METHODS TO REDUCE ADVERSE EVENTS CAUSED BY PHARMACEUTICAL PREPARATIONS COMPRISING PLASMA DERIVED PROTEINS

Abstract

The instant invention provides a method to reduce adverse events caused by a pharmaceutical preparation derived from a plasma fraction wherein the method comprises contacting the plasma fraction with heparin or a heparin-like substance thereby reducing the activity of at least one activated serine protease per ml of the plasma fraction.

Claims

1.-15. (canceled)

16. A method of removing activated serine proteases from a plasma fraction comprising antithrombin III, wherein the method comprises adsorbing the plasma fraction to an anion exchange (AEX) matrix and contacting the matrix-adsorbed plasma fraction with heparin or a heparin-like substance, wherein the activity of at least one activated serine protease per ml of the plasma fraction is reduced.

17. The method according to claim 16, wherein the heparin or heparin-like substance is covalently bound to a matrix

18. The method according to claim 16, wherein the heparin or heparin-like substance is added to the plasma fraction in a soluble form.

19. The method according to claim 16, wherein the plasma fraction is an 8% ethanol supernatant I obtained from a Cohn/Oncley or Kistler/Nitschmann plasma fractionation.

20. The method according to claim 16, wherein the plasma fraction comprises an intermediate of a therapeutic plasma protein preparation.

21. The method according to claim 20, wherein the intermediate is cryo-poor plasma.

22. The method according to claim 21, wherein adsorbing the cryo-poor plasma to the AEX matrix facilitates isolation of proteins of the Prothrombin complex and/or allows adsorption of c1-esterase inhibitor to the AEX matrix.

23. The method according to claim 16, wherein the AEX matrix is DEAE or QAE.

24. The method according to claim 16, wherein the AEX matrix is an anion exchange membrane.

25. The method according to claim 16, wherein the activated serine protease is kallikrein, FXIa, or FXIIa.

26. The method according to claim 16, further comprising preparing a pharmaceutical preparation from the plasma fraction contacted with heparin or a heparin-like substance, wherein the pharmaceutical preparation has reduced adverse events compared to a pharmaceutical composition prepared without contacting the plasma fraction with heparin or a heparin-like substance, wherein the adverse events comprise one or more of thrombosis, skin reactions, bronchospasms, hypoxia, severe rigors, tachycardia, stomach aches, and raised blood pressure.

27. The method according to claim 16, wherein the plasma fraction is an intermediate for preparation of an immunoglobulin preparation.

28. The method according to claim 16, wherein the plasma fraction is an intermediate for preparation of an albumin preparation.

29. A method of removing activated serine proteases from a plasma fraction comprising antithrombin III, wherein the method comprises contacting the plasma fraction with heparin or a heparin-like substance covalently bound to a matrix, wherein the activity of at least one activated serine protease per ml of the plasma fraction is reduced.

30. The method according to claim 29 wherein the plasma fraction is an 8% ethanol supernatant I obtained from a Cohn/Oncley or Kistler/Nitschmann plasma fractionation.

31. The method according to claim 29, wherein the activated serine protease is kallikrein, FXIa, or FXIIa.

32. The method according to claim 29, further comprising preparing a pharmaceutical preparation from the plasma fraction contacted with heparin or a heparin-like substance, wherein the pharmaceutical preparation has reduced adverse events compared to a pharmaceutical composition prepared without contacting the plasma fraction with heparin or a heparin-like substance, wherein the adverse events comprise one or more of thrombosis, skin reactions, bronchospasms, hypoxia, severe rigors, tachycardia, stomach aches, and raised blood pressure.

33. The method according to claim 29, wherein the plasma fraction is an intermediate for preparation of an immunoglobulin preparation.

34. The method according to claim 29, wherein the plasma fraction is an intermediate for preparation of an albumin preparation.

Description

FIGURES

[0059] FIG. 1: Coagulation cascade.

[0060] FIG. 2: Schematic of a modified Cohn/Oncley industrial plasma fractionation.

[0061] FIG. 3: Schematic of a modified Kistler/Nitschmann industrial plasma fractionation.

[0062] FIG. 4: Processing alternatives for manufacturing to the fraction II/III stage in Cohn/Oncley industrial plasma fractionation schemes.

[0063] FIG. 5: Analytical Results (Predicted Response Graph) for coagulation related serine protease activity as a function of the level of removal of either antithrombin III (AT III), c1-esterase inhibitor (C1) or prothrombin complex (PT) from a pharmaceutical preparation, SC Immunoglobulin. The statistical analysis of testing results was performed by using the computer program Cornerstone Version 5.0 (Applied Materials Co.).

[0064] FIG. 6: Correlation between Prekallikrein-Ag and Kallikrein-like activity (a); correlation between Factor XI-Ag and Factor XI-like activity (b).

EXAMPLES

Example 1

Example 1.1

Introduction

[0065] The analysis of levels of kallikrein and FXIa in multiple batches of subcutaneous immunoglobulin suggested higher levels related to batches in which a greater proportion of the plasma fraction was subjected to PT and C1-INH adsorption and only a small amount of the plasma fraction was adsorbed to the heparin affinity resin. While the Vitamin-K dependent factors of the clotting system are adsorbed to DEAE-Sepharose, the factors involved in the contact activation system remain in the Ig-fraction. These factors, namely high molecular weight kininogen complexed to prekallikrein or FXI and FXII are known to be activated by negatively charged surfaces (McMillin C R, et al.: The secondary structure of human Hageman factor (factor XII) and its alteration by activating agents. J Clin Invest. 1974; 54, 1312-22). Surprisingly however the batch analysis suggested that positively charged materials such as anion exchange resins (eg. DEAE & QAE resins) could also activate these serine proteases. Thus the steps of preabsorbing a plasma fraction to an anion exchange (AEX) matrix and then contacting the plasma fraction with heparin or a heparin-like substance particularly where this substance can be subsequently removed from the plasma fraction provides an ideal means to ensure that the resulting pharmaceutical preparations are essentially free of activated coagulation factors such as FXIa. To investigate this further the following studies were conducted.

[0066] An analytical investigation of an immunoglobulin for subcutaneous administration (SC Immunoglobulin) was performed. Various analytical methods were applied with regard to the potential presence of trace amounts of activated clotting factors and proteolytic activity in the SC Immunoglobulin.

[0067] The evaluation of analytical data revealed that the SC Immunoglobulin batches contain levels of procoagulant activity in correlation to applied variations of the adsorption scheme. A comparison of adsorption schemes of individual batches revealed that higher levels of procoagulant activity are correlated to high Prothrombin complex (PT) and low antithrombin (AT III) adsorption levels during the plasma fractionation process steps.

[0068] Based on the finding of procoagulant activity the production process was adapted to include maximum AT III adsorption. The subsequent examples provide strong evidence that a high level of AT III adsorption leads to a significant decrease in the procoagulant activity of the SC Immunoglobulin product.

Example 1.2

Manufacturing Process for a SC Immunoglobulin

[0069] The drug substance was prepared by a modified Cohn Fractionation (Cohn E J, Strong L E, et al. Preparation and properties of serum and plasma proteins; a system for the separation into fractions of the protein and lipoprotein components of biological tissues and fluids. J Am Chem Soc 1946; 68:459-75). Plasma was thawed, the formed cryoprecipitate was separated and contained fibrinogen and antihemophilic Factor VIII/von Willebrand factor complex. With the supernatant, cryo-depleted plasma (also known as cryo-poor plasma), optional batch adsorption of the prothrombin complex (PT adsorption) and C1 esterase inhibitor (C1 adsorption) could be optionally performed (see FIG. 4). Subsequently, ethanol was added to the cryo-depleted plasma or filtrate from previous adsorption(s) to adjust an ethanol concentration of 8%. The precipitate, Cohn Fraction I, mainly contained fibrinogen and factor XIII and was separated by filtration. With the 8% ethanol (Fraction I) supernatant an optional batch adsorption of antithrombin III could be performed.

[0070] 60 to 100 mL heparin affinity resin per liter cryo-depleted plasma was suspended with the same quantity of Fraction I supernatant in a chromatography column. The obtained suspension is added to the residual quantity of the batch for fractionation. The pH value is adjusted to 6.5 (0.1) with hydrochloric acid while stirring. The total stirring time is 45 to 60 min at a product temperature of 0 (2) C. Subsequently, the product solution is filtered through a filter bag and the filtrate is transferred for further plasma fractionation.

[0071] The Fraction I supernatant or flow through fraction from previous AT III adsorption was precipitated at an ethanol concentration of 25%. The resulting precipitate, Cohn Fraction II/III, was obtained by centrifugation and contained mainly immunoglobulins. Fraction II/III is frozen and stored at 20 C. or below.

[0072] After dissolution in an aqueous glycine solution the fraction II/III was further precipitated at 10% ethanol concentration in the presence of 0.5% fatty alcohol (also referred to as 10% pre-precipitation because it precedes the main 20% precipitation). The precipitate containing mainly IgM, IgA and lipoproteins was removed by filtration.

[0073] The supernatant was further precipitated at an ethanol concentration of 20%. The formed precipitate which consisted mainly of IgG (Gammaglobulin paste) was obtained by filtration. Crude Gammaglobulin paste was frozen. Afterwards, it was dissolved and subjected to adsorption by using an ion exchange resin and activated carbon to remove residual albumin and fatty alcohol. Impurities bound to the resin and activated carbon were removed by filtration, respectively. The filtrate was subsequently stabilized with sucrose and glycine. The stabilized solution was pasteurized as an effective virus reduction step. After completion of pasteurization, the stabilizers were removed by ultrafiltration (dialysis). The solution was then concentrated to obtain the drug substance, the immunoglobulin ultraconcentrate.

[0074] After the pooling process of the immunoglobulin ultraconcentrate lots, bulk adjustment was performed and the adjusted bulk solution was then filtered through clarification cartridge filters followed by sterilizing filtration. Immediately after completion of the filling process, vials were automatically stoppered and sealed with crimp caps.

Example 1.2

Analytical Methods

1.2.1 Factor XIa-Like Activity (aPTT Approach in FXI Depleted Plasma)

[0075] The activated partial thromboplastin time (aPTT) is a coagulation test that encompasses all steps of the intrinsic pathway of blood coagulation from the activation of the contact phase system to fibrin formation. During the pre-incubation phase of the aPTT assay, Factor XII was activated by negatively charged surfaces (e.g. Pathromtin SL) and activated Factor XI to Factor XIa in the presence of high molecular weight kininogen. The result of this initial step was to produce FXIa. The clot measurement phase of the aPTT assay took place after re-calcification during which FXIa activated FIX, thus continuing the cascade through FXa to thrombin.

[0076] Factor XI-deficient plasma was applied and the presence of activated coagulation factor XI in the sample especially led to a decrease in the coagulation time. The sample was considered as activated with lower clotting times caused by FXIa-like activity in the sample. A longer clotting time indicated a lower pro-coagulant acivity.

[0077] Factor XI-deficient plasma and Pathromtin SL reagent were incubated for 6 minutes at +37 C. Pathromtin SL is a reagent consisting of phospholipid and a surface activator (silicon dioxide particles) used to activate the factors of the intrinsic coagulation system. Subsequently, a sample was added, together with 25 mM CaCl2 solution, which triggers the coagulation process. The time between CaCl2 addition and clot formation was measured. Buffer was used as control sample and as diluent for product sample preparation. The buffer used for FXIa testing experiments consisted of purchased imidazole buffer and 1% human albumin. Factor XIa reference material was used for quantification purposes and the test data were presented as FXIa equivalence.

1.2.2 Kallikrein-Like Activity (Chromogenic Substrate S-2302)

[0078] Kallikrein-like activity was estimated by means of the cleavage of the chromogenic substrate H-D-Pro-Phe-Arg-pNA (chromogenic substrate S-2302, Chromogenix Co.)

[0079] and absorbance measuring of pNA at 405 nm. S-2302 is a chromogenic substrate which mainly reacts with plasma kallikrein, and therefore is used for the determination of kallikrein-like activity.

[0080] After addition of the chromogenic substrate solution, the samples were incubated at +37 C. for 30 minutes. The active kallikrein in the sample is able to cleave the substrate in a concentration dependent manner. This led to a difference in absorbance (optical density) between the pNA formed and the original substrate which was measured photometrically at 405 nm. Moreover, the evaluation was performed on the basis of a standard curve by applying commercial standard reference material of kallikrein.

1.2.3 Proteolytic Activitiy (Chromogenic Substrates)

[0081] The colorimetric determination of proteolytic activity in samples was performed by applying chromogenic substrates. After addition of the chromogenic substrate solution, the samples (1:20 diluted) were incubated at +37 C. for 30 minutes. Proteolytic activity in the sample is able to cleave the substrate in a concentration dependent manner. The method for the determination of activity is based on the difference in absorbance (optical density) between the pNA formed and the original substrate. The rate of pNA formation, i.e. the increase in absorbance per second at 405 nm, is proportional to the enzymatic activity and was determined.

[0082] The following table (Table 1) provides an overview of the substrates applied within this study and the respective specificity.

TABLE-US-00001 TABLE 1 Overview of chromogenic substrates applied Label (Chromogenix Co.) Chromogenic substrate mainly for* S-2302 Kallikrein-like activity S-2366 Activated protein C, FXIa S-2238 Thrombin S-2765 FXa S-2251 Plasmin, streptokinase-activated plasminogen S-2288 Broad spectrum of serine proteases, several proteases with arginine specificity *according to Chromogenix Co., Italy

1.2.4 Factor XI ELISA

[0083] Human FXI antigen in SC Immunoglobulin samples was quantitatively determined by using commercially available paired antibodies (sandwich-style ELISA), e.g. supplied by Coachrom Diagnostika Co. A polyclonal antibody to FXI was coated onto wells of a microtitre plate to capture FXI in the sample or in the standard reference solution. Afterwards, a horseradish peroxidase conjugated antibody to FXI (polyclonal) was added to the wells of the microtitre plate. After removal of unbound antibodies by several washing steps, a peroxidase reactive substrate solution was added which leads to a coloration in a concentration dependent manner.

[0084] The coloration was formed in proportion to the amount of FXI present in the sample. This reaction was terminated by the addition of acid and is measured photometrically at 450 nm by utilizing BEPII or BEPIII systems (Siemens Co.). Moreover, a standard curve was applied by using standard human plasma (Siemens Co.).

[0085] Human FXI was detected as well as human FXIa due to the cross-reactivity of both with the polyclonal paired antibodies applied.

1.2.5 Prekallikrein ELISA

[0086] Human prekallikrein antigen in SC Immunoglobulin samples was quantitatively determined by using commercially available paired antibodies (sandwich-style ELISA) supplied by Affinity Biologicals Co. A polyclonal antibody to prekallikrein is coated onto wells of a microtitre plate to capture prekallikrein in the sample or in the standard reference solution. Afterwards, a horseradish peroxidase conjugated antibody to prekallikrein (polyclonal) was added to the wells of the microtitre plate. After removal of unbound antibodies by several washing steps, a peroxidase reactive substrate solution was added which led to a coloration in a concentration dependent manner.

[0087] The coloration was formed in proportion to the amount of prekallikrein present in the sample. This reaction was terminated by the addition of acid and the color produced quantified by photometric measurement at 450 nm. BEPII or BEPIII systems (Siemens Co.) were used for the determination.

[0088] Human prekallikrein was detected as well as human kallikrein due to the cross-reactivity of both with the polyclonal paired antibodies applied.

1.2.6 Factor XII ELISA

[0089] Human FXII antigen in SC Immunoglobulin samples was quantitatively determined by using commercially available paired antibodies (sandwich-style ELISA), e.g. supplied by Kordia Co. The test approach applied is comparable to the determination of FXI and PK by ELISA technology as mentioned above.

1.3 Test Results

[0090] 29 lots of SC Immunoglobulin drug product manufactured at CSL Behring Marburg (Germany) were analyzed for procoagulant activity. The lots were chosen on the basis of their adsorption scheme. The listed percentage of the adsorption rate per SC

[0091] Immunoglobulin lot is the result of mixing various fraction II/III intermediate lots with different adsorption levels.

[0092] For example, the total amount (100%) of fraction II/III pastes used for SC Immunoglobulin sample no. 13 was PT adsorbed, whereas 59.8% was also subjected to the C1 esterase inhibitor adsorption step and 1.8% to the AT III adsorption.

[0093] Supplementary testing activities and analyses for SC Immunoglobulin with regard to the potential presence of trace amounts of activated clotting factors and proteolytic activity in SC Immunoglobulin drug product were also initiated. For the identification and quantification of residual clotting factors several complementary approaches were performed: [0094] Trace amounts of FXI and FXIa were measured by a modified aPTT test performed with FXI-deficient plasma. [0095] Kallikrein-like activity was measured by applying the chromogenic substrate S-2302 (Chromogenix Co.) due to being generally supposed as major impurities of immunoglobulin preparations. [0096] The potential presence of proteolytic activity was investigated by using chromogenic substrates characterizing a wide range of proteases. [0097] ELISA technology was used for the determination of FXI-, PK- and FXII-antigen, respectively.

[0098] The results of SC Immunoglobulin drug product investigated by FXIa-like activity, Kallikrein-like activity and proteolytic activity are summarized in FIG. 5. The statistical analysis of testing results was performed using the computer program Cornerstone Version 5.0 (Applied Materials Co.). The evaluation comprises 29 lots of SC Immunoglobulin in total.

TABLE-US-00002 TABLE 1 Selected lots of SC Immunoglobulin drug product for further analytical evaluation FXIa-like Kallikrein- activity like Adsorption scheme FXIa activity [%] equiv. (S-2302) Sample no. PT C1 AT III [g/mL] [g/mL] Lots without any adsorption steps: 1 0 0 0 0.06 <0.8 2 0.03 <0.8 3 0.18 <0.8 4 0.08 <0.8 Lots with both 100% PT and AT III adsorption, but differing amounts of C1 adsorbed material: 5 100 41.3 100 <0.01 <0.8 6 43.4 <0.01 <0.8 7 60.5 <0.01 <0.8 Lots with 100% PT and without almost any AT III adsorption, but differing amounts of C1 adsorbed material: 8 100 0 0 14.14 20.0 9 8.8 0 6.56 11.9 10 13.1 0 11.10 15.9 11 30.8 0 12.90 19.5 12 34.8 0 17.96 18.8 13 59.8 1.8 23.98 23.7 Lots with 100% PT and 70 to 80% C1 adsorbed material, but differing amounts of AT III adsorbed material: 14 100 76.4 15.1 14.94 21.6 15 72.3 59.2 3.73 12.3 Additional lots randomly chosen: 16 41.1 41.1 63.2 0.13 <0.8 17 4.6 0 86.4 <0.01 <0.8 18 77.7 25.0 100 <0.01 <0.8 19 86.5 13.5 0.3 2.13 6.0 20 77.4 0 30.3 1.51 3.0 21 100 6.2 55.0 0.33 5.0 22 3.3 11.9 3.42 9.8 23 26.8 29.6 5.66 11.3 24 14.1 58.4 2.67 7.6 25 0.9 1.8 8.11 14.4 26 1.9 4.0 5.40 13.4 27 30.7 40.4 8.19 10.1 28 17.6 22.6 5.57 15.9 29 20.4 10.1 3.71 17.5

[0099] SC Immunoglobulin sample no. 13 was selected as a batch with a high level of procoagulant activity whereas sample no. 7 represents SC Immunoglobulin drug product with a low level of procoagulant activity as determined in the analytical testing. Both lots were compared and the test results are shown in Table 3. Both batches differ in the manufacturing process of fraction II/III (25% precipitate) used as starting intermediate fraction for the further manufacturing process of the respective SC Immunoglobulin drug product. The total amount (100%) of fraction II/III pastes used for both batches was PT adsorbed and about 60% passed the C1 esterase inhibitor adsorption in both cases. However, 100% of fraction II/III used for sample no. 7 was AT III adsorbed whereas only an insignificant amount of fraction II/III passed the AT III adsorption step (1.8%) which was subsequently manufactured into lot no. 13.

[0100] The data demonstrate that drug product with a high AT III adsorption rate in the process (sample no. 7) contains very low levels of activated clotting factors and proteolytic activity in the drug product (see Table 3) in comparison to a product manufactured with very little AT III adsorption (sample no. 13).

[0101] The method used for the determination of proteolytic activity in SC Immunoglobulin drug product was performed by applying chromogenic substrates (S-2765, S-2238, S-2251 and S-2288) and indicated a significantly lower effect in the drug product if an AT IIIAT III adsorption step was subsequently performed. Due to a relatively low reaction by using substrate S-2251, the presence of plasmin seems to be less relevant for SC Immunoglobulin drug product. Moreover, an increased depletion of FXI-Ag (factor of 3.2), PK-Ag (factor of 6.6) and FXII-Ag (factor of 1.2) measured by ELISA was determined and correlated to the processed AT III adsorption on a high level. The above data were supported by analytical results of intermediate fractions obtained before and after the AT III adsorption step (Fraction I supernatant prior to AT III adsorption vs. After the AT III adsorption step) as presented in the following table, which shows a significant decrease in FXIa-like activity, as well as FXI, FXII and PK antigen content.

TABLE-US-00003 TABLE 2 Comparison of intermediate fractions prior and after the AT III adsorption step FXIa- AT III FXI- like FXII- PK- Intermediate content ELISA activity ELISA ELISA fraction [IU/mL] [g/mL] [ng/mL] [mIU/mL] [g/mL] Fraction I 0.7 3.4 295 601 6.3 supernatant prior to AT III adsorption After the AT III 0.1 0.3 <10 16 4.5 adsorption step

TABLE-US-00004 TABLE 3 Comparison of SC Immunoglobulin samples (7 vs. 13) SC Immunoglobulin lot No. 7 13 Adsorption scheme PT [%] 100 100 C1 [%] 60.5 59.8 AT III [%] 100 1.8 Analytical methods FXIa-like activity <0.01 23.98 [FXIa equiv. g/mL] Kallikrein-like activity (S-2302) [g/mL] <0.8 23.7 S-2765 [mOD/min] 0.3 26.8 S-2238 [mOD/min] 0.7 38.2 S-2251 [mOD/min] 0.1 3.9 S-2366 [mOD/min] 0.7 51.7 S-2288 [mOD/min] 1.2 53.2 Factor XI-Ag (ELISA) [g/mL] 5.9 19.1 Prekallikrein-Ag (ELISA) [g/mL] 8.2 54.5 Factor XII-Ag (ELISA) [mIU/mL] 29.3 35.7

[0102] The test results revealed that a high level of AT III adsorption leads to a significant decrease in the procoagulant activity of SC Immunoglobulin drug product. To further detail the effect of the AT III adsorption the content of specific clotting factors in drug product was measured by ELISA. The strong correlation between both prekallikrein-Ag and kallikrein-like activity and FXI-Ag and FXIa-like activity is shown in FIG. 6.

[0103] Increasing AT III adsorption led to a depletion of FXI, PK and FXII measured as antigen by ELISA as well as a reduction of FXIa- and kallikrein-like activity as shown in Table 2, Table 3 and FIG. 5.

[0104] The analysis revealed that SC Immunoglobulin lots manufactured with high level of AT III adsorption exhibit low procoagulant activity. These lots reveal lower concentrations of FXI-like activity (in FXI-depleted plasma) as well as a lower kallikrein-like activity values (PKA blank value). The determination of proteolytic activity in SC Immunoglobulin drug product via applying various chromogenic substrates (S-2765, S-2238, S-2251 and S-2288) indicated a significantly lower proteolytic activity in the drug product when AT III adsorption level is high.

[0105] Increasing AT III adsorption led to a depletion of FXI, PK and FXII measured as antigen by ELISA as well as a reduction of FXIa- and kallikrein-like activity as shown in FIG. 5.

[0106] A strong correlation between both prekallikrein-Ag and kallikrein-Ag and FXI-Ag and FXIa-like activity was shown. It was shown that procoagulant activity detected in SC Immunoglobulin is mainly caused by the content of kallikrein and FXIa. The data generated within this study provide strong evidence that a high level of AT III adsorption leads to a significant decrease in the procoagulant activity of SC Immunoglobulin drug product.

Example 2

[0107] Analysis revealed that increasing AT III adsorption during processing of subcutaneous immunoglobulins leads to a depletion of FXI, prekallikrein and FXII antigens as well as a reduction of FXIa and kallikrein-like activity. A strong correlation between both kallikrein-antigen and FXI-antigen and FXIa-like activity was shown. FXIa and Kallikrein were identified as relevant impurities. Based on the finding of procoagulant activity the production process was adapted to include maximum adsorptionthat is essentially 100% of the plasma fraction is exposed to the heparin affinity resin.

[0108] The removal of AT III from product intermediates for reduction of activated factors activation appears initially paradoxical, because AT III is known to inhibit activated coagulation factors. In fact, ATIII inhibits to a certain extent activated coagulation factors. Further, it is known that heparin accelerates the activity of ATIII by a factor of 1000 (Rosenberg R D: Role of heparin and heparin-like molecules in thrombosis and atherosclerosis. Fed Proc. 1985; 44(2), 404-9). Therefore, the following analysis was performed. In the first experiment, a drug product known to contain FXIa and kallikrein-like activities was measured by NaPPT, FXIa-like activity and by reactivity towards chromogenic substrate (S2302) (kallikrein-like activity). Then the drug product was treated with 2 U/mL ATIII or with 2 U/mL ATIII plus 10 U/mL heparin. Clotting parameters were determined again (Table 4).

TABLE-US-00005 TABLE 4 Depletion of activated coagulation factors by AT III and ATIII/heparin. FXIa-like Chromogenic activity substrate NaPTT FXIa equiv. (S-2302) Sample description (sec) (g/mL) (mOD/min) Pharmaceutical 41 6.11 604 preparation Pharmaceutical 120 0.06 29 preparation + AT III (2 U/mL) Pharmaceutical No clot formed <0.01 20 preparation + AT III (2 U/mL) + heparin (10 U/mL)

[0109] While the untreated drug product revealed a shortened NaPTT, 6.11 g/mL FXIa equivalents and elevated reactivity towards S2302 (604 mOD/min), the AT III treated sample displayed a 3-fold prolonged NaPTT, a hundred fold decreased FXIa concentration and a 30-fold lesser reactivity towards S2302. When heparin was added, this inhibitory effect was even stronger. There was no clot formed during NaPTT, the FXIa content was below detection limit and reactivity towards S2302 was even further reduced.

[0110] Those observations are comparable to the situation In vivo. The physiological AT III concentration counterbalances the activated coagulation factors up to a certain limit and reaction time, still allowing thrombus formation. Heparin treatment shifts the balance towards anticoagulation and the likelihood of thrombus formation is markedly reduced. Thus the adsorption of AT III to heparin Fractogel is expected to increase the AT III inhibitory capacity and assures that activated factors are inactivated and stable inactive complexes formed. These can then be removed from the plasma fraction by simply removing the heparin affinity resin. This ensures the pharmaceutical preparation will contain essentially no activated serine proteases and will therefore exhibit a reduced adverse event profile.

[0111] If however the activated coagulation factors are not removed from the plasma fraction then further processing steps in preparing subcutaneous immunoglobulin preparations will not necessarily lead to the removal of the activated coagulation factors such as FXIa. As such it is a requirement that the fractionation process required to prepare the pharmaceutical preparation includes at least one of the plasma fractions to be contacted by heparin or a heparin like substance (eg. heparin affinity resin). Furthermore the use of a heparin affinity resin or similar is advantageous over soluble forms of heparin or heparin like substances as it enables the proteases to be physically be removed from the fraction containing the drug substance. In contrast the addition of ATIII and soluble forms of heparin or heparin like substances will likely lead to complex formation however not necessarily removal as it is possible that given time or subsequent processing steps that the protease/ATIII/heparin complexes may dissociate resulting in the reintroduction of activated serine proteases such as FXIa. Our data indicate that only a removal of the complexes would be effective, instead of an inactivation, and if this removal step is not completed then there is the possibility for proteolytic activity and activated coagulation factor XI to be present in the the final product.

[0112] Furthermore it is of note that where AT III is not removed from the plasma fraction by a heparin affinity batch adsorption step that subsequent fractionation steps do nevertheless remove it such that the final subcutaneous immunoglobulin pharmaceutical product is essentially free of AT III. This provides the possibility of a pharmaceutical preparation containing activated proteases but no ATIII and hence accentuates the possibility of adverse events in such products.

Example 3

[0113] This example provides evidence that the use of heparin affinity resins can be added to other intermediate plasma fractions which contain ATIII in order to remove contaminating activated serine proteases such as FXIa.

[0114] The removal of activated coagulation factors was investigated for the intermediate fraction, cryo-poor plasma at laboratory scale. The heparin affinity resin (0.5 g) was incubated with the cryo-poor plasma (19.5 mL) at room temperature for 30 minutes with stirring. Afterwards the resin was separated from the plasma fraction by centrifugation (Heraeus Co., Multifuge 3SR+ at 1700 rpm for 10 minutes at room temperature). The levels of ATIII, total Factor XI antigen (see method described at 1.2.4 above) and Factor XIa-like activity (see method described at 1.2.1 above) were measured in the cryo-poor plasma before and after exposure to the heparin affinity resin (Table 5). The ATIII activity was approximately 1.1 IU/mL in cryo-poor plasma and this was reduced to 0.6 IU/mL after exposure to the heparin affinity resin. The Factor XI (FXI) levels were reduced from 5.5 g/mL to 0.3 g/mL whilst activated Factor XI like activity equivalents were reduced from 2.1 g/mL to below the assay detection limit of <0.01 g/mL. These results suggest that the heparin affinity resin adsorption step is effective at treating plasma fractions comprising ATIII such as cryo-poor plasma.

TABLE-US-00006 TABLE 5 Levels of ATIII, Factor XI antigen and FXIa-like activity in cryo- poor plasma before and after exposure to heparin affinity resin. FXIa-like ATIII Factor XI- activity content Ag (ELISA) FXIa equiv. Sample description [IU/mL] [g/mL] [g/mL] Cryo-depleted plasma prior 1.1 5.5 2.1 to ATIII adsorption Cryo-depleted plasma after 0.6 0.3 <0.01 ATIII adsorption

[0115] The study revealed depletion ratios of 10.2 g FXI antigen per adsorbed IU of ATIII. The total depletion was about 203 g FXI antigen per gram of resin.

[0116] Additionally the study suggests that the heparin affinity resin can remove both FXIa and FXI. It is known that FXI molecule contains heparin binding sites and presumably this contributes to the heparin affinity resins ability to remove both the activated and non-activated FXI.