Universal Calibration Method for Assaying Enzymatic Inhibitors

Abstract

The present invention relates to a universal calibration method of use for assaying inhibitors of the same enzyme, for example for assaying inhibitors of an enzyme of blood coagulation. The invention also relates to the use of this universal calibration in a method for assaying a reversible or irreversible inhibitor of the enzyme in a biological sample. The invention also relates to the use of the universal calibration in a method for screening inhibitors of the enzyme.

Claims

1. A method for obtaining a universal calibration for assaying an inhibitor of an enzyme, the method comprising the following steps: a) determining the residual enzymatic activity in the stationary state for each of a plurality of mixtures containing the enzyme E and a substrate S that is labeled and specific to the enzyme, wherein: the substrate specific to the enzyme is labeled with a label having a detectable physical property, in each of the mixtures, the substrate is present in excess relative to the enzyme, the mixtures contain the same initial substrate concentration, [S].sub.0, the mixtures have the same total volume V and the same reaction medium M.sub.R, the mixtures have known and decreasing initial enzyme concentrations, the highest initial enzyme concentration being [E].sub.0, or have known and decreasing initial enzyme activities, the highest initial enzyme activity being A.sub.0, and the residual enzymatic activity in the stationary state of a mixture is determined by the following steps: a1) mixing a solution of the enzyme E and a solution of the substrate S to obtain a mixture of known initial enzyme concentration, or of known initial enzyme activity, and of initial substrate concentration [S].sub.0, a2) measuring the value of the detectable physical property of the label and plotting, on a graph, the value of this physical property as a function of time in order to obtain a curve, the curve having a rectilinear portion corresponding to the stationary state, and a3) calculating the gradient of the rectilinear portion of the curve obtained in a2), wherein the gradient obtained in a3) is the residual enzymatic activity in the stationary state of the mixture; b) for each of the mixtures of step a), converting the initial enzyme concentration of the mixture into anti-enzyme activity expressed as a percentage by standardizing said initial enzyme concentration of the mixture relative to the highest initial enzyme concentration [E].sub.0, or converting the initial enzyme activity of the mixture into anti-enzyme activity expressed as a percentage by standardizing said initial enzyme activity of the mixture relative to the highest initial enzyme activity A.sub.0, c) creating a universal calibration curve by plotting, on a graph, for each mixture, the anti-enzyme activity determined in step b) as a function of the residual enzymatic activity in the stationary state obtained in step a).

2. The method according to claim 1, wherein the enzyme belongs to the class of hydrolases, to the class of lyases, or to the class of isomerases.

3. The method according to claim 1, wherein the inhibitor is a reversible direct inhibitor, a reversible indirect inhibitor, an irreversible direct inhibitor or an irreversible indirect inhibitor.

4. The method for obtaining a universal calibration according to claim 1, wherein in step b), for a mixture with an initial enzyme concentration [E], the anti-enzyme activity expressed as a percentage is calculated by the equation:
AntiEnzyme.sub.%=1([E]/[E].sub.0), and, for a mixture with an initial enzyme activity A, the anti-enzyme activity expressed as a percentage is calculated by the equation:
AntiEnzyme.sub.%=1(A/A.sub.0).

5. The method for obtaining a universal calibration as according to claim 1, wherein, in step c), the calibration curve is a straight line with the equation: ${\mathrm{AntiEnzyme}}_{\left(\%\right)}=1-\left(\frac{1}{v_0}v\right)$ wherein: AntiEnzyme.sub.(%) is the anti-enzyme activity expressed as a percentage, v is the residual enzymatic activity in the stationary state, 1/v.sub.0 is the gradient of the calibration curve, and v.sub.0 is the residual enzymatic activity in the stationary state observed in the absence of an inhibitor.

6. The method for obtaining a universal calibration according to claim 1, further comprising a step d) consisting in creating a conversion chart specific to an inhibitor of the enzyme E and which makes it possible to convert the anti-enzyme activity, determined for a sample to be tested, into the amount or concentration of inhibitor.

7. The method for obtaining a universal calibration according to claim 6, wherein step d) comprises the following sub-steps: d1) determining the residual enzymatic activity in the stationary state for at least two standardization mixtures each containing the inhibitor at a known initial concentration, the enzyme E at the initial concentration [E].sub.0 or at the initial activity A.sub.0, and the labeled substrate S specific to the enzyme at the concentration [S].sub.0, wherein: in each of the standardization mixtures, the enzyme is present in excess relative to the inhibitor, the standardization mixtures have the same volume V and the same reaction medium M.sub.R, and the residual enzymatic activity in the stationary state of a mixture is determined by the following steps: d1) mixing the inhibitor with a solution of the enzyme E and a solution of the substrate S in order to obtain a standardization mixture with a known initial concentration of inhibitor, d1) measuring the value of the detectable physical property of the label and plotting, on a graph, the value of this physical property as a function of time in order to obtain a curve, the curve having a rectilinear portion corresponding to the stationary state, and d1) calculating the gradient of the rectilinear portion of the curve obtained in d1), wherein the gradient obtained in step d1) is the residual enzymatic activity in the stationary state of the standardization mixture; d2) for each of the standardization mixtures, using the universal calibration curve obtained in step c) of the universal calibration method in order to determine the anti-enzyme activity of the mixture from the residual enzymatic activity in the stationary state measured in step d1) for the standardization mixture; and d3) creating a standard curve or chart, by plotting on a graph, for each standardization mixture, the initial concentration of inhibitor of the standardization mixture as a function of the anti-enzyme activity determined in step d2).

8. The method for obtaining a universal calibration according to claim 7, wherein the volume V is identical to the volume V.

9. The method for obtaining a universal calibration according to claim 7, wherein step d) also comprises the sub-step d4) consisting in determining the equation of the standard curve by a regression of the pairs (anti-enzyme activity, concentration of inhibitor) plotted on the graph in step d3).

10. The method for obtaining a universal calibration according to claim 7, wherein if the inhibitor is a reversible direct inhibitor, the equation of the standard curve is the following equation: ${\left[I\right]}_0=i\frac{{\mathrm{AntiEnzyme}}_{\left(\%\right)}}{1-{\mathrm{AntiEnzyme}}_{\left(\%\right)}}+e{\mathrm{AntiEnzyme}}_{\left(\%\right)}$ wherein: AntiEnzyme.sub.(%) is the anti-enzyme activity, [I].sub.0 is the concentration of inhibitor present in the sample to be tested, and i and e are two fixed constants specific to the inhibitor; and if the inhibitor is an irreversible indirect inhibitor, the equation of the standard curve is the following equation:
[I].sub.0=eAntiEnzyme.sub.(%) wherein: AntiEnzyme.sub.(%) is the anti-enzyme activity in the sample to be tested, [I].sub.0 is the concentration of inhibitor present in the sample, and e is a fixed constant specific to the inhibitor.

11. A method for assaying an inhibitor of an enzyme in a biological sample, comprising the following steps: 1) determining the residual enzymatic activity in the stationary state for a mixture to be tested of reaction medium M.sub.R, of volume V, and containing an aliquot of the biological sample containing the inhibitor, the enzyme E at an initial concentration [E].sub.0 or at an initial activity A.sub.0, and a labeled substrate S specific to the enzyme at an initial concentration [S].sub.0, wherein: the substrate specific to the enzyme is labeled with a label having a detectable physical property, and the residual enzymatic activity in the stationary state of the mixture to be tested is determined by the following steps: i) mixing the aliquot of the biological sample with a solution of the enzyme E and a solution of the substrate S to obtain a mixture of initial enzyme concentration, or [E].sub.0, or of initial enzyme activity A.sub.0, and an initial substrate concentration [S].sub.0, ii) measuring the value of the detectable physical property of the label and plotting, on a graph, the value of this physical property as a function of time in order to obtain a curve, the curve having a rectilinear portion corresponding to the stationary state, and iii) calculating the gradient of the rectilinear portion of the curve obtained in step ii), wherein the gradient obtained in step iii) is the residual enzymatic activity in the stationary state of the mixture to be tested; and 2) using the universal calibration curve obtained in step c) of the calibration method according to claim 1 in order to determine the anti-enzyme activity of the biological sample from the residual enzymatic activity in the stationary state measured for the mixture to be tested.

12. The assaying method according to claim 11, wherein the biological sample is a sample of blood, of plasma, of platelet-rich plasma, of platelet-poor plasma, or of plasma containing platelet or erythrocyte microparticles or any other cell.

13. The assaying method according to claim 12, wherein the biological sample is a platelet-poor plasma sample.

14. The method according to claim 11, wherein the enzyme is an enzyme of blood coagulation selected from coagulation factors, kallikrein and plasmin, preferably selected from factor IIa, factor Xa and plasmin.

15. The method according to claim 14, wherein the inhibitor is selected from antithrombin, heparin cofactor II, alpha-2-macroglobulin, hirudin, lepirudin, desirudin, rivaroxaban, apixaban, edoxaban, betrixaban, dabigatran, bivalirudin, argatroban, unfractionated heparins, low-molecular-weight heparins, pentasaccharides and danaparoid sodium.

16. The assaying method according to claim 11, wherein the volume V is identical to the volume V.

17. The assaying method according to claim 11, further comprising the following step: 3) converting the anti-enzyme activity expressed as a percentage and obtained in step 2) into a concentration of inhibitor using the chart specific to the inhibitor which was obtained in step d) of the calibration method according to claim 6.

18. The assaying method according to claim 17, wherein the volume V is identical to the volume V.

19. (canceled)

20. A method for estimating the hemorrhagic risk in a patient treated with a direct oral anticoagulant, the method comprising the following steps: determining the amount or the concentration of direct oral anticoagulant in a biological sample from the patient, using an assaying method according to claim 11; comparing this amount or concentration with a predetermined threshold; and concluding that there is a hemorrhagic risk if the amount or concentration of direct oral anticoagulant is greater than the predetermined threshold.

21. The method for estimating the hemorrhagic risk in a patient treated with a direct oral anticoagulant according to claim 20, further comprising the following step: determining the amount of anti-inhibitor compound to administer to the patient as a function of the amount or concentration of direct oral anticoagulant measured in the biological sample from the patient.

22. The method according to claim 20, wherein the patient treated with the direct oral anticoagulant is a patient suspected of having received an overdose of direct oral anticoagulant or is a patient recently having undergone a change in treatment from an antivitamin K medicament to the direct oral anticoagulant, or is a patient about to undergo a surgical intervention.

23. A screening method for identifying an inhibitor of an enzyme, comprising the following steps: 1) determining the residual enzymatic activity in the stationary state for a mixture of reaction medium M.sub.R, of volume V, and containing a test compound at an initial concentration [C], the enzyme E at an initial concentration [E].sub.0 or at an initial activity A.sub.0, and a labeled substrate S specific to the enzyme at an initial concentration [S].sub.0, wherein: the substrate specific to the enzyme is labeled with a label having a detectable physical property, the enzyme is present in the mixture in excess relative to the test compound, and the residual enzymatic activity in the stationary state of the mixture is determined by the following steps: i) mixing the test compound with a solution of the enzyme E and a solution of the substrate S to obtain a mixture of initial enzyme concentration [E].sub.0, or of enzyme activity A.sub.0, an initial substrate concentration [S].sub.0 and a concentration [C] of test compound, ii) measuring the value of the detectable physical property of the label and plotting, on a graph, the value of this physical property as a function of time in order to obtain a curve, the curve having a rectilinear portion corresponding to the stationary state, and iii) calculating the gradient of the rectilinear portion of the curve obtained in step ii), wherein the gradient obtained in step iii) is the residual enzymatic activity in the stationary state of the compound to be tested; 2) using the universal calibration curve obtained in step c) of the universal calibration method in order to determine the anti-enzyme activity of the test compound from the residual enzymatic activity in the stationary state measured for the mixture; and 3) comparing the anti-enzyme activity of the test compound determined in step 2) with the anti-enzyme activity determined under the same conditions for a standard inhibitor of the enzyme at a concentration [C], or comparing the anti-enzyme activity of the test compound determined in step 2) with a predetermined threshold, wherein the test compound is identified as an inhibitor of the enzyme if the anti-enzyme activity of the test compound is greater than the anti-enzyme activity of the standard inhibitor of the enzyme or if the anti-enzyme activity of the test compound is greater than the predetermined threshold.

24. The screening method according to claim 23, wherein the volume V is identical to the volume V.

25. The screening method according to claim 23, wherein the enzyme belongs to the class of the hydrolases or to the class of the lyases.

26. A kit for identifying an inhibitor of an enzyme E, comprising the enzyme E, a labeled substrate S specific to the enzyme, and instructions for carrying out a screening method according to claim 23.

27. A kit for assaying at least one of the inhibitors of an enzyme E in a biological sample, the kit comprising the enzyme E, a labeled substrate S specific to the enzyme, at least one inhibitor of the enzyme, and instructions for carrying out an assaying method according to claim 11.

28. The kit according to claim 27, further comprising at least one other inhibitor of the enzyme.

Description

FIGURE LEGENDS

[0284] FIG. 1: Universal calibrations. The graphs give the anti-Xa activity (%) as a function of the corresponding OD/min observed at 10 experimental points. (A) Broad-range methodology: the linear regression of these experimental points gives the equation of the straight line y=1.0530.999 x with a coefficient of determination R.sup.2=0.994. (B) Narrow-range methodology: the linear regression of these experimental points gives the equation of the straight line y=1.0341.081 x with a coefficient of determination R.sup.2=0.996.

[0285] FIG. 2: Conversion charts. These charts make it possible to convert the anti-Xa activity (%) measured into concentration expressed in ng/ml for each of the three direct oral anticoagulants (DOAs). (A) Broad-range methodology: non-linear regressions of equation 2 at several experimental points give, respectively, for rivaroxaban (.circle-solid.) i=0.73 and e=16.43, for apixaban (.box-tangle-solidup.) i=1.35 and e=14.66 and for edoxaban (.circle-solid.) i=1.14 and e=15.66. (B) Narrow-range methodology: non-linear regressions of equation 2 at several experimental points give, respectively, for rivaroxaban (.circle-solid.) i=1.42 and e=15.36, for apixaban (.box-tangle-solidup.) i=2.91 and e=13.23 and for edoxaban (.circle-solid.) i=1.98 and e=14.96.

[0286] FIG. 3: Rivaroxaban. (A) Broad-range methodology: the comparisons of the results of the assays of the levels of rivaroxaban measured on 20 overloads using the universal calibration principle according to the invention to the theoretical levels give the equation of the straight line y=6.91+0.98 x with a coefficient of determination R.sup.2=0.998. (B) Narrow-range methodology: the comparisons of the results of the assays of the levels of rivaroxaban measured on 8 overloads using the universal calibration principle according to the invention to the theoretical levels give the equation of the straight line y=1.63+0.96 x with a coefficient of determination R.sup.2=0.995.

[0287] FIG. 4: Apixaban. (A) Broad-range methodology: the comparisons of the results of the assays of the levels of apixaban measured on 24 overloads using the universal calibration principle according to the invention to the theoretical levels give the equation of the straight line y=10.41+0.97 x with a coefficient of determination R.sup.2=0.999. (B) Narrow-range methodology: the comparisons of the results of the assays of the levels of apixaban measured on 10 overloads using the universal calibration principle according to the invention to the theoretical levels give the equation of the straight line y=2.79+0.97 x with a coefficient of determination R.sup.2=0.999.

[0288] FIG. 5: Edoxaban. (A) Broad-range methodology: the comparisons of the results of the assays of the levels of edoxaban measured on 24 overloads using the universal calibration principle according to the invention to the theoretical levels give the equation of the straight line y=0.99 x+0.83 with a coefficient of determination R.sup.2=0.996. (B) Narrow-range methodology: the comparisons of the results of the assays of the levels of edoxaban measured on 10 overloads using the universal calibration principle according to the invention to the theoretical levels give the equation of the straight line y=1.00x0.24 with a coefficient of determination R.sup.2=0.997.

[0289] FIG. 6: Improving the regression of the universal calibration. The graphs present the anti-Xa activity (%) as a function of the corresponding OD/min observed at 10 experimental points using narrow-range methodology. (A) Linear regression: the linear regression of these experimental points gives the straight line with the equation: y=1.0341.081 x with a coefficient of determination R.sup.2=0.9963. (B) Regression with a second-order polynomial: the regression of these experimental points gives the polynomial of equation y=0.9780.789 x0.281 x.sup.2 with a coefficient of determination R.sup.2=0.9998.

[0290] FIG. 7: Dilution in plasmauniversal calibrations. The graphs present the anti-Xa activity (%) as a function of the corresponding OD/min observed at 10 experimental points using broad-range (.circle-solid.) and narrow-range (.box-tangle-solidup.) methodologies. (A) Dilution in buffer: the linear regressions give, in broad-range methodology (.circle-solid.), the equation of the straight line y=1.0530.999 x with a coefficient of determination R.sup.2=0.994 and, in narrow-range methodology (.box-tangle-solidup.), the equation of the straight line y=1.0341.081 x with a coefficient of determination R.sup.2=0.996. (B) Dilution in plasma: the linear regressions give, respectively, in broad-range methodology (.circle-solid.), the equation of the straight line y=1.0221.173 x with a coefficient of determination R.sup.2=0.998 and, in narrow-range methodology (.box-tangle-solidup.), the equation of the straight line y=1.0201.198 x with a coefficient of determination R.sup.2=0.998.

[0291] FIG. 8: Dilution in plasma: rivaroxaban. (A) Broad-range methodology: the comparisons of the results of the assays of the levels of rivaroxaban measured on 20 overloads using the universal calibration principle according to the invention with dilution in plasma to the theoretical levels give the equation of the straight line y=1.03+0.96 x with a coefficient of determination R.sup.2=0.999. (B) Narrow-range methodology: the comparisons of the results of the assays of the levels of rivaroxaban measured on 8 overloads using the universal calibration principle according to the invention with dilution in plasma to the theoretical levels give the equation of the straight line y=1.15+0.95 x with a coefficient of determination R.sup.2=0.995.

[0292] FIG. 9: Universal calibration. The graph gives the anti-Xa activity (%) as a function of the corresponding OD/min observed at 10 experimental points. The linear regression of these experimental points gave the equation of the straight line y=0.0940.518 x with a coefficient of determination R.sup.2=0.996.

[0293] FIG. 10: Conversion charts. The charts presented in this figure make it possible to convert the anti-Xa activity (%) measured into concentration in IU/ml for (A) unfractionated heparin: the linear regression of the experimental points gave the equation of the straight line y=2.354 x0.02569 with a coefficient of determination R.sup.2=0.992, and (B) low-molecular-weight heparin: the linear regression of these experimental points gave the equation of the straight line y=1.621 x0.02827 with a coefficient of determination R.sup.2=0.979.

[0294] FIG. 11: (A) Unfractionated Heparin. The comparison of the results of the assays of the levels of unfractionated heparin measured on 11 overloads using the universal calibration principle to the theoretical levels gave the equation of the straight line y=1.025 x0.01107 with a coefficient of determination R.sup.2=0.979. (B) Low-Molecular-Weight Heparin. The comparison of the results of the assays of the levels of low-molecular-weight heparin measured on 7 overloads using the universal calibration principle to the theoretical levels gave the equation of the straight line y=0.9764 x+0.01489 with a coefficient of determination R.sup.2=0.985.

EXAMPLE 1: REVERSIBLE DIRECT INHIBITORS

[0295] Vitamin K antagonists (VKA) are historically the first and only class of anticoagulants that are administered orally. The large amount of variability in the response of patients to this type of treatment, to which can be added numerous food and drug interactions, necessitate regular in vitro monitoring of the treatment, involving numerous blood samples being taken from the patient. This observation led the pharmaceutical industry to develop a new family of anticoagulants referred to as direct oral anticoagulants (DOAs) which, theoretically, do not require regular monitoring and only have relatively few food and drug interactions. This family of anticoagulants is split into two classes: [0296] anti-Xas, direct inhibitors of factor Xa; and [0297] anti-IIas, direct inhibitors of thrombin.

[0298] As indicated above, there are certain situations wherein assaying direct oral anticoagulants (DOAs) is useful, such as, for example, prior to an emergency surgical intervention, in suspected cases of overdose, or in the case of hemorrhage of unknown origin. The study presented here focuses on the family of the anti-Xas, the main molecules of which are: [0299] rivaroxaban, sold by Bayer/Janssen Pharmaceutical under the name Xarelto; [0300] apixaban, sold by Bristol-Myers Squibb/Pfizer under the name Eliquis; and [0301] edoxaban, sold by Daiichi Sankyo under the name Savaysa or Lixiana.

I. Experimental Protocols

1. Materials

[0302] The plasma samples (Etablissement Francais du Sang, La Plaine Saint-Denis, France) originated from healthy patients who were not following procoagulant or anticoagulant treatment. Different pools of plasma, dated January 2011, March 2012, April 2014, November 2014 and February 2015, were used for the experiments. Before producing the pools from plasma bags, tests were carried out in order to verify that the values of the prothrombin levels (PT), cephalin times with activator (CTA) and also the concentrations of coagulation factors were consistent. The plasma bags were subsequently thawed for 50 minutes at 37 C. and left at room temperature for 30 minutes to stabilize. A pool was produced and the plasma was agitated at 153 revolutions/min for 35 minutes before being divided up. The fractions were stored at approximately 70 C. Before use, they were thawed at 37 C. for 5 minutes.

[0303] The overloads of anticoagulant were produced from solutions of apixaban (Eliquis, Bristol-Myers Squibb, New York, United States) concentrated at 400 g/ml, solutions of edoxaban tosylate (Lixiana, Daiichi-Sankyo, Tokyo, Japan) concentrated at 500 g/ml and solutions of rivaroxaban (Xarelto, Bayer, Leverkusen, Germany) concentrated at 393 g/ml.

[0304] The reagents from the commercial kit STALiquid Anti-Xa (Diagnostica Stago, Asnires sur Seine, France), F.Xa (factor Xa of bovine origin) and substrate (MAPA-Gly-Arg-pNA) were used. These commercial reagents and also the pool of normal human plasma Pool Norm (Diagnostica Stago, Asnieres sur Seine, France), the Owren Koller buffer (OKB) (pH 7.35, STAOwren-Koller, Diagnostica Stago, Asnieres sur Seine, France) and the desorption solution (DU) STADesorb U (Diagnostica Stago, Asnieres sur Seine, France) were regenerated at room temperature for 30 minutes, as recommended by the manufacturer.

[0305] Dimethylsulfoxide (DMSO) (Carlo Erba, Val de Rueil, France) diluted to 5% in demineralized water was used for the experiments.

[0306] All the measurements were carried out on the STA automated device (Diagnostica Stago, Asnieres sur Seine, France) AUT00460 (Hasting number). The software version for the automated device used in these experiments was version v3.04.05.

2. Methods

[0307] Anti-Xa Enzymatic Assays.

[0308] The kinetics were monitored by colorimetry at 405 nm, detecting the release of para-nitroanilide (pNA). The optical density (OD) was measured every two seconds. Two methodologies were optimized for assaying the three direct oral anti-Xa anticoagulants: the broad-range methodology and the narrow-range methodology. This is because the broad-range methodology makes it possible to assay the anticoagulants over the whole of the desired range; however, since the measurement of low concentrations of these anticoagulants requires great precision a narrow-range methodology was therefore developed.

[0309] In broad range, 6.25 l of biological sample (plasma) were diluted with OKB in a final volume of 50 l. The microcuvette was incubated for 240 seconds at 37 C. in the presence of 150 l of substrate. The reaction was started by the addition of 150 l of F.Xa, incubated beforehand at 37 C. In narrow range, 25 l of sample were diluted with 25 l of OKB. The experimental conditions for addition of substrate and enzyme are the same as for the broad-range methodology. The gradients were determined between 50 and 80 seconds.

[0310] Universal Calibration.

[0311] The calibration was carried out from a range of F.Xa composed of 10 points ranging from 10% to 100% F.Xa, corresponding to an anti-Xa activity ranging from 90% to 0%. The range was created from reagents from the kit STA Liquid Anti-Xa. The F.Xa (2 ml, 1.8 ml, . . . , 0.2 ml) was diluted in OKB (0 ml, 0.2 ml . . . 1.8 ml) for a final volume of 2 ml. The pool of healthy plasmas dated 04/2014 was used for this experiment. 5% DMSO was diluted to 1/20th in the plasma. The results (measurements of the OD/min between 50 and 80 seconds) were determined in two replicates.

[0312] The gradients of the kinetics measured for the 10 calibration points make it possible to plot the graph of anti-Xa activity (%) as a function of the OD/min. The equation of the line obtained is used to convert the OD/min of the samples tested into anti-Xa activity as percentage. The equation of the line is given by a polynomial regression of first order or of second order.

[0313] Creation of the Conversion Charts.

[0314] The parameters of the chart, specific to each direct oral anticoagulant (DOA), were determined for both of the methodologies (broad-range methodology and narrow-range methodology) by a non-linear regression of equation 2 on a few experimental points. In broad range, the samples overloaded at 0, 100, 200, 300 and 400 ng/ml of DOA were used to create the chart associated with rivaroxaban. The samples overloaded at 0, 100, 200, 300, 400 and 500 ng/ml were used to create the charts associated with apixaban and with edoxaban. In narrow range, the samples overloaded at 0, 30, 65, and 100 ng/ml of DOA were used to create the chart associated with rivaroxaban. The samples overloaded at 0, 50, 100, and 150 ng/ml of DOA were used to create the charts associated with apixaban and with edoxaban. The OD/min was measured for each sample between 50 and 80 seconds.

[0315] Plasma Overloads.

[0316] The pool of plasma dated 04/2014 was used to produce the overloads. The stock solutions of direct oral anticoagulants (DOAs) were diluted in DMSO %. An intermediate solution at 10 g/ml was used in order to produce the dilution ranges for the different overloads. The dilution of the DOAs in the plasma was to 1/20th. The concentrations of the overloads were 0, 10, 20, 30, 40, 50, 65, 75, 100, 125, 150, 175, 200, 225, 250, 275, 300, 325, 350, 375, 400 ng/ml for rivaroxaban and 0, 10, 20, 30, 40, 50, 75, 100, 125, 150, 175, 200, 225, 250, 275, 300, 325, 350, 375, 400, 425, 450, 475, 500 ng/ml for apixaban and edoxaban. The final volume of the overloads was 1 ml. The solutions containing edoxaban were stored in darkness. The results (measurements of the OD/min between 50 and 80 seconds) were determined for each overload in three replicates.

[0317] Configuration of the Tests for the Universal Calibration.

[0318] A test configuration was created by F.Xa dilution for the broad-range and narrow-range methodologies. The experimental conditions are the same as those described in the section anti-Xa enzymatic assays.

[0319] Configuration of the Tests for the Creation of the Charts.

[0320] As above, the different measurements were obtained according to the broad-range and narrow-range methodologies described in the section anti-Xa enzymatic assays.

[0321] Configuration of the Tests for the Plasma Overloads.

[0322] The measurements on the plasmas overloaded with DOAs were carried out according to the broad-range and narrow-range methodologies described in the section anti-Xa enzymatic assays.

[0323] Determining the Limit of Blank.

[0324] The limit of blank (LoB) was determined according to a standard procedure. The five different plasmas used were those dated January 2011, March 2012, April 2014, November 2014 and February 2015. Four replicates were carried out each day for three days, on the same automated device. The tests were carried out for the broad-range and narrow-range methodologies.

[0325] Determining the Limit of Detection.

[0326] The limit of detection (LoD) was determined according to a standard procedure. Low-analyte plasmas were prepared. The overloads were carried out to 1/20th. The plasma dated 04/2014 was used. When the regression of the universal calibration was a first-order polynomial, the concentrations of analytes in broad range were 10 ng/ml of rivaroxaban and of apixaban and 15 ng/ml for edoxaban. In narrow range, the concentrations of the overloads were 5 ng/ml for the three anticoagulants. Five overloads were produced for a final volume of 900 l. When the regression of the universal calibration was a second-order polynomial, the concentrations of analytes in broad range and in narrow range were 2 ng/ml for the three direct oral anticoagulants (DOAs). The final volume of the overloads was 1000 l. Each overload was separated into three tubes. The samples were frozen at 70 C. until their use. The analyses were carried out over three days on the same automated device with four analytical replicates per day and per overload.

[0327] Anti-Xa Enzymatic Assays with Dilution in Plasma.

[0328] The experimental conditions are identical to those described in the preceding sections, except for the fact that the dilution was not carried out in buffer but in plasma. The diluent used was the Pool Norm. The universal calibration was nonetheless obtained by diluting F.Xa in OKB under the same conditions as those described in the section Universal calibrations.

II. Results

1. Universal Calibrations

[0329] FIG. 1 presents examples of universal calibrations according to the invention, obtained respectively in broad-range and narrow-range methodologies, which express the results in anti-Xa activity (%) as a function of the OD/min measured. The linear regressions of the different experimental points give the equations:

y=1.0530.999x(R.sup.2=0.994), and y=1.0341.081x(R=0.996).

[0330] The theoretical equation of these universal calibrations is given by the straight line of equation 1 as described above in the description; the results obtained here are in perfect agreement with this expression, that is to say that a linear regression is obtained having a coefficient of determination very close to 1 and an intercept point close to 1. It is interesting to note that the equations of these two straight lines are different. This is explained by the fact that the dilution of the sample is different between the two methodologies (1/8 for the broad-range methodology and 1/2 for the narrow-range methodology); a matrix effect is observed.

2. Conversion Charts

[0331] FIG. 2 illustrates the charts which make it possible to convert the anti-Xa activity (%) measured into concentration expressed in ng/ml for each of the three direct oral anticoagulants (DOAs). These charts were created by carrying out a non-linear regression of equation 2 on samples for which the pairs (anti-Xa activity, concentration of anticoagulant) were known. In broad-range methodology, these non-linear regressions have, respectively, coefficients of determination R.sup.2 of 0.997 for rivaroxaban, 0.995 for apixaban and 0.993 for edoxaban. In narrow-range methodology, these non-linear regressions have, respectively, coefficients of determination R.sup.2 of 0.993 for rivaroxaban, 0.997 for apixaban and 0.992 for edoxaban. All these coefficient of determination values highlight the relevance of the mathematical model developed. In addition, at equal concentrations, rivaroxaban has the highest in vitro anti-Xa activity whereas edoxaban has the lowest anti-Xa activity: this confirms that the concentration expressed in ng/ml is not a universal unit for assaying anti-Xa DOAs.

3. Assays of Plasmas Overloaded with Direct Oral Anticoagulants (DOAs)

[0332] In this section, the results of the assays of the levels of plasmas overloaded with DOAs (rivaroxaban, apixaban and edoxaban) were obtained using the universal calibration principle according to the invention and compared to the theoretical levels. The results are deemed to be satisfactory when the gradient of the linear regression is between 0.9 and 1.1 and when the coefficient of determination R.sup.2 is greater than or equal to 0.95.

[0333] Rivaroxaban.

[0334] FIG. 3 presents the results of the comparisons of the assays of the rivaroxaban levels measured using the universal calibration principle in broad-range methodology (20 overloads) and in narrow-range methodology (8 overloads) and compared to the theoretical levels. These comparisons gave, respectively, a gradient of 0.98 and 0.96 and a coefficient of determination of 0.998 and 0.995, thereby validating the results.

[0335] Apixaban.

[0336] FIG. 4 presents the results of the comparisons of the assays of the apixaban levels measured using the universal calibration principle in broad-range methodology (24 overloads) and in narrow-range methodology (10 overloads) and compared to the theoretical levels. These comparisons gave, respectively, a gradient of 0.97 and 0.97 and a coefficient of determination of 0.999 and 0.999, thereby validating the results.

[0337] Edoxaban.

[0338] FIG. 5 presents the results of the comparisons of the assays of the edoxaban levels measured using the universal calibration principle in broad-range methodology (24 overloads) and in narrow-range methodology (10 overloads) and compared to the theoretical levels. These comparisons give, respectively, a gradient of 0.99 and 1.00 and a coefficient of determination of 0.996 and 0.997, thereby validating the results.

4. Limits of Blank and Limits of Detection

[0339] In this section, the limits of blank (LoB) and the limits of detection (LoD) measured were obtained using the universal calibration principle in broad-range and narrow-range methodology. Moreover, a method for improving these values in narrow-range methodology is proposed. As a reminder, the decision threshold for allowing a surgical intervention in a patient taking a direct oral anticoagulant (DOA) treatment is a level less than or equal to 30 ng/ml (as a minimum for rivaroxaban, Pernod et al., Annales franaises d'anesthsie et de reanimation, 2013, 32(10): 691-700).

[0340] Broad-Range Methodology.

[0341] Table 1 below lists the limits of blank and the limits of detection obtained in broad-range methodology using the universal calibration principle according to the invention. The LoD values observed were respectively 21.5 ng/ml for rivaroxaban, 21.8 ng/ml for apixaban and 28.8 ng/ml for edoxaban. These results indicate that the broad-range methodology does not make it possible to measure low levels of anti-Xa DOAs with the precision required in the context of an emergency surgical intervention.

TABLE-US-00001 TABLE 1 Broad-range methodology - limits of blank and limits of detection. LoB (ng/ml) LoD (ng/ml) rivaroxaban 16.897 21.496 apixaban 16.613 21.723 edoxaban 20.802 28.734

[0342] Narrow-Range Methodology.

[0343] Table 2 below lists the limits of blank and the limits of detection obtained in narrow-range methodology using the universal calibration principle according to the invention. The LoD values observed were respectively 6.9 ng/ml for rivaroxaban, 6.9 ng/ml for apixaban and 8.7 ng/ml for edoxaban. These results demonstrate that the narrow-range methodology has sufficiently good performance to precisely measure very low levels of anti-Xa DOAs.

TABLE-US-00002 TABLE 2 Narrow-range methodology - limits of blank and limits of detection. LoB (ng/ml) LoD (ng/ml) rivaroxaban 5.352 6.849 apixaban 5.450 6.882 edoxaban 6.802 8.657

[0344] Optimization in Narrow-Range Methodology.

[0345] It has been demonstrated and validated experimentally that the universal calibration follows a straight line given by equation 1. However, empirically, a regression of the universal calibration by a second-order polynomial should express the experimental points better than a linear regression. This is confirmed experimentallysee FIG. 6. The regression of the universal calibration by a second-order polynomial makes it possible to improve the LoB and LoD in narrow-range methodology (cf. table 3 below) but does not impact the quality of the comparisons of the levels measured to the theoretical levels (results not shown).

TABLE-US-00003 TABLE 3 Optimization in narrow-range methodology - limits of blank and limits of detection. LoB (ng/ml) LoD (ng/ml) rivaroxaban 2.854 4.202 apixaban 2.759 4.001 edoxaban 3.574 4.955

5. Dilution in Plasma

[0346] The difference in the sample dilution value between the broad-range methodology (dilution=1/8) and narrow-range methodology (dilution=1/2) induces a matrix effect. A method is proposed here in order to overcome this. For this purpose, in the universal calibration, the sample is no longer diluted in buffer (OKB) but it is diluted in plasma (Pool Norm) with the factors of dilution remaining unchanged. FIG. 7 compares the appearance of the universal calibrations when the sample is diluted in buffer to the appearance of the universal calibrations when the sample is diluted in plasma. The linear regressions of the experimental points in broad-range methodology and in narrow-range methodology give, respectively, the equations of straight lines:

y=1.0530.999x and

y=1.0341.081x+

when the sample is diluted in buffer, and the equations of straight lines:

y=1.0221.173x and

y=1.0201.198x

when the sample is diluted in plasma.

[0347] These results indicate that, when the sample is diluted in plasma, the equations of the two straight lines of the universal calibrations are superimposed on one another. In this configuration, it is no longer necessary to carry out two calibrations but simply a single one that is common to the two methodologies (broad-range methodology and narrow-range methodology). It should also be highlighted that the comparison of the theoretical levels of rivaroxaban to the levels measured in overloaded plasmas using the universal calibration principle with dilution of the sample in plasma gives a gradient of 0.96 and a coefficient of determination of 0.999 in broad-range methodology and a gradient of 0.95 and a coefficient of determination of 0.995 in narrow-range methodology, see FIG. 8: these results demonstrate that it is possible to measure in vitro the levels of direct oral anticoagulants in samples diluted in plasma.

EXAMPLE 2: IRREVERSIBLE INDIRECT INHIBITORS

I. Experimental Protocols

1. Materials

[0348] The plasma samples (Etablissement Francais du Sang, La Plaine Saint-Denis, France) and the reagents in example 2 are the same as those used in example 1.

[0349] The overloads of heparins were produced from solutions of heparin calcium (Cari-Parine, Sanofi-Aventis, Paris, France), concentrated at 5000 IU/0.2 ml and of tinzaparin sodium (Innohep, Leo Pharma, Ballerup, Denmark) at 10 000 IU/0.5 ml.

2. Methods

[0350] Anti-Xa Enzymatic Assays.

[0351] The kinetics were monitored as in example 1. The methodology was optimized for the assay of unfractionated heparins and low-molecular-weight heparins. Two (2) l of plasma were diluted in OKB in a final volume of 100 ml. The microcuvette was incubated for 690 seconds at 37 C. in the presence of 100 ml of F.Xa. The measurement phase was triggered by the addition of 100 ml of substrate, incubated beforehand at 37 C. The gradients were determined between 20 and 50 seconds.

[0352] Universal Calibration.

[0353] The calibration was carried out as in example 1, except for the fact that the measurements were carried out between 20 and 50 seconds.

[0354] Creation of the Conversion Charts.

[0355] The parameters of the chart, specific to each heparin, were determined by a linear regression of equation 3 on a few experimental points. The samples overloaded at 0, 0.3, 0.6, and 1.0 IU/ml were used to create the chart associated with heparin calcium. The samples overloaded at 0, 0.2, 0.4, and 0.6 IU/ml were used to create the chart associated with tinzaparin sodium. The OD/min is measured for each sample between 20 and 50 seconds.

[0356] Plasma Overloads.

[0357] A pool of plasma was used to produce the overloads. The stock solutions of heparins were diluted in physiological serum. The concentrations of the overloads were 0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, and 1.0 IU/ml for heparin calcium and 0, 0.1, 0.2, 0.3, 0.4, 0.5, and 0.6 IU/ml for tinzaparin sodium. The final volume of the overloads was 0.5 ml. The results (measurements of the OD/min between 20 and 50 seconds) were determined for each overload in three replicates.

[0358] Configuration of the Tests for the Universal Calibration, Configuration of the Tests for Creating the Charts, and Configuration of the Tests for the Plasma Overloads.

[0359] The experimental conditions are the same as those described in the section anti-Xa enzymatic assays.

II. Results

1. Universal Calibration

[0360] FIG. 9 illustrates the universal calibration which expresses the results in anti-Xa activity (%) as a function of the OD/min measured. The linear regression of the different experimental points gives the equation y=0.9940.518 x=(R.sup.2=0.996). The theoretical equation of this calibration is given by the straight line of equation 1. The results obtained are in perfect agreement with this expression, that is to say a linear regression with a coefficient of determination very close to 1 and an intercept point close to 1.

2. Conversion Charts

[0361] FIG. 10 presents the charts created for converting the anti-Xa activity (%) measured into concentration in IU /ml for each of the two families of heparins. These charts were created by carrying out a linear regression on samples for which the pairs (anti-Xa activity, concentration of anticoagulant) were known. The linear regressions of the experimental points gave the equation of the straight line y=2.354 x0.01569 for unfractionated heparin and the equation of the straight line y=1.621 x0.028247 for low-molecular-weight heparin. These equations are in perfect harmony (straight line with an intercept close to zero) with equation 2. Moreover, equation 3 is theoretically valid when the heparin concentration is less than or equal to the antithrombin concentration. In the opposite case, the heparin is no longer measurable, given that it completely saturates the plasma antithrombin. This behavior can be seen in the charts of FIG. 10. Indeed, the linear regression only applies to anti-Xa activities less than or equal to 52.75% for unfractionated heparin and 52.85% for low-molecular-weight heparin. These values are extremely consistent, although they were determined independently. Thus, all the results obtained confirm the relevance of the mathematical model developed here.

3. Assays of Plasmas Overloaded with Heparins

[0362] The results of the assays of the levels of plasmas overloaded with unfractionated heparin and with low-molecular-weight heparin obtained using the universal calibration principle according to the present invention are presented here by comparison to the theoretical levels. The results were deemed to be satisfactory when the gradient of the linear regression is between 0.9 and 1.1 and when the coefficient of determination R.sup.2 is greater than or equal to 0.95.

[0363] Unfractionated Heparin.

[0364] FIG. 11(A) presents the results obtained using 11 unfractionated heparin overloads. The comparison of the levels measured using the universal calibration principle to the theoretical levels gave a gradient of 1.025 and a coefficient of determination of 0.987, which validates the results.

[0365] Low-Molecular-Weight Heparin.

[0366] FIG. 11(B) presents the results obtained using 7 low-molecular-weight heparin overloads. The comparison of the levels measured using the universal calibration principle to the theoretical levels gave a gradient of 0.9764 and a coefficient of determination of 0.985, which validates the results.