Method for Obtaining Low Molecular Weight Heparins by Tangential Flow Filtration

Abstract

Method for obtaining low molecular weight heparins (LMWH) with a weight average molecular weight distribution between 3.0 and 5.0 kDa comprising at least one concentration step by tangential flow filtration (TFF). The method is particularly useful for the preparation of bemiparin and enoxaparin without the use of fractional precipitation nor the use of alcoholic solutions. In particular, the preparation of LMWH is obtained by depolymerization of heparin and filtration (TFF ultrafiltration and/or diafiltration) of the depolymerized heparin without the use of fractional precipitation and without an alcoholic solution.

Claims

1. A method for preparing low molecular weight heparins (LMWH) with an average molecular weight distribution in the range of approximately 3.0 to approximately 5.0 kDa, the method comprising: a) providing a crude depolymerized heparin solution with an oligosaccharide chain weight average molecular weight distribution in the range of approximately 0.6 to approximately 10 kDa and a heparin concentration of up to approximately 4% w/v; b) performing at least one concentration step by aqueous phase tangential flow filtration (TFF) using a nominal cut-off 1 kDa membrane to provide a heparin concentrate having a heparin concentration of up to approximately 25% w/v, thereby obtaining said LMWH.

2. The method according to claim 1, wherein a first concentration step (b) is performed to provide a heparin concentration of up to approximately 10% w/v and a second concentration step (b) is performed to provide a heparin concentration in the range of approximately 10% to approximately 25% w/v.

3. The method according to claim 2, wherein the second concentration step is performed to provide a heparin concentration of in the range of approximately 12% to approximately 22% w/v.

4. The method according to claim 1 further comprising the step of clarifying the crude depolymerized heparin solution of step (a).

5. The method according to claim 1 further comprising performing at least one depth filtration.

6. The method according to claim 5, wherein said depth filtration is performed prior to a first TFF concentration step if only one is performed, or after the first TFF concentration step if more than one TFF concentration step is performed.

7. The method according to claim 1 further comprising at least one step of diafiltration with water.

8. The method according to claim 7, wherein said diafiltration step is performed prior to a first TFF concentration step if only one is performed or after the first TFF concentration step if more than one TFF concentration step is performed.

9. The method according to claim 1 further comprising at least one step of treating with H.sub.2O.sub.2.

10. The method according to claim 10, wherein a step of treating with H.sub.2O.sub.2 is performed after a first TFF concentration step if only one is performed, or after the first TFF concentration step if more than one TFF concentration step is performed.

11. The method according to claim 1 further comprising the step of lyophilizing said heparin concentrate.

12. The method according to claim 1, wherein the crude depolymerized heparin excludes a crude depolymerized heparin that has been prepared by fractional precipitation.

13. The method according to claim 1, wherein said method excludes a step of fractional precipitation of heparin or depolymerized heparin.

14. The method according to claim 1, wherein the weight average molecular weight (Mw) of said LMWH is in the following ranges TABLE-US-00011 Fraction M1 Fraction M2 Fraction M3 Mw, Da <2000 Da, % 2000-8000 Da, % >8000 Da, % 3800-5000 12.0-20.0 68.0-82.0 ≤18.0.

15. The method according to claim 1, wherein the weight average molecular weight (Mw) of said crude depolymerized heparin is in the following ranges TABLE-US-00012 Fraction M1 Fraction M2 Fraction M3 Mw, Da <2000 Da, % 2000-8000 Da, % >8000 Da, % 3000-5000 ≤25 60-80 ≤20.

16. The method according to claim 1, wherein the weight average molecular weight (Mw) of said LMWH obtained is in the following ranges TABLE-US-00013 Fraction M1 Fraction M2 Fraction M3 Mw, Da <2000 Da, % 2000-6000 Da, % >6000 Da, % 3000-4200 <35.0 50.0-75.0 ≤15.0.

17. The method according to claim 1, wherein the weight average molecular weight (Mw) of said crude depolymerized heparin is in the following ranges TABLE-US-00014 Fraction M1 Fraction M2 Fraction M3 Mw, Da <2000 Da, % 2000-6000 Da, % >6000 Da, % 2500-5000 <40 50-75 <25.

18. The method according to claim 1, wherein the concentration by tangential flow filtration (TFF) employs a 0.7 to 1 kDa membrane.

19. The method according to claim 18, wherein a 0.9 to 1 kDa membrane or 1 kDa membrane is employed.

20. The method according to claim 1, wherein the heparin is enoxaparin sodium.

21. The method according to claim 1, wherein the process excludes a step of fractional precipitation with alcohol.

22. The method of claim 21, wherein the LMWH product of the process excludes alcohol.

23. The method of claim 1, wherein the process excludes a) precipitation of the LMWH or salt thereof, b) a chromatographic purification of the LMWH or salt thereof, c) treatment of the LMWH with ion exchange resin; d) an organic solvent extraction of the LMWH; e) precipitation of the LMWH or salt thereof; f) treatment of the LMWH with enzyme; g) use of organic solvent in the process; or h) any combination thereof.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0050] The accompanying drawings, which are appended herein and form part of the specification, illustrate one or more embodiments of the present invention and, together with the description, also contribute to the explanation of the principles of the present invention and to allow a person skilled in the art to reproduce and use the invention. The following drawings are provided for illustration purposes only and in no way limit the full scope of the present invention.

[0051] FIG. 1: Comparative schematic for normal flow filtration and tangential flow filtration.

[0052] FIG. 2: General schematic of the tangential filtration process.

[0053] FIG. 3: Comparison of concentration and diafiltration processes.

[0054] FIGS. 4A and 4B: Schematics of the purification process according to the invention.

[0055] FIGS. 5A-5D: Comprise various graphs representing the variation in both average molecular weight (FIG. 5A) and distribution of molecular weights (FIG. 5B: <2000 kDa; FIG. 5C: >8000 kDa; FIG. 5D: 2000-8000 kDa) is linear in the second concentration from 10% to 20% of nominal concentration of the product in the retentate.

[0056] FIG. 6: Nominal cut-off: definition.

DESCRIPTION OF THE INVENTION

[0057] In the present invention “tangential flow filtration” or “TFF” is understood as the filtration technique in which the solution to be filtered passes tangentially over the surface of the filter, such that the pressure difference that is generated allows components that are smaller than the pore size to pass through same (permeate). Larger components are retained over the filter surface and returned to the feed tank (retentate).

[0058] In the present invention “clarification” is understood as the filtration performed to eliminate particles in suspension present in the solution, such as filtration performed by filters from 1-60 microns, preferably from 1-25 microns.

[0059] In the present invention, “depth filtration” is understood to mean filtration in which a multi-step labyrinth filter medium is used, which helps retain the particles. The larger particles will be retained on the surface and the finer ones follow their path towards the inside of the filter medium being trapped in the inner layers, so that the turbidity of the dissolution is reduced. In a specific embodiment, this is a filtration performed with filters of 1-5 microns, preferably of 2-4 microns. The filtration can be performed with water or buffered solution.

[0060] In the present invention, “concentration” is understood as the tangential filtration step in which the retained product increases its concentration in the solution by removing permeate (see FIG. 3).

[0061] In the present invention, “diafiltration” is understood as the tangential filtration step in which, while the permeate is removed, the solution is fed with the same buffer solution of water flow rate, so that the concentration of the retentate in the solution is not modified (see FIG. 3). In this case, a membrane such as that used in concentration by TFF can be used; that is, a membrane with a nominal cut-off of approximately 51 kDa, preferably approximately 0.7 to approximately 1 kDa, more preferably approximately 0.9 to approximately 1 kDa, and even more preferably approximately 1 kDa.

[0062] As used herein, “molecular weight” is “weight average molecular weight”.

[0063] In one embodiment, the heparin (after depolymerization) is a sodium salt of heparin, e.g. enoxaparin sodium or bemiparin sodium.

[0064] Crude enoxaparin sodium can be obtained by alkali (e.g. NaOH) depolymerization of the benzyl ester of heparin obtained from pig intestinal mucosa.

[0065] In one specific embodiment, the product obtained after depolymerization of enoxaparin sodium corresponds to a solution which, in addition to containing raw enoxaparin sodium, contains impurities corresponding to the saponification in alkaline medium of the benzyl ester of heparin, in addition to salts corresponding to the pH adjustments made during the breaking process. According to one embodiment of the invention, the TFF process is carried out on this raw enoxaparin sodium solution, so that the concentration of this solution is carried out with the aim, on the one hand, of eliminating low molecular weight impurities, and on the other, of reaching the appropriate concentration to carry out the bleaching treatment with hydrogen peroxide. Alternatively, bleaching treatment with H.sub.2O.sub.2 could be performed prior to the step of concentration by TFF. These steps can be performed without the use of fractional precipitation of the crude product (crude depolymerized heparin).

[0066] Alternatively, bleaching treatment with H.sub.2O.sub.2 could be performed prior to the step of concentration by TFF.

[0067] Additionally, a diafiltration process (optional) can be carried out for a thorough removal of low molecular weight impurities, before or after the concentration step. At the end of this step, optionally a second concentration is performed in order to remove the generated saline impurities and adjust the content in low molecular weight oligosaccharide chains, for which the average molecular weight of the solution is monitored; once the optimal value is reached, it is lyophilized to obtain enoxaparin sodium of the appropriate purity.

[0068] Therefore, in one specific embodiment the process of the invention comprises: [0069] a) providing a depolymerized heparin solution with an oligosaccharide chain distribution range of between 0.6 and 10 kDa and an enoxaparin sodium concentration of approximately 4% w/v; [0070] b) carrying out a concentration step by TFF in aqueous phase using a membrane with nominal cut-off ≤1 kDa until obtaining a heparin concentration of up to approximately 25% w/v, of up to approximately 10% w/v, or preferably between approximately 5% and approximately 10% w/v; [0071] c) optionally, performing a diafiltration step with water (e.g. Non-buffered water) before or after step b), [0072] d) performing a treatment step with H.sub.2O.sub.2 before or after step b), [0073] e) optionally, performing a single concentration step by aqueous phase TFF using a 51 kDa nominal cut-off membrane to achieve a heparin concentration of up to 25% w/v, preferably between approximately 12% and approximately 25% w/v; and [0074] f) performing a lyophilization step on the product obtained.

[0075] In a specific embodiment, the process of the invention comprises: [0076] a) providing a depolymerized enoxaparin sodium solution with an oligosaccharide chain distribution range of between approximately 0.6 and approximately 10 kDa and an enoxaparin sodium concentration of up to 4% w/v; [0077] b) carrying out a concentration step by TFF in aqueous phase using a membrane with nominal cut-off ≤1 kDa until obtaining a heparin concentration of up to approximately 25% w/v, of up to approximately 10% w/v, or preferably between approximately 5% and approximately 10% w/v; [0078] c) performing a treatment step with H.sub.2O.sub.2 on the product obtained in step b), [0079] d) performing a single concentration step by aqueous phase TFF using a 51 kDa nominal cut-off membrane to achieve a heparin concentration of up to 25% w/v, preferably between approximately 12% and approximately 25% w/v; and [0080] e) performing a lyophilization step on the product obtained. [0081] Preferably, the process of this invention includes a clarification and/or depth filtration step before step b).

[0082] In a specific embodiment, the process of the invention comprises: [0083] a) providing a depolymerized enoxaparin sodium solution with an oligosaccharide chain distribution range of between approximately 0.6 and approximately 10 kDa and an enoxaparin sodium concentration of up to 4% w/v; [0084] b) carrying out a concentration stage by TFF in aqueous phase using a membrane with nominal cut-off ≤1 kDa until obtaining a heparin concentration of up to approximately 25% w/v, of up to approximately 10% w/v, or preferably between approximately 5% and approximately 10% w/v; [0085] c) performing a diafiltration step with water on the product obtained in step b), [0086] d) performing a treatment step with H.sub.2O.sub.2 on the product obtained in step c), [0087] e) optionally, performing a depth filtration step on the product obtained in step d), [0088] f) performing a single concentration step by aqueous phase TFF using a 51 kDa nominal cut-off membrane to achieve a heparin concentration of up to 25% w/v, preferably between approximately 12% and approximately 25% w/v; and [0089] g) performing a lyophilization step on the product obtained. [0090] Preferably, the process of this invention includes a clarification and/or depth filtration step before step b).

[0091] In a specific embodiment, the process of the invention comprises: [0092] a) providing a depolymerized enoxaparin sodium solution with an oligosaccharide chain distribution range of between 0.6 and 10 kDa and an enoxaparin sodium concentration of up to 4% w/v; [0093] b) performing a single concentration step by aqueous phase TFF using a 51 kDa nominal cut-off membrane to achieve a heparin concentration of up to 25% w/v, preferably between approximately 5% and approximately 10% w/v; [0094] c) performing a depth filtration step on the product obtained in step b), [0095] d) performing a treatment step with H.sub.2O.sub.2 on the product obtained in step c), [0096] e) performing a single concentration step by aqueous phase TFF using a 51 kDa nominal cut-off membrane to achieve a heparin concentration of up to 25% w/v, preferably between approximately 12% and approximately 25% w/v; and [0097] f) performing a lyophilization step on the product obtained. [0098] Preferably, the process of this invention includes a clarification and/or depth filtration step before step b).

[0099] In a specific embodiment, the process of the invention comprises: [0100] a) providing a depolymerized enoxaparin sodium solution with an oligosaccharide chain distribution range of between 0.6 and 10 kDa and an enoxaparin sodium concentration of up to 4% w/v; [0101] b) performing a diafiltration step with water on the solution of step a), [0102] c) performing a treatment step with H.sub.2O.sub.2 on the product obtained in step c), [0103] d) performing a single concentration step by aqueous phase TFF using a 51 kDa nominal cut-off membrane to achieve a heparin concentration of up to approximately 25% w/v, preferably between approximately 5% and approximately 20% w/v; and [0104] e) performing a lyophilization step on the product obtained. [0105] Preferably, the process of this invention includes a clarification and/or depth filtration step before step b).

EXAMPLES OF THE INVENTION

[0106] The following specific examples provided below serve to illustrate the nature of the present invention. These examples are included for illustrative purposes only and are not to be construed as limitations to the invention claimed herein. Crude enoxaparin sodium and bemiparin sodium were prepared by depolymerization of heparin without performing fractional precipitation.

[0107] The process is described below for obtaining crude enoxaparin sodium, the starting product used in described examples 1, 2, 3, 4 and 5. Dissolve 10 g of heparin sodium in purified water and under stirring add benzethonium chloride solution, forming benzethonium heparinate. Wash the formed product several times with water to remove excess chlorides and finally the dry the product by lyophilization. Dissolve the benzethonium heparinate in methylene chloride and adjust the temperature. Add benzyl chloride and allow to react. The product obtained is heparin benzyl ester. Dissolve the heparin benzyl ester in water and add sodium hydroxide. At the end of the reaction, neutralize the solution; the product obtained is crude enoxaparin sodium. After obtaining the crude enoxaparin sodium, the examples described below were performed.

Example 1

[0108] The method of the invention was performed with the following main steps: [0109] a) first concentration by TFF to obtain a product with a heparin concentration of 4% to 10% w/v; [0110] b) diafiltration and treatment with H.sub.2O.sub.2; [0111] c) second concentration by TFF to obtain a final product with a heparin concentration of 10% to 20% w/v.

[0112] Initially, crude enoxaparin sodium with a product concentration of 40 g/L and an oligosaccharide chain distribution between 0.6 and 10 kDa was used as the starting product. This initial product was prefiltered with a 3.0 μm Clarigard® filter. The product of the heparin depolymerization process is crude enoxaparin sodium.

[0113] Next, the first concentration step was carried out by TFF, intended to increase the concentration of enoxaparin to a value between 4% and 10%, as well as to reduce the concentration of contaminants (mainly salts with a molecular weight <0.5 kDa and other small products resulting from previous manufacturing processes). To do this, a Millipore® regenerated cellulose membrane with a nominal cut-off of ≤1 kDa was used. The concentration step was started with about 2005 g of product with a transmembrane pressure (TMP) of 3.25 bar prior to passing the permeate flow to a separate container, the system was kept in total recirculation for about 15 minutes.

[0114] During concentration, permeate flow ranged from 9 initial to 4.8 final LMH (L/m.sup.2/h) (53% of initial flow), for an average flow of 6.2 LMH. 2× permeate samples were collected, one in VCF (volume concentration factor)=1.26× (P1) and the other of the permeate volume at the end of concentration (P2). A sample of the volume of material retained at the end of concentration (R1), as well as a sample of the initial feed, was also taken after prefiltration by the 3.0 μm filter (B1). The main data of the concentration study are presented in Table 1.

TABLE-US-00003 TABLE 1 LMWH Concentration Study (4 .fwdarw. 10%) Mean Initial permeate Flow rate volume Extractable VCF flow T Time ΔPressure Tangential (L) volume (L) (X) (LMWH) (° C.) (min) (bar) (LMM) 2.005 1.2 2.49 6.2 23.6 106 0.5 .fwdarw. 0.7 5.3 .fwdarw.25.7

[0115] Subsequently, a diafiltration step was carried out using purified water, intended to clarify the resulting product. During diafiltration, permeate flow continuously decreased from an initial 8.4 to a final 2.4 LMH (approx. 29%), for an average flow of 5.03 LMH. 3× permeate samples were collected, each at the end of each diavolume (D1, D2 and D3 respectively) and one from the volume of material retained at the end of diafiltration (D5). The conditions under which the diafiltration study was carried out can be seen in Table 2:

TABLE-US-00004 TABLE 2 LMWH diafiltration study Mean Initial Diafiltration permeate Flow rate volume Extractable volumes flow T Time ΔPressure Tangential (L) volume (L) (N) (LMWH) (° C.) (min) (bar) (LMM) 0.798 2.3 3 5.03 26.5 390 0.3 .fwdarw. 0.8 5.3 .fwdarw.27.8

[0116] The approximately 0.8 L of product obtained after diafiltration was subsequently subjected to chemical reaction with H.sub.2O.sub.2.

[0117] Before continuing with the second concentration step by TFF, an additional (optional) clarification step was carried out, again using a 3.0 μm Clarigard® filter, to remove any particles that may have decanted at the bottom of the vessel.

[0118] Next, a second concentration step was carried out by TFF, this time aimed at achieving a concentration of enoxaparin of between 10% and 20% w/v, also using a Millipore® regenerated cellulose membrane of ≤1 kDa nominal cut-off.

[0119] During concentration, permeate flow ranged from 3 initial to 0.75 final LMH (25%), for an average flow of 1.5 LMH. 2× permeate samples were collected, one in VCF=1.34× (P3) and the other in the permeate volume at the end of concentration (P4). A sample was also taken of the volume of material retained after concentration (R2), after depolarizing the membrane by leaving the system running at a low TMP (1.2 bar) for 10 minutes.

[0120] Table 3 below shows the molecular weights of the samples taken during the process described above, analyzed according to the method established by the European Pharmacopoeia (EP).

TABLE-US-00005 TABLE 3 Molecular weights of samples obtained at different stages of the process Mw, <2000 2000-8000 >8000 Sample Da Da, % Da, % Da, % Raw enoxaparin sodium 3737 23.8 71.1 8.1 B1 (Clarigard ® clarification) 3723 24.0 71.1 5.0 P1 (1st permeate sample, 1st <600 — — — concentration) P2 (2nd permeate sample, 1st <600 — — — concentration) R1 (retentate after first concentration) 3645 25.2 69.9 4.8 D1 (permeate after first diavolume) 1196 96.5 3.5 0.0 D2 (permeate after second diavolume) 1491 84.5 15.5 0.0 D3 (permeate after third diavolume) 1646 78.1 21.9 7.0 D5 (retentate after completion of 4158 14.6 78.4 7.0 diafiltration) P3 (1st permeate sample, 2nd 1866 67.3 32.7 0.0 concentration) P4 (2nd permeate sample, 2nd 1991 60.3 39.7 0.0 concentration) R2 (retentate after second 4444 8.1 84.0 7.9 concentration)

[0121] As shown in the percentages of the MW fractions of the target product (less than 2000 Da, 2000 to 8000 Da and more than 8000 Da) in the table above, it appears that the first concentration step does not negatively affect the product profile (there is no loss of any fraction in the permeate as seen in samples P1 and P2).

[0122] For samples taken during diafiltration, D1, 02, 03 and 05 indicate that: [0123] basically there is no loss of the highest MW fraction in the permeate, along the diafiltration; [0124] there is some loss of the smallest and average MW fractions in the permeate, where the highest loss is always relative to the lowest fraction; [0125] the rate of loss of the smallest and medium fractions decreases, respectively, along the diafiltration; [0126] in general, there is some reduction and enrichment along the diafiltration of smaller fractions and medium-high fractions respectively (according to the ≤1 kDa membrane cut-off).

[0127] The numbers for the instantaneous permeate samples P3, P4 and sample R2 at the end of the second concentration step indicate that: [0128] basically there is no loss of the highest MW fraction in the permeate, throughout this step; [0129] there is some loss of the smallest and medium MW fractions in the permeate, where the highest loss is always relative to the smallest fraction (as in diafiltration); [0130] the rate of loss of the medium fraction increases in this second concentration step (as in diafiltration).

Example 2

[0131] The method of the invention was performed with the following main steps: [0132] a) First concentration by TFF to obtain a product with a heparin concentration from 4% to 10% w/v; [0133] b) H.sub.2O.sub.2 treatment; [0134] c) Second concentration by TFF to obtain a final product with a heparin concentration from 10% to 20% w/v.

[0135] Approximately 1936 g of starting product was transferred to the tank and the system was operated in total recirculation at a TMP=3.2 bar for 10 minutes at a cross-flow of 5.1 LMM.

[0136] During concentration, permeate flow ranged from an initial 9.6 to a final 4.2 LMH (44%), for an average flow of 6 LMH. A sample of instantaneous permeate was collected at VCF=1.43× (P5), another permeate sample after completion of the concentration (P6) and finally a sample of the retentate (R3) at the end of the concentration, after leaving the filter in total recirculation at TMP=0.6 bar for 10′ (membrane depolarization). A sample of the starting solution (B2) was also taken before transfer to the tank.

[0137] Approximately 794 g of product was transferred to the tank after being treated with H.sub.2O.sub.2 and proceeded to the second concentration step.

[0138] During concentration, permeate flow ranged from 5.7 initial to 0.9 final LMH (16.0%), for an average flow of 2.5 LMH. 1× instantaneous permeate sample was collected at VCF=1.33× (P7) and another of the permeate volume at the end of concentration (P8). A sample of the retentate volume at the end of concentration (R4) was also collected, after depolarizing the membrane by leaving the system running at a low TMP (0.8 bar) for 10 minutes.

TABLE-US-00006 TABLE 4 Molecular weights in the process steps Mw, <2000 Da, 2000-8000 >8000 Da, Sample Da % Da, % % Raw enoxaparin sodium 3646 24.9 70.6 4.5 B2 (Clarigard ® clarification) 3742 23.5 71.5 4.9 P5 (1st permeate sample, 1st <600 — — — concentration) P6 (2nd permeate sample, 1st <600 — — — concentration) R3 (retentate after first 3784 23.1 71.5 5.4 concentration) P7 (1st permeate sample, 2nd 1311 90.8 9.2 0.0 concentration) P8 (2nd permeate sample, 2nd 1476 84.4 15.6 0.0 concentration) R4 (retentate after second 4308 13.0 79.2 7.8 concentration)

[0139] Table 4 shows the molecular weights of the samples taken during the process described above, analysed according to the method established by the European Pharmacopoeia (EP).

[0140] As the numbers in the table above show, it appears that the first concentration step does not negatively affect the product profile (without loss of any fraction in the permeate, samples P5 and P6, as in samples P1 and P2).

[0141] The numbers relating to permeate samples P7, P8 and the retentate volume sample R4 at the end of the second concentration step indicate that: [0142] there is no loss of the highest fraction of MW in the permeate, throughout said step; [0143] there is some loss of the smallest and medium MW fractions in the permeate, the highest loss is always relative to the smallest fraction (as in the first diafiltration test); [0144] the rate of loss of the medium fraction increases in the second concentration (as in the diafiltration process).

[0145] Finally, the retentate solution is lyophilized to obtain dry enoxaparin sodium.

Example 3

[0146] The method of the invention was performed with the following main steps: [0147] a) First concentration by TFF to obtain a product with a heparin concentration from 4% to 10% w/v; [0148] b) Depth filtration; [0149] c) H.sub.2O.sub.2 treatment; [0150] d) Second concentration by TFF to obtain a final product with a heparin concentration from 10% to 20% w/v.

[0151] Approximately 2000 g of starting product (turbidity >1000 NTU) was transferred to the tank and the system was operated in full recirculation at a TMP=3.2 bar for 15 minutes at a cross-flow of 5.2 LMM.

[0152] During concentration, permeate flow ranged from an initial 9.6 to a final 3.9 LMH (41% of initial flow), for an average flow of 5.8 LMH. A permeate sample was collected at VCF=1.43× (P1′) and another at the end of concentration (P2′) as well as the retentate (R1′) at the end of concentration after leaving the filter in total recirculation at TMP=0.7 bar for 15′ (membrane depolarization). A sample of the initial feed (B1′) was also taken prior to transfer to the tank.

[0153] The R1′ solution was passed through a Millistak+® HC Pro COSP depth filter, reducing turbidity to 0.57 NTU, to be subsequently treated with H.sub.2O.sub.2. Approximately 0.8 litres of product were transferred to the tank after being treated with H.sub.2O.sub.2 and the second concentration step was performed.

[0154] This second concentration was performed by sequentially aliquoting both the retentate and the permeate from the initial nominal concentration of 10% (samples R′3 and P′3, respectively) to the final nominal concentration of 20% (samples R′12 and P′12, respectively), passing through the intermediate concentrations of 11, 12, 13, 14, 15, 16, 17, 18 and 19%.

[0155] Table 5 below shows the molecular weights of the samples taken during the process described above, analyzed according to the method established by the European Pharmacopoeia (EP).

TABLE-US-00007 TABLE 5 Molecular weights in the process steps Mw, <2000 2000-8000 >8000 Sample Da Da, % Da, % Da, % B1′ (Clarigard ® clarification) 3464 29.6 65.95 4.50 P1′ (1st permeate sample, 1st conc.) <600 — — — P2′ (2nd permeate sample, 1st conc.) <600 — — — R1′ (retentate after first concentration) 3416 31.90 63.37 4.70 P3′ (1st permeate sample 11%, 2nd 1054 100.00 0.00 0.00 conc.) P4′ (1st permeate sample 12%, 2nd 945 99.20 0.77 0.00 conc.) P5′ (1st permeate sample 13%, 2nd 999 98.40 1.59 0.00 conc.) P6′ (1st permeate sample 14%, 2nd 1060 96.90 3.10 0.00 conc.) P7′ (1st permeate sample 15%, 2nd 1139 94.20 5.76 0.00 conc.) P8′ (1st permeate sample 16%, 2nd 1197 92.20 7.82 0.00 conc.) P9′ (1st permeate sample 17%, 2nd 1255 90.00 10.04 0.00 conc.) P10′ (1st permeate sample 18%, 2nd 1288 88.70 11.29 0.00 conc.) P11′ (1st permeate sample 19%, 2nd 1378 85.40 14.63 0.00 conc.) P12′ (1st permeate sample 20%, 2nd 1330 87.20 12.84 0.00 conc.) R3′ (1st retentate sample 11%, 2nd 3744 24.6 69.9 5.6 conc.) R4′ (1st retentate sample 12%, 2nd 3757 24.3 70.0 5.7 conc.) R5′ (1st retentate sample 13%, 2nd 3795 23.5 70.8 5.8 conc.) R6′ (1st retentate sample 14%, 2nd 3827 22.6 71.6 5.9 conc.) R7′ (1st retentate sample 15%, 2nd 3881 21.5 72.4 6.1 conc.) R8′ (1st retentate sample 16%, 2nd 3937 20.2 73.5 6.3 conc.) R9′ (1st retentate sample 17%, 2nd 3994 18.9 74.7 6.4 conc.) R10′ (1st retentate sample 18%, 2nd 4005 18.4 75.1 6.4 conc.) R11′ (1st retentate sample 19%, 2nd 4079 16.7 76.6 6.7 conc.) R12′ (1st retentate sample 20%, 2nd 4116 15.9 77.3 6.8 conc.)

[0156] The variation both in average molecular weight and in the distribution of molecular weights is linear during the 2nd concentration from 10% to 20% nominal concentration of the product in the retentate, so that, by adjusting the final value of the concentration of the product in the solution of the retentate, it is possible to define a certain molecular weight profile and distribution of oligosaccharide chains for obtaining enoxaparin sodium. This fact can be seen in FIG. 5. Finally, the retentate solution is lyophilized to obtain dry enoxaparin sodium.

Example 4

[0157] The product obtained in the previous example was analysed to determine its anti-FXa and anti-FIIa activity. The results obtained were as follows:

TABLE-US-00008 Anti-FXa activity, IU/mg (dried Anti-FIIa activity, IU/ mg Ratio substance) (dried substance) aFXa/a FIIa 112 29.5 3.8

[0158] This quality attribute adequately fulfils the ranges defined by both European Pharmacopoeia and US Pharmacopoeia for enoxaparin sodium: [0159] Anti-FXa activity: 90-125 IU/mg (dried substance) [0160] Anti-FIIa activity: 20.0-35.0 IU/mg (dried substance) [0161] Ratio aFXa/aFIIa: 3.3-5.3

Example 5

[0162] The method of the invention was performed with the following main steps: [0163] a) Diafiltration; [0164] b) H.sub.2O.sub.2 treatment; and [0165] c) Concentration by TFF to obtain a final product with a heparin concentration from 4% to 15% w/v.

[0166] Approximately 3010 g of starting product was transferred to the tank and the system was operated in total recirculation at a TMP=4.9 bar for 15 minutes at a cross-flow of 2.0 LMM.

[0167] During diafiltration, permeate flow ranged from an initial 17.7 to a final 13.1 LMH, for an average flow of 12.6 LMH and for 6 diavolumes. After each diafiltered volume samples were collected of both the permeate (DP1 to DP6) and the retentate (DR1 to DR6). The permeate was treated with H.sub.2O.sub.2 and approximately 2720 g of treated product were transferred to the tank.

[0168] During concentration, permeate flow ranged from 12.7 initial to 1.28 final LMH (89.9% reduction), for an average flow of 5.1 LMH. Concentration was performed from 4 to 15%, collecting both a permeate sample (CP1 to CP6) and a retentate sample (CR1 to CR6) from 10%. 1× instantaneous permeate sample was collected at VCF=1.33× (P7) and another of the permeate volume at the end of concentration (P8). A sample of the retentate volume at the end of concentration (R4) was also collected, after depolarizing the membrane by leaving the system running at a low TMP (0.8 bar) for 10 minutes.

[0169] Table 6 below shows the molecular weights of the samples taken during the process described above, analyzed according to the method established by the European Pharmacopoeia (EP).

TABLE-US-00009 TABLE 6 Molecular weights in the process steps MW, <2000 2000-8000 >8000 Sample Da Da, % Da, % Da, % DR1 (retentate 1 diavolume) 3703 27.40 66.42 6.20 DR2 (retentate 2 diavolumes) 3770 26.10 67.51 6.40 DR3 (retentate 3 diavolumes) 3807 25.00 68.51 6.50 DR4 (retentate 4 diavolumes) 3841 24.10 69.36 6.50 DR5 (retentate 5 diavolumes) 3856 23.50 69.93 6.50 DR6 (retentate 6 diavolumes) 3884 22.90 70.40 6.70 CR1 (retentate concentration 10%) 3889 22.20 71.24 6.60 CR2 (retentate concentration 11%) 3912 21.60 71.77 6.60 CR3 (retentate concentration 12%) 3954 20.70 72.54 6.80 CR4 (retentate concentration 13%) 3967 20.10 73.08 6.80 CR5 (retentate concentration 14%) 3995 19.30 73.78 6.90 CR6 (retentate concentration 15%) 4029 18.50 74.54 7.00

[0170] The variation both in average molecular weight and in the distribution of molecular weights is linear from 10% to 15% nominal concentration of the product in the retentate, so that, by adjusting the final value of the concentration of the product in the solution of the retentate, it is possible to define a certain molecular weight profile and distribution of oligosaccharide chains for obtaining enoxaparin sodium. The retentate solution is lyophilized to obtain dry enoxaparin sodium.

Example 6

[0171] The product obtained in the previous example was analysed to determine its anti-FXa and anti-FIIa activity. The results obtained were as follows:

TABLE-US-00010 Anti-FXa activity, IU/mg (dried Anti-FIIa activity, IU/ mg Ratio substance) (dried substance) aFXa/a FIIa 104 26.8 3.9

[0172] In view of the preceding description and the examples, a person skilled in the art would arrive at the invention as claimed without needing to resort to undue experimentation. The above will be better understood with reference to the preceding examples, which describe certain processes for the preparation of embodiments of the present invention. All the references made to these examples are for illustration purposes only. The examples must not be considered limiting and are only illustrations of some of the many possible embodiments considered by the present invention.

[0173] As used in the present document, the term “approximately” means±10%, ±5% or ±1% of the specified value, preferably ±10%. Moreover, all the ranges specified in the present document include the limits of the range and all the whole and fractional values, particularly according to the definition of the term “approximately”.