0022· 1767/9I/1474·1352S02.00/0
`THP. JOtJNNAI. or IMMHttoa..ocv
`Copyright IQ 1991 by The American Assoclallon of lmmunol<igtsls
`
`Vol. 147. 1352-1359. No. 4. Augus1 15. 1991
`Prlnred Ln U.S.A.
`
`REDUCED IMMUNOGENICITY AND IMPROVED PHARMACOKINETICS OF
`HUMANIZED ANTI-Tac IN CYNOMOLGUS MONKEYS
`
`JOHN HAKIMI. 1 • RICHARD CHIZZONITE.' DAVID R. LUKE.* PHILIP C. F AMILLETTI. 8
`PASCAL BAILON,1 JO A. KONDAS,• ROBERTS. PILSON,* PING LIN,* DAVID V. WEBER,1
`CHERYL SPENCE,1 LISA J. MONDINI.* WEN-HUI TSIEN.• JAMES L. LEVIN.' VON H. GALLATI,**
`LAURENCE KORN.tt THOMAS A. WALDMANN.'• CARY QUEEN ... AND WILLIAM R. BENJAMIN*
`
`From the Departments of •1mrnunopharmacology. 'Molecular Genetics. 'Drug Metabolism. 'Bloprocess Development. and
`Roche Research Center. Hoffmann-La Roche. Inc .. Nutley. NJ 07110: 'TS/ Mason Research Institute.
`1Proteln Biochemistry.
`Worchester, MA 01608: ••Central Research. F. Hq[fmarm-l#a Roche Ltd .. 4002 Basel. Switzerland;
`"Protein Design Labs. Inc ..
`
`Mountain Vtew. CA 94043: "Metabolism Branch. National Cancer Institute. National Institute of Health. Bethesda. MD 20892
`
`The anti-Tac mAb has been shown to bind to the
`of at least two polypeptide chains that can independently
`p55 chain of the IL-2R, block IL-2 binding and in­
`bind IL-2: the p55. IL-2R a chain. or Tac peptide (2. 3).
`
`hibit T cell proliferation. A humanized form of anti­
`and the more recently discovered p75 or IL-2R f3 chain (4.
`Tac (HAT) has been constructed that retains the
`5). Study of the p55 peptide was facilitated by the devel­
`
`binding properties of murine anti· Tac (MAT). These
`opment of a mAb, MAT, which binds to human p55 (2).
`two mAb were evaluated in cynomolgus monkeys to
`The Tac peptide ts expressed on the surface of Ag- or
`compare relative immunogenicity and pharmacoki­
`mltogen-actlvated T cells but not on resting T cells. More­
`
`netic properties. Monkeys treated with HAT daily
`over, treatment of human T cells with MAT strongly
`for 14 days exhibited anti-HAT antibody titers
`inhibits their proliferative response to Ag or to IL-2 by
`which were 5-to 10-fold lower than their MAT­
`preventing binding of IL-2 to p55 (3. 6).
`treated counterparts and these antibodies devel­
`High levels of p55 are expressed on malignant cells of
`oped later than in the MAT-treated monkeys. Two
`some lymphoid cancers such as adult T cell leukemia.
`of four monkeys receiving a single injection of MAT
`
`cutaneous T cell lymphoma and Hodgkin's disease (1).
`developed anti-MAT antibodies, whereas none of
`Increased or abnormal IL-2R expression Is also associated
`four monkeys developed antibodies after a single
`with many autoimmune conditions Including rheumatoid
`treatment with HAT. In monkeys injected with
`arthritis, SLE. organ transplant rejection. and graft-vs­
`either HAT or MAT daily for 14 days, the anti-anti­
`host disease (1). Hence, the IL-2R is a potentially useful
`body titers induced were inversely related to the
`and versatile therapeutic target. Agents that specifically
`
`amount of anti-Tac administered. Antibodies that
`eliminate Tac-expressing malignant cells or activated T
`
`developed against MAT were both anti-isotypic and
`ceJJs involved in an autoimmune response could be effec­
`
`anti-idiotypic. whereas those developed against
`tive against those disorders without harming normal Tac­
`HAT appeared to be predominantly anti-idiotypic.
`negative T cells. These agents would potentially be more
`
`The pharmacokinetic properties, that is the half-life
`and area under the curve values. of HAT were also
`selective than other immunosuppressants such as anti­
`
`significantly different from those of MAT. The area
`bodies against the C03 antigenic epitope (i.e .. OKT3). In
`under the curve values for HAT in naive monkeys
`the case of autoimmune conditions. It might in fact only
`were approximately twofold more than those for
`be necessary to suppress T cell proliferation by IL-2R
`
`MAT. and the mean serum half-life of HAT was 214
`blockade. without destroying the T cells, to achieve ther­
`h. approximately four-to fivefold more than MAT.
`apeutic benefit.
`These pharmacokinetic values were reduced in
`Antl-IL-2R antibodies have been effective in animal
`
`monkeys previously sensitized with HAT or MAT
`models as well as in early human trials. In vivo admin­
`
`suggesting that the presence of anti-antibodies al­
`istration of anti-1L-2R antibodies greatly prolonged sur-
`tered these parameters.
`vival of heart allografts in mice and rats (7. 8) and alle­
`viated lnsulltis In nonobese diabetic mice and lupus ne­
`phritis in NZB x NZW mice (9). MAT llself was highly
`effective In prolonging survival of allografts in cynomol­
`gus monkeys (10) with Improved efficacy observed with
`HAT (1 1 ). Jn phase I clinical trials for kidney transplan­
`tation. prophylactic administration of MAT significantly
`reduced the Incidence of rejection episodes. without as­
`sociated toxicity (12). Another antl-IL-2R antibody was
`also effective In this setting (13). Treatment with MAT
`induced temporary partial or complete remission In 7 of
`20 patients with adult T cell leukemia (14) (T. A. Wald­
`mann. unpublished observations).
`Several major problems limit the effectiveness of a
`murine mAb such as MAT when used In human patients.
`1352
`
`Received for publlcaUon December 19. 1990.
`Accepted for publlcaUon May 30. 1991.
`The costs of publlcallon of this article were defrayed In part by the
`payment of page charges. This article must therefore be hereby marked
`aduertlSement In accordance with 18 U.S.C. Section 1734 solely to lndl·
`cate this fact.
`' Address correspondence and reprint requests to Dr. John Hakimi.
`Hoffmann-LaRoche. Department of lmmunopharmacology. 340 Kings·
`land SI.. Nutley. NJ 07110.
`2 Abbreviations used In this paper: IL-2R. IL-2 complex: MAT. mouse
`anU-Tac: Tac. p55 subunit of the human IL·2R: sIL·2R. soluble rIL-2R:
`HAT. humanized anti-Tac: HRP. horseradJsh peroxidase: AUC. area under
`the curve.
`
`The cellular receptor for IL-2 plays an Important role
`In regulation of immune function ( 1 ). The IL-2R2 consists
`
`1 of 8
`
`BI Exhibit 1037
`
`

`

`IMMUNOGENICITY AND PHARMACOKINETICS OF HUMANIZED ANTI-TAC
`
`1353
`
`obtained from Dr. F. Khan. Bloprocess Development. Hoffmann-La
`Roche Inc .. Nutley. NJ.
`HRP-labeled II..-2. HAT and MAT were prepared using a modifi­
`cation of a previously described method (29). A total of 20 mg of
`HRP. grade I (Boehringer-Mannheim. Indianapolis. IN) In 6 ml of
`distilled water was activated by adding 1.0 ml of 0 . 1 M NalO, for 20
`min at room temperature (20-2$0C) and subsequently quenched with
`1.0 ml of 0.5 M ethylene glycol. The activated HRP was dialyzed
`against 5 mM sodium acetate buffer. pH 4.5. and brought up to a
`final volume of I 0 ml. Five mg of protein were dialyzed against 0. I
`M NaHC03• pH 8.0. and added to the activated HRP and diluted with
`IO ml of 0.5 M sodium carbonate buffer. pH 9.5. After 2 h at room
`temperature. 3 ml of 0.1 M NaBH. were added and incubated In the
`dark for 4 to 6 h at 4°C. The HRP-conjugated proteins were dialyzed
`against 0.1 M sodium phosphate buffer. pH 6.5. and then diluted
`with an equal volume of 0.2 M sodium phosphate buffer. 20 mg/ml
`BSA. 1 mg/ml Thlmersol. and 2 mg/ml phenol.
`Monkeys and experimental protocol. Eight groups of four 4 to 6
`kg cynomolgus monkeys (two males and two females: Mason Re­
`search Institute. Worchester. MA) were treated daily on days I
`through 14 (Table I). Groups l and 5 received PBS as a vehicle
`control. Monkeys in groups 2. 3. and 4 received HAT at doses of
`0.05. 0.5. or 5.0 mg/kg. respectively. and groups 6. 7. and 8 received
`MAT at doses of 0.05. 0.5. or 5.0 mg/kg. respectively. On day 42.
`groups 1 to 4 and groups 5 to 6 received a single 5 mg/kg dose of
`HAT or MAT. respectively. Test samples were administered via
`venous catheters surgically placed in the femoral vein attached to a
`vascular port. Samples were administered as single bolus Injections
`within several seconds. Blood samples were obtained by venipunc­
`ture throughoul the 55-day study. Monkeys were tranquilized with
`intramuscular ketamlne HCI before administration of test samples
`and collection of scrum samples.
`Jmmunosorbant assays. To measure the serum levels of monkey
`antlbodJes against HAT or MAT. Nunc-lmmuno MaxlSorp (Nunc.
`Naperville. IL) wells were coated with 100 ng of either HAT or MAT
`in 200 µI of PBS overnight (20-24 h) at 4°C. To each well. LOO µl of
`I % fatty acid and globulin-free BSA (Sigma Chemical Co .. St. Louis.
`MO) in PBS were added for 1 h at room temperature. followed by
`washing with PBS containing 0.05% Tween 20. Wells were incu­
`bated wllh 200 µl of goat standards or test samples. plus 50 µl of
`HRP-HAT or HRP-MAT at a final dilution of 1/4000 overnight at
`4°C. Samples were diluted In 25 mM sodium phosphate. 75 mM
`NaCl. 0.05% Tween 20. 0.01 % BSA. 50 µg/ml phenol red, pH 7.4.
`The initial concentration of the unknowns in the assay was I /3 with
`subsequent threefold dilutions. The plates were washed and then
`developed with I mM 2.2'-a:iinobis (3-ethylbenzthiazollne-sulfonic
`acid) (Sigma) In 0.1 M citrate buffer. 0.03% H202• pH 4.2. for 30 min.
`The absorbance at 405 nm was determined with a Vmax Kinetic
`Microplate reader (Molecular Devices. Menlo Park. CA]. The color
`intensity Is directly proportional to the antibody concentration In
`the serum samples. The relative concentrations of anti-HAT and
`anti-MAT anl!bodles in the monkey serum samples were calculated
`from a goat antibody standard curve titrated on each plate. The
`values expressed are apparent antibody levels. because the detection
`of anl!bodies In this assay is dependent on concentration. affinity.
`and presence of blocking agents such as anti-Tac and slL-2R. The
`assay primarily detects free monkey antibodies: however. some an­
`tibody from antibody-anti-Tac complexes would be detected if a
`reequilibrlum of the antibody interactions was established In the
`wells during the overnight incubation.
`Serum concentrations of 1-lAT and MAT were determined In an
`IL-2 lmmunosorbant receptor assay (27). Plates were coated with I 6
`ng of slL-2R in 200 µI of PBS overnight at 4°C and then blocked with
`1 % BSA as described above. Wells were washed and Incubated with
`200 µl of sample overnight at 4°C. Typically. the initial serum In the
`assay was diluted 1/10 with subsequent 1/2 dilutions. The Initial
`sample concentration varied depending on which treatment group
`
`The mouse antibody is immunogenic in humans and
`provokes a neutralizing antibody response. and may not
`be as efficient as a human antibody at recruiting human
`immune effector functions. In addition. mouse antibodies
`have a much shorter circulating half-life In humans than
`do natural human antibodies (15).
`Problems associated with the therapeutic use of murine
`antibodies have been partially addressed by the genetic
`construction of chimeric antibodies. which combine the
`V region binding domain of a mouse antibody with hu­
`man antibody C regions (1 6). However, because chimeric
`antibodies retain the whole mouse V region. they may
`still be immunogenic. Data on the treatment of human
`patients with chimeric antibodies are only beginning to
`accumulate (15. 17).
`To further reduce the immunogenicity of murine anti­
`bodies. Winter and colleagues (1 8-21) constructed "hu­
`manized" antibodies, in which only the minimum neces­
`sary parts of the mouse antibody. the CDR. were com­
`bined with human V region frameworks and C regions.
`Based on this approach, we have recently constructed a
`humanized anti-Tac antibody (22). The humanized anti­
`Tac antibody (HAT) retains several key mouse framework
`residues, predicted by computer modeling. which are re­
`quired to maintain high affinity binding for p55. In ad­
`dition, the humanized antibody mediates antibody-de­
`pendent cellular cytotoxiclty against T cell leukemia cells
`(23). Previously. it was demonstrated in cynomolgus mon­
`keys with cardiac allografts that HAT appeared less im­
`munogenlc than MAT ( 1 1 ). In this study. cynomolgus
`monkeys were given MAT and HAT to further evaluate
`the relative immunogcnicity and pharmacoklnetics of the
`two mAb. To provide a stringent test of HAT. we applied
`a dosing schedule of f requent injections that would reveal
`any immunogenicity.
`
`MATE:RIALS AND METHODS
`
`Cells. MAT was produced in tissue culture as described previously
`(2). HAT was produced from SP2/0 cells transfccted with the genes
`encoding for the H and L chains of the humanized antibody (22. 23).
`Cells were optimized for antibody secretion by limiting dilution clon­
`ing. Production of HAT was performed in a 3-liter continuous per­
`fusion bloreactor (Bellco Biotechnology. Vineland. NJ) equipped with
`a glass cylinder matrix as previously described (24. 25). The cells
`were grown at 37°C in lscoves's modified Dulbecco's medium (JRH
`Biosciences. Lenexam. KS) supplemented with 5% FCS (JRH Bios­
`ciences). 100 U/ml penicillin G. 100 µg/ml streptomycin. and 25 mM
`HEPES buffer. pH 6.9 lo 7.0. During the production phase of the
`fermentation. days 9 to 83. the medium flow rate was maintained at
`416 ml/h and the conditioned medium contained approximately 8
`mg/liter of HAT.
`Proteins. HAT and MAT were purified on separate II..-2R affinity
`chromatography columns with capacities of I 25 and 300 mg. re­
`spectively (26). Briefly. purified recombinant s1I..-2R (27) was Im­
`mobilized on NuGel P-AF Poly-N-hydroxysucclnlmlde (Separation
`Industries. Metuchln. NJ). Antibodies eluted and concentrated from
`the receptor column were further purified on two serially Jinked
`Sephacryl S-300 columns (60 x 11 .3 cm. Pharmacia Fine Chemicals.
`Piscataway. NJ) In Dulbecco's PBS (Whittaker Bioproducts. Walkers­
`v!lle. MD). All purification steps were carried out at 4°C. and buffers
`were prepared with ultra pure water (Hydro. Research Triangle Park.
`NC). The final products were stcr!llzcd through a 0.2 µM Corning
`filter (Corning Glass Works. Corning. NY) and found to contain less
`than 10 endotoxln units/mg (28). Purity was determined by SDS­
`PAGE under reducing and nonreducing conditions and found to be
`more than 99%.
`Anti-HAT and anti-MAT standards were prepared by immunizing
`goals with the respective proteins in CFA. The goat lgG standards
`were isolated on protein A-SE>pharose CL-4b (Pharmacia) and affinity
`purified on HAT or MAT AffiGel-l 0 affinity columns (Bio-Rad. Rich­
`mond. CAJ. Purified human rlL-2 expressed in Escherichia coll was
`
`Group
`
`TABL.E I
`Jmmunogenicity study treatment groups
`Dally Dose
`Chal!enge Dose
`Days I to 14
`Day 42
`HAT. 5 mg/kg
`Vehicle control
`I
`HAT. 5 mg/kg
`HAT. 0.05 mg/kg
`2
`HAT. 0.5 mg/kg
`HAT. 5 mg/kg
`3
`HAT. 5 mg/kg
`HAT. 5.0 mg/kg
`4
`5
`MAT. 5 mg/kg
`Vehicle control
`MAT. 0 05 mg/kg
`MAT. 5 mg/kg
`6
`7
`MAT. 0.5 mg/kg
`None•
`MAT. 5.0 mg/kg
`8
`None•
`•Monkeys not challenged with MAT due to the anaphylaxls observed
`In group 6.
`
`Response
`
`None
`None
`None
`Anaphylaxis 1/4
`None
`Anaphylaxis 4/4
`
`2 of 8
`
`BI Exhibit 1037
`
`

`

`1354
`
`IMMUNOCENIC!TY AND PHARMACOKJNET!CS OF HUMANIZED ANTI-TAC
`
`A. Group 6
`
`B. Group 7
`
`was studied. Wllhout washJng the samples from the wells. 50 µI of
`
`HRP-IL-2 was added to a final dilution of 1/2000. Af te r 3 h at room
`temperature. wells were washed and developed as described above.
`The color Intensity Is Inversely proportional to the anti-Tac concen­
`tration In the samples. HAT and MAT concentrations In the serum
`were calculated from a standard curve of purified HAT and MAT
`titrated on each plate. In this assay. only the anti-Tac available to
`bind to the slL-2R would be detected. As discussed above for the
`lmmunogenlctty ELISA. a new equilibrium of the antibody complexes
`In the serum could oc cur In the wells during the 24 h Incubation.
`Pharmacoklnetlcs. The AUC and t,12 values for HAT and MAT
`were estimated to reflect the total body burden of the antibody within
`
`
`the lntravascular pool as well as the serum die-away curve. respec­
`tively. Serum concentra tions of the an tibod ies were plotted vs time
`..... 0
`
`on a log-linear graph and the AUC values were calculated by trape­
`O>
`:J 100
`zoidal rule (30). The apparent elimination t112 from a single dosing
`-
`was esllmated by linear regression analysis of the terminal portion
`of the curve from a minimum of four data points.
`t- 211 <
`For multiple dose pharm acoklnetlcs, the maximum serum concen­
`
`trations and Ume lo reach maximum serum concentrations were
`�
`obtained visually from the serum concentration-time graphs. The
`
`apparent t,12 after mulllple dosing was approximated from a mini­
`180
`0 -
`
`mum of three serum concentration-time points obtained after the
`final dose.
`
`1f
`
`-E
`
`tu
`
`160
`
`RESULTS
`
`c. Group 8
`
`16
`
`226
`
`76
`
`10
`
`48
`
`>.
`"'O
`0
`0 :. :. -
`Study design and cltntcal observations. A cynomol­
`J:l
`- 100
`gus monkey study was designed to evaluate the relative
`c: <
`
`
`
`lmmunogenlclty and pharrnacoklnetlc properties of MAT
`
`and HAT. A schematic representation of the study design
`ls shown in Figure 1 and details of the treatment groups
`
`are described In Table I. During the study. the monkeys
`
`
`remained behaviorally and clinically normal with the
`160
`
`
`
`following exceptions. On day 42 one female monkey In
`
`group 4 exhibited an apparent anaphylactlc response
`with 5 mg/kg of HAT. This monkey was
`posttreatment
`
`
`
`treated with epinephrine. dexamethasone. Benadryl. and
`
`was hydrated with saline. The monkey gradually im­
`0 :. :. -
`0
`6
`
`proved and by day 44 appeared normal. All four monkeys
`Time (day)
`0.05 mg/kg/day MAT initially,
`In group 6 that received
`
`
`also exhibited an apparent anaphylactlc response post­
`Figure 2. Time-dependent development of anti-MAT antibodies In In­
`treatment with 5 mg/kg MAT. The monkeys responded
`dividual monkeys administered A. 0.05. 8. 0.50. C. 5.0 mg/kg/day MAT
`for I 4 days. Anll-MAT concentrallons were determined In an ELISA using
`
`to epinephrine and fluids. The animals in the MAT treat­
`
`an affinity purified goat anti-MAT antibody as a standard.
`ment groups 7 and 8 were not challenged
`on day 42. In
`various ELISA systems. no increase in total monkey IgE
`HAT ln Figure 3 (note differences In the ordinate scales).
`
`
`
`was observed, nor was the presence of Ag-specific antl­
`Prebleed sera from all 32 monkeys and sera from day 0
`
`antl-Tac IgE detected (data not shown). The cause of this
`to 42 from control monkeys in groups 1 and 5 showed no
`
`anaphylactic response remains unknown.
`groups. 9 of 12
`
`activity In the ELISA. In the MAT -treated
`Immunogenlcity characterization. Monkey anttglob­
`
`monkeys developed antibodies during the Initial 14 day
`
`ulln levels (i.e .. antibodies to HAT and MAT) were evalu­
`
`treatment period. usually by day 12. In contrast. anti­
`ated in an Ag-bridging ELISA. which can be used to detect
`
`
`HAT antibodies In all but one of the 12 HAT-treated
`
`antibodies of various species and lsotypes using the same
`monkeys were not detected until at least 5 to 10 days
`
`
`reagents. Affinity-purified goat anti-HAT and goat anti­
`
`after the final dose of HAT was administered. In addition.
`
`MAT antibody standards were similarly detected In the
`
`the HAT-treated monkeys showed dramatically lower
`range of 100 to 1000 ng/ml In thelr respective
`assays
`
`
`serum antlglobulin concentrations than the MAT-treated
`(data not shown).
`groups.
`The time-dependent development of antibodies In In­
`
`
`
`The antibody titer developed to HAT as well as MAT
`dividual monkeys is shown for MAT in Figure 2 and for
`
`was In general Inversely related to the protein dose ad­
`4 which was treated with 5 mg/kg/
`
`ministered. In group
`
`
`by day HAT. only one monkey had detectable antibodies
`
`day 42. This was the only HAT-treated monkey that
`
`
`
`
`exhibited an anaphylactic response upon challenge with
`
`HAT on day 42 (Table I). even though monkeys from
`
`
`other groups had apparently higher serum antibody lev­
`
`els on day 42. All monkeys in group 6 that received 0.05
`mg/kg/day MAT exhibited an anaphylactlc response
`
`on day 42. Monkeys In groups 7 and 8
`upon rechallenge
`
`were not challenged (Table I). Thus. four monkeys treated
`
`f'Rl!ATMENT
`U OAYS
`.. !J...,ll . .. !
`10
`so
`20 I .
`.J. J ..
`I
`81.000 COt.LECTION DAYS
`
`CHALLENGE
`DAV •2
`I
`
`40
`
`50
`d. : I I I
`/ ..........
`t'Ge"f OAY 4f
`,. .... , ......
`u., •. ,. ...
`F'lgure l. lmmunog enlclty and pharmacoklnetfc study design for eval­
`uation of antf·Tac antibod ies In cynomolgus monkeys. See Table I for
`addll!onal dctatl.
`
`3 of 8
`
`BI Exhibit 1037
`
`

`

`IMMUNOGENIC!TY AND PHARMACOKINETICS OF HUMANIZED ANTI-TAC
`
`1355
`
`A. Group 2
`
`10000 A. Contro l 8. 0.05 mg/kg
`
`c. 0.5 mg/kg
`
`D. 5.0 mg/kg
`
`�o��-o��-o
`LOLO--o-__
`-6--�
`.����-+-----+
`B. Group 3
`
`- - -
`
`1000
`
`e
`......
`0
`2
`>- 100
`"O
`0
`.D
`·.;::;
`c
`<
`
`10
`
`I
`o.s
`
`�
`
`..
`
`..
`42
`4t u
`Time (day)
`
`4l
`
`51
`
`Figure 4. Primary and secondary Immune responses to HAT and MAT.
`Anti-HAT (0) and anti-MAT (6) antibody concentrations from Individual
`monkeys on day 42 before high-dose challenge and on day 55. Data In A
`represents the anti-MAT response on day 55 In two naive monkeys from
`group 5. No anU·HAT antibodies developed In any monkeys from group
`I. Before challenge animals received multiple doses or B. 0.05. C. 0.50,
`D, 5.0 mg/kg/day of anti-Tac antibody for 14 days. Monkeys dosed with
`MAT In C and D were not rechallengcd with MAT on day 42 (dala not
`shown).
`
`-
`E
`0 --
`....... 0 10
`:J
`I- 10
`<
`I
`0 - 10
`
`0
`0
`.Q
`- 10
`c:
`<
`
`10
`
`•o
`
`40
`
`20
`
`40
`
`10
`
`40
`
`to
`
`>-"C
`
`c. Group 4
`
`0 - - - -
`-
`0
`•
`10
`
`�+-+
`-
`ao
`••
`40
`
`45
`
`1•
`20 H
`Time (day)
`Figure 3. Time-dependent development of anli-HAT anUbodles In In·
`dlvldual monkeys administered A. 0.05. 8. 0.50. C. 5.0 mg/kg/day HAT
`for 14 days. AnU-HAT concentrallons were determined In an ELISA using
`an affinity purified goat anti-I IA T antibody as a standard.
`
`with MAT developed an anaphylactic response at a dose
`100-fold lower than the individual high dose HAT-treated
`monkey.
`A comparison of day 42 (prechallenge) and day 55
`serum antlglobulln levels from all challenged monkeys
`In groups l to 6 Is shown in Figure 4. A primary immune
`response to the single treatment with MAT was observed
`In two naive monkeys in group 5, although the same
`treatment with HAT to group 1 monkeys resulted In no
`antibodies (Fig. 4A). A secondary immune response was
`observed in all animals previously treated with antibody.
`No secondary response was observed in groups 7 and 8.
`because they were not challenged. The greatest second­
`ary responses were observed In the monkeys receiving
`either 0.05 mg/kg MAT or HAT (Fig. 48).
`The specificity (i.e .. anti-Id or antHsotype) of the anti­
`HAT and anti-MAT responses was determined in a com­
`petitive ELISA assay (Fig. 5). Inhibition of antibody bind­
`ing In the ELISA by HAT. MAT, as well as slL-2R indi­
`cates the presence of anti-CDR or antl-idiotyplc antibod­
`ies. because these proteins specifically compete for or
`block recognition of the CDR regions of anti-Tac. Com­
`petition by irrelevant human and mouse IgG proteins
`Indicates the presence of antHsotyplc antibodies. The
`
`Figure 5. Characlerlzallon of anti-HAT and anti-MAT responses In
`monkeys on day 35. A shows the anll·MAT from a monkey In group 6
`and B shows the anll·HAT response from a monkey In group 2. A fixed
`amount of antiserum was Incubated tn the presence of various concen·
`tratlons of HAT (0). MAT (6). slL-2R (OJ. human lgG (V). or mouse lgG (0).
`These data are representative of the data obtained from all oft he monkeys
`with anttglobulln antibodies on day 35.
`
`goat anti-HAT and anti-MAT antibodies were partially
`Inhibited by all of the competitors (data not shown).
`indicating that HAT and MAT administered to goats in­
`duced both an antl-ldiotypic and antHsotypic response.
`Serum from all MAT-treated monkeys was completely
`Inhibited with excess MAT. demonstrating that the
`ELISA assay Is specific for MAT (Fig. 5A). HAT and slL-
`2R were the next most effective Inhibitors followed by
`mouse IgG. Thus. the monkey response to MAT was a
`mixture of antHsotyplc and antl-ldiotyplc antibodies
`
`4 of 8
`
`BI Exhibit 1037
`
`

`

`1356
`
`IMMUNOGENICITY AND PHARMACOKINETICS OF HUMANIZED ANTI-TAC
`
`similar to the goal anti-MAT and anU-HAT responses.
`Complete Inhibition of monkey anti-HAT antibodies was
`achieved with HAT. again demonstrating assay specific­
`ity (Fig. 5B). Human and mouse IgG had little effect
`Indicating that the anti-HAT response In monkeys Is not
`an antl-lsotyplc response. MAT and sIL-2R were almost
`as eff ectlve as HAT In Inhibiting anti-HAT binding. Thus.
`the monkey anti-HAT antibodies are directed toward de­
`terminants shared by MAT and HAT. and blocked by sIL-
`2R. I.e .. toward the CDR regions. This supports our con­
`clusion that the anti-HAT response Is anti-ldlotyplc in
`monkeys.
`Pharmacoktnetic characterization. The pharmacokl­
`netic characteristics of HAT and MAT were determined
`in a competitive immunosorbant receptor assay. In this
`assay. both proteins inhibited IL-2 binding within twofold
`of each other. The detectlon limit was In the range of 125
`to 500 ng/ml. Serum concentrations of HAT and MAT
`were measurable only in monkeys receiving doses of 0.5
`and 5 mg/kg/day of antibody (Figs. 6 and 7, respectively.
`note the differences of the ordinate scales). In general.
`HAT concentrations were increased over the dosing
`period. suggesting that equilibrium was not achieved with
`receptor sites or the extravascular space. The mean max­
`imum concentrations after dosing with 0.5 and 5 mg/kg/
`day of HAT for 14 days were 57 ± 20 (mean± SD) and
`726 ± 115 µg/ml. respectively. In contrast. maximum
`concentrations of 0.5 and 5.0 mg/kg/day of MAT were
`26 ± 9 and 311 ± 57 µg/ml. respectively. but occurring
`at approximately 7 to 9 days after the Initiation of ther­
`apy. The mean time course of decline or t112 values of
`HAT from the serum after 14 days of dosing were highly
`
`100
`
`Figure 7. Serum concentration profile of A. MAT or B. HAT In lndlvld­
`ua.I monkeys recelv1ng 5.0 mg/kg/day of antibodies for 14 days. See Figure
`6 for additional details.
`
`variable. ranging from approximately 47 to 432 h, and
`Independent of dose. The t1/'l of MAT was not calculated
`due to the rapid decline of serum concentrations even
`during the 14-day dosing regimen.
`In control naive animals. the serum concentration-time
`profiles of HAT were significantly different from the
`profiles with MAT (Table Il). The Individual AUC and t1f'l
`values after a single i.v. dose of 5 mg/kg of HAT or MAT
`to control monkeys in groups I and 5. respectively. on
`day 42 are shown in Table II. The mean AUC was ap­
`proximately twofold more In the HAT-treated control
`monkeys when compared to the MAT-treated control
`counterparts. 26,657 ± 6237 vs 11.442 ± 3563 µg · h/ml,
`respectively. A four- to fivefold difference was observed
`In the mean t112 values between HAT and MAT (213.6 ±
`58.8 and 47.8 ± 9.04 h, respectively) (Fig. 8).
`The pharmacokineUc profiles in the multiple-dosed
`groups were significantly altered. Only four MAT-treated
`monkeys were rechallcnged on day 42 due to the observed
`anaphylactlc response. Three of the monkeys had no
`detectable serum MAT levels. whereas In the fourth mon­
`key. levels were detectable but not within the quantlta­
`tlon limits of our assay governed by the standard curve
`(data not shown). Clearly. the elimination of MAT from
`group 6 animals that were treated with 0.05 mg/kg/day
`MAT was significantly enhanced compared to group 5
`animals that had not received MAT previously. Kinetic
`parameters of all but two of the HAT-treated monkeys In
`groups 2 to 4 were estimated (Table II). The AUC values
`In groups 2 and 3 were lower than those In naive animals
`[group I). In animals treated with 5 mg/kg/day (group 4),
`the t112 and AUC values were higher than the values
`obtained for group 2 and 3 monkeys. One monkey In
`group 4 had greater values than the naive group animals.
`
`A.
`
`Group 7
`
`B. Group 3
`
`75
`
`.....
`e
`.......
`0
`::I -
`I- 50
`<
`�
`
`25
`
`'-0
`
`100
`
`
`I- 0
`<
`:I:
`-
`0
`(/)
`"ii
`>
`�
`E
`
`50
`
`75 l
`6/<>,<>
`::I r '-
`
`Q)
`(/)
`
`15
`
`0
`
`0
`
`5
`
`10 15 10 25 30 35 40 45
`Time (day)
`Flgure6. Serum concentration profile of A. MAT or B. HAT In Individ­
`ual monkeys receiving 0.50 mg/kg/day of antibodies for 14 days. Anti­
`MAT and anti-HAT concentrations were determined In a competitive IL-
`2 lmmunosorbant receptor assay using the respective purified anti-Tac
`mAb as the standard.
`
`5 of 8
`
`BI Exhibit 1037
`
`

`

`IMMUNOGENICITY AND PHARMACOKINETJCS OF HUMANIZED ANTI-TAC
`
`1357
`
`TABLE II
`Jndluldual AUC (µg-h/ml}and t., (hi values In monkeys gluen single t.u. dose of 5 mg/kg of HAT or MAT on day 42
`Group 5
`Group 3
`Group 4
`Group 2
`Group l
`HAT Controls
`r,.
`AUC
`259
`35.251
`270
`22.140
`167
`21.961
`200
`27.276
`0 NE. Not evaluable.
`
`HAT 0.05 mg/kg
`
`JiAT 0.5 mg/kg
`
`AUC
`
`Iv,
`
`8.004
`NE"
`6.705
`2.848
`
`19.9
`NE
`26.9
`9.7
`
`AUC
`
`4.883
`5.489
`2.050
`677
`
`Iv,
`
`34.5
`26.7
`18.8
`8.9
`
`AUC
`
`HAT 5 mg/kg
`fv,
`115
`318
`28
`NE
`
`29.485
`99.870
`6.850
`NE
`
`M AT Controls
`
`AUC
`
`9.325
`8.424
`16.388
`11.633
`
`1.,
`53.1
`38. l
`57.6
`42.6
`
`•
`
`10
`
`�
`e
`....
`0 .a
`I- 100 90
`0
`<
`0
`::!:
`0
`0
`I-
`<
`x
`0
`.,
`4i
`>
`.£
`E :J
`� I/)
`
`3
`0
`
`!:I
`
`0
`
`0
`
`:J
`
`0
`
`50 100 150 200 250
`Time !houri
`Figure 8. HAT (OJ and MAT (0) serum concentration In naive monkeys
`administered a single bolus of 5 mg/kg anti-Tac. See Figure 7 for addi­
`tional details.
`
`This may be explained by elevated serum concentrations
`of HAT on day 42 from the multiple dosing regimen (Fig.
`6). In addition. the t112 values of each HAT-treated animal
`after rechallenge were significantly reduced compared to
`values after steady-state dosing (data not shown).
`An association between anti-HAT serum concentra­
`tions and either AUC or t1;2 values was observed (Fig. 9).
`These pharmacokinetic parameters were inversely re­
`lated to the serum anti-HAT concentrations. indicating
`that elevated antibody levels contribute to the accelerated
`elimination of HAT. Correlation of the kinetic parameters
`and antibody levels for the HAT-treated monkeys in
`groups 2 and 3 appeared similar. The group 4 values
`appeared more typical of the parameters measured for
`the control group 1 monkeys suggesting that antibody
`effects on HAT pharmacokinetics were similar between
`groups. Taken together. the survival of HAT was mean­
`ingfully longer than that of MAT in naive monkeys. Fur­
`thermore. the development of antibodies to the adminis­
`tered mAb was associated with a reduced serum t112•
`
`DISCUSSION
`
`In this report, we demonstrate reduced immunogeniclty
`and improved pharmacokinetics of humanized anti-Tac
`relative to Its murine counterpart in cynomolgus mon­
`keys. MAT-treated monkeys developed anti-MAT anti­
`bodies during the primary 14-day treatment period at all
`doses tested (Fig. 2). Development of anti-MAT antibodies
`correlated with the rapid reduction in serum MAT con­
`centrations with time (Fig. 7). A similar anti-MAT re­
`sponse was observed in cynomolgus monkeys receiving
`allografts and treated with MAT (11). Furthermore. a
`single high dose of MAT administered to the control
`
`E 1001'
`�
`:i
`.!.
`ii
`:c
`
`0
`It O •
`• 0
`
`10
`
`5
`
`10000011
`
`•
`
`A
`
`0
`
`B
`
`e .....
`.a:;
`0
`z 10000 J • •
`
`0 0
`
`(.)
`::::>
`<
`
`0
`
`•
`
`1000
`0
`&Oo'-���-'- ��'--���-'-�--'
`60
`40
`0
`10
`io
`30
`50
`
`Anti-HAT (ug/ml)
`
`Figure9. Associations between antl-HAT concentrations vs AUC (A)
`and half-life (B) values after a single 5 mg/kg rechallenge dose on day 42
`in monkeys In group 1 (0. four animals). group 21•. three animals). group
`3 (0. four animals). and group 4 {II. three animals). See Table II.
`
`monkeys on day 42 was sufficient to evoke an antibody
`response within 13 days {Fig. 4). MAT was similar to
`other murlne antibodies studied in primates (31. 32) In
`being recognized as a foreign Ag and inducing both antl­
`idlotypic and anti-tsotypic responses.
`HAT clearly proved to be less immunogentc than MAT.
`Anti-HAT antibody titers were 5- to 10-fold lower in their
`respective dosing groups. and the antibodies were not
`detected until several days after the dosing regimen was
`completed (Fig. 3). Similar results were observed with
`anti-Tac-treated monkeys with cardiac allografts ( 11 ).
`Reduced lmmunogenicity correlated with an improved
`serum-time profile (Fig. 6). Unlike MAT. a single bolus of
`HAT. did not induce a measurable antibody response.
`Inasmuch as HAT and MAT as well as sIL-2R can
`