(12) United States Patent
`Relton
`
`US006252055B1
`US 6,252,055 B1
`Jun. 26, 2001
`
`(10) Patent N0.:
`(45) Date of Patent:
`
`(54) CONCENTRATED ANTIBODY
`PREPARATION
`
`(75) Inventor: Julian Marcus Relton, Sevenoaks
`(GB)
`(73) Assignee: Glaxo Wellcome Inc., Triangle Park,
`NC (US)
`
`5,766,947 * 6/1998 Rittershaus et al. .
`
`FOREIGN PATENT DOCUMENTS
`
`684 164
`42 11 169
`0 064 210
`0 661 060
`94 15640
`
`7/1994 (CH).
`6/1993 (DE) .
`11/1982 (EP) .
`7/1995 (EP) .
`7/1994 (WO).
`
`( * ) Notice:
`
`(21) Appl. No.:
`(22) PCT Filed:
`
`Subject to any disclaimer, the term of this
`patent is extended or adjusted under 35
`U.S.C. 154(b) by 0 days.
`09/180,485
`May 22, 1997
`
`(86) PCT No.:
`§ 371 Date:
`
`PCT/EP97/02595
`Nov. 12, 1998
`
`§ 102(e) Date: Nov. 12, 1998
`
`(87) PCT Pub. No.: WO97/45140
`
`PCT Pub. Date: Dec. 4, 1997
`Foreign Application Priority Data
`
`(30)
`
`May 24, 1996
`
`(GB) ................................................ .. 9610992
`
`(51) Int. Cl.7 ...................... .. C07K 16/00; A61K 39/395;
`C12N 5/20
`(52) US. Cl. ................. .. 530/414; 424/130.1; 424/177.1;
`435/326; 530/3881; 530/3871; 530/3905
`(58) Field of Search ............................... .. 435/3432, 326;
`530/3882, 388.75, 414, 390.5, 388.1, 387.1;
`424/1331, 174.1, 177.1, 130.1
`
`(56)
`
`References Cited
`
`U.S. PATENT DOCUMENTS
`
`5,668,003 * 9/1997 Kim et al..
`
`OTHER PUBLICATIONS
`
`VeleZ et al. Biotech. & Bioeng.; vol. 33; p. 938—940, 1989*
`
`Database WPI Week 8949 DerWent Publications Ltd., Lon
`don, GB; AN 89—359879 XP002023849 & JP 01 268 646 A
`(Meiji Milk Products KK), Oct. 26, 1989 cited in the
`application see abstract.
`
`Biotechnology and Bioengineering, vol. 33, No. 7, Feb. 20,
`1989, NeW York, NY, USA, pp. 938—940, XP000005263 D.
`VeleZ et al.: Use of tangential ?oW ?ltration in perfusion
`propagation of hybridoma cells for production of mono
`clonal antibodies.: see the Whole document.
`
`* cited by examiner
`
`Primary Examiner—David Saunders
`Assistant Examiner—Amy DeCloux
`(74) Attorney, Agent, or Firm—Nixon & Vanderhye PC.
`
`(57)
`
`ABSTRACT
`
`Concentrated monoclonal antibody preparations for admin
`istration to humans are described in Which the antibody is
`present at a concentration of greater than 100 mg/ml and as
`high as 350 mg/ml.
`
`5 Claims, No Drawings
`
`Ex. 1018 - Page 1 of 7
`
`AMGEN INC.
`Exhibit 1018
`
`

`
`US 6,252,055 B1
`
`1
`CONCENTRATED ANTIBODY
`PREPARATION
`
`This application is a national stage ?ling under 35 USC
`371 from PCT/EP97/02595, ?led May 22, 1997.
`The present invention relates to a concentrated antibody
`preparation, pharmaceutical formulations containing such a
`preparation, its use in human therapy and processes for its
`preparation.
`Most commercially available immunoglobulins pro
`duced at high concentration are derived from human serum
`and produced by the blood products industry. The ?rst
`puri?ed human immunoglobulin G (IgG) preparation used
`clinically Was immune serum globulin Which Was prepared
`in the 1940’s (Cohn, E. J. et al ‘Preparation and properties
`of serum and plasma proteins’. J. Am. Chem. Soc. pg68,
`459—475 (1946) and Oncley, J. L et al ‘The separation of
`antibodies, isoagglutinins, prothrombin, plasminogen and
`[3-lipoproteins into sub-fractions of human plasma.’ J. Am.
`Chem. Soc. 71, 541—550 (1949)).
`The next generation of puri?ed IgG’s Were developed in
`the 1960’s, and focused on preparations suitable for intra
`venous administration (Barandun, S.et al ‘Intravenous
`administration of human y-globulin.’ Vox. Sang. 7, 157—174
`(1962)).The ?rst of these—IgG intravenous preparation
`(Gamimune®, Cutter Biological), Was formulated as a 5%
`(50 mg/ml) IgG solution in 0.2 M glycine, 10% maltose, pH
`6.8. This solution Was stable for at least 2.5 years at 5° C.
`Key criteria for the acceptance of intravenous IgG (IVIG)
`products Were that the IgG had undergone little fragmenta
`tion and that no high molecular Weight aggregates Were
`present.
`Today, human therapeutic immunoglobulin products are
`available for either intramuscular (IMIG) or intravenous
`(IVIG) administration. IMIG are used principally for hepa
`titis A prophylaxis and sometimes for the treatment of
`agammaglobulinaemic patients. IVIG are used in the treat
`ment of primary immunode?ciencies and idiopathic throm
`bocytopenic purpura, as Well as for secondary immune
`de?ciencies, various infections, haematological and other
`autoimmune diseases. In general IMIG products are mar
`keted as 16% (W/v) (160 mg/ml) solutions and IVIG prod
`ucts as 5% (W/v) solutions (50 mg/ml).
`Manufacturers experience With IVIG has shoWn that
`these preparations are unstable in relatively dilute solutions
`(<10% (W/v)), and the instability is manifested by the
`formation of insoluble particles by a process knoWn as
`‘shedding’ When the material is stored at room temperature
`(Fernandes, P. M. and Lundband, J. L. ‘Preparation of a
`stable intravenous gamma-globulin: process design and
`scale up.’ Vox. Sang. 39, 101—112 (1980)). Commercially
`available 16.5% y-globulin is usually stabilised in a buffered
`glycine-saline solution. The use of maltose at 5—10% as a
`stabiliser has been shoWn to be effective in protecting 5%
`IVIG from particulate formation (Fernandes et al supra).
`In addition to shedding, concentrated (16.5%) solutions
`of IVIG have a tendency to aggregate during long term
`storage. As much as 10—30% (W/W) of the IVIG solution
`could be comprised of aggregates (Gronski, P.et al,‘On the
`nature of IgG dimers. I. Dimers in human polyclonal IgG
`preparations: kinetic studies.’ Behring Inst. Mitt. 82,
`127—143 (1988)).
`
`10
`
`15
`
`20
`
`25
`
`30
`
`35
`
`40
`
`45
`
`55
`
`60
`
`65
`
`2
`The majority of these aggregates are dimers produced by
`complexes of idiotypic and anti-idiotypic antibodies. Since
`monoclonal antibodies prepared from tissue culture super
`natants do not contain anti-idiotype antibodies, these sort of
`dimers are absent. HoWever, dimer formation in these prepa
`rations can be caused by complexation betWeen partially
`denatured monomeric antibody molecules. Mechanical
`stress such as that encountered during tangential ?oW ultra
`?ltration used for concentrating antibody preparations can
`also lead to an increase in aggregation (Wang, Y.-C. J. and
`Hanson, M. A. ‘Parenteral formulations of proteins and
`peptides: stability and stabilisers.’ J. Parenteral Sci. Technol.
`42, Suppl. S3—S26 (1988)).
`Concentrated (>100 mg/ml) preparations of immunoglo
`bulins are therefore available but to date these are polyclonal
`antibody preparations derived from the blood processing
`industry, and are stabilised by the addition of various excipi
`ents such as glycine and maltose.
`It is therefore surprising that monoclonal antibody prepa
`rations have been obtained at a concentration >100 mg/ml in
`the absence of excipients and Without a concomitant
`increase in aggregates.
`The DerWent Abstract of JP01268646A (AN89-359879)
`reports that the application describes an injection prepara
`tion of an IgG3 monoclonal antibody having a concentration
`of 0.1 pg to 100 mg/ml. Subject matter disclosed in these
`publications is outside the scope of the instant invention.
`The present invention therefore provides a monoclonal
`antibody preparation for administration to a human charac
`terised in that the antibody in said preparation is at a
`concentration of 100 mg/ml or greater, preferably greater
`than 100 mg/ml. Above a concentration of 350 mg/ml the
`preparation can be very viscous and recovery rates become
`unacceptably loW. The ideal concentration is betWeen 100
`and 300 mg/ml.
`Preparations according to the invention are substantially
`free from aggregate. Acceptable levels of aggregated con
`taminants Would be less that 5% ideally less than 2%. Levels
`as loW as 0.2% are achievable, although approximately 1%
`is more usual. The preparation is also preferably free from
`excipients traditionally used to stabilise polyclonal
`formulations, for example glycine and/or maltose.
`The present invention therefore provides a monoclonal
`antibody preparation for administration to a human charac
`terised in that the antibody in said preparation is at a
`concentration of 100 mg/ml or greater, preferably greater
`than 100 mg/ml and the preparation is substantially free
`from aggregate.
`Recombinant antibodies by their very nature are pro
`duced in a synthetic and unnatural cell culture environment.
`Expression systems Which are used to generate suf?cient
`quantities of the protein for commercialisation are routinely
`based on myeloma or chinese hamster ovary (CHO) host
`cells.
`In order to culture such cells, complex synthetic media
`Which are devoid of contaminating animal protein have been
`devised resulting in glycosylation patterns of the protein
`Which Would not be expected to arise in nature. It is therefore
`all the more surprising that a complex glycoprotein pro
`duced under such synthetic conditions can be prepared at
`concentrations several times greater than Would occur in
`normal human serum With all its buffering capabilities.
`
`Ex. 1018 - Page 2 of 7
`
`

`
`US 6,252,055 B1
`
`3
`The present invention therefore provides a monoclonal
`antibody preparation for administration to a human charac
`terised in that the antibody in said preparation is a recom
`binant antibody and is at a concentration of 100 mg/ml or
`greater, preferably greater than 100 mg/ml. The preparation
`is preferably substantially free from aggregate.
`During the production of puri?ed antibodies Whether for
`therapeutic or diagnostic use, it is important that the anti
`body is sufficiently stable on storage and various chemical
`entities may have an adverse effect on the stability of the
`antibody. For example, trace amounts of copper (Cu++) are
`noW knoWn to have a destabilising effect on immunoglobu
`lin molecules on storage (WO93/08837), and that this effect
`can be eliminated by formulating the immunoglobulin mol
`ecule With a suitable chelator of copper ions, for example
`EDTA or citrate ion.
`The present invention is applicable to a preparation of
`immunoglobulins of all classes, i.e. IgM, IgG, IgA, IgE and
`IgD, and it also extends to a preparation of Fab fragments
`and bispeci?c antibodies. The invention is preferably
`applied to a preparation of immunoglobulins of the class
`IgG, Which includes the sub-classes lgGl, IgG2, IgG3 and
`IgG4. The invention is more preferably applied to a prepa
`ration of immunoglobulins of the class IgG4 and IgG1, most
`preferably lgGl.
`The invention ?nds particular application in the prepa
`ration of recombinant antibodies, most particularly chimae
`ric antibodies or humanised (CDR-grafted) antibodies. Par
`ticular examples of these include chimaeric or humanised
`antibodies against CD2, CD3, CD4, CD5, CD7, CD8,
`CD11a, CD11b, CD18, CD19, CD23, CD25, CD33, CD54,
`and CDW52 antigen. Further examples include chimaeric or
`humanised antibodies against various tumour cell markers
`eg 40 kd (J.Cell Biol. 125 (2) 437—446 (1994)) or the
`antigens of infectious agents such as hepatitis B or human
`cytomegalovirus. Particularly preferred examples include
`chimaeric or humanised antibodies against CDW52, CD4
`and CD23 antigen.
`Immunoglobulins intended for therapeutic use Will gen
`erally be administered to the patient in the form of a
`pharmaceutical formulation. Such formulations preferably
`include, in addition to the immunoglobulin, a physiologi
`cally acceptable carrier or diluent, possibly in admixture
`With one or more other agents such as other immunoglobu
`lins or drugs, such as an antibiotic. Suitable carriers include,
`but are not limited to, physiologic saline, phosphate buffered
`saline, glucose and buffered saline, citrate buffered saline,
`citric acid/sodium citrate buffer, maleate buffer, for example
`malic acid/sodium hydroxide buffer, succinate buffer, for
`example succinic acid/sodium hydroxide buffer, acetate
`buffer, for example sodium acetate/acetic acid buffer or
`phosphate buffer, for example potassium dihydrogen
`orthophosphate/disodium hydrogen orthophosphate buffer.
`Optionally the formulation contains Polysorbate for stabili
`sation of the antibody. Alternatively the immunoglobulin
`may be lyophilised (freeZe dried) and reconstituted for use
`When needed by the addition of Water and/or an aqueous
`buffered solution as described above.
`The preferred pH of the pharmaceutical formulations
`according to the invention Will depend upon the particular
`route of administration. HoWever, in order to maximise the
`
`10
`
`15
`
`25
`
`35
`
`45
`
`55
`
`65
`
`4
`solubility of the antibody in the concentrated solution, the
`pH of the solution should be different from the pH of the
`isoelectric point of the antibody.
`Thus, according to a further aspect the invention pro
`vides a monoclonal antibody preparation for administration
`to a human characterised in that the antibody in said prepa
`ration is at a concentration of 100 mg/ml or greater and the
`pH of the preparation is different from the pH of the
`isoelectric point of the antibody.
`Routes of administration are routinely parenteral, includ
`ing intravenous, intramuscular, and intraperitoneal injection
`or delivery. HoWever, the preparation is especially useful in
`the generation of sub-cutaneous formulations Which must be
`loW in volume for example approximately 1 ml in volume
`per dose. To ensure that therapeutic dosage can be achieved
`in such a formulation, a concentrated preparation Will invari
`ably be necessary. Preferred concentrations for sub
`cutaneous preparations are for example in the range of 100
`mg/ml to 200 mg/ml, for example 150 mg/ml to 200 mg/ml.
`Asub-cutaneous preparation has the advantage that it can be
`self-administered thus avoiding the need for hospitalisation
`for intravenous administration.
`Preferably, sub-cutaneous formulations according to the
`invention are isotonic and Will be buffered to a particular pH.
`The preferred pH range for a sub-cutaneous formulation Will
`in general range from pH 4 to pH 9. The preferred pH and
`hence buffer Will depend on the isoelectric point of the
`antibody concerned as discussed above. Thus, in the case of
`sub-cutaneous preparations containing anti-CD4 antibodies
`the pH Will preferably be in the range of pH 4 to pH 5.5, for
`example pH 5.0 to pH 5.5 e.g. pH 5.5, and in the case of
`anti-CD23 antibodies in the range of pH 4 to pH 6.5. Thus,
`preferred buffers for use in sub-cutaneous formulations
`containing anti-CD4 antibodies are maleate, succinate,
`acetate or, more preferably phosphate buffers. Buffers are
`preferably used at a concentration of 50 mM to 100 mM.
`Sub-cutaneous formulations according to the invention
`may also optionally contain sodium chloride to adjust the
`tonicity of the solution.
`Thus, according to a further aspect of the invention
`provides a monoclonal antibody preparation for sub
`cutaneous administration to a human characterised in that
`the antibody in said preparation is at a concentration of 100
`mg/ml or greater and the pH of the preparation is different
`from the pH of the isoelectric point of the antibody.
`In a further aspect of the invention the monoclonal
`preparation is envisaged for use in human therapy. Various
`human disorders can be treated such as cancer or infectious
`diseases for example those mentioned above, and immune
`disfunction such as T-cell-mediated disorders including
`severe vasculitis, rheumatoid arthritis, systemic lupis, also
`autoimmune disorders such as multiple sclerosis, graft vs
`host disease, psoriarsis, juvenile onset diabetes, Sjogrens’
`disease, thyroid disease, myasthenia gravis, transplant
`rejection, in?ammatory boWel disease and asthma.
`The invention therefore provides the use of a concen
`trated monoclonal antibody preparation as described herein
`in the manufacture of medicament for the treatment of any
`of the aforementioned disorders. Also provided is a method
`of treating a human being having any such disorder com
`prising administering to said individual a therapeutically
`effective amount of a preparation according to the invention.
`
`Ex. 1018 - Page 3 of 7
`
`

`
`US 6,252,055 B1
`
`5
`The dosages of such antibody preparations Will vary With
`the conditions being treated and the recipient of the
`treatment, but Will be in the range 50 to about 2000 mg for
`an adult patient preferably 100—1000 mg administered daily
`or Weekly for a period betWeen 1 and 30 days and repeated
`as necessary. The doses may be administered as single or
`multiple doses.
`An antibody preparation may be concentrated by various
`means such as cross ?oW (tangential) or stirred
`ultra?ltration, the preferred route is by tangential ?oW
`ultra?ltration. LoW recovery rates and precipitate formation
`can be a problem When concentrating antibody. The present
`invention solves this particular problem by a method of
`concentration Which involves reducing shear stresses of
`cross ?oW ultra?ltration at high circulation rates (500
`ml/min). Reducing the recirculation for example to 250
`ml/min leads to successful concentration of antibody to
`>150 mg/ml and to the high recovery of material.
`The invention therefore provides a process for the prepa
`ration of a concentrated antibody preparation as described
`herein. The recovery of the antibody in the concentrated
`preparation is preferably greater than 70% but is routinely
`greater than 90%.
`Concentrated antibody preparations prepared according
`to the above process may contain additional ingredients such
`as buffers, salts, Polysorbate and/or EDTA. These additional
`agents may not be required in the ?nal pharmaceutical
`formulation in Which case they can be removed or
`exchanged using dia?ltration according to conventional
`methods knoWn in the art. For example, concentrated anti
`body preparations containing citrate buffer and EDTAcan be
`converted into concentrated antibody preparations contain
`ing phosphate or maleate buffer using this method.
`The invention also provides a novel concentrated anti
`body preparation obtainable by such methods.
`The folloWing are non-limiting examples of the inven
`tion.
`
`6
`The system Was ?ushed out With PBS, and the concentrate
`stored overnight at 4° C. After day 2, 0.01% (W/v) Thiomer
`sal Was added to the concentrate before overnight storage to
`prevent microbial contamination. At the end of the
`concentration, the system Was ?ushed out With 500 ml of
`PBS, then a further 500 ml ?ush of PBS Was recirculated
`around the retentate side of the membranes for 30 min. The
`concentration of antibody in these ?ushes Was determined
`by measuring the absorbance at 280 nm.
`The total time taken to concentrate Campath-1H from
`16.4 mg/ml to 257 mg/ml using only 2 plates in the Minitan
`Was 17.25 hours. Table 1(a) shoWs the change in concen
`tration of Campath-1H over this time. The concentration
`
`10
`
`15
`
`TABLE 1(a)
`
`Time (h)
`
`0
`
`5
`
`6.5 9.5 11.5
`
`12.5
`
`14.5
`
`16
`
`17
`
`17.25
`
`20
`
`Conc
`(mg/ml)
`
`16 34 41 79
`
`106
`
`136
`
`190
`
`230
`
`301
`
`257
`
`25
`
`30
`
`35
`
`increased in an exponential manner to a peak of 300 mg/ml
`after 17 hours. The ?nal concentration Was slightly loWer
`than this peak value; the discrepancy probably due to the
`dif?culty in obtaining a representative sample from a very
`viscous liquid. Table 1(b) shoWs that the concentration of
`Campath-1H Was accompanied by a reciprocal decrease in
`the How rate of the permeate. This
`
`TABLE 1(b)
`
`Conc
`(mg/ml)
`
`Flow
`(ml/min)
`Viscosity
`(cPs)
`
`16 34
`
`41
`
`79
`
`106
`
`136
`
`189
`
`301
`
`257
`
`4
`
`3
`
`2.5
`
`1.5 1.25
`
`0.9
`
`0.5
`
`1
`
`1
`
`1
`
`1
`
`1
`
`1
`
`0.96
`
`4.85
`
`8.1
`
`EXAMPLE 1
`Concentration of Campath 1 H
`The Minitan ultra?ltration rig (Minitan XX42 ASY MT
`Ultra?ltration System, Millipore) Was assembled With 2
`polysulphone 30K NMWCO ?lter plates (Minitan PTTK
`30K NMWCO,Millipore), and the tubing and plates Were
`sanitised for 30 min With 0.1 M NaOH according to manu
`facturers instructions (Minitan Ultra?ltration System:
`Assembly, Operation, Maintenance Instructions, Millipore
`Corporation, P15076). The sanitant Was removed by ?ush
`ing With 1—2 liters of phosphate buffered saline (PBS), pH
`7.2.
`Campath-1H (a humanised antibody against the CDW52
`antigen: Reichmann et al Nature, 332,323—327 (1988))(2200
`ml at 16.4 mg/ml in 50 mM sodium citrate, pH 6.0), Was
`circulated through the retentate side of the membranes at a
`?ux rate of 600 ml/min at a back pressure of 2—2.5 bar. The
`back pressure Was maintained at this value throughout the
`remainder of the experiment, and the permeate ?ux rate
`measured at various time intervals. Samples of antibody
`Were removed from the retentate vessel at various time
`points and assayed for antibody concentration, turbidity, %
`aggregate and viscocity.
`Because the ?ltration rate Was so sloW in this experiment
`it Was neccessary to carry out the concentration over 3 days.
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`Table also shoWs there Was a dramatic increase in the
`viscocity of the remaining concentrate above a concentration
`of 189 mg/ml.
`Table 1(c) shoWs that the recovery Was high up to a
`concentration of 190 mg/ml, but started to decline markedly
`above this concentration as the viscocity increase led to
`material sticking to glassWare and tubing and being lost
`during ?ushing prior to overnight storage. The ?nal recovery
`of 257 mg/ml material
`
`TABLE 1(c)
`
`Conc
`(mg/ml)
`
`Rec (%)
`
`16
`
`100
`
`41
`
`97
`
`106
`
`97
`
`190
`
`85
`
`257
`
`63
`
`(excuding material removed during sampling and lost in
`Washes) Was 63.4%. A further 14.6% Was recovered in the
`?rst PBS Wash of the system and 0.5% in the second,
`recirculated PBS Wash. In total, therefore, 78.5% of the
`initial material Was recovered at the end of the experiment
`(excuding material removed during sampling and lost in
`Washes), leaving a loss of 21.5% mainly due to viscous
`material sticking to glassWare and plastics.
`Turbidity of the Campath-1H solution during concentra
`tion Was calculated. The absorbance of suitably diluted 1.0
`ml aliquots of antibody samples at 650 nm Was used as a
`
`Ex. 1018 - Page 4 of 7
`
`

`
`US 6,252,055 B1
`
`7
`measure of turbidity. Table 1 (d) shows that there Was no
`increase.
`
`TABLE 1(d)
`
`Conc
`(mg/ml)
`
`16
`
`41
`
`79
`
`106
`
`136
`
`190
`
`301
`
`257
`
`Rec (%)
`Aggregate (%)
`
`1.11
`1.01
`1
`0.91
`1.1
`1.16
`0.96
`0.002 0.015 0.023 0.032 0.042 0.032 0.035
`
`1.01
`
`Samples for aggregate determination Were diluted to a
`protein concentration of 1 mg/ml using PBS and 50 pl or 100
`pl aliquots injected onto a TSK-GEL G3000SWXL siZe
`exclusion HPLC column. The column Was developed With
`0.05% NaN3 and 0.1 M NaZSO4 in 0.1 M phosphate buffer,
`pH 6.7 at a How rate of 1.0 ml/min. The amount of aggregate
`Was determined by integrating the peaks of absorbance at
`280 nm and Were found to remain around 1% throughout.
`
`15
`
`20
`
`25
`
`30
`
`8
`
`TABLE 2(a)
`
`Time (h)
`
`0
`
`2
`
`3.5
`
`5
`
`6
`
`Cone
`mg/ml
`Viscosity
`cPs
`
`13.9
`
`47.2
`
`83
`
`112.8
`
`252
`
`1
`
`1
`
`1
`
`1
`
`9.7
`
`At concentrations above 83 mg/ml there Was a noticable
`opalescence in the concentrated material, and this caused a
`precipitate to form as the concentration increased above this
`value. This led to the decrease in ?ux rates of the permeate
`shoWn in Table 2(b) and also to the rise in turbidity shoWn
`in Table 2(c). The level of aggregate remained very loW at
`all concentrations, being less than 0.2% throughout (see
`Table 2(c)). Table 2(b) shoWs that recoveries Were high until
`the viscocity increased and the precipitate occured, Where
`they fell dramatically to a ?nal recovery in the retentate after
`removal from the rig of 50%.
`
`TABLE 2(b)
`
`13.9
`
`47.2
`
`83
`
`112.8
`
`252
`
`100.0
`13.5
`
`100.0
`3.2
`
`100.0
`2.0
`
`113.0
`2.5
`
`51.4
`
`TABLE 2(C)
`
`13.9
`
`47.2
`
`83
`
`112.8
`
`252
`
`0.011
`
`0.012
`
`0.030
`
`0.185
`
`0.140
`
`0.150
`
`0.150
`
`0.160
`
`0.170
`
`Conc
`mg/ml
`
`Rec (%)
`Flow
`(ml/min)
`
`Conc
`mg/ml
`
`Turb
`A650 nm
`Agg (%)
`
`This poor recovery Was due to the high viscocity of the
`concentrated Anti-CD4 antibody making it stick to the
`tubing and membranes of the ultra?ltration system. All the
`Anti-CD4 antibody lost in this Way could be subsequently
`recovered by ?ushing out the system With buffer. Table 2(d)
`shoWs the recovery of Anti-CD4 antibody in successive 50
`ml Wash fractions during the ?ushing out of the Minitan rig
`at the end of the experiment. The ?rst fraction contains 11.7
`g of Anti-CD4 antibody at a concentration of 235 mg/ml, so
`this could be pooled With the 12.6 g of concentrate initially
`recovered from the rig at 252 mg/ml Without signi?cantly
`diluting the overall concentration. The remaining Wash
`fractions contained a total of 5.1 g of Anti-CD4 antibody, but
`this Was at a concentration of less than 57 mg/ml, so could
`not be pooled With the concentrated material. The overall
`recovery in the concentrate and the ?rst Wash fraction Was
`90%. It Was noticed that after storage of the ?nal concen
`trated Anti-CD4 antibody overnight at 4° C. led to some of
`the precipitate redissolving.
`
`EXAMPLE 2
`Concentration of Anti-CD4 Antibody—Method A
`The Minitan ultra?ltration rig Was assembled and sani
`tised as in Example 1 except that 8 polysulphone 30K
`NMWCO ?lter plates Were used instead of 2, and the Whole
`rig Was placed in a sterile hood. Anti-CD4 antibody (2142
`ml at 13.9 mg/ml in 50 mM sodium citrate, pH 6.0) Was
`circulated through the retentate side of the membranes at a
`?ux rate of 190 ml/min at a back pressure of 2—2.5 bar. The
`back pressure Was maintained at this value throughout the
`remainder of the experiment, and the permeate ?ux rate
`measured at various time intervals. Samples of antibody
`Were removed from the retentate vessel at various time
`points and assayed for antibody concentration, CD4 binding,
`turbidity, % aggregate and viscocity.
`At the end of the experiment the retentate Was pumped out
`of the Minitan rig and the retentate side of the membranes
`Was ?ushed With 500 ml of 50 mM sodium citrate, 0.05 mM
`EDTA, pH 6.0 and 50 ml fractions of the ?ush Were
`collected. Finally, the system Was ?ushed by recirculating
`500 ml of 50 mM sodium citrate, 0.05 mM EDTA, pH 6.0
`around the retentate side of the membrane for 30 min. The
`antibody concentration of the ?ush fractions Was determined
`by measuring their absorbance at 280 nm.
`The results are shoWn in Tables 2(a)—(d). The increase in
`the number of plates used for the concentration led to a
`decrease in the time taken to achieve a concentration of 250
`pg/ml—250 mg/ml to 6 h compared to the 17.25 h for the
`Campath-1H concentration (see Table 2(a)). Table 2(a) also
`shoWs that the viscocity of the Anti-CD4 antibody did not
`measurably increase until a concentration of 113 mg/ml Was
`achieved. Above this concentration the viscocity increased
`dramatically.
`
`35
`
`40
`
`45
`
`50
`
`55
`
`TABLE 2(d)
`
`50
`ml
`mg 11727
`
`100
`2828
`
`150
`866
`
`200
`379
`
`250
`245
`
`300
`202
`
`350
`175
`
`400
`151
`
`450
`132
`
`500
`125
`
`Ex. 1018 - Page 5 of 7
`
`

`
`US 6,252,055 B1
`
`9
`An experiment Was therefore set up to determine the
`concentration at Which the precipitated Anti-CD4 antibody
`Was completely resolubilised. A 10 ml aliquot of Anti-CD4
`antibody at 250 mg/ml Was progressively diluted by the
`addition of 50 mM sodium citrate, 0.05 mM EDTA, pH 6.0.
`The absorbance of suitably diluted 1.0 ml aliquots of anti
`body samples at 650 nm Was used as a measure of turbidi
`ty.The results are shoWn in Table 3.
`
`10
`of citric acid and the pH of the ?nal solution Was adjusted
`With NaOH. This material Was sterile ?ltered and divided
`into 2 equal aliquots Which Were then separately concen
`trated in the Filtron ultra?ltration device using a recircula
`tion rate of 250 ml/min. The results are shoWn in Table 4.
`Table 4: Concentration of Anti-CD4 to Greater than 100
`mg/ml in a Cross FloW Ultra?ltration Cell at 250 ml/min
`Recirculation Rate
`
`TABLE 3
`
`10
`
`TABLE 4
`
`Cone
`mg/ml
`
`Turb
`(A650 nm)
`
`237.7
`
`149.5
`
`110.4
`
`88.6
`
`76.3
`
`60.3
`
`Concentration of Anti-CD4 to greater than 100 mg/ml in a
`Cross FloW Ultra?ltration Cell at 250 ml/min Recirculation Rate
`
`1.17
`
`0.096
`
`0.082
`
`0.074
`
`0.03
`
`0.027
`
`15
`
`Parameter
`
`Before
`Concentration
`
`Concentration
`1
`
`Concentration
`2
`
`The precipitate redissolved, but the turbidity and opales
`cence did not disappear completely until a Anti-CD4 anti
`body concentration of about 80 mg/ml Was reached. It Was
`above this concentration that the opalescence Was ?rst
`observed during the concentration, so the precipitate seems
`to be reversible and to be concentration dependant.
`
`20
`
`Maximum
`Concentration
`achieved (mg/ml)
`Concentration of
`?nal product
`(mg/m1)
`Recovery after
`concentration (%)
`Time taken for
`concentration (h)
`Aggregate (%) *
`Turbidity (A650 nm)
`Osmolality
`(mos/kg)
`CD4 Binding
`(mg/m1)
`
`—
`
`169
`
`156
`
`14.4
`
`106.4
`
`100.5
`
`—
`
`—
`
`4.14
`0.003
`281
`
`90
`
`11
`
`95
`
`9
`
`3.95
`0.018
`288
`
`3.97
`0.037
`306
`
`20
`
`98.8
`
`68.3
`
`*Aggregate analysis by Size Exclusion HPLC
`
`Samples for aggregate determination Were diluted to a
`protein concentration of 1 mg/ml using PBS and 50 ,ul or 100
`pl aliquots injected onto a TSK-GEL G3000SWXL siZe
`exclusion HPLC column. The column Was developed With
`0.05% NaN3 and 0.1 M NaZSO4 in 0.1 M phosphate buffer,
`pH 6.7 at a How rate of 1.0 ml/min. The amount of aggregate
`Was determined by integrating the peaks of absorbance at
`280 nm.
`Both concentrations achieved a maximum concentration
`of >150 mg/ml in the ultra?ltration apparatus With no
`deleterious affects on antibody solubility. The concentra
`tions took 9—11 h. The ?nal concentrations of ~100 mg/ml
`Were a result of dilution With the Washes required to maxi
`mise recovery from the ultra?ltration apparatus. Overall
`recoveries Were 90—95%, and no visible precipitate or
`increase in levels of aggregate Were observed. The slight rise
`in turbidity after concentration as measured by the absor
`bance at 650 nm caused a slight opacity of the ?nal
`concentrate, but this Was removed on formulation With
`Polysorbate 80 and sterile ?ltration and Was not considered
`signi?cant.
`The CD4 binding activity for concentration 1 Was almost
`100 mg/ml as expected, but a much loWer value Was
`obtained for concentration 2. The ?nal osmolality of the
`pooled material from concentrations 1 and 2 Was approxi
`mately 297 mOs/kg, and the pool Was a clear, bright solution
`that could easily pass through a sterile 0.2 pm ?lter.
`Anti-CD4 Antibody Concentration in Stirred Cell
`A 330 ml aliquot of Anti-CD4 antibody (as above) Was
`concentrated at 5° C. in an Amicon stirred ultra?ltration cell
`(?tted With YM30 membrane Amicon) to a ?nal concentra
`tion of 170 mg/ml by applying a pressure of 1.5 bar using
`nitrogen gas. The Anti-CD4 antibody in the ultra?ltration
`
`EXAMPLE 3
`Concentration of Anti-CD4 Antibody-Method B
`Buffer Adjustment of Anti-CD4 Antibody
`The Anti-CD4 antibody (1460 ml; 24 g) Was prepared in
`50 mM sodium citrate, 0.05 mM EDTA, pH 6.0. This buffer
`Was made up to ~100 mM sodium citrate, 0.05 mM EDTA,
`pH 6.0 by adding solid citric acid to the antibody preparation
`and adjusting the pH to 6.0 With NaOH. The resulting
`preparation Was sterile ?ltered through a 0.22 ,um ?lter and
`stored as 2 aliquots of ~12 g.
`Anti-CD4 Antibody Concentration in Filtron Ultrasette
`The Filtron Mini-Ultrasette and Watson-MarloW pump
`Were placed in a cold room. The Mini-Ultrasette (30 K
`cut-off cross-?oW ultra?lter Filtron) Was ?ushed With Water
`then sanitised for 30 min With 0.1 M NaOH according to
`manufacturers instructions. (Mini Ultrasette Tangential
`FloW Device Operating Instructions & Mini Ultrasette Care
`and Use Manual., Filtron Technology Corporation). The
`sanitant Was removed by ?ushing With sterile Water folloWed
`by 1—2 liters of sterile PBS, pH 7.2 until the pH of the
`effluent Was 7.2. Anti-CD4 Was circulated through the
`retentate side of the membranes at a ?ux rate of 250 ml/min
`throughout.
`After concentating the Anti-CD4 antibody to ~150 mg/ml
`the retentate Was pumped out of the Mini-Ultrasette and the
`retentate side of the membranes Was ?ushed With 3><20 ml
`of 50 mM sodium citrate, 0.05 mM EDTA, pH 6.0 and each
`20 ml fraction of the ?ush Was collected. The antibody
`concentration of the ?ush fractions Was determined by
`measuring their absorbance at 280 nm as described in
`Example 1.
`The results suggested that reducing the retentate ?ux rate
`may provide a method for concentrating anti-CD4 to ~150
`mg/ml by cross ?oW ultra?ltration and avoiding any pre
`cipitation. This Was tested using the Filtron Mini-Ultrasette
`and a retentate circulation rate of 250 ml/min.
`A suitable isotonic buffer for this Work Was 100 mM
`sodium citra