPHEN™ FVIII:C (Hyphen) employs reagents and procedures similar to those of Revohem™ FVIII Chromogenic, but uses 1:40 diluted plasma for high range of FVIII and 1:10 diluted plasma for low range. Our results suggested that BIOPHEN™ FVIII:C with two different dilution rates cannot be used for evaluation of emicizumab because: 1) hCSA with different dilutions yields slightly different values for surrogate FVIII activity, and 2) the dilution rate cannot be determined in advance when emicizumab concentration is unknown. However, because BIOPHEN™ FVIII:C using 1:20 diluted plasma showed good accuracy over a wide range (0.9 to 150 IU/dL) of FVIII:C [
      • Novembrino C.
      • Boscolo Anzoletti M.
      • Mancuso M.E.
      • Shinohara S.
      • Peyvandi F.
      Evaluation of an automated chromogenic assay for factor VIII clotting activity measurement in patients affected by haemophilia A.
      ], a single dilution rate (1:20) is applicable for BIOPHEN™ FVIII:C. Second, the amount of FIXa or FX can be a limiting factor for emicizumab-induced coagulative reaction, since emicizumab is known to produce small amounts of FIX- emicizumab-FX ternary complexes [
      • Kitazawa T.
      • Esaki K.
      • Tachibana T.
      • et al.
      Factor VIIIa-mimetic cofactor activity of a bispecific antibody to factors IX/IXa and X/Xa, emicizumab, depends on its ability to bridge the antigens.
      ,
      • Masato B.
      Clinical validation of chromogenic substrate assay as a novel coagulation test for evaluating coagulation function in hemophilia a patients without inhibitor during treatment with emicizumab. Japan Registry of Clinical Trials.
      ]. Especially, based on our SPR results of the higher binding affinity of FIXa to emicizumab (KD = 0.23 μM for FIXa vs. 1.65 μM for FX), FIXa is more likely to be depleted, compared with FX, under high concentrations of emicizumab. To evaluate the FIXa-depleting effects, we used different levels of human coagulation factors (1 and 90 nM of FIXa) in the Chromogenix Coatest SP Factor VIII containing bovine factors, and incorporated various plasma dilutions. In this part of the study, although reduction of FIXa (1 nM) decreased FXa generation irrespective of emicizumab concentration, no bell-shaped curve was noted. Third, since emicizumab does not require activation by thrombin, a short reaction time or measurement time can result in overestimation of FXa generation by emicizumab, compared with that by FVIII. However, a bell-shape curve was not observed under any time setting (data not shown).
      In comparison, using emicizumab at a concentration range of 0.01–2 μM, which covered the entire range of emicizumab concentrations used in previous clinical trials [
      • Schmitt C.
      • Adamkewicz J.I.
      • Xu J.
      • et al.
      Pharmacokinetics and pharmacodynamics of emicizumab in persons with hemophilia a with factor VIII inhibitors: HAVEN 1 study.
      ], both hCSA and TGA showed a linear correlation in log-log plots of surrogate FVIII activity and emicizumab concentration. Based on these findings, the surrogate FVIII activity by hCSA can be corrected to that of TGA using the equation; log10 (FVIII by TGA) = 0.7890 x log10 (FVIII by hCSA) - 0.0483. With this equation, hCSA-based FVIII can be easily approximated to TGA-based FVIII by using 50 % of the value (13.9 vs. 27.3 IU/dL in TF-TGA and hCSA, respectively) of therapeutic emicizumab concentration (i.e., 340 nM). For in vivo application of hCSA, further studies of clinical samples are needed to determine how far hCSA-based FVIII deviates from TGA-based FVIII in real-world patients.
      The ad|min1| measured by aPTT-based CWA showed bell-shaped curve in relation to emicizumab concentration, with overall curve properties resembling those of the peak thrombin by TGA. In addition, since the FVIII activity surrogate to emicizumab was about 6–7 times of that by TF-triggered TGA, aPTT-based CWA can also be easily approximated to TGA. However, ad|min1| was markedly higher when emicizumab was added to rFVIII (even when the amount of the added FVIII was as little as 1 IU/mL). This finding suggests that CWA is also problematic, similar to OSA, with regard to the evaluation of the additive effect of FVIII on emicizumab. In addition to the above drawback, CWA is also not easily available for routine daily clinical use.
      It remains controversial whether the concomitant use of FVIII with emicizumab confers additive coagulative activity or not. A previous study using modified CWA reported that the combination of emicizumab and FVIII provides additive benefits [
      • Nogami K.
      • Matsumoto T.
      • Tabuchi Y.
      • et al.
      Modified clot waveform analysis to measure plasma coagulation potential in the presence of the anti-factor IXa/factor X bispecific antibody emicizumab.
      ,
      • Nogami K.
      • Soeda T.
      • Matsumoto T.
      • Kawabe Y.
      • Kitazawa T.
      • Shima M.
      Routine measurements of factor VIII activity and inhibitor titer in the presence of emicizumab utilizing anti-idiotype monoclonal antibodies.
      ]. Another study based on TF-TGA reported that the additive effect of emicizumab diminished with increases in FVIII concentrations from 0.1 to 1 IU/mL or higher [
      • Bravo M.I.
      • Raventos A.
      • Perez A.
      • Costa M.
      • Willis T.
      Non-additive effect on thrombin generation when a plasma-derived factor VIII/von Willebrand factor (FVIII/VWF) is combined with emicizumab in vitro.
      ]. In our study, hCSA, CWA, and TF-TGA showed consistent results, where the additive effect decreased with increases in FVIII concentrations. Specifically, the additive effects of FVIII diminished from 0.5 IU/mL of additional FVIII, and completely disappeared at 1 IU/mL. Surprisingly, the effects decreased paradoxically following the addition of >1 IU/mL of FVIII. These findings raise concern on the currently available protocols used for hemostatic management of serious bleeding (e.g., intracranial, gastrointestinal) or major surgeries, where 0.8–1 IU/mL of additional FVIII is needed, in addition to emicizumab. For precise assessment of the additive effect of FVIII on emicizumab, careful and large clinical studies based on the pharmacodynamics relationship between the two are needed.
      The underlying mechanism of the fall in the additive effect of FVIII is not clear. Given that emicizumab has a significantly lower binding affinity than FVIII [
      • Kitazawa T.
      • Esaki K.
      • Tachibana T.
      • et al.
      Factor VIIIa-mimetic cofactor activity of a bispecific antibody to factors IX/IXa and X/Xa, emicizumab, depends on its ability to bridge the antigens.
      ], some studies speculated that high doses of additional FVIII may competitively inhibit the binding of emicizumab to FIXa or FX [
      • Bravo M.I.
      • Raventos A.
      • Perez A.
      • Costa M.
      • Willis T.
      Non-additive effect on thrombin generation when a plasma-derived factor VIII/von Willebrand factor (FVIII/VWF) is combined with emicizumab in vitro.
      ]. More interestingly, all assays, apart from OSA, consistently showed that >1 IU/mL of FVIII in combination with emicizumab was paradoxically associated with lower coagulation activity compared to when the same dose was used alone. Furthermore, a previous study using PT/APTT-based CWA showed that high emicizumab concentrations were associated with loss of the additive effect even at small doses of FVIII [
      • Nogami K.
      • Matsumoto T.
      • Tabuchi Y.
      • et al.
      Modified clot waveform analysis to measure plasma coagulation potential in the presence of the anti-factor IXa/factor X bispecific antibody emicizumab.
      ]. The results of these studies highly suggest that emicizumab can inhibit the binding of FVIII to FIXa or FX. Despite the lower affinity, since emicizumab binds to FIXa and FX faster than FVIII, a small proportion of emicizumab may competitively inhibit the binding of FVIII to FIXa or FX, especially in assays using intrinsic factors.
      The present study has certain limitations. First, although the binding affinity of FX to emicizumab is relatively lower than that to FIXa, the residual amount of unbound FX can be a reaction-limiting factor. Nevertheless, we could not evaluate the dose-effect of FX. Second, the additive effect of emicizumab and FVIII was evaluated only at a single concentration of emicizumab. To assess whether emicizumab conversely inhibit FVIII binding to FIXa or FX, the additive effect should be evaluated using low and high emicizumab concentrations. Third, although hCSA and TGA consistently showed gradual decrease in the additive effect of FVIII, we did not compare in detail the values between the two assays. The correction method of hCSA to TGA in the emicizumab/FVIII combination needs careful and further investigation.
      In conclusion, we provided preliminary validation data of a simple easy-to-access and real-time monitoring assay for emicizumab based on hCSA. When hCSA used 1:20 diluted plasma, FVIII-surrogate activity of emicizumab was easily estimated by halving hCSA-reported value over a wide range of emicizumab concentration (0.01–2 μM). The additive effect of FVIII on emicizumab is diminished at higher doses. Further clinical studies are needed to determine the clinical applicability of hCSA. In this regard, a clinical study is currently being conducted to compare individual variation of surrogate FVIII activity in hCSA and TGA [
      • Masato B.
      Clinical validation of chromogenic substrate assay as a novel coagulation test for evaluating coagulation function in hemophilia a patients without inhibitor during treatment with emicizumab. Japan Registry of Clinical Trials.
      ].
      The following are the supplementary data related to this article.
      Supplementary Fig. 1
      Supplementary Fig. 1Waveform patterns of modified clot waveform analysis (CWA) for rFVIII (0.01–6 IU/ mL) (A-C) and emicizumab (0.01–10 μM) (D-F), obtained by measuring APTT by the thrombocheck APTT-SLA reagent. Representative clot waveforms (left panels) and the first derivative curves (middle panels) are illustrated as modified waveform patterns. The derived Ad|min1| values are plotted according to the concentration of rFVIII and emicizumab (right panels). Data are mean ± SEM. rFVIII, recombinant factor VIII; Emi, emicizumab; ad|min1|, adjusted-|min1|.
      Supplementary Fig. 2
      Supplementary Fig. 2Binding kinetics of emicizumab to human factors. (A) FX; (B) FXa; (C) FIX; (D) FIXa. Fitted data are shown in red lines. FX, factor X; FXa, activated factor X; FIX, factor IX, FIXa, activated factor IX.

      CRediT authorship contribution statement

      TY, HI, KS, EK were involved in the concept and design of the study. TY conducted all experiments and data analysis and prepared the manuscript. HI was involved in CSA, OSA and CWA and supervised all experiments. SN and KT were involved in SPR. AM and KS were involved in TGA and supervised all experiments. MB and KA provided advice on all experiments. EK supervised the study design, experiments and the manuscript, and secured funding for the study. All authors critically revised the manuscript, commented on drafts of the manuscript, and approved the final manuscript.

      Funding

      This research was supported by Chugai Pharmaceutical Co. Emicizumab used in this study was kindly donated by Chugai Pharmaceutical. Chugai Pharmaceutical Inc. had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

      Declaration of competing interest

      HI received research grant from Sysmex, consulting fee from CSL-Behring and speaker honoraria from Sanofi, Bayer and CSL-Behring. KS was an endowed Assistant Professor funded by Baxter/Baxalta/Shire and CSL-Behring (until March 2020). KA received research grant from KM Biologics, consulting fee from Chugai Pharmaceutical, and honoraria from Chugai Pharmaceutical, Sanofi, Bayer, Takeda Pharmaceutical, Novo Nordisk, CSL Behring, KM Biologics, Pfizer, Fujimoto Pharmaceutical Corporation and Japan Blood Products Organization. EK received research grants from Chugai Pharmaceutical and CSL-Behring, and honoraria from Chugai Pharmaceutical, Sanofi, Bayer, Takeda Pharmaceutical, Novo Nordisk, Fujimoto Pharmaceutical Corporation and CSL Behring. All other authors declare no conflict of interest.

      Acknowledgment

      The authors acknowledge Daisuke Nosaka, Tetsuhiro Soeda, Suguru Harada, Yuki Oguchi and Takehisa Kitazawa, from Chugai Pharmaceutical Co., for their help in providing technical advice and interpretation of the significance of the results. We also thank Yuko Harada, Ryui Miyashita, Yushi Chikasawa, Takashi Muramatsu, and Takeshi Hagiwara, members of Department of Laboratory Medicine, Tokyo Medical University, for the excellent technical advice.

      References

        • Shima M.
        • Hanabusa H.
        • Taki M.
        • et al.
        Factor VIII-mimetic function of humanized bispecific antibody in hemophilia A.
        N. Engl. J. Med. 2016; 374: 2044-2053https://doi.org/10.1056/NEJMoa1511769
        • Callaghan M.U.
        • Negrier C.
        • Paz-Priel I.
        • et al.
        Long-term outcomes with emicizumab prophylaxis for hemophilia A with or without FVIII inhibitors from the HAVEN 1–4 studies.
        Blood. 2021; 137: 2231-2242https://doi.org/10.1182/blood.2020009217
        • Oldenburg J.
        • Mahlangu J.N.
        • Kim B.
        • et al.
        Emicizumab prophylaxis in hemophilia A with inhibitors.
        N. Engl. J. Med. 2017; 377: 809-818https://doi.org/10.1056/NEJMoa1703068
        • Young G.
        • Liesner R.
        • Chang T.
        • et al.
        A multicenter, open-label phase 3 study of emicizumab prophylaxis in children with hemophilia A with inhibitors.
        Blood. 2019; 134: 2127-2138https://doi.org/10.1182/blood.2019001869
        • Mahlangu J.
        • Oldenburg J.
        • Paz-Priel I.
        • et al.
        Emicizumab prophylaxis in patients who have hemophilia A without inhibitors.
        N. Engl. J. Med. 2018; 379: 811-822https://doi.org/10.1056/NEJMoa1803550
        • Pipe S.W.
        • Shima M.
        • Lehle M.
        • et al.
        Efficacy, safety, and pharmacokinetics of emicizumab prophylaxis given every 4 weeks in people with haemophilia A (HAVEN 4): a multicentre, open-label, non-randomised phase 3 study.
        Lancet Haematol. 2019; 6: e295-e305https://doi.org/10.1016/S2352-3026(19)30054-7
        • Schmitt C.
        • Adamkewicz J.I.
        • Xu J.
        • et al.
        Pharmacokinetics and pharmacodynamics of emicizumab in persons with hemophilia a with factor VIII inhibitors: HAVEN 1 study.
        Thromb. Haemost. 2021; 121: 351-360https://doi.org/10.1055/s-0040-1717114
        • Shinohara S.
        • Saito T.
        • Noguchi-Sasaki M.
        • Ishiwata T.
        • Morris M.
        Evaluation of emicizumab calibrator and controls with a modified one-stage FVIII assay on an automated coagulation analyzer. Int Soc Thromb Haemost.
        • Nogami K.
        • Matsumoto T.
        • Tabuchi Y.
        • et al.
        Modified clot waveform analysis to measure plasma coagulation potential in the presence of the anti-factor IXa/factor X bispecific antibody emicizumab.
        J. Thromb. Haemost. 2018; 16: 1078-1088https://doi.org/10.1111/jth.14022
        • Müller J.
        • Pekrul I.
        • Pötzsch B.
        • Berning B.
        • Oldenburg J.
        • Spannagl M.
        Laboratory monitoring in emicizumab-treated persons with hemophilia a.
        Thromb. Haemost. 2019; 119: 1384-1393https://doi.org/10.1055/s-0039-1692427
        • Kitazawa T.
        • Esaki K.
        • Tachibana T.
        • et al.
        Factor VIIIa-mimetic cofactor activity of a bispecific antibody to factors IX/IXa and X/Xa, emicizumab, depends on its ability to bridge the antigens.
        Thromb. Haemost. 2017; 117: 1348-1357https://doi.org/10.1160/TH17-01-0030
        • Leksa N.C.
        • Aleman M.M.
        • Goodman A.G.
        • Rabinovich D.
        • Peters R.
        • Salas J.
        Intrinsic differences between FVIIIa mimetic bispecific antibodies and FVIII prevent assignment of FVIII-equivalence.
        J. Thromb. Haemost. 2019; 17: 1044-1052https://doi.org/10.1111/jth.14430
        • Kizilocak H.
        • Marquez-Casas E.
        • Malvar J.
        • Carmona R.
        • Young G.
        Determining the approximate factor VIII level of patients with severe haemophilia A on emicizumab using in vivo global haemostasis assays.
        Haemophilia. 2021; 27: 730-735https://doi.org/10.1111/hae.14359
        • Muto A.
        • Yoshihashi K.
        • Takeda M.
        • et al.
        Anti-factor IXa/X bispecific antibody (ACE910): hemostatic potency against ongoing bleeds in a hemophilia A model and the possibility of routine supplementation.
        J. Thromb. Haemost. 2014; 12: 206-213
        • Sampei Z.
        • Igawa T.
        • Soeda T.
        • et al.
        Identification and multidimensional optimization of an asymmetric bispecific IgG antibody mimicking the function of factor VIII cofactor activity.
        PLoS One. 2013; 8e57479https://doi.org/10.1371/journal.pone.0057479
        • Kitazawa T.
        • Igawa T.
        • Sampei Z.
        • et al.
        A bispecific antibody to factors IXa and X restores factor VIII hemostatic activity in a hemophilia A model.
        Nat. Med. 2012; 18: 1570-1574https://doi.org/10.1038/nm.2942
        • Uchida N.
        • Sambe T.
        • Yoneyama K.
        • et al.
        A first-in-human phase 1 study of ACE910, a novel factor VIII-mimetic bispecific antibody, in healthy subjects.
        Blood. 2016; 127: 1633-1641https://doi.org/10.1182/blood-2015-06-650226
        • Waters E.K.
        • Hilden I.
        • Sorensen B.B.
        • Ezban M.
        • Holm P.K.
        Thrombin generation assay using factor XIa to measure factors VIII and IX and their glycoPEGylated derivatives is robust and sensitive.
        J. Thromb. Haemost. 2015; 13: 2041-2052https://doi.org/10.1111/jth.13134
        • Muto A.
        • Yoshihashi K.
        • Takeda M.
        • et al.
        Anti-factor IXa/X bispecific antibody ACE910 prevents joint bleeds in a long-term primate model of acquired hemophilia A.
        Blood. 2014; 124: 3165-3171https://doi.org/10.1182/blood-2014-07-585737
        • Srivastava A.
        • Santagostino E.
        • Dougall A.
        • et al.
        WFH guidelines for the management of hemophilia, 3rd edition.
        Haemophilia. 2020; 26: 1-158https://doi.org/10.1111/hae.14046
        • Hartmann R.
        • Feenstra T.
        • Knappe S.
        • Schrenk G.
        • Scheiflinger F.
        • Dockal M.
        A bispecific antibody demonstrates limited measurability in routine coagulation assays.
        Blood Coagul. Fibrinolysis. 2020; 31: 353-365https://doi.org/10.1097/MBC.0000000000000921
        • Lowe A.
        • Kitchen S.
        • Jennings I.
        • Kitchen D.P.
        • Woods T.A.L.
        • Walker I.D.
        Effects of emicizumab on APTT, FVIII assays and FVIII inhibitor assays using different reagents: results of a UK NEQAS proficiency testing exercise.
        Haemophilia. 2020; 26: 1087-1091https://doi.org/10.1111/hae.14177
        • Adamkewicz J.I.
        • Chen D.C.
        • Paz-Priel I.
        Effects and interferences of emicizumab, a humanised bispecific antibody mimicking activated factor VIII cofactor function, on coagulation assays.
        Thromb. Haemost. 2019; 119: 1084-1093https://doi.org/10.1055/s-0039-1688687
        • Novembrino C.
        • Boscolo Anzoletti M.
        • Mancuso M.E.
        • Shinohara S.
        • Peyvandi F.
        Evaluation of an automated chromogenic assay for factor VIII clotting activity measurement in patients affected by haemophilia A.
        Haemophilia. 2019; 25: 521-526https://doi.org/10.1111/hae.13746
        • Bravo M.I.
        • Raventos A.
        • Perez A.
        • Costa M.
        • Willis T.
        Non-additive effect on thrombin generation when a plasma-derived factor VIII/von Willebrand factor (FVIII/VWF) is combined with emicizumab in vitro.
        J. Thromb. Haemost. 2020; 18: 1934-1939https://doi.org/10.1111/jth.14887
        • Okuda M.K.N.
        • Uemura Y.
        Development of new APTT reagent using synthetic phospholipids.
        J. Jpn. Soc. Lab. Med. 2002; 3: 124-131
        • Hemker H.C.
        • Giesen P.
        • Al Dieri R.
        • et al.
        Calibrated automated thrombin generation measurement in clotting plasma.
        Pathophysiol. Haemost. Thromb. 2003; 33: 4-15https://doi.org/10.1159/000071636
        • Kanda Y.
        Investigation of the freely available easy-to-use software 'EZR' for medical statistics.
        Bone Marrow Transplant. 2013; 48: 452-458https://doi.org/10.1007/s40262-017-0616-3
        • Yoneyama K.
        • Schmitt C.
        • Kotani N.
        • et al.
        A pharmacometric approach to substitute for a conventional dose-finding study in rare diseases: example of phase III dose selection for emicizumab in hemophilia A.
        Clin. Pharmacokinet. 2018; 57: 1123-1134https://doi.org/10.1007/s40262-017-0616-3
        • Masato B.
        Clinical validation of chromogenic substrate assay as a novel coagulation test for evaluating coagulation function in hemophilia a patients without inhibitor during treatment with emicizumab. Japan Registry of Clinical Trials.
        • Nogami K.
        • Soeda T.
        • Matsumoto T.
        • Kawabe Y.
        • Kitazawa T.
        • Shima M.
        Routine measurements of factor VIII activity and inhibitor titer in the presence of emicizumab utilizing anti-idiotype monoclonal antibodies.
        J. Thromb. Haemost. 2018; 16: 1383-1390https://doi.org/10.1111/jth.14135