WO2023066867A1 - Method for analyzing properties of biomedical membranes - Google Patents

Abstract

The invention relates to a method for analysing the properties of biomedical membranes, in particular for the removal of active substances over time by adsorption, dialysis, filtration, or diafiltration.

Description

METHOD FOR ANALYZING THE PROPERTIES OF BIOMEDICAL MEMBRANES

The invention relates to the field of analyzing the properties of biomedical membranes, in particular dialysis, filtration or diafiltration membranes. It relates in particular to the study of the pharmacokinetics of drugs (and in particular the elimination of active substances over time) in continuous replacement methods such as continuous extrarenal purification, cardiopulmonary assistance, hepatic replacement , the specific and non-specific adsorption of substances present in the blood.

Since the 1950s, a number of medical devices requiring extracorporeal circulation (MDEC) have been developed and advances in technology have allowed their use to be extended to intensive care units. MDECs include continuous renal replacement therapy, extracorporeal membrane oxygenation, liver support in acute failure, and specific and nonspecific hemoperfusion. MDECs can induce complications and potentially dangerous side effects.

In particular, adsorption (sequestration or uptake) of drugs needed by critically ill patients by the membrane is an expected adverse effect that can lead to treatment failure, which is relatively understudied and poorly documented. Knowledge of MDEC-induced drug adsorption has been provided mainly by case reports.

The method described here proposes a method for evaluating the adsorption properties of membranes used in MDEC devices, by a perfectly controlled in vitro process, which makes it possible to accurately simulate different methods of supplementation, with different active substances. This method is of interest for emergency physicians or practicing in intensive care, who must be aware that the adsorption of drugs in MDECs can have a major impact on the clinical evolution of patients, but also for the pharmaceutical and manufacturers of MDECs, which must be able to characterize their products. The methods described here are likely to optimize and reduce the costs of pharmacological treatments in intensive care units, improving patient outcomes.

Among the extracorporeal methods used, we can cite: Dialysis, replacement treatment in case of degraded renal function. The principle is to rid the blood of waste or toxins and excess accumulated water in the body. By diffusion effect, the small molecules will cross the membrane, while the large molecules (often macromolecules) will be retained on one side. This method uses concentration gradients between the initial solution (blood) and the effluent.

Filtration or haemofiltration, method of extra-renal purification by convective transport (gradient of hydrostatic pressure). In this method, the blood is ultrafiltered and the collected product (the ultrafiltrate) is discarded along with the waste it contains. A reinfusion liquid compensates for the part of the removed plasma. This method makes it possible to concentrate certain solutes, in particular if the volume of the reinjection liquid is lower than the volume of blood extracted.

Hemodialysis, which combines dialysis and hemofiltration and uses diffusion and convection.

Hepatic dialysis, for subjects with hepatic insufficiency and aimed at ridding the blood of toxins which accumulate there due to the poor functioning of the liver. i) Oxygenation by extracorporeal membrane (acronym ECMO in English), in which the blood passes through an exchange compartment in which the CO ₂ dissolved in the blood is extracted from it, and O ₂ is supplied to it. We can cite ECMO via the veno-arterial (VA) route or the veno-venous (VV) route. In both cases, the blood drawn from the venous system is oxygenated outside the body, but in the case of VA, ECMO provides cardiorespiratory support, whereas it only provides respiratory support in the VV.

The different systems can be associated with each other during simultaneous use on the patient using two modes: in series (same blood circulation which passes through different devices in a row) or in parallel (each device has its own extracorporeal circulation).

For continuous extrarenal purification (CRTE) (filtration dialysis adsorption), the blood flow can range from 50 ml/min to 250 ml/min), and the diafiltration flow rates range from 500 to 8000 ml/h, the surface of exchange being adapted to the patient (child vs. adult): 0.5 to 1.5 m ² For ECMO, the flow rates are of the order of 3 to 5 I/min, with an exchange surface of 1 to 4 m ² . Thus, the ECMO has an exchange surface identical to that of the EERC but which can reach approximately 3 times this one, and a blood flow 20 times higher than those of the EERC.

In these different methods, the patient's blood is sent to an exchange compartment and passes on one side of a membrane, a fluid (fluid for dialysis or filtration, gas in the case of ECMO) passing from the other side of the membrane. Substances pass to one side or the other of the membrane depending on the substitution method considered. A problem can arise when the patient has been treated with a drug (active substance), because the membrane is likely to adsorb part of the drug and therefore reduce its concentration quickly, and to distort the predictions of manufacturers and doctors as to elimination of the product, which can lead to under-dosage of the drug in the blood and a poor clinical course.

It is therefore important to implement a method to evaluate or measure the adsorption properties of the membranes used, which will allow the clinician better monitoring of treatments.

Some authors have proposed methods using blood as a vector liquid; however, these methods have the disadvantage of being able to be implemented only for short durations (and not reflecting the durations of implementation in the clinic) due to the problems of blood coagulation in these in vitro models. Furthermore, the use of blood can lead to the complexation of the active ingredient with blood proteins or the penetration of the active ingredient into erythrocytes. This can pose problems for measuring the specific capacities of a given membrane.

Moreover, the models described are generally not interested in adsorption, and have not been developed to measure this parameter reliably and reproducibly.

In particular, the measurement of adsorption in the implementation of ECMO (in this technique, the drop in the quantity of drugs (in particular of the free form, which is the active form in the patient) is only due to adsorption, because there is no effluent) has not really been studied. US 2019/292288 aims to develop a polymer to cover the surface of a separation membrane, in order to prevent the adhesion of proteins and platelets ([0012], [0010]). The tests are therefore carried out on bovine blood ([0163]) and the reduction in aqueous permeability is evaluated, due to the reduction in the size of the pores due to the adhesion of blood proteins and platelets to the membrane ([0161] ). Blood is therefore circulated without active principle in the module containing the membrane, but it can be seen that all the elements taken out of the module (1) return to the tank (4) and are reinjected into the module (1). On reading [0161]-[0163], it can be noted that the aqueous permeability coefficient is measured by passing water through the circuit before and after having passed blood (and after washing). On reading [0169]-[0175], one can see a Sieving coefficient measurement for albumin.

US 2019/292288 does not use crystalloid solution

US 2019/292288 does not use a reject compartment, the "circulation beaker" being reused in circulation. However, the claimed method requires the use of a reject compartment, in order to reflect clinical practice and step ii cannot be reduced to the simple measurement of the concentration of the permeate (even if also carried out in US 2019/ 292288) Moreover, US 2019/292288 does not take into account the various measurements carried out on the active substance (extraction coefficient, quantity eliminated by dialysis or filtration), which is not surprising since this document is aimed at a completely different technical purpose.

In fact, the methods described below aim to determine the intrinsic properties of the membranes with respect to an active substance in order to avoid potential problems. It can be noted that in the implementation of these methods a solution of crystalloids is used, in order to avoid any risk of interaction between blood proteins and the active substance.

Thus, and in particular by not measuring extraction coefficients or the effects of dialysis or filtration, US 2019/292288 cannot achieve the same result as the claimed method.

It is thus the object of the invention to propose a generic method which can make it possible to measure the adsorption of an active substance by a given membrane, whether or not there is the presence of a permeate (term designating effluents, fluids removing impurities in dialysis, filtration or diafiltration systems). In this method, different concentrations are measured over time, the volumes are very precisely measured and controlled. In particular, consideration is given to the addition or not of fresh diluent (with or without solute) to replace diluent losses due to filtration or sampling over time. By this control of the conditions of implementation, the inventors were able to demonstrate underestimated effects of membranes due to their own charge and the charge of drugs (for example positively charged aminoglycosides).

In particular, the proposed method is characterized by an identical methodology whatever the device or the active principle studied.

1) the central compartment (container/bag of diluent containing the active ingredient) is prepared by controlling exactly both the volume of the diluent and the concentration of the active ingredient. To do this, the container of diluent and the active principle can be weighed before addition to the central compartment. Such an initial check makes it possible to know precisely the exact quantity of active principle which will be present in the circuit. A first solution (or central solution) is thus obtained, which will circulate in the central compartment.

2) The central container is then connected to a pump via a catheter and then sent into the exchange compartment at a constantly controlled rate, the rate being determined to reflect the conditions observed in the clinical setting (see above).

3) the exchange compartment contains a membrane separating the exchange compartment into two sub-compartments. The first solution circulates in a sub-compartment. A fluid (gas in case of simulation of ECMO, liquid in case of simulation of dialysis, filtration, diafiltration) passes into the other sub-compartment. The direction of passage (same direction or opposite direction of the first solution) depends on the simulated extracorporeal circulation method. When the fluid is a liquid, the permeate (the effluents) is recovered at the outlet of the exchange compartment, and they are sent to a discharge compartment. This was preferably weighed empty before the start of the experiment, in order to know the exact mass of recovered effluent (from which the volume can be calculated).

4) we are therefore in the presence of a second circuit associated with the exchange fluid (the first circuit being that of the central compartment) comprising a first reservoir containing the exchange fluid, a pump for sending the exchange fluid into the exchange compartment, and send the permeate to the rejection compartment. In the event of gas exchange simulation, no permeate is obtained.

In order to avoid the problems associated with the use of blood, a physiological solution of crystalloids is used, as the first solution, in which the active principle is dissolved, in the central compartment, as well as as a fluid of exchange (if the latter is liquid, [this second crystalloid solution (exchange fluid) containing no active substance to allow an exchange between the first crystalloid solution and the second crystalloid solution through the membrane). The use of a solution of crystalloids makes it possible to overcome the risks of introducing drugs into the erythrocytes, and to avoid drug/protein bonds.

A physiological crystalloid solution means that the crystalloid solution:

- has the same osmolarity as plasma (about 300 mosmol/l). This solution is therefore isotonic with blood. It can contain all the so-called physiological solutes. It is possible to use a crystalloid solution of the Ringer lactate type, or an isotonic electrolytic crystalloid solution. However, an isotonic saline solution (0.9 g/L of NaCl) can be adapted for the implementation of the methods described here, or even an isotonic glucose solution. It is also possible to use, within the solution of crystalloids, potassium chloride (KOI), calcium chloride (CaCl ₂ ), and/or magnesium sulphate (MgSO ₄ ) in the solution.

- optionally has a composition close to that of plasma ultrafiltrate. Thus, the ionic composition of plasma includes: Na ⁺ (142 mmol/l), K ⁺ (5 mmol/l), CP (103 mmol/l), Ca ²⁺ (2.5 mmol/l), Mg ²⁺ (1 mmol/l), HCO ₃ ' (27 mmol/l), lactate (2 mmol/l). It is preferentially balanced (chloride ion at a concentration of 100 to 112 mmol/l).

- preferably does not contain colloids. In particular, it is desired to avoid albumin, which is known to bind drugs or other colloids capable of modulating the transport capacity of drugs across the membrane (dextrans, hydroxy-ethyl-starches). In fact, the method seeks to determine the intrinsic properties of the membranes with respect to an active substance, and it is thus desired to reduce the external risks of disturbance of the relationship between the membrane and the active substance. In some embodiments, however, colloids (including albumin) may be added to simulate the dynamics that may exist in a patient. The choice of crystalloid solution composition is dictated by the solubility/stability restrictions reported in the drug's SmPC. As an illustration, a solution containing bicarbonate will be avoided to test the properties of a membrane with respect to piperacillin, which is degraded in the presence of bicarbonate.

The implementation of the method is done with large volumes (about 5 liters for the central compartment and more for the exchange fluid and the rejection compartment, in particular in diafiltration). It is perfectly suitable to use pharmaceutical grade preparations, which are already marketed by various manufacturers (Praxisdienst, Baxter, Fresenius...).

In general, it is preferred to use the same base for the crystalloid solution in all the compartments (central compartment and exchange fluid), in order to avoid a Gibbs-Donnan effect. abbreviations and concepts used

Nomenclature

Permeate or liquid effluent eliminated at the outlet of the exchange compartment. In clinical conditions, it is dialysis or filtration fluid that is loaded with toxins or waste, or diafiltered and eliminated fluid.

Liquid retentate being reinjected at the outlet of the exchange compartment. In clinical conditions, it is the blood that is returned to the patient.

Ciniet: concentration of the active substance in the circuit of the central compartment at the entrance to the exchange compartment (instantaneous concentration, measured at each different moment)

Coutiet: concentration of the active substance at the exit of the exchange compartment (instantaneous concentration, measured at each different time), before reinjection into the central compartment

Ceffi instant: concentration of the active substance at the outlet of the exchange compartment (instantaneous concentration, measured at each different moment), before injection into the discharge compartment (concentration in the effluents or the permeate)

Ccumui efn _u : concentration of active substance in the release compartment at the end of the predetermined period. Quantities (generally expressed in millimoles (biological substances) or milligrams (drugs, molecules))

Qccinitiai: quantity of active substance introduced into the central compartment at the start of the experiment

Qccfinai: quantity of active substance present in the central compartment at the end of the experiment (after the predetermined duration)

Qcceiiminated: total quantity of active substance eliminated from the central compartment Qcumui effiu - total quantity of active substance present in the permeate at the end of the predetermined period

Qdia: quantity of active substance eliminated by dialysis

Qfiitr: quantity of active substance eliminated by filtration

Qadsor: quantity of active substance eliminated by adsorption on the membrane AUCciniet: area under the curve representing the variation of C _in iet over time (report of values for each different moment)

AUCcoutiet: area under the curve representing the variation of C _or tiet over time (report of values for each different moment)

Other abbreviations (ratios, percentages, per unit)

Sieving coefficient: SC = ratio of the instantaneous concentration in the permeate (effluent) to the instantaneous concentration at the inlet: (C _e ffi instant) / (C _in iet) - This parameter is essentially linked to the free form of the drug (if the membrane has no other properties (adsorption)). It also makes it possible to describe certain electrostatic properties of the membrane with respect to the ionic properties of the substance.

Extraction coefficient (expressed as a percentage): CE = ratio of the difference between the instantaneous concentration at the inlet and the instantaneous concentration at the outlet (towards the central compartment), to the instantaneous concentration at the inlet: CE = [ (Ciniet - Coutiet)/Cjniet] x 100. This coefficient measures the decrease in quantity of drug due to the concentration gradient between the patient circuit (central compartment) and the dialysis circuit.

The invention thus relates to a method for determining the adsorption properties of a biomedical membrane, this membrane being able to be used in a medical device for extracorporeal circulation (dialysis membrane, filtration, diafiltration and/or ECMO gas exchange), for a given active substance after a predetermined duration, comprising a) the implementation, during the predetermined duration, of steps consisting of: i. Passing a first crystalloid solution containing the active substance from a central compartment to an exchange compartment containing the biomedical membrane, and into which an exchange fluid is injected, in order to obtain a retentate and a permeate, when the exchange fluid is a liquid, and measure the instantaneous concentration C _in iet of the active substance in the solution at the entrance to the exchange compartment at various times chosen during the predetermined duration, ii. Reinject the retentate into the central compartment, and measure the instantaneous concentration C _or tiet of the active substance in the retentate, at the outlet of the exchange compartment, and before optionally completing the first solution with a loss compensation solution, at the different times chosen during the predetermined duration iii. If a permeate is obtained, reject the permeate in a rejection compartment, and measure the instantaneous concentration C _e ffi instant of the active substance in the permeate, at the outlet of the exchange compartment, before injection into the rejection compartment at the different moments chosen during the predetermined duration (the sampling frequency is adapted to the duration of the phenomenon studied) b) After the predetermined duration, and for each moment chosen i. Determine the overall quantity of active substance eliminated in the System. Qceliminated ^— QcCinitial " QcCfinal ii. If a permeate is obtained, determine the sieving coefficient SC; iii. Determine the extraction coefficient CE iv. If a permeate is obtained, determine the quantity of active substance Q _cum uiefflu eliminated by dialysis and /or filtration v. Determine the quantity of active substance adsorbed on the membrane by the formula Q _adS or = Qeiiminée cc - Qcumul eff vi. Determine the adsorption of the membrane by the formula (Q _ad sor I Qc removed) X 100.

In a particular embodiment, the method comprises, before a), the implementation of the method for preparing an extracorporeal circulation circuit as described below.

The experimental device therefore contains two circuits: a circuit called the central compartment, simulating the extracorporeal circuit from the patient (simulated blood), with a flow rate corresponding to the simulated blood flow, and a second circuit dedicated to the exchange fluid, with a flow rate corresponding to the type of extracorporeal circulation that one wishes to simulate. The exchange compartment is the place through which the two circuits pass, separated by the membrane to be tested. The direction of the exchange fluid in the exchange compartment (same direction as that of the simulated blood or opposite direction) depends on the type of extracorporeal circulation that one wishes to simulate.

The predetermined duration depends on the type of extracorporeal circulation that one wishes to simulate.

For medical devices aimed at eliminating the drug (expected effect), such as dialysis or diafiltration devices, the duration must make it possible to obtain the elimination of 90% or more of the initial quantity to correctly describe the pharmacokinetic phenomena. This initial elimination phase may be followed by a release study phase. The total duration of a study is therefore generally 8 hours or less for the study of elimination and rarely more than 24 hours for the study of elimination and release.

For medical devices not intended to eliminate drugs (for which adsorption is an adverse effect (in particular ECMO or certain filtration devices), adsorption will be studied over a period of approximately 6 hours followed by a release study over a period of about 18 hours resulting in a study duration of about 24 hours.

A predetermined duration of 24 hours thus constitutes a cycle with clinical relevance which can be repeated as often as necessary to determine the adsorption over the entire lifetime of the device: 24 hours for the adsorbent columns, 72 hours for the filters of extrarenal purification, 14 days for ECMO membranes. Different concentrations are therefore measured at selected times during the determined period, generally on a regular basis (for example every 15 minutes, or every 20 minutes, or every 30 minutes, or every hour). By way of illustration, for a predetermined experiment duration of 24 h, measurements can be taken at the start of the experiment, then every 15 minutes during the first hour or the first two hours, then every hour. It is interesting that the times chosen are closer together at the start of the experiment, when the concentration of the active substance in the central circuit is the greatest, and when the adsorption on the membrane risks being the greatest.

For each moment chosen, there will therefore be 2 to 4 concentrations that can be measured. However, the samples can be taken at different locations (central compartment (entrance and exit of the exchange compartment), exchange fluid (exit of the exchange compartment) at different times, but this does not then allow the coefficients to be calculated. of sieving and extraction over time, these being calculated on concentrations measured at the same time. It is thus possible to draw the curve of concentrations over time. The interest of the method is that it makes it possible to reflect the clinical situation, in particular the number and duration of passage of the active ingredient on the membrane.

The concentration measurement is carried out by taking a small volume (preferably less than or equal to 3ml, but sufficient to be analyzed and to be able to measure the concentration). In order to compensate for this loss of volume in the central compartment, it can be supplemented by adding the volume taken from a crystalloid solution at the outlet of the exchange compartment, after the sample taken to measure the concentration C _or tiet- This solution of supplement does not contain the active ingredient studied.

This volume adjustment is particularly justified in the event of a reduction in the volume of the central reservoir to adapt to pediatric conditions, in which a volume of the central reservoir of 500 mL can be used instead of the 5000 mL in adults. In fact, in this case of a study over 8 hours where 12 samples of 2 mL would be taken from a central reservoir of 500 mL, which would correspond to approximately 5% of the central compartment, a value above the variability of the method and therefore justifying the volume correction. On the other hand, in a situation of adults, 12 samples of 3 mL only represent a decrease of 0.7% of the central reservoir, a non-significant variation which can be accepted.

Thus, the quantity of active principle remains constant (or is considered as such) in the overall circuit (first and second circuits) and is that determined at the start of the experiment.

At the end of the experiment, after the predetermined duration, it is possible to calculate the overall quantity of active substance eliminated Qccéiiminée from the system (central compartment). This value is obtained by the difference between the initial quantity of active substance supplied to the system in the central compartment (Qccinitiai) and the quantity of active substance remaining in the system (central compartment) after the predetermined duration (Qcctinai)-

Thus, Qccéiiminée = Qccinitiai - Qcctinai- These quantities can be calculated with precision because of the control maintained throughout the experiment on the parameters of the system (and in particular the monitoring of volume variations).

Quantities are calculated by multiplying concentrations by volumes. If it is clear that the initial quantity (Qccinitiai) is perfectly defined (one can control exactly the quantity of active substance brought, by weighing it before dissolution in the central compartment), it is necessary to know the concentration of active substance at the outcome of the experiment (it is measured by methods known in the art in a sample taken from the central compartment), but also the volume of the central compartment.

To do this, we prefer to proceed as follows: i) We measure exactly the initial volume of the central compartment (for example by weighing the central compartment, because the precision of the scales is generally better than the precision of the devices measuring the volume, or easier to implement). Generally, the initial volume is of the order of five liters, corresponding to the average blood volume in an adult patient. It can be lower if one wants to simulate the conditions observed in pediatrics, or if the drug is rare or extremely expensive; thus, a CC with a volume of the order of 0.5 to 1 L can be used. A specific procedure (by weighing) to determine the exact volume is explained below. ii) The volumes taken at the chosen times are measured exactly (these volumes are of the order of 3 ml), at the entrance and at the exit of the exchange compartment (to measure the instantaneous concentrations C _in iet and Coutiet) - iii) Optionally, one completes (just downstream of the sampling position for calculation of C _or tiet) the solution of the central compartment by a volume of crystalloid solution without active ingredient identical to that taken at the inlet and at the outlet of the exchange compartment for the chosen time considered, in order to keep the volume in the central compartment constant iv) The final volume of the central compartment at the end of the experiment is calculated by Final volume = Initial volume - (sum of the volumes withdrawn) + (sum volumes possibly added). v) One can thus obtain the quantity of active substance at the end of the Qcctinai experiment, by multiplying the final concentration by the final volume.

The amount of material removed was by membrane adsorption, filtration and/or dialysis. The method also makes it possible to determine the share of each of these elimination pathways, bearing in mind that the only elimination pathway is adsorption, when simulating medical devices without liquid effluate such as ECMO.

When filtration, dialysis or diafiltration membranes are used, the method thus also makes it possible to determine the diffusion, dialysis and/or diafiltration properties of the membrane.

In these eventualities, a certain number of parameters can be calculated:

- The sieving coefficient (Sieving coefficient in English) SC, is essentially related to the crossing of the membrane by the free form of the drug, if the membrane does not have other properties (such as adsorption) which lead to an interaction of it with the drug. By using the sieving coefficient, information can be obtained about certain electrostatic properties of the membrane in relation to the ionic properties of the active substance.

The sieving coefficient is calculated at each chosen moment by calculating the ratio of the instantaneous concentration in the permeate by the instantaneous concentration at the entrance to the exchange compartment.

SC — Ceffl instant / Cjnlet- - The extraction coefficient CE, measures the decrease in quantity of drug due to the concentration gradient between the patient circuit (central compartment) and the dialysis circuit. It is expressed as a percentage and represents the ratio of the difference in concentration of the active principle in the central circuit, between the inlet and the outlet of the exchange compartment by the concentration at the inlet.

CE = (Ciniet - Coutiet) / C _in iet (possibly multiplied by 100 to obtain the percentage).

- The quantity of active substance Q _cum ui eftiu eliminated by dialysis and/or filtration. This quantity can be calculated in different ways: a) from the curve representing the concentrations in the instantaneous effluents (C _e ffi instant), taken at the various chosen times. This curve can be integrated by any method known in the art, and by using appropriate software. As a good approximation, this area under the curve can be measured by the trapezium method, known in the art, and which is based on a linear interpolation by intervals (the chosen moments). This area is then also the sum of the areas of each trapezium between two consecutive chosen moments. It is easily understood that the number of samples can be increased to improve precision.

The quantity of active substance Q _cum uiefflu is then calculated by the formula

Qcumui effiu= Sum (AUC C _e ffi instant) x exchange fluid flow rate (generally the diafiltration flow rate). b) from the final volume of permeate, and the final concentration of the active substance in this final permeate.

Thus, Qcumuieftiu = Ccumui ettiu x Volume of the reject compartment.

In practice, the effluents are discharged into several pockets. Indeed, the commercial bags have a volume of 5 L (Baxter Gambro) or 10 L (Fresenius). Given the prescribed diafiltration flow rate (for example 4 l/h), a 5 L bag is filled in 74 minutes, and several bags are therefore needed for an experiment lasting a few hours. The flow rate of the exchange fluid can be precisely calculated by the methods described below.

We then calculate, for each pocket, the quantity of active substance in the pocket, and we obtain the total quantity of active substance eliminated by dialysis and/or filtration by the formula Q _cum ui eftiu = sum for all the rejection pockets (C _cu mui etnu in a reject pocket x Volume of the reject pocket) (or Q _CU mui effl _U = (Ccumui etfiu x Volume of the reject compartment)). It is then possible to determine the quantity of active substance adsorbed on the membrane by the formula Q _ad sor = Qéiiminée cc - Qcumui en (difference between the quantity of substance eliminated in the central compartment and the quantity of active substance eliminated by dialysis and filtration, present in the effluent).

It is also possible to determine the quantities of active substance eliminated by dialysis or filtration. i) In general, the quantity of active substance eliminated by dialysis is determined by measuring the difference in the areas under the curve (AUC) of the instantaneous concentrations at the inlet or the outlet of the exchange compartment, multiplied by the simulated blood flow

Qdia = (AUC Ciniet - AUC C _or tiet) x simulated blood flow (central compartment flow). The areas under the instantaneous concentration curve at the inlet or outlet of the exchange compartment are calculated in particular by the trapezium method. The simulated blood flow is a parameter determined by the experimenter at the start of the experiment. ii) in a pure filtration model, the quantity of active substance eliminated by filtration is the total quantity of active substance present in the effluents. Thus, Qfjitr ^— Qcumul efflu- iii) in a diafiltration model, the quantity of active substance eliminated by dialysis is determined by the formula given above:

Qdia = (AUC Ciniet - AUC C _or tiet) x simulated blood flow (central compartment flow), and the amount of active substance removed by filtration is determined by

Qfiltr ^— Qcumul efflu " Qdia-

We can then calculate the respective shares of dialysis, filtration and adsorption in the elimination of the active substance:

- Share of dialysis = (Q _dia / Qccemia) x 100

- Filtration share = (Qfii _tr / Qccéüminée) x 100

- Share of adsorption = (Q _adsor / Qccumin) x 100).

We can verify that the sum is equal to 100%. This method is implemented with the membranes used for extracorporeal circulations.

Thus, in one embodiment, the membrane is a dialysis membrane. In this embodiment, a second crystalloid solution (of identical composition to the first crystalloid solution, but without active substance) is used as the exchange fluid. In this embodiment, the quantity of active substance eliminated by dialysis can be determined.

In another embodiment, the membrane is a filtration membrane (i.e. pure filtration is simulated). In this embodiment, the exchange fluid is preferably a second crystalloid solution (of identical composition to the first crystalloid solution, but without active substance), and the method can also also include the determination of the quantity of active substance eliminated by filtration.

In another embodiment, the membrane is a diafiltration membrane. In this embodiment, it is preferred that the exchange fluid be a second crystalloid solution (identical in composition to the first crystalloid solution, but without active substance), and that the method also includes determining the amount of substance active substance eliminated by dialysis and the determination of the quantity of active substance eliminated by filtration.

In another embodiment, the membrane is an extracorporeal oxygenation membrane (simulating ECMO). The exchange fluid is then a gas (air-type medical gas or oxygen-enriched mixture) containing oxygen. No liquid permeate is obtained by this method and only the adsorption of the membrane is calculated.

In a particular embodiment, the device used mirrors the device used for continuous extra-renal purification (CRRT). The method therefore simulates an EERC. It is also possible to calculate the percentage of active substance eliminated from the central compartment during the duration of the study (the predetermined duration) under normal diafiltration flow conditions. This percentage is calculated by the formula:

% substance eliminated from CC = (Qcc _eliminated / Qccinitiai) x 100.

This assessment can be used to check the quality of the work insofar as the pharmacokinetics is considered well described when 90% and more of the drug have been eliminated from the central reservoir under the usual conditions of use of the device.

The invention also relates to a method for preparing an extracorporeal circulation circuit for the implementation of a method as described above, comprising: i) The provision of a bag containing a solution of crystalloid in physiological solution, preferably balanced, with a volume of between 4.5 and 5 liters ii) The measurement of the exact volume of the solution included in the bag, preferably by weighing the bag (empty bag and filled bag) , and taking into consideration the solutes present in the crystalloid solution and the weight of the empty bag (in particular as explained in the present application). Thus, given the precision of the scales (compared to those of conventional instruments for measuring volumes), an accuracy of less than 0.05% can be obtained. iii) The provision of a quantity of active substance to be tested (drug), and the dilution of this active substance in the near crystalloid, using a defined volume of the crystalloid solution (preferably coming from the bag) ; it is preferred to obtain a concentration of the active substance in the pocket corresponding to the clinically relevant concentration for this active substance. iv) The provision of an extracorporeal circulation device comprising o A first circuit to which the bag above is connected (also called the central circuit), said first circuit containing a pump allowing the loop circulation of the crystalloid solution in the central circuit, the crystalloid solution passing through an exchange compartment containing the membrane to be tested for the active substance, sampling sites being located (relative to the flow of the crystalloid solution) at the inlet (upstream ) and at the exit (downstream) of the exchange compartment. It is therefore possible to take samples (of the order of 3 mL) of the crystalloid solution from the first circuit at the sampling sites at any chosen time. o A second circuit in which passes an exchange fluid (gas in the case of ECMO or crystalloid solution identical to the solution of the bag above in the case of dialysis, filtration or diafiltration), said second circuit containing a pump allowing the circulation of the exchange fluid in the exchange compartment.

In the case of a liquid exchange fluid, the second circuit also has a sampling site at the outlet of the exchange compartment in order to sample a volume of the exchange fluid (of the order of 3 mL), and one or more collection pockets to recover the exchange fluid after passing through the exchange compartment.

In the case where the exchange fluid is a gas, it is preferred when the device also contains a second pocket of the same volume as the pocket containing the crystalloid solution and the active substance (the first pocket). This second bag contains a crystalloid solution of similar composition, but containing no active principle. The principle is to connect the close second to the first circuit when the first pocket containing the crystalloid solution and the active substance is disconnected from it. This makes it possible to study the release of the active substance by the membrane in the event of adsorption.

Thus, the first pocket P1 contains the drug to be studied. The method is implemented with P1 connected to the first circuit. After a certain time (6 hours is an acceptable duration, since it can be considered that the major part of the adsorption will have taken place, this can be confirmed a posteriori by the disappearance of differences in concentrations of the drug between the sites of samples at the inlet and outlet of the first circuit), P1 is disconnected from the first circuit (the pche P1 is “clamped”) and the second bag P2 is connected to the first circuit to circulate the crystalloid solution of P2 in the first round. As indicated, this makes it possible to study the release of the drug for an adequate duration (it can be done for the next 18 hours), starting from the hypothesis (generally confirmed) that the release time is always longer than the adsorption time.

These 24-hour cycles with daily membrane exposure can be repeated to mimic actual exposure as would result from normal use of the device In one embodiment, this method comprises adjusting the flow rate in the first circuit from 50 ml/min to 250 ml/min, preferably around 200 ml/min, and adjusting the second circuit between 500 to 8000 ml/h , preferably around 2500 ml/h. This makes it possible to simulate continuous extrarenal purification (EERC, filtration dialysis adsorption).

In another embodiment, this method includes adjusting the flow rate in the first circuit from 3 to 5 I / min, the second fluid being air. This simulates ECMO.

The invention also relates to a computer program product, comprising a set of instructions which, when it is executed by processing means, is capable of implementing a method as described above, for providing the adsorption of a given active substance on a membrane after a predetermined time.

The invention also relates to a non-transitory computer-readable storage medium on which is stored a computer program comprising program instructions, the computer program being loadable into a data processing unit and being adapted to bring the data processing unit to execute a method and a method as described above.

In another embodiment, the invention relates to a microprocessor comprising a computer algorithm for carrying out a method for determining the adsorption properties of a membrane, with respect to an active substance (medicine) in which one provides the values read as indicated above, and who performs the calculations as mentioned to provide the adsorption of the membrane, as well as the various other values mentioned above.

In particular, we provide i) The elements relating to the initial concentration of the drug (in particular the volume of the pocket of the central compartment, the initial mass of drug) ii) The various instantaneous concentrations measured over time, as well as the times of measurement iii) The flow rates (or elements allowing the calculation of the flow rates) of the central compartment and the dialysis circuit (effluate) The results obtained include i) The extraction coefficients over time ii) The sieving coefficients over time iii) The amount of drug removed from the central compartment iv) The amount of drug removed by filtration v) The amount of drug removed by dialysis vi) The amount of drug removed by adsorption vii) The relative shares of dialysis, filtration, adsorption

The method can be used to calculate the Michælis-Menten parameters Vimax and Km, taking into account the fact that balances exist between the active substances and the membranes, giving rise to exchanges (capture or sequestration / release or release) during patient treatment.

By analogy with the kinetics of reactions catalyzed by enzymes, one can relate the rate of elimination of the active substance (corresponding to the rate of enzymatic reaction) to the concentration of active substance in the central compartment (corresponding to the concentration of the substrate for enzymatic reactions) as well as constant parameters, characteristic of the membrane and its interaction with the drug (V _imax and Km).

Vimax OR Vimax or Vmax or v _ma x corresponds to the original maximum clearance rate (elimination by filtration, dialysis and/or diafiltration as appropriate) measured for a saturating concentration of active substance (for example in pmol/min) .

Km corresponds to the Michaelis constant. It is the concentration of active substance for which the initial rate of the reaction is equal to half of the maximum initial rate.

It is understood that, for an enzymatic reaction in which an enzyme generally has a single substrate, Km and V _im ax are specific for the enzyme. In the present case, Km and V _im ax can be calculated which will be specific to a pair (active substance, membrane). Indeed, a given active substance is likely to react (to be sequestered and released) differently with different membranes, and a given membrane will not adsorb different active substances in the same way.

It is recalled that, according to the model of Michælis and Menten, the equation to be taken into account is the following:

with

Vj initial rate (i.e. in the absence of onset of clearance) of dialysis/filtration/diafiltration for a concentration of active substance [S] (e.g. in pmol/min)

[S]: Substrate concentration (in mol/L)

V _max and Km defined above for the membrane and the active substance.

There are several methods to determine the Michælis-Menten parameters. In practice, a series of measurements of the original velocity are made for different values of the concentration [S].

The constants can then be determined by methods similar to those described for enzymes (Lineweaver-Burk representation, Hanes-Woolf representation, Eadie-Hofstee representation, Eisenthal and Cornish-Bowden representation).

We can also plot the graph representing the Vj as a function of the concentration [S]. According to the Michaelis equation, we can indeed deduce that Vj asymptotically approaches v _max when [S] increases. Moreover, when [S]=Km, Vi=v _max /2. We can therefore graphically determine v _max , then Km.

The invention thus relates to a method for calculating the v _imax (maximum initial speed) and Km corresponding to the interaction of a given active substance with a given biomedical membrane, comprising: i) Repeating the method described above for different concentrations of the given active substance and determine the adsorption values for each concentration of given active substance, as well as the quantity of active substance eliminated by dialysis, filtration and/or diafiltration, and determine the percentage thereof, with respect to the quantity initially supplied ii) Calculate the parameters v _imax and Km using the data obtained in i).

DESCRIPTION OF FIGURES

Figure 1: Stability of vancomycin in crystalloid solution Figure 2: Vancomycin elimination kinetics during a dialysis session. The curves represented are the concentrations in the central compartment (CC), at the entrance (inlet) and at the exit (outlet) of the exchange compartment.

Figure 3: Gentamicin elimination curves in a diafiltration model, Concentrations are measured in the central compartment (CC), at the inlet and at the outlet of the exchange compartment.

Figure 4: Elimination curves of amikacin in a diafiltration model, The concentrations are measured in the central compartment (CC), at the entrance and at the exit of the exchange compartment.

Figure 5: curve making it possible to determine the V _imax and Km for gentamicin on the ST® filter from Baxter-Gambro. The dotted line shows Km (V _imax / 2)

EXAMPLES

The examples above present particular embodiments of implementation.

Example 1. Setting up the system

This example describes generic elements to take into account for the implementation of the methods as described.

1. Generic conditions, and specific elements.

Central compartment: The central compartment (CC) is a main element of the system, simulating the patient. The fluids in the central compartment therefore simulate the patient's blood system. Thus, the fluids of the central compartment circulate in a closed environment, being reinjected into the central compartment after passing through the exchange compartment containing the membrane to be tested.

Preferably, 5 liters of a crystalloid solution in balanced physiological solution are used for the central compartment. The physiological term refers to the isotonicity of the solutions with respect to plasma (300 mosmol/L), the balanced term refers to the chloride ion concentration of a value comparable to that of plasma (95-105 mmol/L) .

There is a range of pharmaceutical grade crystalloid solutes with these properties.

As explained above, an important element in the choice of the crystalloid solution comes from the stability of the drug (of the molecule to be tested) in it during the duration of the study (a preliminary study is preferentially carried out, and one can choose, as acceptance criterion, a stability corresponding to a variation of less than 10% over the duration of the study is accepted, i.e. a duration of 8 hours for continuous hemodiafiltration (CRRT, Continuous Renal Replacement Therapy) and 24 hours for ECMO) (Figure 1 for the vancomycin stability study).

Drug Dose For each molecule of interest, a study can be performed by loading the CC to the maximum recommended therapeutic concentration. This mode of exposure mimics a bolus administration. This mode of implementation is preferred, because it harmonizes the mode of exposure of the drug to the filter and facilitates comparisons between drugs, between membranes and between manipulations.

The method can however be implemented by simulating administration by infusion. Thus, the administration of aminoglycosides (gentamicin, tobramycin, amikacin) can be carried out with an electric syringe pump for 30 min, in order to simulate the real conditions of administration of these active ingredients in the clinic.

Effluate flow rate You can select different effluent flow rates (also called effluent): i) Low: 1 L/h ii) medium and recommended: 2.5 L/h (corresponding to a dose of dialysis or filtration or of diafiltration of 35 mL/kg/h in a 70 kg man) iii) high: 4 L/h iv) very high: 6 to 8 L/h.

Flow of

simulated, a constant flow rate is used, preferably set at 200 mL/min.

Location and time of sampling on the entire circuit simultaneously:

In the CC: at the entrance (C _in iet) and at the exit (C _or tiet) of the exchange compartment

In the effluate: at the exit of the exchange compartment (C _e ffi instant) -

The samples are taken simultaneously at the different sites.

We generally always choose the same times, determined to facilitate the study of adsorption: T0 (before the start of the session) then after its start at +15, +30, +45, +60 minutes and then every hour during 6 to 8 hours to study as completely as possible the phenomena of adsorption, dialysis and filtration then after for 18 to 16 hours the release of drugs in the event of significant adsorption.

Collection of accumulated effluents: this is done in bags dedicated to this collection: we can cite the 5 L bags at Baxter-Gambro and the 10 L bags at Fresenius.

2. Calibration

The method was first implemented using substances known not to be adsorbed by any membrane (urea, creatinine or potassium).

The recovery rates of urea and creatinine during 10 sessions combining dialysis, filtration and diafiltration were 104 + 19 and 105 + 15%, respectively.

As expected, in pure filtration mode, the average urea sieving coefficient was 1.01 + 0.01 and 0.99 + 0.02, respectively.

The potassium recovery rate in the effluent was 100+ 2%.

It can therefore be shown that there is a correlation between the dia/filtration flow rates and the clearances of creatinine and urea, in linear regression with R ² of 0.968 for urea and 0.959 for creatinine.

A study of the effects of exchange liquid flow rates was carried out for dialysis and filtration. Indeed, there is a very wide range of flow rates that can be delivered by these medical devices, ranging in adults from 0.5-0.8 L/h, minimum values found in the literature, but which can reach 6 or even 8 L/h in high-speed purification.

It was possible to show a close linear correlation between potassium clearance and diafiltration rate (especially for rates of 1 to 4 L/h), which begins to diverge (for urea, creatinine and potassium) for the very high rate of 8 L /h.

Thus, the implementation of the method has made it possible to reveal a phenomenon of mass transfer resistance by dialysis, by which the fluid circulates in the capillaries (each circuit) which are at lower resistance, which limits the exchanges blood- dialysate with higher resistance. This phenomenon of mass transfer resistance, demonstrated by the in vitro method, is never taken into account in clinical studies using high (50 mL/Kg/h) or even very high hemofiltration rates (100 mL /Kg/h).

3. Parameter control

It is important to precisely control the parameters implemented during the experiment.

In order to precisely measure the volumes and concentration of active ingredient (drug) in the central compartment, it is preferable to carry out a weight measurement, which is more precise (imprecision of 2 g for a measurement of 5,000 to 10,000 g on a Stadler scale) than the volumetric measurement. (for volumes of the order of 5L). Good reproducibility can thus be observed: mass of the full pockets (n = 23) = 5260 + 24 g (CV = 0.45%), mass of the empty pockets (n = 17) = 60.35 + 0.79 g ( CV = 1.3%), with a precision and reproducibility that cannot be achieved by volume measurement methods.

The volume of solution is obtained by i) removing the weight of the container (measurement of the weight of the empty pockets) ii) taking into account the density of the crystalloid solution. Preferably, bags of pharmaceutical grade are used, the composition of which is known (it is therefore possible to know the quantities of substances added to obtain a balanced physiological (iso-osmotic) solution (with a chlorine concentration close to that of plasma)).

Drugs are diluted and injected into the bag using the solution in the bag, to avoid any addition or withdrawal of fluid.

Example 2. Study of a dialysis model

Vancomycin with a molecular weight (MW) of approximately 1450 Da is a medium-sized molecule close to the sieving coefficient of old membranes (vitamin B12 with a MW of approximately 1355 Da) and the Prismaflex ST150® membrane (Baxter-Gambro) which has a sieving coefficient for vitamin B12 of 1, whereas that of inulin (MW 6179 Da approximately) is only 0.96.

Also the elimination of vancomycin by filtration seems more effective than by dialysis.

Vancomycin removal was studied by the ST150® filter in pure dialysis mode over a period of 6 h at a dialysate flow rate of 2.5 L/h with a flow rate of blood simulates 200 mL/min. Two sessions were performed, the results were calculated by the average of the two sessions.

First, we checked the stability of the substance studied in the crystalloid solution and under the temperature and light conditions in which the dialysis sessions will take place (Figure 1).

A stability study in the sampling tubes is then carried out under the operating conditions:

[Table 1]

Thus, the stability of the molecule under the operating conditions allows its study.

Session duration was limited to 6 h and not 8 h (because 90% or more of CC vancomycin is eliminated after 6 h). This duration is therefore acceptable to describe the elimination of almost the entire dose as the free fraction of the drug.

The table below gives the detailed results of the elimination pathways of vancomycin by the ST150® filter during two pure dialysis sessions at a dialysis flow rate of 2.5 L/h without its own filtration flow rate.

[Table 2]

Thus, the free fraction of vancomycin (which, it should be remembered, is fixed at 50% on plasma proteins in the blood under clinical conditions) is easily eliminated by dialysis with the ST150® as evidenced by the rate of elimination of vancomycin from CC and the clearance value of the free form of CC which has a value close to that of the dialysis flow.

Figure 2 shows the elimination kinetics of vancomycin during a dialysis session. The concentration curves in the CC are in black dots and dashed lines. The concentration curves measured at the entrance to the membrane are in red dots and dashed lines. The curves of the concentrations at the outlet of the membrane are in dots and broken blue lines.

Throughout the session the mean value of CE (Cjniet-Coutiet)/Ciniet) remained stable (15 + 1%), confirming first-order kinetics with total clearance of CC calculated from simulated blood flow data from 1.8 L/h compared to 2.2 L/h measured from the effluents.

In conclusion, the method made it possible to show:

1) That vancomycin is effectively removed by dialysis during a 6 h session.

2) That there is no detectable adsorption of the free fraction of vancomycin by the ST150® filter. Thus, the demonstration, by the method described here, of the absence of interaction between the free form of vancomycin and the ST150® filter could be added as such as a property of the product, in the Summary of Product Characteristics ( RCP) of vancomycin, specifically compared to the filter tested ST150®. Thus, if downward variations in vancomycin concentrations are observed with this filter, they relate to other mechanisms to be elucidated than the properties of the filter.

This study gives results that contradict those reported previously.

Tian et al (Artif Organs. 2008;32(1):81-4) reported a quantity of vancomycin adsorbed by a polyacrylonitrile (PAN) filter of 10 mg for an initial dose of 36 mg, i.e. approximately 25%. However, it can be noted that the authors used a closed system, without clinical relevance because without dialysis or simultaneous filtration, which leads to overestimating the real role of adsorption. The use of a whole blood model does not shed any light on the free form, the only active form for the phenomena of tissue diffusion and physiological elimination or caused in particular by CRRT. A possible hypothesis would be that the difference in adsorption results from the use, by Tian et al, of a pure polyacrylonitrile membrane with a low sieving coefficient, measured by the clearance of vitamin B12 (molecular weight 1355 Da) whereas that of vancomycin is 1450 Da. It can be noted that the sieving coefficient of the ST150® is 40,000 Da, which explains the problem-free dialysis of vancomycin with these high permeability membranes.

Onichimowski et al (J Artif Organs. 2021 Mar;24(1):65-73) studied the elimination of vancomycin in a system using the same ST150® membrane, in a porcine blood model. The CRRT was in pure filtration and post-dilution mode. The filtration rate was 0.6 L/h with a blood flow of 100 mL/min Min. The concentration of vancomycin studied by the authors, of 1000 mg/L, is unrelated to the upper limit of therapeutic concentrations of 50 mg/L. The publication also shows a very high variability of concentrations in the vancomycin stability study, which raises questions about the stability of the model. In addition, the authors only measured vancomycin concentrations at the inlet of the filter and not also at its outlet. They conclude with an adsorption rate of 33% of vancomycin, but with a considerable standard deviation (figure 2 of this document) and which presents an overlap with the standard deviations of the controls suggesting statistically non-significant differences.

Example 3. Study of filtration

A study was made in pure filtration with fluconazole.

We observed the elimination of more than 90% of the initial amount of CC, for sessions in pure filtration mode. The entire eliminated dose was found in the effluents leading to the conclusion that there was no adsorption.

The curves i) concentration in the CC ii) concentration measured at the inlet of the exchange compartment iii) concentration at the outlet of the exchange compartment iv) concentration in the instantaneous effluent were drawn and overlapped. This superimposition of the curves in filtration mode without adsorption is very characteristic of this mode of elimination. Indeed the exchanges are those of an ultrafiltrate by pressure gradient. It is then normal to have the same concentrations in the CC and the C _in iet, the C _or tiet- The concentration in the effluent reflects the free fraction of the drug, studied in the model. Thus, the concentration in the effluent is identical to that of the incoming concentration, which is observed. Example 4. Demonstration of adsorption and relative quantification with respect to dialysis, filtration or diafiltration carried out simultaneously.

The following works were carried out:

Membrane: ST®150, derived from PAN and covered with PEI reputed to be very sorbent

Drugs: three drugs from the same pharmacological class, aminoglycosides, were compared under identical experimental conditions: gentamicin and tobramycin at the recommended initial concentration of 40 mg/L and amikacin at the recommended initial concentration of 80 mg/L

Size of the central compartment: 5 liters of Hemosol®, isotonic and balanced crystalloid solution

Central compartment flow rate: 200 ml / h

Diafiltration flow: total flow of 4 L/h divided into 2.5 L/h of dialysis and 1.5 L/h of filtration

Effluate Flow: Since the net loss was set to 0 L/h, the effluent flow was 4 L/h.

Duration of the experiment: 6 hours insofar as previous studies had shown that during this period of time 90% and more of the initial dose was eliminated

The samples were taken at: tO, tO+15 min, +30, +45 min, +1 h, +2, +3, +4, +5 and +6 h.

The summary table shows the main results of this study.

[Table 3]

These results show that within the same pharmacological class, the elimination pathways are specific to the molecule. The most stable part being that linked to dialysis, the most variable parts being filtration and adsorption. Noting that adsorption takes place at the expense of filtration. But the very high clearance values of CC show that adsorption does more than replace filtration and adds a specific clearance term of a value much higher than the theoretical maximum value delivered to the system which should have been identical to the maximum flow rate. total diafiltration, i.e. 4 L/h. Note that for the least adsorbed drug, amikacin, the clearance value is almost identical to the expected theoretical value.

Example 5. Example of ECMO.

In vitro study of the adsorption of therapeutic concentrations of Gentamicin and Amikacin using the circuit and the H LS Advanced 7.0 filter-pump block, GETINGE.

The literature reports problems with the adsorption of drugs administered intravenously in intensive care units in the circuits of medical devices with arterio-venous and venovenous peripheral circulatory assistance (ECMO). Adsorption can be of such importance that it diminishes and even eliminates the effectiveness of drugs adsorbed onto the ECMO membrane.

However, the question of the adsorption, by the ECMO circuits, of aminoglycosides and in particular of amikacin and gentamicin which are the two most prescribed aminoglycosides in adult resuscitation, has never been addressed, to the knowledge of the inventors.

The method described above was implemented using a crystalloid solution, Phoxilium®, Baxter-Gambro. The simulated blood flow had been set at a theoretical value of 3 L/min. It was 3.1 ± 0.2 L/min.

It was shown, by equality of the areas under the concentration curve at the inlet and at the outlet of the filter-pump block, the absence of retention and therefore the absence of detectable adsorption in this block.

Study of the adsorptive capacities of the filter-pump unit

Comparison of the concentrations measured at the entrance and at the exit of the block Conclusion 1: Analysis by Student's t test for pairwise measurements reveals no significant difference in the concentrations of gentamicin and amikacin between entry and exit from the block.

Calculation of the instantaneous extraction coefficient (CEjnst) and its average value in the filter-pump unit:

During the duration of the session the mean (+/- standard deviation) EC _ins t values for gentamicin and amikacin were

Gentamicin = 1.2 ± 5%

Amikacin: = - 0.2 ± 4%

Conclusion 2: these extraction coefficients of gentamicin and amikacin can be qualified as negligible.

Calculation of the areas under the concentration curve at the inlet (AllCjn) and at the outlet (AUCout) of the system:

Amikacin

AUC Ciniet = 25388 mg.min/L

AUC Coutiet = 25670 mg.min/L

Gentamicin

AUC Ciniet = 24843 mg.min/L

AUC Coutiet = 25127 mg.min/L

Conclusion 3: The differences in AUC C _in iet and AUC C _or tiet of amikacin and gentamicin are of the order of 1% and are therefore negligible.

Calculation of pump filter block clearance

The clearance of the pump filter unit is the product of the extraction coefficient by the flow rate of crystalloid solution according to the equation:

Amikacin: pump filter unit clearance = 3.1 x 0.011 = 0.034 L/h

Gentamicin: pump filter unit clearance = 3.1 x 0.011= 0.034 L/h

Conclusion 4: the clearance values are negligible compared to the pump flow rate.

Clearance of the complete circuit: tubing and filter-pump unit

The calculation of the clearance from the central compartment makes it possible to evaluate the clearance provided by the entire system: catheters-tubings-pump filter unit.

Clearance is given by the ratio of the initial dose in the CC divided by the AUCcc. Amikacin

Claiicc = 410 mg / 25863 mg.min/L = 0.9 L/h Gentamicin:

Claiicc = 244 mg 1 15078 mg.min/L = 1.0 L/h

These clearances can be compared to the flow rate of crystalloid solution delivered at the same time: 3.1 x 60 = 186 L/h

Clearances from the central compartment for amikacin and gentamicin represent only less than 1% of the total flow and are therefore negligible.

General conclusion:

The analysis carried out in ECMO according to the method described here makes it possible to conclude that no element shows a significant adsorption (> 10%) of gentamicin and amikacin in the circuit comprising the filter-pump unit HLS Advanced 7.0, GETINGE .

Example 6. Advantages of the method

The use of a solution of crystalloids makes it possible to specifically study the free fraction of the drug, which is the determining element of its fate in the body both in terms of its therapeutic effects and its elimination. Additionally, testing can be performed under a variety of operating conditions, allowing comparisons when all else is equal.

The method makes it possible to determine the importance of adsorption, which begins as soon as a drug is administered intravenously, and induces a reduction in the value of the peak concentration.

This method makes it possible to determine the respective roles of filtration, dialysis and adsorption which can be expressed both in mg and % of the dose eliminated from the CC over the time of the study (chosen to correspond to the elimination of 90% of the initial dose) and thus allows a complete description of the phenomenon, which is not possible with studies of 1 h to 3 h. thus, the method could be used for 72 h with the same drug and the same filter. The method described here has thus made it possible to show the absence of detectable adsorption for a large number of molecules: phenobarbital, salicylate, vancomycin, lithium.

It also made it possible to show the rapid and complete inactivation of a drug class, the echinocandins, which are very strongly bound to plasma proteins but whose free form disappears in less than 2 hours for drugs administered intravenously at a single daily dose. . The results were subsequently confirmed clinically.

The method also made it possible to highlight two phenomena that had never been described. To. Dialysis induced by adsorption in pure filtration mode

Example 3 showed that in pure filtration, with zero water-sodium balance (no water-sodium loss inducing haemoconcentration), a close superimposition of the curves is observed in the different sampling sites.

However, in the presence of adsorption, an arteriovenous gradient was observed which allows the calculation of a very high extraction coefficient and decreasing very rapidly over time in the hour following the start of exposure. drug to filter.

It can be hypothesized that adsorption is an internal process in the membrane resulting in an instantaneous decrease in drug concentrations in the effluate. This results in the creation of a drug flow aimed at compensating for this concentration gradient between the crystalloid solution and the effluate (principle of dialysis). Thus, the study of the elimination of echinocandins including caspofungin and micafungin shows an adsorption rate of 100% in less than two hours of these two antifungals, the creation of an arteriovenous gradient decreasing over time and concentrations in the cumulative effluents below the limit of quantification.

Thus, adsorption may be of such importance that it prevents the occurrence of detectable concentrations in effluents. b. Limitation of the elimination of ionized drugs by charge identity with the filter

The study of the elimination of aminoglycosides suggests such a mechanism, which had never been described. The aminoglycosides studied are very hydrophilic and ionized drugs at physiological pH in the form of cations. The Baxter-Gambro ST® filter is made of a naturally ionized electronegative polyacrylonitrile (PAN) support, which has biological consequences. Thus, to modify the charge of this membrane, a thin layer of polyethyleneimmine (PEI) is deposited on the PAN structure.

Polyethyleneimines (PEI), or polyaziridines (organic polymers with the chemical formula H[CH ₂ -CH ₂ -NH-] _n H) have amines (primary, secondary, and/or tertiary depending on their linear or branched structure) and are therefore polycationic in nature.

Thus, the ST® Baxter-Gambro membranes used are highly electropositive and cationic.

There is therefore an identity of charge of the drug and the membrane when using this membrane with aminoglycosides. A study was therefore carried out with a highly cationic aminoglycoside, gentamicin.

Visual analysis of the elimination curves, whether in filtration or diafiltration of gentamicin, shows rapid early elimination followed by a long plateau phase which represents 2/3 of the total session time. At 120 min, 84% of the dose has been eliminated (Figure 3).

This visual analysis of the absence of elimination induced by gentamicin in the late phase contrasts with the progressive reduction of amikacin with superposition of the curves indicating elimination by pure filtration (figure 4).

The analysis of the results made it possible to highlight a sieving coefficient greater than 1, which is normally impossible, because the concentration in the effluent should at most be equal to the concentration at the inlet of the filter. Such a value greater than 1 requires that energy has been supplied, which the model, using passive mechanisms (dialysis, filtration and adsorption) does not provide. On the other hand, it is known that an energy proportional to the inverse square of the distance is generated when two positive charges approach each other.

It can therefore be hypothesized that anomalies in the elimination curves of ionized substances with SC values greater than 1 result from a charge identity effect of the drug and the filter.

The method thus implemented thus made it possible to highlight the consequences of the load of the filters on the elimination of certain drugs. We can thus think that the membrane/drug charge identity promotes an early and transient extrusion of the drug at high concentrations (SC > 1) while it prevents its elimination at low concentrations (plateau of concentrations during the secondary phase of elimination of gentamicin.

of the here in a filter derived from the

The adsorption of gentamicin in filters has been regularly demonstrated.

The interaction can be described in terms of amount sequestered (mg or % of initial dose) or clearance. However, these assessments do not respond to the general principle of pharmacokinetics. The sequestration phenomenon leads to zero-order kinetics.

The zero-order kinetics (Michaelis-Menten) is defined by

Vimax: maximum initial speed of the phenomenon

Km: concentration/dose resulting in half of V _imax

It was assumed that the V _imax and Km of the gentamicin/ST150 filter interaction could be assessed using the model described above.

Material and methods

The ST®, Baxter-Gambro filter was used in diafiltration mode at a flow rate set at 2.5 L/h. Simulated blood flow and weight loss: set to 0.

5 liter bag of Hemosol crystalloid solution: used in the compartments.

Gentamicin concentrations were measured by immunoenzymatic method.

The elimination of gentamicin was evaluated by the method described above using increasing initial concentrations (introduced into the central compartment)

The concentrations tested ranged from 4 to 240 mg/L (initial doses ranging from 20 to 1200 mg).

Twelve 6-hour sessions were performed, resulting in >90% removal.

Results

The sequestration of gentamicin in the ST® filter is well described by a Michaelis-Menten equation. The targeted therapeutic concentrations of gentamicin are around the Vmax, which may explain the difficulties in defining the optimal dose.

These results show that the gentamicin clearance provided by the ST® filter should no longer be considered as a constant parameter of gentamicin elimination but rather a concentration-dependent parameter.

Figure 5 shows the curve for determining the V _imax and Km. The dotted line shows Km (V _imax / 2).

Claims

CLAIMS Method for determining the adsorption properties of a biomedical membrane for a given active substance after a predetermined duration, comprising: a) The implementation, during the predetermined duration, of steps consisting of: i. Passing a first crystalloid solution containing the active substance from a central compartment to an exchange compartment filtration compartment containing the biomedical membrane and into which an exchange fluid is injected in order to obtain a retentate and potentially a permeate, and measuring the instantaneous concentration C _in iet of the active substance in the solution at the entrance to the exchange compartment at different times chosen during the predetermined duration, ii. Reinject the retentate into the central compartment, and measure the instantaneous concentration of the active substance in the retentate before reinjection (C _or tiet) of the loss compensation solution and into the central compartment at the various times chosen during the predetermined period iii. If a permeate is obtained, reject the permeate in a rejection compartment, and measure the instantaneous concentration C _e ffi instant of the active substance in the permeate before reinjection into the rejection compartment at the different times chosen during the predetermined period b) After the predetermined duration, i. Determine the overall quantity of active substance eliminated in the system: Qccéiiminée = Qccinitiai - Qcctinai by the difference between the quantity of active substance initially brought into the central compartment (Qccinitiai) and the quantity of active substance present in the central compartment at the end the predetermined duration (Qcctinai) ii. If a permeate is obtained, determine, for each moment, the sieving coefficient SC, in particular by the formula SC = (C _e fti instant) /(Cj _n iet) measuring the ratio of the concentration in the permeate at each instant / concentration at the inlet: iii. Determine the extraction coefficient, expressed as a percentage, by calculating the ratio of the difference of the concentration at the entrance to the exchange compartment and the concentration at the exit (towards the central compartment)) to the concentration at the exchange compartment entrance); iv. If a permeate is obtained, determine the quantity of active substance Qcumuieffiu eliminated by dialysis and/or filtration v. Determine the quantity of active substance adsorbed on the membrane by the formula Q _adS or = Qeiiminée cc - Qcumul efflu vi. Determine the adsorption of the membrane by the formula Q _adS or /

2. Method according to claim 1, characterized in that the membrane is a dialysis membrane, and that the exchange fluid is a second crystalloid solution without active substance, the method also comprising determining the quantity of active substance eliminated by dialysis.

3. Method according to claim 1, characterized in that the membrane is a filtration membrane, and that the exchange fluid is a second solution of crystalloid without active substance, the method also comprising the determination of the quantity of active substance eliminated by filtration.

4. Method according to claim 1, characterized in that the membrane is a diafiltration membrane, and that the exchange fluid is a second crystalloid solution without active substance, the method also comprising the determination of the quantity of active substance eliminated by dialysis and the determination of the quantity of active substance eliminated by filtration.

5. Method according to claim 1, characterized in that the membrane is an extracorporeal oxygenation membrane, and that the exchange fluid is a gas containing oxygen [said method not comprising obtaining a permeate liquid].

6. Method according to claim 2, in which the quantity of active substance eliminated by dialysis is determined by measuring the difference in the areas under the curve (AUC) of the instantaneous concentrations at the inlet or the outlet of the exchange compartment, multiplied by the simulated blood flow Q _dja = (AUC Cjniet-AUC Coutiet) x simulated blood flow.

7. Method according to claim 3, in which the quantity of active substance eliminated by filtration is determined by Qfii _tr = Qcumui effiu-

8. Method according to claim 4, in which the quantity of active substance eliminated by dialysis is determined by Q _dja = (AUC Cmiet-AUC C _or tiet) x simulated blood flow and the quantity of active substance eliminated by filtration is determined by Qfntr ^— Qcumui effiu - Qdia-

9. Method according to one of claims 1 to 4 and 6 to 8, in which the quantity of active substance Qc _Um uieffiu eliminated by dialysis and/or filtration is calculated by measuring the area under the curve (AUC) representing the concentrations in the instantaneous effluents (Ceffi instant) ) Qcumui ettiu- Sum (AUC Cetti instant) x flow rate of the exchange fluid.

10. Method according to one of claims 1 to 4 and 6 to 8, in which the quantity of Ocumuiettiu active substance eliminated by dialysis and/or filtration is calculated by measuring the final concentration (C _cu mui ettiu) in the compartment of discharge Qcumui eftiu= (C _cu mui eftiu x Volume of the discharge compartment).

11 . Method according to one of claims 1 to 4, and 6 to 10, in which the method simulates continuous extra-renal purification (CRRT), and that the percentage eliminated from the central compartment during the duration of the study is calculated under normal diafiltration flow conditions, % eliminated of CC = (QCeliminated / QcCinitial) X 100.

12. Method according to one of claims 1 to 11, in which the respective shares of dialysis, filtration and adsorption in the elimination of the active substance are calculated:

- Share of dialysis = (Q _dia / Qcceiiminée) x 100 - Filtration share = (Qfii _tr I Qccéiiminée) x 100

- Share of adsorption = (Qadsor / Qcceimined) x 100).

13. Method for calculating the v _imax (maximum initial speed) and Km (Michaelis constant) corresponding to the interaction of a given active substance with a biomedical membrane, comprising i) repeating the method according to one of claims 1 to 12 for different concentrations of the given active substance and determine the adsorption values for each concentration of given active substance, as well as the quantity of active substance eliminated by dialysis, filtration and/or diafiltration, and determine the percentage thereof, relative to the quantity initially supplied ii) Calculate the parameters v _imax and Km using the data obtained in i).

14. Computer program product, comprising a set of instructions which, when executed by processing means, is capable of implementing the method according to one of claims 1 to 12, for providing the adsorption of a given active substance on a membrane after a predetermined time.

15. A non-transitory computer-readable storage medium on which is stored a computer program comprising program instructions, the computer program loadable into a data processing unit and being adapted to cause the data processing unit to carrying out a method according to any one of claims 1 to 12.

16. Method for preparing an extracorporeal circulation circuit for the implementation of a method according to one of claims 1 to 12, comprising:

The provision of a bag containing a crystalloid solution in physiological solution, with a volume of between 4.5 and 5 liters The measurement of the exact volume of the solution included in the bag The provision of a quantity of active substance to be tested (drug), and the dilution of this active substance in the crystalloid near, using a defined volume of the crystalloid solution (preferably from the bag), in order to determine the concentration of the active substance in the pocket. The provision of an extracorporeal circulation device comprising o a first circuit to which the bag containing the crystalloid solution and the active substance is connected, said first circuit containing a pump allowing the loop circulation of the crystalloid solution in the first circuit, the crystalloid solution passing through an exchange compartment containing the membrane to be tested for the active substance, sampling sites being located at the inlet and at the outlet of the exchange compartment. o A second circuit in which passes an exchange fluid, said second circuit containing a pump allowing the circulation of the exchange fluid through the exchange compartment.

17. Method according to claim 15, in which the exchange fluid is a liquid and the second circuit has a sampling site at the outlet of the exchange compartment, and one or more pockets for collecting the exchange fluid after passage in the exchange compartment.

18. Method according to claim 15, in which the exchange fluid is a gas and the device further contains a second pocket of the same volume as the pocket containing the crystalloid solution and the active substance, containing a crystalloid solution of similar composition, but containing no active principle, said second pocket being able to be connected to the first circuit when the pocket containing the crystalloid solution and the active substance is disconnected therefrom.