AU5838899A - Platelet-specific chimeric immunoglobulin and methods of use therefor - Google Patents

Description

-1-

AUSTRALIA

PATENTS ACT 1990 DIVISIONAL APPLICATION NAME OF APPLICANTS: The Research Foundation of State University of New York AND Centocor, Inc.

ADDRESS FOR SERVICE: DAVIES COLLISON CAVE Patent Attorneys 1 Little Collins Street Melbourne, 3000.

INVENTION TITLE: Platelet-specific chimeric immunoglobulin and methods of use therefor The following statement is a full description of this invention, including the best method of performing it known to us: Q:\OPER\MJC\63323-96.DIV 9/11/99 PLATELET-SPECIFIC CHIMERIC IMMUNOGLOBULIN AND METHODS OF USE THEREFOR This application is a divisional of Australian Patent Application No. 63323/96, the entire contents of which is incorporated herein by reference.

Background of the Invention Platelet aggregation is an essential event in the 10 formation of blood clots. Under normal circumstances, blood clots serve to prevent the escape of blood cells from the vascular system. However, during certain disease states, clots can restrict or totally occlude blood flow resulting in cellular necrosis.

15 For example, platelet aggregation and subsequent S thrombosis at the site of an atherosclerotic plaque is an important causative factor in the genesis of conditions such as angina, acute myocardial infarction, and reocclusion following successful thrombolysis and angioplasty. Heart attack patients are typically treated with thrombolytic agents such as tissue plasminogen S* activator or streptokinase, which dissolve the fibrin component of clots. A major complication associated with fibrinolysis is reocclusion based on platelet aggregation which can result in further heart damage. Since glycoprotein (GP)IIb/IIIa receptors are known to be responsible for platelet aggregation, reagents which block these receptors are expected to reduce or prevent reocclusion following thrombolytic therapy and to accelerate the rate of thrombolysis. Such reagents are also expected to be useful in therapy of other vaso-occlusive and thromboembolic disorders.

One approach to blocking platelet aggregation involves monoclonal antibodies specific for GPIIb/IIIa receptors.

A

murine monoclonal antibody, designated 7E3, that inhibits platelet aggregation and appears useful in the treatment of -2human thrombotic diseases is described in published BUZopan Patent Application Mo0n. 205,207 and 206,532. It in kniown in the art that murine antibodies have characteristics which may severely limit their use in human s therapy. As foreign proteino, nurino antibodies may elicit imunie react ions that reduce or des tray their therapeutic.

efficacy and/or evoke allergic or hypersensitivity react ions in patients. The need for readmini stration of so uch therapeutic modaities in thru moic disorders 10 ,Increases the likelihood of these types of iiume reactions.

too.* Chimeric antibodies consisting of non-hwumaz binding 0..regions Joined to human constant regions have been suggested as a means to circumvent the imamunoreactivity problems of murine antibodies. See Proa Natl. Acad. fics.

MU#. fil:6B51 (1984) and PCT Publication No. 1(0 86/01533.

Since the constant region is largely responsible for the iuuunoreativity of some antibody molecules, chimeric antibodies with constant regioms of human origin have been surmised to be less likely to evoke an anti-urine response in humans. However, it is unpredictable whether the joining of a human constant region to a wurine binding region of a desired specificity will reduce innunreactivity the extent and/or incidence of iuiuuuogenicity) and/or alter the binding capability of the, resulting chimeric antibody.

sugmaryZ of the invention This invention pertains to a platelet-specific chime~ric iuuunoglobulin comprising a variable or antigen binding region of non-human origin and a constant region of human origin. The chiuieric iunuoglobulins can be specific for GP11b/Ila receptor or other platelet coumonents.

These antibodies bind to platelets and can block platelet aggregation and thus are useful an antithroazbotic agents, -3in the prevention or reduction of occlusion or reocclusion in a variety of clinical situations following thrombolytic therapy, concomitant with angioplasty), and in the prevention of stenosis and/or restenosis. In another embodiment, an agent which binds GPIIb/IIIa and the vitronectin receptor is used to reduce or prevent occlusion, reocclusion abrupt closure), stenosis and/or restenosis. The anti-platelet antibodies of the present invention are also useful in imaging.

Brief Description of the Figures Figure 1 is a schematic diagram of the results of a Northern analysis of heavy chain and light chain mRNAs for the 7E3 monoclonal antibody using cloned variable regions as probes.

15 Figures 2A-2B are schematic representations of the plasmids p7E3Vhc. (Fig. 2A) and p7E3VhC m (Fig. 2B), which carry the chimeric gene constructs encoding the light and heavy chains, respectively, of a chimeric 7E3 immunoglobulin.

Figure 3 shows the binding of the chimeric 7E3 immunoglobulin encoded by vectors p7E3Vhc. and p 7 E3VHhC, to platelets.

Figure 4 shows the inhibition of platelet aggregation by a chimeric 7E3 (c7E3) immunoglobulin.

Figure 5 is a graph of the plasma antibody concentration (ng/mL) versus time (days), which illustrates the rapid initial clearance of c7E3 Fab K) from the plasma in three patients with stable coronary disease, following a 0.20-mg/kg dose of c7E3 Fab administered intravenously as a five minute infusion.

Figures 6A-6C are illustrations summarizing the effect on platelet activity of a single bolus dose of chimeric 7E3 Fab (0.15 mg/kg, 0.20 mg/kg or 0.25 mg/kg) 2 hours after -4administration of antibody (Ty, A dose response is evident when platelet activity is assayed in terms of receptor blockade (Figure 6A), platelet aggregation (Figure 6B), and bleeding time (Figure 6C). The lines represent median values.

Figures 7A-7C are illustrations of the duration of anti-platelet effect of chimeric 7E3 Fab administered prior to angioplasty in a bolus dose of 0.25 mg/kg. The lines indicate the median values from time zero at baseline through 24 hours for receptor blockade (Figure S. 7A), platelet aggregation (Figure 7B), and bleeding time (Figure 7C).

Figures 8A-8C are illustrations summarizing the 9 anti-platelet activity of a 0.25 mg/kg bolus dose followed 15 by a 12 hour continuous infusion (10 pg/minute) of chimeric S7E3 Fab a) in 11 patients. The lines represent median values determined for percent receptor blockade (Figure 8A), percent of pre-dose (baseline at time zero) platelet aggregation (Figure 8B), and bleeding times (Figure 8C).

•20 Figure 9 is an illustration of the absolute change in hematocrit from baseline to a time 24 hours following the end of infusion for 47 patients described in Example 4.

Figure 10 is a Kaplan-Meier plot illustrating the .o probability of no urgent repeat percutaneous revascularization procedures from the time of randomization for the three treatment groups.

Figure 11 is a graph displaying the odds ratios and confidence intervals for key subgroups (listed at right) entered in the trial. Data are presented for the primary efficacy endpoint (death, nonfatal infarction, urgent angioplasty or surgery, or placement of coronary stent or intraaortic balloon pump for refractory ischemia).

In addition, the absolute event rates for the primary endpoint for each subgroup are tabulated at the left (Event Rates Figure 12 is a graph illustrating the fraction of all patients with no event over the 6 month follow-up period.

Figure 13 is a graph illustrating the fraction of patients with no event over the 6 month follow-up period among those patients who had a successful intervention and no events until after 30 days.

Figure 14 is a graph illustrating the fraction of patients with no event over the 6 month follow-up period considering events after the first 48 hours among patients with an initially successful intervention.

Figure 15 is a graph illustrating the fraction of all 15 patients with no attempt to revascularize the procedure related artery (PRA, procedure related artery) over the 6 month follow-up period.

Figure 16 is a representative graph illustrating saturation binding of '~I-c7E3 Fab to unstimulated

HUVEC.

S 20 Saturation data were used to generate the Scatchard plots shown in Figures 17A-17E.

Figures 17A-17E are illustrations of a Scatchard 0" analysis of saturation binding of 5 I-c7E3 Fab to: unstimulated HUVEC (Figure 17A); HUVEC stimulated for 4 hours with 50 units/ml of TNFa, (Figure 17B); HUVEC stimulated for 24 hours with 50 units/ml of TNFa,(Figure 17C); unstimulated HUVEC in serum-free media (Figure 17D); unstimulated HUVEC in the presence of 0.02% azide to prevent antibody capping and internalization (Figure 17E).

HUVEC were incubated with increasing concentrations of I-c7E3 Fab in the presence or absence of a 100-fold excess of cold c7E3 Fab to define non-specific binding.

Bound 25 I-c7E3 Fab is plotted on the abscissa and the amount bound divided by the concentration of free antibody -6is plotted on the ordinate. Linear regression through the curves yielded an equation which is shown on each graph.

The slope is defined as the K value. The intercept with the Y-axis is the B. or maximal amount of antibody bound. Data points in each graph represent the means of triplicate determinations.

Figure 18 is an illustration of a Scatchard analysis of '"I-LM609 binding to endothelial cells. HUVEC were incubated with increasing concentrations of the vitronectin receptor-specific antibody 'II-LM609 in the presence or absence of a 100-fold excess of cold c7E3 Fab to define non-specific binding. Bound 2 I-LM609 is plotted on the abscissa and the amount bound divided by the concentration of free antibody is plotted on the ordinate. Linear 15 regression through the curves yielded the following equation: y 1.0831e+10 1.2188e+9x R 2 0.888. The slope is defined as the K. value. The intercept with the Y-axis is the or maximal amount of antibody bound.

Data points represent the means of triplicate determinations.

Figure 19 is a graph illustrating competitive binding i of antibodies with 2 I-c7E3 Fab for binding to endothelial cells. HUVEC were incubated with 1 ig/ml s5I-c7E3 in the presence of increasing concentrations of unlabeled 25 competitor. Anti-CD51 is a monoclonal antibody which recognizes the a-chain of the vitronectin receptor; Anti-IIIa is a monoclonal antibody which reacts with GPIIIa; vitronectin is the native protein isolated from human plasma; c7E3 Fab is chimeric 7E3 Fab fragment; m7E3 IgG is murine 7E3 IgG; anti-7E3 is a rabbit, variable region specific anti-7E3 antibody; LM609 is a monoclonal antibody which binds to complexed aB 3 (vitronectin receptor) but does not bind to GPIIb/IIIa; 10E5 is a monoclonal antibody which reacts with GPIIb/IIIa but does -7not recognize endothelial cell GPIIb/IIIa; chimeric MT412 is an anti-CD4 antibody used c: an isotype-matched chimeric Fab fragment control. Data points represent the means of triplicate determinations.

Figures 20A-20B are histograms illustrating adhesion protein expression on endothelial cells after treatment with c7E3 Fab. E-selectin expression on HUVEC after 4 hour treatment with c7E3 Fab and ICAM-1 expression on HUVEC after 24 hour treatment with c7E3 Fab was monitored by tI-anti-E-selectin binding (Figure 20A) or 1ZI-anti-ICAMbinding (Figure 20B). HUVEC were treated with the indicated concentration of chimeric 7E3 Fab or chimeric MT412 Fab (an anti-CD4, isotype-matched negative control antibody) antibody for either 4 or 24 hours. TNFa was used 15 as a positive control to increase E-selectin and ICAM-1 expression. Data points represent the means SEM of triplicate determinations.

Figures 21A-21B are histograms illustraing PMN adhesion to endothelial cells treated with c7E3 Fab. HUVEC were treated with 100 pg/ml chimeric 7E3 Fab or 100 pg/ml chimeric MT412 Fab antibody for either 4 (Figure 21A) or 24 hours (Figure 21B). TNFa was used as a positive control to increase E-selectin and ICAM-1 expression and therefore Sadhesiveness for PMN. Data represent the means SEM of 25 triplicate determinations.

Detailed Description of the Invention The chimeric immunoglobulins of the present invention are comprised of individual chimeric heavy and light immunoglobulin chains. A chimeric heavy chain comprises a nonhuman antigen-binding region derived from the heavy chain of a nonhuman antibody specific for platelets, such as an antibody specific for the GPIIb/IIIa receptor) linked to a human heavy chain constant region. A chimeric -8light chain comprises a nonhuman antigen binding region derived from the light chain of the non-human antibody) linked to a human light chain constant region.

The present immunoglobulins can be monovalent, divalent or polyvalent. Monovalent immunoglobulins are dimers (HL) formed of a chimeric heavy chain associated through disulfide bridges with a chimeric light chain.

Divalent immunoglobulins are tetramers (HL 2 formed of two dimers associated through at least one disulfide bridge.

Polyvalent immunoglobulins can also be produced, for example, by employing a heavy chain constant region that aggregates p heavy chain constant regions).

Chimeric immunoglobulin fragments such as Fab, Fab' or F(ab') 2 can also be produced.

15 The non-human antigen binding regions of the chimeric t immunoglobulin can be derived from immunoglobulins specific for platelets. Preferred immunoglobulins are specific for platelet GPIIb/IIIa receptors and can block ligand binding to the glycoprotein IIb/IIIa receptor complex.

Thrombosis begins with the adhesion of platelets at sites of vessel wall injury. The adhesion of platelets is mediated by platelet surface receptors which bind to extracellular matrix proteins in the exposed subendothelium, such as von Willebrand factor, collagen, 25 fibronectin, vitronectin, and laminin. Platelet adhesion results in a monolayer of platelets. Subsequently, platelet activation occurs in response to agonists such as epinephrine, ADP, collagen, and thrombin. Activation leads to the exposure of the glycoprotein IIb/IIIa receptor (GPIIb/IIIa) on the platelet surface. GPIIb/IIIa on activated platelets is then available to bind to fibrinogen, which can mediate platelet aggregation. The binding of GPIIb/IIIa to other adhesive proteins, such as von Willebrand factor can also cause platelet cross-linking -9and aggregation. Thus, the binding of adhesive molecules, such as fibrinogen or von Willebrand factor, to GPIIb/IIIa to cause aggregation of platelets is a common step in thrombus formation, making the GPIIb/IIIa receptor an attractive target for therapeutic agents which can interfere with the interaction of GPIIb/IIIa with these molecules. Furthermore, by use of an anti-GPIIb/IIIa chimeric antibody, the aggregation of activated platelets is expected to be inhibited, without interfering with the initial adhesion of platelets. This selective inhibition of platelet aggregation may be desirable because platelet adhesion, without aggregation, may contribute to maintaining hemostasis.

*0 Examples of suitable antibodies specific for platelets include 7E3 and 10E5. See European Patent Application Nos.

EP 0,205,207, EP 0,206,532, and EP 0,206,533, the teachings of which are incorporated herein by reference. The 7E3 antibody (or antibody reactive with the same or a functionally equivalent epitope) is especially preferred S 20 because it is specific for the complexed form of the GPIIb/IIIa receptor. Other antibodies specific for the GPIIb/IIIa receptor (antigen recognized by 7E3), such as those specific for either the IIb or IIIa components, can also be used. Antibodies specific for other platelet antigens can be employed. For example, antibodies reactive with platelet a granule membrane protein GMP-140, such as S12 antibody Biol. Chem., 252:9799-9804 (1984); U.S.

Patent No. 4,783,330) can be used.

The antigen binding region of the chimeric antibody can be derived from an immunoglobulin of nonhuman origin.

Preferably, the antigen binding region is of murine origin because murine antibodies against platelets, and particularly GPIIb/IIIa receptors, are available or can be produced in murine systems. Other animal or rodent species provide alternative sources of antigen binding regions (see Newman et al., Bio/technolov, Af: 1455-1460 (1992)).

In one embodiment, the antigen binding region of the chimeric immunoglobulin comprises at least a portion of a platelet-specific immunoglobulin of nonhuman origin sufficient for specific or selective antigen binding, such as one or more complementarity determining regions or portions thereof derived from the nonhuman immunoglobulin (see Winter, U.S. Patent No. 5,225,539, European Patent No. EP 0,239,400, U.K. Patent No. 2,188,638; Adair et al., WO 91/09967; Jolliffe et al., WO 91/09966). In another embodiment, the chimeric immunoglobulin comprises at least one chimeric heavy chain comprising a variable region derived from the heavy chain of a nonhuman immunoglobulin, linked to at least a portion of a human 15 heavy chain constant region, and at least one chimeric light chain comprising a variable region derived from a light chain of the nonhuman immunoglobulin covalently linked to at least a portion of a human light chain constant region. Other combinations of variable and constant regions are also possible (see U.S. Patent No. 5,169,939).

The constant regions of the chimeric antibodies are derived from human immunoglobulins. The heavy chain constant region can be selected from any of the five 25 isotypes alpha, delta, epsilon, gamma or mu. Further, heavy chains of various subclasses (such as the IgG subclasses of heavy chains) are responsible for different effector functions and thus, by choosing the desired heavy chain constant region, chimeric antibodies with desired effector function can be produced. Preferred constant regions are gamma 1 (IgG1), gamma 3 (IgG3) and gamma 4 (IgG4). The light chain constant region can be of the kappa or lambda type.

In general, the chimeric antibodies can be produced by preparing, for each of the light and heavy chain components -11of the chimeric immunoglobulin, a fused gene comprising a first DNA segment that encodes at least the functional portion of the platelet-specific variable region of nonhuman origin functionally rearranged variable region with joining segment) linked to a second DNA segment encoding at least a part of a human constant region. Each fused gene is assembled in or inserted into an expression vector, yielding an expression vector containing a fused gene in expressible form. In one embodiment, DNA comprising the antigen binding region is covalently joined to the constant region via an intervening sequence. In another embodiment, constructs lacking one or more intervening sequences can be constructed or obtained.

Recipient cells capable of expressing the gene products are 15 then transfected with the genes. The transfected recipient cells are cultured under conditions that permit expression of the incorporated genes and the expressed immunoglobulins or immunoglobulin chains are recovered.

Genes encoding the variable region of Ig light and 20 heavy chains can be obtained from lymphoid cells that produce the platelet-specific antibodies. For example, the hybridoma cell lines that produce antibody against the GPIIb/IIIa receptor provide a source of immunoglobulin S. variable region for the present chimeric antibodies. Other rodent cell lines are available. Cell lines can be produced by challenging a rodent with a human platelet or a GPIIb/IIIa receptor-containing component or fraction of platelets, forming fused hybrid cells between antibody-producing cells and a myeloma cell line, cloning the hybrid and selecting clones that produce antibody against platelets or glycoprotein IIb/IIIa receptor.

Constant regions can be obtained from human antibody-producing cells by standard cloning techniques.

Alternatively, because genes representing the two classes of light chains and the five classes of heavy chains have been cloned, constant regions of human origin are readily available from these clones. Chimeric antibody binding fragments such as and Fab fragments can be prepared by designing a chimeric heavy chain gene in truncated form.

For example, a chieric gene encoding a F(ab') 2 heavy chain portion would include DNA sequences encoding the CHI domain and hinge region of the heavy chain. Such fragments can also be obtained by enzymatic cleavage of a chimeric immunoglobulin. For instance, papain or pepsin cleavage can generate Fab or fragments, respectively.

Preferably, the fused genes encoding the light and heavy chimeric chains (or portions thereof) are assembled in two different expression vectors that can be used to cotransfect a recipient cell. Each vector contains two 15 selectable genes--one for selection in a bacterial system and one for selection in a eukaryotic system--each vector having a different pair of genes. These vectors allow production and amplification of the fused genes in bacterial systems, and subsequent cotransfection of 20 eukaryotic cells and selection of the cotransfected cells.

Examples of selectable genes for the bacterial system are •the genes that confer ampicillin resistance and the gene that confers chloramphenicol resistance. Two selectable genes for selection of eukaryotic transfectants are 25 preferred: the xanthine-guanine phosphoribosyltransferase gene (gpt), and (ii) the phosphotransferase gene from Tn5 (designated neo).

Selection with gpt is based on the ability of the enzyme encoded by this gene to use xanthine as a substrate for purine nucleotide synthesis; the analogous endogenous enzyme cannot. In a medium containing xanthine and mycophenolic acid, which blocks the conversion of inosine monophosphate to xanthine monophosphate, only cells expressing the gpt gene can survive. The product of the -13neo gene blocks the inhibition of protein synthesis in eukaryotic cells caused by the antibiotic G418 and other antibiotics of its class. The two selection procedures can be used simultaneously or sequentially to select for the expression of immunoglobulin chain genes introduced on two different DNA vectors into a eukaryotic cell.

The preferred recipient cell line is a myeloma cell.

Myeloma cells can synthesize, assemble and secrete immunoglobulins encoded by transfected Ig genes. Further, they possess the mechanism for glycosylation of the 'immunoglobulin. A particularly preferred recipient cell is an Ig-non-producing myeloma cell line such as SP2/0. These cell lines produce only the immunoglobulin encoded by the transfected immunoglobulin genes. Myeloma cells can be 15 grown in culture or in the peritoneum of mice where secreted immunoglobulin can be obtained from ascites fluid.

Other lymphoid cells such as B lymphocytes or hybridoma cells can serve as suitable recipient cells.

Several methods exist for transfecting lymphoid cell 20 with vectors containing immunoglobulin encoding genes. A preferred way of introducing DNA into lymphoid cells is by electroporation. In this procedure recipient cells are subjected to an electric pulse in the presence of the DNA to be incorporated. See Potter et Proc. Natl.

Acad. Sci. USA, 81:7161 (1984). Another way to introduce DNA is by protoplast fusion. In this method, lysozyme is used to strip cell walls from bacteria harboring the recombinant plasmid containing the chimeric Ig gene. The resulting spheroplasts are fused with myeloma cells with polyethylene glycol. After protoplast fusion, the transfectants are selected and isolated. Another technique that can be used to introduce DNA into many cell types is calcium phosphate precipitation.

The chimeric immunoglobulin genes can also be expressed in nonlymphoid cells such as bacteria or yeast.

-14- When expressed in bacteria, the immunoglobulin heavy chains and light chains can become part of inclusion bodies.

Thus, the chains must be isolated and purified and then assembled into functional immunoglobulin molecules. Other strategies for expression in E. coll are available (see Pluckthun, Bio/Technolo., 1:545-551 (1991); Skerra, A. et al., Bio/Technoloy, 1:273-278 (1991)), including secretion from E. coli as fusion proteins comprising a signal sequence.

Utility of Platelet-specific Chimeric Immunoalobulin chimeric platelet-specific antibodies of this invention are useful as antithrombotic therapeutic agents.

For example, the chimeric antibodies (or fragments thereof) can be used to inhibit platelet aggregation and thrombus 15 formation in patients having a thrombus or at risk of thrombus formation. The antibodies can also be used to inhibit flow variations cyclic flow variations) which are caused by platelet aggregation, and which may precede thrombus formation or reformation. The antibodies 20 can be used in a variety of situations where thrombus :.-formation or reformation (reocclusion) is to be prevented or minimized. Furthermore, the antibodies can be used in a variety of situations where stenosis or restenosis is to be inhibited (reduced, delayed or prevented).

For example, patients at risk of or having coronary artery disease can benefit from the administration of an effective amount of a chimeric anti-platelet antibody or antibody fragment of the present invention a chimeric anti-GPIIb/IIIa antibody or preferably a fragment thereof, such as chimeric 7E3 Fab or F(ab') 2 to inhibit occlusion, reocclusion, stenosis and/or restenosis of vessels.

For example, the antibody can be administered to an individual a mammal such as a human) to prevent thrombosis in pulmonary embolism, transient ischemic attacks (TIAs), deep vein thrombosis, coronary bypass surgery, surgery to insert a prosthetic valve or vessel in autologous, non-autologous or synthetic vessel graft) or deployment of a vascular (coronary or peripheral) stent. The antibodies of the present invention can also be administered to an individual to prevent platelet aggregation and thrombosis before, during and/or after a coronary artery intervention procedure angioplasty, placement of a stent, angioplasty with stent placement, vascular graft) or other vascular intervention procedures deployment of a peripheral stent, insertion of a prosthetic valve or vessel in autologous, non-autologous or synthetic vessel graft)). For example, the antibodies can be administered to an individual undergoing angioplasty procedures performed by balloon, coronary atherectomy, laser angioplasty or other suitable 20 methods. Antibody can be administered prior to the angioplasty procedure (pre-angioplasty), during angioplasty, and/or post-angioplasty. Such treatment can prevent thrombosis and thereby reduce the rate of thrombotic complications following angioplasty, such as death, myocardial infarction, or recurrent ischemic events necessitating PTCA or coronary bypass surgery (acute ischemic events). In addition, such treatment can yield a longer-term benefit by reduction of ischemic events or complications of a coronary artery intervention procedure angioplasty, placement of a stent, angioplasty with stent placement, vascular graft), such as death, myocardial infarction, or recurrent ischemic events necessitating PTCA or coronary bypass surgery (revascularization procedures), indicative of reduction, delay or prevention of stenosis or restenosis. A long-term reduction or prevention of ischemic events or complications due to other vascular intervention procedures deployment of a peripheral stent, insertion of a prosthetic valve or vessel in autologous, non-autologous or synthetic vessel graft)) can also be achieved by administering an antibody of the present invention before, during and/or after the procedure.

For instance as shown in Example 4, administration of a chimeric anti-platelet antibody (chimeric 7E3 Fab fragment) as adjuvant therapy prior to angioplasty (percutaneous transluminal coronary angioplasty, PTCA) .increased bleeding times and reduced agonist-induced platelet aggregation as assayed by ex vivo platelet aggregation assays. The results of the experiments 15 reported in Examples 4 and 5 also suggest that blockade of platelet GPIIb/IIIa and inhibition of aggregation by c7E3 antibody (an Fab fragment) translates into in vivo antithrombotic efficacy in humans.

The results of a randomized, double-blind, placebocontrolled study of administration of a chimeric 7E3 antibody fragment are presented in Examples 6 and 7. The data presented in Example 6 reveal that administration of Schimeric 7E3 antibody fragment to patients undergoing angioplasty and at high risk of abrupt closure 25 (reocclusion) can prevent abrupt closure (reocclusion), decreasing the incidence of acute ischemia. As is further illustrated in Example 7, administration of a chimeric 7E3 antibody fragment to patients undergoing angioplasty and at high risk of abrupt closure (reocclusion) can reduce, delay and/or prevent restenosis at later times.

The long-term benefit as shown by reduction of nonacute ischemic complications observed upon administration of a chimeric anti-GPIIb/IIIa antibody fragment (see Example 7) is unprecedented. Additional compounds or agents which selectively bind to the GPIIb/IIIa receptor -17can be used to prevent or reduce stenosis or restenosis, with the clinical benefit of reducing the occurrence of non-acute ischemic complications. Upon administration, these compounds can additionally prevent occlusion or reocclusion. These compounds include GPIIb/IIIa antagonists, immunoglobulin or non-inmunoglobulin peptides or proteins synthetic, recombinant), analogs thereof, and nucleic acids or nucleic acid analogs, for example. For example, patients at risk of or having coronary artery disease can benefit from the administration of an effective amount of a compound which selectively binds to the GPIIb/IIIa receptor thereby inhibiting occlusion, reocclusion, stenosis and/or restenosis of vessels.

15 In one embodiment, stenosis and/or restenosis of vessels can be inhibited by administering, prophylactically or therapeutically, a compound or agent GPIIbjIIIa antagonists, immunoglobulin or non-immunoglobulin peptides or proteins synthetic, recombinant), analogs thereof, and nucleic acids or nucleic acid analogs) which binds GPIIb/IIIa and the vitronectin receptor designated a~.3 Platelet glycoprotein GPIIb/IIIa (also referred to as CD41/CD61) belongs to a family of integrin receptors which share structural and immunological characteristics. An 25 integrin closely related to GPIIb/IIIa is the vitronectin receptor (av 3 also referred to as CD51/CD61), which uses the same P subunit as GPIIb/IIIa but has a different a subunit. The vitronectin receptor is expressed on cells such as endothelial cells and vascular smooth muscle cells (and to a lesser extent, on platelets), and mediates adhesion to a variety of extracellular matrix proteins vitronectin, fibronectin, von Willebrand Factor, fibrinogen, osteopontin, thrombospondin, collagen, perlecan). The homology between GPIIb/IIIa and the -18vitronectin receptor is sufficient so that monoclonal antibody 7E3, directed against GPIIb/IIIa, also binds to the vitronectin receptor expressed on endothelial cells (Example Injury of the vessel wall leads to the release of a variety of mediators of cellular activation and proliferation. Platelet aggregation, platelet degranulation, inversion, and platelet surface events involved in coagulation lead to thrombosis and the release of other factors growth factors and cytokines, such as platelet-derived.growth factor), which stimulate cellular proliferation and migration at the injury site.

Inflammatory cytokines may induce the production of matrix proteins collagen, osteopontin, vitronectin), which 15 accumulate in the area. Cellular migration is triggered, and vascular smooth muscle cells, endothelial cells, macrophages, fibroblasts, and other inflammatory cells migrate to the site, leading to a lesion atheroma), which narrows the vessel lumen (stenosis or restenosis).

The etj3 integrin or vitronectin receptor is implicated in the migration of dells endothelial cells) to the site of injury. a~, 3 binds to extracellular matrix proteins, such as vitronectin, osteopontin or other matrix proteins, present in atherosclerotic lesions. Cross- 25 linking of aX3 receptors, such as when they bind to vitronectin, can initiate a migration/activation signal, as well as the production of substances which promote migration. Restenosis, which narrows the vessel lumen, leads to thrombotic events.

While not wishing to be bound by a particular theory, it is reasonable to conclude that the administration of an agent an antibody such as c7E3 Fab) which reacts with both GPIIb/IIIa and a,83, and inhibits their function, can inhibit platelet aggregation, degranulation, and -19inversion, and can passivate the vessel vall preventing thrombus formation and acute clinical events. Other events, such as platelet surface events involved in coagulation, can thereby be reduced or prevented, leading to a reduction in the amount of thrombin formed and of other factors released growth factors and cytokines), and an inhibition of cellular proliferation, migration, and lesion formation. Based on the unprecedented long-term benefit, as shown by reduction of non-acute ischemic complications observed upon administration of a chimeric anti-GPIIb/IIIa antibody fragment (see Example the administration of an agent an antibody such as c7E3 Fab) which reacts with platelet GPIIb/IIIa and integrin present on endothelial cells, and inhibits their function, can reduce or prevent non-acute ischemic complications of a coronary artery intervention procedure late clinical events from about three to six months or later following the procedure). For example, an antibody or antibody fragment, such as Fv, Fab, Fab' and F(ab') 2 fragments, which binds GPIIb/IIIa and the vitronectin receptor, can be administered.

Suitable antibodies can be polyclonal or monoclonal, and the term antibody or immunoglobulin is intended to encompass both polyclonal and monoclonal antibodies. The term antibody or immunoglobulin is also intended to encompass single chain antibodies, chimeric, humanized or primatized (CDR-grafted) antibodies, as well as chimeric or CDR-grafted single chain antibodies, comprising portions from more than one species. See, Cabilly et al., U.S. Patent No. 4,816,567; Cabilly et al., European Patent No. 0,125,023 Bl; Winter, U.S. Patent No. 5,225,539; and Winter, European Patent No. 0,239,400 Bl. See also, Newman, R. et al., BioTechnology, 10: 1455-1460 (1992), regarding primatized antibody, and Ladner et al., U.S.

Patent No. 4,946,778 and Bird, R.E. et al., Science, 242: 423-426 (1988)) regarding single chain antibodies.

For example, murine 7E3 or chimeric 7E3 antibody, which bind GPIIb/IIIa and the vitronectin receptor, can be administered. In another embodiment, the antibody to be administered is reactive with the same (or a functionally equivalent) epitope on GPIIb/IIIa and the vitronectin receptor as that bound by the 7E3 antibody. For example, an antibody which blocks the binding of the 7E3 monoclonal antibody to GPIIb/IIIa and to the vitronectin receptor can be used. In a preferred embodiment, such a cross-reactive ;antibody or portion thereof an Fab fragment) has a high affinity for the GPIIb/IIIa receptor on platelets and 15 for the vitronectin receptor on endothelial cells, for C example, at least about 5.0 x 10 7

M

1 or more preferably at least about 1.0 x 10' M'.

Such antibodies can be raised against an appropriate immunogen platelets, isolated and/or purified 20 GPIIb/IIIa or or their component chains, especially $3 chain, portions of the foregoing or synthetic molecules, such as synthetic peptides). Preparation of immunizing antigen, and polyclonal and monoclonal antibody production can be performed using any suitable technique. A variety of methods have been described (see U.S. Patent No.

5,336,618 (Coller); Kohler et al., Nature, 256: 495-497 (1975) and Eur. J. Immunol. 6: 511-519 (1976); Milstein et al., Nature 266: 550-552 (1977); Koprowski et al., U.S.

Patent No. 4,172,124; Harlow, E. and D. Lane, 1988, Antibodies: A Laboratory Manual, (Cold Spring Harbor Laboratory: Cold Spring Harbor, NY); Current Protocols In Molecular Biology, Vol. 2 (Supplement 27, Summer '94), Ausubel, F.M. et al., Eds., (John Wiley Sons: New York, NY), Chapter 11, (1991)).

-21- Administration The aggregation of platelets activates the coagulation cascade and produces a more stable fibrin meshwork and occlusive clot, which can be lysed by thronbolytic agents.

The antibody of the present invention or compounds which selectively bind to the GPIIb/IIIa receptor can be administered to an individual a human) alone or in conjunction with a thrombolytic agent, such as a plasminogen activator tissue plasminogen activator, 10 urokinase, streptokinase, recombinant tissue plasminogen activator), or an anticoagulant an antithrombin agent) or anti-platelet agent, such as aspirin, heparin, hirulog, hirudin, or a coumarin anticoagulant warfarin), to prevent or reduce reocclusion that can occur after thrombolysis and to accelerate clot lysis, for example. The compounds, antibody or fragment thereof can also be administered before, along with or subsequent to administration of a thrombolytic agent, antithrombin agent, :anticoagulant or anti-platelet agent, in amounts sufficient 20 to prevent platelet aggregation that can result in occlusion or reocclusion and/or to delay or prevent stenosis or restenosis.

An effective amount an amount sufficient to achieve the desired therapeutic effect, such as an amount sufficient for inhibition of platelet aggregation and thereby of inhibition of thrombus formation or reformation, an amount sufficient to reduce/delay or prevent stenosis or restenosis or ischemic events) of a compound or an antibody or antibody fragment can be given parenterally, preferably intravenously, in a pharmaceutically acceptable vehicle such as sterile saline. Buffered media may be included.

The antibody formulation can contain additional additives, such as a stabilizer Polysorbate 80, USP/NF). The antibody can be administered in a single dose, continuously, or in multiple infusions a bolus -22injection, followed by continuous infusion).

Alternatively, the compound or antibody could be administered by a controlled release mechanism by a polymer or patch delivery system) or by another suitable method. The amount to be administered will depend on a variety of factors such as the clinical symptoms, weight of the individual, whether other drugs thrombolytic agents) are administered.

During repeat therapy with anti-platelet antibodies, drug-induced thrombocytopenia may occur; this may be a result of the body recognizing the antibody-coated platelets as foreign proteins, raising antibodies against them, and then clearing them via the reticuloendothelial system more rapidly than normal. Because of the uniquely 15 high density of the GPIIb/IIIa receptor on the platelet surface (-80,000 receptors per platelet) and the large number of platelets in the circulation (-0.25 -0.5 X per 1 thrombocytopenia may be an important complication of treatment with anti-platelet antibodies. The use of a 20 chimeric anti-platelet anti-GPIIb/IIIa) antibody can avoid this problem. The chimeric anti-platelet antibodies of the present invention may minimize (reduce or prevent) the thrombocytopenia which might otherwise occur on administration of an anti-platelet antibody. For example, minimal thrombocytopenia was observed on administration of chimeric 7E3 Fab (see Examples 6 and 7).

Secondly, upon administration to humans, chimeric 7E3 Fab antibody fragment displays a surprisingly reduced incidence of induced immunogenicity as compared with its murine counterpart (see Examples 4 and 7), particularly in view of the immunogenicity of the variable region of murine 7E3 Fab. Upon binding platelets, the majority of the murine component of the anti-platelet chimeric antibody will be bound to the platelet surface, -23via the GPIIb/IIIa receptor, and thus will be inaccessible to the immune system, rendering the chimeric antibody functionally indistinguishable from a human antibody directed against the same epitope. Therefore, other chimeric anti-platelet antibodies of the present invention may be similarly non-immunogenic in spite of their murine antigen binding region.

The platelet-specific chimeric immunoglobulins of this invention are also useful for thrombus imaging. For this purpose, antibody fragments are generally preferred. As escribed above, a chimeric heavy chain gene can be designed in truncated form to produce a chimeric immunoglobulin fragment Fab, Fab', or F(ab') 2 for S. immunoscintigraphic imaging. These molecules can be labeled either directly or through a coupled chelating agent such as DTPA, with radioisotopes such as 3'Iodine, 25 Iodine, "mTechnetium or "'Indium to produce radioimmunoscintigraphic agents. Alternatively, a radiometal binding (chelating) domain can be engineered 20 into the chimeric antibody site to provide a site for labeling. Thus, a chimeric immunoglobulin can be designed as a protein that has a nonhuman platelet-specific variable region, a human constant region (preferably truncated), and a metal binding domain derived from a metal binding protein, such as metallothionein.

The platelet-specific chimeric immunoglobulin or fragment thereof is administered to a patient suspected of having thrombus. After sufficient time to allow the labeled immunoglobulin to localize at the thrombus site, the signal generated by the label is detected by means of a photoscanning device such as a gamma camera. The detected signal is then converted to an image of the thrombus. The image makes it possible to locate the thrombus in vivo and to devise an appropriate therapeutic strategy.

-24- The invention is further described by the following examples, which are not intended to be limiting in any way.

EXEMPLIFICATION

Exampl Production of chimeric platelet specific IaG4 AL General Strategy The strategy for cloning the variable regions for the heavy and light chain genes from the 7E3 hybridoma was based upon the linkage in the genome between the variable 10 region and the corresponding J (joining) region for functionally rearranged (and expressed) Ig genes. J region eoDNA probes can be used to screen genomic libraries to isolate DNA linked to the J regions; DNA in the germline configuration (unrearranged) would also hybridize to J 15 probes but is not linked to a variable region sequence and can be identified by restriction enzyme analysis of the isolated clones.

The cloning strategy, therefore, was to isolate variable regions from rearranged heavy and light chain genes using J. and JK. These clones were tested to see if their sequences were expressed in the 7E3 hybridoma by Northern analysis. Those clones that contained expressed sequences were put into expression vectors containing human constant regions and transfected into mouse myeloma cells to determine if an antibody was produced. The antibody from producing cells was then tested for binding specificity and functionality in comparison with the 7E3 murine antibody.

A deposit of cell line derivative, murine hybridoma 7E3, was made with the American Type Culture Collection, 12301 Parklawn Drive, Rockville, MD 20852, on May 30, 1985.

The accession number HB 8832 was assigned after successful viability testing.

A. Materials and Methods All parts and percentages are by weight, and degrees are Celsius unless otherwise stated.

Heavy Chain Genomic Library Construction To isolate the heavy chain variable region gene from the 7E3 hybridoma, a size-selected genomic library was constructed using the phage lambda vector gtlO. Southern analysis of EcoRI digested 7E3 DNA using a Ju probe revealed a single 3.5 kb band corresponding to a rearranged heavy chain locus. It was likely that this fragment contained the 7E3 heavy chain variable region gene. High molecular weight DNA was isolated from 7E3 hybridoma cells 15 and digested to completion with restriction endonuclease EcoRI. The DNA was then fractionated on a 0.7% agarose gel and the DNA of size range 3-4 kb was isolated directly from the gel. After phenol/chloroform extraction and Sephadex gel filtration, the 3-4 kb fragments were ligated with 20 lambda gtlO arms (Promega Biotech, Inc.) and packaged into phage particles in vitro using Packagene from Promega Biotech. This library was screened directly at a density of approximately 30,000 plaques per 150 mm petri dish using a 2 P-labeled JH probe. Plaque hybridizations were carried out in 5X SSC, 50% formamide, 2X Denhardt's reagent, 200 Ag/ml denatured salmon sperm DNA at 42 degrees C for 18-20 hours. Final washes were in 0.5X SSC, 0.1% SDS at degrees. Positive clones were identified after autoradiography.

Light Chain Genomic Library Construction To isolate the variable region gene for the 7E3 light chain, a genomic library was constructed in the lambda vector EMBL-3. High molecular weight DNA was partially digested with restriction endonuclease Sau3A and size-fractionated on a 10-40% sucrose density gradient.

DNA fragments of 18-23 kb were ligated with EMBL-3 arms and packaged into phage particles in vitro using Packagene.

This library was screened at a density of 30,000 plaques per 150 mm plate using a J. probe. Hybridization and wash conditions were identical to those used for the heavy chain library.

Probes The mouse heavy chain JH probe is a 2 kb BamHI/EcoRI 15 fragment containing both J3 and J4 segments. The mouse light chain J. probe is a 2.7 kb HindIII fragment containing all five J. segments. 32 P-labeled probes were prepared by nick translation using a kit obtained from Amersham, Inc. Free nucleotides were removed by 20 centrifugation through a Sephadex G-50 column. The specific activities of the probes were approximately 109' cpm/Ag.

S

*ft %too-* Northern Analysis pg total cellular RNA was subjected to electrophoresis on 1% agarose/formaldehyde gels (Maniatis, et al, Molecular Cloning) and transferred to nitrocellulose. Blots were hybridized with nick translated DNA probes in 50% formamide, 2X Denhardt's solution, SSC, and 200 Ag/ml denatured salmon sperm DNA at 42 degrees for 10 hours. Final wash conditions were 0.5X SSC, 0.1% SDS at 65 degrees.

-27- DNA Transfection usina ElectroDoration Plasmid DNA to be transfected was purified by centrifuging to equilibrium in ethidium bromide/cesium chloride gradients two times. 10-50 pg of plasmid DNA was added to 8 x 10' SP2/0 cells in PBS on ice and the mixture placed in a Biorad electroporation apparatus.

Electroporation was at 200 volts and the cells were plated out in 96 well microtiter plates. Appropriate drug selection was applied after 48 hours and drug resistant 10 colonies were identified after 1-2 weeks.

Ouantitation of Antibody production Tissue culture supernatant was analyzed for IgG protein content by particle concentration fluorescence immunoassay (Jolley, M.E. et Al., (1984) J. Immunol. Meth.

15 67:21) using standard curves generated with purified IgG.

Concentration of chimeric 7E3 antibody with human constant regions was determined using goat antihuman IgG Fc antibody-coated polystyrene beads and fluorescein conjugated goat anti-human IgG Fc antibody. The assay was carried out with an automated instrument (Pandex Laboratories, Inc.).

Purification of Platelet-sDecific Chimeric IaG4 Antibody Tissue culture supernatant was concentrated with a Diaflo YM100 ultrafiltration membrane (Amicon), and loaded onto a protein A-sepharose column. The chimeric antibody was eluted from the protein A column with a sodium citrate pH gradient from pH 6.5 to pH 3.5. The purified antibody was concentrated using a Diaflo YM100 ultrafiltration membrane. Antibody concentration was measured by determining the absorbance at 280 nm.

-28- Binding inhibition Assay Purified antibody (either murine 7E3 or chimeric 7E3) was used to compete with radioiodinated 7E3 antibody for binding to human platelets. Platelet-rich plasma (PEP) was prepared by centrifugation of citrated whole human blood at 1875 rpm for 3.5 minutes. 2 5 -labeled 7E3 antibody (150,000 cpu) was added to the appropriate dilution of the purified test antibody and the reaction was initiated by the addition of 150 Al PEP. Incubation was for 1-2 hours at room temperature and the platelets with bound antibody ***.were separated from free antibody by centrifugation through sucrose at 12,000 g for 4 minutes in a 0.4 ml microfuge tube. The tube tip containing the platelet/antibody pellet was cut off and counted in a gamma counter. The competition for binding to platelets between iodinated 7E3 and chimeric 7E3 was compared to the competition between iodinated 7E3 and unlabeled 7E3 IgG.

Inhibition of -Platelet Aggregation *fot Purified 7E3 or chimeric 7E3 antibody was added to citrated whole human blood and incubated at 37 degrees for minutes. The rate of platelet aggregation was measured after activation with collagen or ADP using a whole blood aggregometer (Chronolog Corp.).

Cloning of the Platelet-specific Variable Gene Regrions Several positive clones were isolated from the heavy and light chain libraries after screening and approximately one million plaques using the JH and J. probes, respectively. Following at least three rounds of plaque purification, bacteriophage DNA was isolated for each positive clone, digested with either EcoRI (heavy chain -29clones) or HindIII (light chain clones) and fractionated on 1% agarose gels.

The DNA was transferred to nitrocellulose and the blots were hybridized with JH (heavy chain) or J.

nP-labeled DNA probes. For the heavy chain, 2 clones were obtained that contained 3.5 kb EcoRI DNA fragments that hybridized to the JH probe. Two size classes of HindIII fragments of 3.0 and 6.0 kb were identified with the J, probe.

Cloned DNA corresponding to the authentic heavy and light chain variable regions from the 7E3 hybridoma should hybridize to mRNA isolated from the hybridoma.

Non-functional DNA rearrangements at either the heavy or light chain loci should not be expressed. Figure 1 shows a 15 Northern analysis that demonstrates that the 3.5 kb EcoRI putative heavy chain fragment and the 6.0 kb HindIII putative light chain fragment each hybridizes to the appropriate size mRNA from the 7E3 hybridoma. The subcloned fragments were labeled with "P by nick translation and hybridized to northern blots containing total RNA derived from SP2/0 (the fusion partner of the 7E3 hybridoma) or from 7E3, as indicated in Figure 1. The kb EcoRI heavy chain fragment hybridized with a 2 kb mRNA in 7E3 RNA but not in SP2/0 RNA. Similarly, the 6.0 kb 25 light chain HindIII fragment hybridized with a 1250 bp mRNA in 7E3 RNA but not in SP2/0 RNA. These are the correct sizes for heavy and light chain mRNAs, respectively.

Because the cloned DNA fragments contain sequences expressed in the 7E3 hybridoma, these data suggest that the clones contain the correct variable region sequences from the 7E3 hybridoma. The final functional test, however, is the demonstration that these sequences, when combined with appropriate constant region sequences, are capable of directing the synthesis of an antibody with a specificity and affinity similar to that of the murine 7E3 antibody.

Vectors and Expression Systems The putative light and heavy chain V genes cloned from the 7E3 hybridoma were joined to human K and G4 constant region genes in expression vectors described previously (Sun, L. at al., Proc. Natl. Acad. Sci. USA 84:214-218 (1987)). The 17-1A V, HindIII fragment of pSV184AHneol7-1AVhC, was replaced with the 6.0 kb HindIII fragment corresponding to the putative light chain variable region gene from 7E3. Similarly, the 17-1A V, EcoRI fragment of pSV2AHgptl7-1AVH-hCo was replaced with the kb EcoRI fragment corresponding to the putative heavy chain o 'V region gene from 7E3. The structures of the resulting 15 plasmids, designated p7E3VhC,. and p7E3VNhCo, are shown in SFigures 2A-2B.

*To express the chimeric heavy and light chain genes, the two plasmids were cotransfected into the nonproducing mouse myeloma cell line SP2/0. The light chain plasmid S. 20 confers resistance to G418 and the heavy chain plasmid confers resistance to mycophenolic acid, thus allowing a double selection to be used to obtain clones carrying and expressing genes from each plasmid. Colonies resistant to G418 and mycophenolic acid were expanded to stable cell lines and maintained in the presence of both drugs. Tissue culture supernatant from these cell lines was tested for antibody using a particle concentration fluorescence immunoassay with polystyrene beads coated with goat anti-human IgG Fc antibody and the same antibody labeled with fluorescein. Out of the first 10 lines checked, one (designated c-7E3F6) that produced approximately 2pg/ml was selected for further study.

-31- Platelet Binding Activity Assay After purification of c-7E3F6 antibody using a protein A-sepharose column, the antibody was concentrated and compared to murine 7E3 IgG in the platelet binding activity assay. Figure 3 shows that murine 7E3 and c-7E3F6 (the putative chimeric antibody) compete with radiolabeled 7E3 for platelet binding to the same extent; the binding curves are superimposable indicating that the binding characteristics of murine and chimeric 7E3 are identical by this criterion.

Inhibition of Platelet Aggregation by c-7E3F6 Purified c-7E3F6 was compared to murine 7E3 in a functional assay that measures the ability of the test antibody to inhibit aggregation of human platelets. The results of such an assay are shown in Figure 4 and demonstrate that both antibodies inhibit collagen-induced platelet aggregation to the same extent at the same antibody concentration. c-7E3F6 also inhibits ADP-induced platelet aggregation to a similar extent.

20 The results of the platelet binding assay and the inhibition of platelet aggregation assay demonstrate that: the correct variable region genes were indeed cloned from the 7E3 hybridoma; and the substitution of the human constant regions for the murine constant regions has no effect on the binding or functional characteristics of the 7E3 variable regions as measured by these assays.

FibrinoQen-coated Bead Assay The chimeric c-7E3F6 antibody was found positive in a qualitative, functional assay that measures the ability of an antibody to inhibit the agglutination between platelets and fibrinogen-coated beads. Coller, B. et a (1983) J.

Clin. Invest. 73:325-338.

-32- ExamPle 2 Production of Chimeric IaGl and IaG3 The DNA segment encoding the variable region of the heavy chain from the murine 7E3 antibody was linked to the human 71 and 73 constant regions present on the expression vectors pSV2AHgptl7-1AVH-hCol and pSV2AHgpt17-1AVH-hCm (Sun at al., Proc. Natl. Acad. Sci. USA 84:214-218 (1987)), by replacing the 17-1A heavy or light chain variable region fragments with the corresponding 7E3 variable region tragments. The resulting chimeric heavy chain genes were cotransfected with the chimeric light chain gene into cells to generate stable cell lines secreting yl,K, and S3,K antibodies.

Example 3 Initial Studies of Use of Chimeric 7E3 Fab in Humans Preparation of Chimeric 7E3 Fab Fragments The Fab fragment of chimeric 7E3 (c7E3) was produced by enzymatic digestion of purified chimeric 7E3 IgG (gamma 1 heavy chain, kappa light chain) with the proteolytic enzyme papain. The Fab fragment was isolated by a series of purification steps designed to yield a product which was free of other digestion fragments and other contaminating components protein, nucleic acid, viruses). The final product was prepared as a sterile, non-pyrogenic solution containing 2 mg of monoclonal chimeric 7E3 Fab per ml of 0.15 M sodium chloride, 0.01 M sodium phosphate, pH7.2. In certain preparations, polysorbate 80 was included at a final concentration of 0.001% Prior to use, the product was filtered through a 0.22 micron low protein binding filter. The product was stored at 2-8 °C.

-33- Pharmacokinetics: Plasma Clearance of c7E3 Fab in Humans The plasma clearance of chimeric 7E3 (c7E3) Fab fragment was studied in three patients with stable coronary disease. Following a 0.20-mg/kg dose of c7E3 Fab administered intravenously as a five minute infusion, blood samples were taken at various times from two minutes to 72 hours. It was anticipated that a certain portion of the antibody would exist in an unbound state in plasma. To quantify this unbound antibody component, it was necessary to rapidly separate the plasma from the platelets to prevent further binding x yivo. The plasma concentration of free c7E3 Fab was measured by solid-phase enzyme immunoassay (EIA). The assay employed affinity isolated anti-murine 7E3 IgG purified from rabbit antisera for 15 solid-phase capture and a detection system based on a biotinylated derivative of the same rabbit anti-7E3 antibody preparation. The results are presented in Table 1.

*mo• -34- PLASMA CONCENTRATION OF c7E3 Fab IN PATIENTS TREATED WITH A 0.20-MG/KG DOSE c7E3 Fab (pg/mL) imet B Pre-dose ND ND ND 2 min NA 2.554 2.312 min 1.149 1.873 1.411 min 0.713 1.331 1.111 15 min 0.610 0.916 0.852 min 0.499 0.985 0.756 min 0.463 0.815 0.515 45 min 0.340 0.704 0.405 1 hr 0.308 0.436 0.195 15 2 hr 0.288 0.262 0.149 6 hr 0.203 0.095 0.105 12 hr 0.112 0.072 0.064 t 24 hr 0.065 0.058 0.053 48 hr 0.055 0.147 0.174 72 hr 0.196 ND 0.076 *Interval between end of infusion and blood drawing. Note that platelets were in contact with the plasma for an S .additional 2 minutes after the blood was drawn for the time required to separate the plasma by centrifugation).

ND Not detected/below the detectable level of the assay S"(0.025 pg/mL).

NA Not available.

to.

If the entire injected dose of 7E3 were detected as free antibody in plasma, the theoretical maximum antibody concentration would be approximately 5.0 ug/mL (0.20 mg/kg divided by 40 mL of plasma/kg). However, this theoretical maximum concentration would never be attained because of the large component of injected antibody which binds to platelets. In fact, at the earliest measurement time (2 minutes), the average plasma concentration of c7E3 Fab was 2.43 pg/mL; this value was the observed maximum plasma concentration The data obtained at *subsequent post-injection times show a rapid initial decrease in the plasma concentration of c7E3 Fab. By the 1-hour and 24-hour measurements, the administered antibody remaining in the plasma was less than 0.5 gg/mL and 15 0.1 pg/mL, respectively. A plot of the plasma antibody concentration (ng/mL) versus time data, presented in Figure graphically demonstrates the rapid initial clearance of c7E3 Fab from the plasma in all three patients.

A preliminary analysis of the pharmacokinetic characteristics of c7E3 Fab fragment was undertaken.

Several models, including several mixed (random and fixed effects) linear models as well as standard two-compartmental and non-compartmental models, were used to fit the plasma concentration data. The free plasma antibody data did not adequately fit standard pharmacokinetic models. As the site of action for c7E3 Fab is a receptor located on platelets, it is not unexpected that the plasma concentration of free antibody would not be related to its concentration at its site of action in any simple way. The rapid initial clearance of c7E3 Fab from the plasma reflects, in part, the rapid antibody binding to platelet GPIIb/IIIa receptors. Of the models examined, the random effects linear model was shown to best fit the plasma concentration data. Using this model, preliminary -36values for the pharmacokinetic parameters, Cl,, V4, and t,, were determined and are presented in Table 2.

TABLE 2 PHARMACOKINETIC VALUES FOR c7E3 Fab* Parameter Value C1, (proportion/hr) 15.6 Vd 6.8 t4 (hr) 0.1 (6 min) 10 A random effects linear model was used to fit the data.

Ci, Plasma clearance is defined as the rate of decrease in plasma concentration divided by the concentration and is computed as a rate per hour, if the rate at a given time 15 continued for an hour, the computed rate would be the proportion of drug removed in that hour.

Vd Volume of distribution is defined as the dose administered divided by the measured plasma concentration multiplied by the plasma volume. A 3-L plasma volume typical for a 70-kg person was used in the calculations.

t- Elimination half-life.

Urinary Excretion in Humans Urine samples were collected from three patients with stable coronary disease who were treated intravenously with 0.25-mg/kg of c7E3 Fab (plasma clearance data for these same three patients are discussed above). Total urine output was collected for the following post-injection time periods: 0 to 2 hours, 2 to 6 hours, 6 to 12 hours, and 12 to 24 hours. In addition, a sample of predose urine was also collected. Representative samples of the collected -37urine samples were analyzed for free 7E3 Fab using a slight modification of the EIA described above. In all cases, no c7E3 Fab was detected in the urine.

Preclinical Toxicoloav Preclinical toxicology studies have been performed in 18 monkeys (Cyonomolgus and Rhesus), using chimeric 7E3 Fab. Bolus doses of up to 0.6 mg/kg, followed by infusion of up to 0.8 pg/kg/min for 96 hours were administered (includes studies with heparin, aspirin and recombinant tissue plasminogen activator). In all monkeys, at all doses, in all combinations, 7E3 was safe and o well-tolerated, with no significant bleeding complications *or other adverse events.

Dose Escalation of Chimeric 7E3 Fab in Stable Angina Patients A dose escalation study was conducted enrolling 52 stable angina patients (males from 43 to 75 years old) who were off anti-platelet therapy for more than 10 days. A variety of dosing regimens were administered. Patients S 20 received either single intravenous bolus injections of 0.15 to 0.30 mg/kg of chimeric 7E3 Fab (20 patients) or a bolus loading followed by continuous intravenous infusions g/minute) from 12 to 96 hours in duration (32 patients).

Platelet GPIIb/IIIa receptor blockade, platelet aggregation in response to 20 AM ADP (agonist), and bleeding times were determined 2 hours after administration of a bolus dose of c7E3 Fab (0.15-0.30 mg/kg). Receptor blockade and platelet aggregation in response to agonist were determined essentially as described (Gold, H.K. et al., J. Clin. Invest. 86:651-659 (1990)). Bleeding times were determined by the Simplate method. With increasing doses there was a progressive increase in receptor blockade, as indicated by the percent of receptors blocked (determined from the availability of receptor binding sites). The increase in receptor blockade was paralleled by inhibition of platelet aggregation (measured as a percent of the pre-dose value or baseline), and by an increase in bleeding time.

The peak effect in terms of all three parameters was observed at 0.25 mg/kg. This dose corresponds to a plasma concentration of 5 pg/ml--the concentration at which peak inhibition was seen in a platelet-rich plasma from a normal subject which had been incubated for 15 minutes in an aggregometer cuvette in the presence of increasing concentrations of chimeric 7E3 Fab. (The extent of aggregation of the plasma of the normal subject was 15 measured by the percent of light transmitted through the cuvette. Prior to the addition of an agonist, the plasma was relatively opaque and the percent of light transmitted was set at zero. When the agonist ADP was added to a control sample without antibody, the light transmission 20 progressively increased as aggregation progressed.

However, when c7E3 Fab is present, a dose-dependent block of aggregation was observed with complete inhibition at 0• sg/ml c7E3 Fab.) The duration of action in terms of receptor blockade, inhibition of platelet aggregation, and bleeding time was determined. Peak effects on receptor blockade, platelet aggregation, and bleeding time were seen at 2 hours, with gradual recovery over time. Bleeding times returned to near normal values by 6-12 hours.

Because peak receptor blockade and functional inhibition were achieved with 0.25 mg/kg, the duration of platelet inhibition by continuous infusions following this loading dose were assayed to determine if the duration of platelet inhibition could be prolonged. The degree of receptor blockade, inhibition of platelet aggregation, and -39prolongation of bleeding time were maintained for the duration of continuous infusion in five patients who received a 10 pg/minute continuous infusion of chineric 7E3 Fab for 72 hours following the d.25 mg/kg loading dose.

Recovery started as soon as the infusion was discontinued.

Similar results were seen with 12, 24, 48 and 96 hour infusions.

None of the patients experienced a hypersensitivity reaction. There were no significant treatment related trends in hematology or chemistry laboratory values. Nor were there any major bleeding events. Insignificant bleeding events were rare and included transient mild nose bleed and mild gum oozing in patients with periodontal 1: :disease. The results of the trial indicated that chimeric 15 7E3 Fab can be administered to patients using dosing regimens that produce profound inhibition of platelet function for periods as long as several days.

Immunogenicitv Results In trials with murine 7E3 F(ab') 2 and Fab (150 patients), immune responses detected using a sensitive enzyme-linked immunoassay system occurred in 16% (24/150) of the patients. All reactions were of low titer, typically in the range of 1:50 to 1:200. The treatment group included normal volunteers treated with 0.01 0.25 mg/kg murine 7E3 unstable angina patients treated with 0.05 0.20 mg/kg murine 7E3 F(ab') 2 and PTCA patients treated with 0.1 mg/kg murine 7E3 or 0.15 0.35 mg/kg murine Fab, as well as stable angina patients treated with a single bolus intravenous injection of 0.10 0.30 mg/kg of murine 7E3 Fab, a single bolus dose of either 0.25 or 0.30 mg/kg followed by continuous infusion for 12-36 hours (0.15 gg/kg/min or lOg/min) of murine Fab, or with two injections of murine Fab separated by six hours (a single bolus of 0.2 mg/kg 0.30 mg/kg followed by a bolus of 0.05 ag/kg).

Immunogenicity was notably reduced with the human-mouse chimeric 7E3 Fab. None of the 52 patients having stable angina enrolled in the dose escalation study and treated with chimeric 7E3 Fab (see above) showed an immune response to treatment as measured by a similar assay adapted to the chimeric Fab.

Reversibility of Anti-platelet Activity Chimeric 7E3 Fab x) has a slow off rate from platelets and free plasma chimeric 7E3 Fab clears from circulation rapidly (see above). Thus, the antiplatelet effects of chimeric 7E3 are readily reversible by administration of random donor platelets. This reversal or *15 antidote effect by transfusion of platelets has been demonstrated in 2 patients who had received either murine Fab or chimeric Fab and who received random donor platelets during a time when they had nearly complete inhibition of platlet aggregation. Restoration of platelet function was determined by measuring bleeding times. This property is useful in situtations where a bleeding event necessitates restoration of platelet function in a patient.

*"Example 4 Use of Chimeric 7E3 Antibody in the Prevention of Thrombotic Complications of Elective Coronary Ancioplasty Percutaneous transluminal coronary angioplasty (PTCA), by balloon or coronary atherectomy, for example, is an effective method of enlarging the lumen of stenosed coronary arteries. In this procedure, there is an inherent risk of acute coronary occulusion during and after angioplasty. The reported rate of coronary occlusion -41varies from approximately of elective angioplasty cases (Detre, K.M. et al., Circulation 82:739-750 (1991)), and is the major cause of in-hospital morbidity and mortality. In high risk patients, the incidence of major cardiac events caused by thrombosis is between 10-20%.

Acute coronary occlusion during or immediately after coronary angioplasty appears to be caused by the combination of deep arterial wall injury with resultant partially occlusive "intimal flaps" with or without superimposed thrombus formation, or thrombus formation alone at a site of vessel wall injury. In animal models, Teocclusion after successful thrombolysis is preceded by periods of cyclical reductions and restorations in coronary blood flow termed "cyclic flow variations" (CFVs). These CFVs are almost entirely a platelet-mediated phenomenon, o and are due to repetitive accumulation and dislodgement of platelet aggregates at sites of coronary stenosis and endothelial injury. Cyclic flow variations after coronary angioplasty have been described in humans. Chimeric 7E3 antibody can be used to inhibit platelet function during angioplasty thereby preventing platelet aggregation and thrombosis. Chimeric 7E3 antibody is particularly useful in patients at high risk of thrombotic occlusion. These patients can be identified on the basis of anatomic 25 angiographically defined characteristics of a lesion at a site of stenosis) or clinical risk factors myocardial infarction, unstable angina, diabetes, women years or older), which predispose to acute coronary thrombosis and produce the clinical syndromes of acute myocardial infarction, unstable angina or abrupt closure.

Chimeric Anti-platelet Antibody in Elective PTCA The trial was conducted in two stages. The primary objective of the first stage was to determine the safety and optimal dose of single dose chimeric 7E3 Fab in patients undergoing elective percutaneous transluminal coronary angioplasy (PTCA). Stage II was conducted to evaluate the safety and preliminary efficacy of chimeric 7E3 (c7E3) when administered by bolus infusion followed by various continuous infusion durations. The Stage II study comprised elective coronary angioplasty patients who were at risk for ischemic cardiac complications. High risk patients included those with unstable angina or stable coronary disease with Type B or C lesion specific characteristics. Table 3 lists the definitional criteria for high risk patients, and Table 4 lists the.

,angiographically defined lesion-specific characteristics.

Preliminary efficacy was measured as inhibition of platelet function and prevention of thrombotic complications. Men, 15 between 18 and 76 years of age, and women not of child bearing potential, between 18 and 76 years of age, were eligible to enroll in both stages of the trial.

*9 *o 9 9 43 ThliLL2 ENROLLM CRITERIA FOR PATIENTS AT HIGH RISK Fj ISQHM CCOMPICAIQNS Hoderately high risk: 1) Unstable angina with no lesion-specific characteristic defined.

2) A stenosis with a single Type B lesion-specific 1W characteristic.

SW

0 0 Highest risk: 1) A stenosis with 2: two Type B lesion-specific 00* characteristics.

2) Unstable angina with a stenosis with at least one Type B lesion-specific characteristic.

3) Diabetes mellitus with a stenosis with at least 15 one Type B lesion-specific characteristic.

4) Women 2: 65 years of age with a stenosis with at least one Type B lesion-specific characteristic.

A stenosis with at least one Type C lesion-specific characteristic.

-44- TABLE 4

LESION-SPECIFIC

Type A Lesions (high success,, 85%; low risk) -Discrete 10 mm length) -Concentric -Readily accessible -Nonangulated segment, <45* -Smooth contour -Little or no calcification -Less than totally occlusive -Not ostial in location -No major branch involvement -Absence of thrombus .1 Type B Lesions (moderate success, 60-85%; moderate risk) -Tubular (10- 20 mm length) -Moderate to heavy calcification -Eccentric -Total. occlusions <3 months old -Moderate tortuosity of -Ostial in location proximal segment -Moderately angulated -Bifurcation lesions requiring segment, >45% <909 double guide wires -Irregular contour -Some thrombus present Type C Lesions (low success, 60%; high risk) -Dif fuse -Excessive tortuosity of proximal segment -Extremely angulated segments 900 -Total occlusion 3 months old -Inability to protect major side branches -Degenerated vein grafts with friable lesions In Stage I, patients were enrolled in groups receiving a single bolus intravenous injection of chimeric 7E3 a) Fab fragment (prepared and formulated as described in Example A total of 15 patients, (3 women and 12 men) were treated. The median age of patients was 62 years (range 46 years to 76 years). A demographic profile is listed in Tables 5A and 5B for all single dose patients and for patients within the individual dose groups.

Five patients (n 5) each received single doses of 0.15 mg/kg, 0.20 mg/kg or 0.25 mg/kg of c7E3 Fab within about 30 minutes prior to elective PTCA in a dose-escalation protocol. All patients were treated with aspirin (standard dose) and fully anticoagulated with 15 heparin (standard dose) at the time of the procedure.

Although the PTCA procedures were classified as elective for Stage I patients, six of the 15 patients had unstable rest angina. The coronary location of the dilatations is summarized in Table 6 (bottom). Seven of 20 the 15 Stage I patients underwent PTCA of one lesion in a single vessel, 6 underwent multi-lesion PTCA in a single vessel, and 2 patients had multi-vessel PTCA performed (Table 6).

The efficacy criteria for obtaining the optimal single dose of c7E3 were prospectively defined as the minimum dose that achieved median values of the following at 2 hours post-infusion: prolongation of bleeding time of at least 20 minutes; blockade of GPIIb/IIIa receptors such that there were greater than 80% of baseline receptor sites blocked; and an inhibition of platelet aggregation in response to 20 pM ADP to s 20% of baseline.

TABLE SA CHIMERIC 7E3 ANTI-PLATELET

ANTIBODY

Patentclassification of Age, Weight, Height, Sex and Race Continuous, Control Single Dose TOTAL PATIENTS 32 9

AGE

Mean 57.4 54.2 60.1 Median 57.0 56.0 62.0 Minimum 38 37 4 maximum 76 74 46 Std. Dev. 9.7 10.5 976 WEIGHT (kg) Mean 82.8 88.2 85.5 Median 84.8 84.2 84.0 minimum 42.3 67.3 70.5 Maximum 113.0 122.7 107.0 Std. Dev. 16.6 19.0 11.1 J*EIGHT (cm) Mean 171.2 176.7 173.6.

Median 172.8 177o8 175.2 Minimum 152.4 157.5 160.0 Maximum 185.0 1B5.418.

Std. Dev. 7.9 8.8 18

SEX

Female 8 1 3 Male 24 8 12 RACE 0 01 White 26 8 1 Black 5 1 2 Asian 0 0 1 Hispanic 1 0 0 CHIMERIC 7E3 ANI-Ptq Patient Classification of &ge. *eight, HeiSht. Sexand Race 0. 15 mg/kg 0. 20 mg/kg 0. 25 mg/kg 5 0.25 mg/kg 40.25 mg/kg 0.25 mg/kg 7E3* 713* 7E3* 10 mcg/min 10 ucg/mg 10 Mcg/min for 6 hr.s for 12 hra for 24 hre Placeb 11 11 10 9 7E3 7E3 TUTAL, 'AT1za1'Jt 3 0

AGE

Mean Median Minimum maximum Std. Dev.

WEIGHT (kg) Mean Median Minimum maximum Std. Dev.

HEIGHT (cm) Mean Median Minimum Maximum Std. Dev.

SEX

Female Male

RACE

60.6 63.0 47 73 10.5 83.8 76.3 74.5 107.0 13.8 170.8 173.0 160.0 177.8 7.0 3 2 0 3 2 0 0 57.2 56.0 46 68 8.3 80.5 78.0 70.5 95.0 9.5 171.9 175.2 160.8 182.9 8.8 0 5 0 4 0 1 0 62.6 65.0 51 76 11.3 92.2 96.0 82.0 99.1 7.7 178.1 177.8 169.0 188.0 7.0 0 5 1 4 0 0 0 59.1 60.0 40 .76 10.5 85.8 89.0 60.4 113.0 17.8 169.3 167.6 157.5 177.8 6.4 3 8 0 8 3 0 0 55.0 53.0 42 73 8.9 83.8 85.0 42.3 111.8 19.6 170.9 172.*7 152.4 185.0 10.*8 4 7 0 9 2 0 0 58.1 58.5 38 75 10.3 78.3 79.9 62.3 95.0 10.*7 173.7 175.2 165.1 180.3 5.4 54.2 58.0 37 74 10.5 88.2 84.2 67.3 122.7 19.0 176.7 177.8 157.5 185.*4 8.8 White Black Asian Hispanic 0 S 0 0 *0 0 0 S. 00 *0* ~0* *0 00 0 0 0* 0 0 ~0 0*S** 0 0 0***e *0 00 TABLE 6 A ANGIOGRAPHIC CHARACTERISTICS Number of Lesions Attempted: Control c7E3 Fab Stage I Single vessel/single lesion Single vessel/multilesion Multivessel/single lesion Multivessel /multi lesion Unknown Location of Attempted Lesions:

RCA

LCX

LAD

c7E3 Fab Stage II 21 7 1 2 1 *In patients with multivessel disease, both vessels are counted..

RCA

LCX

LAD

Right coronary artery Left circumflex coronary artery Left anterior descending coronary artery -49- Stage II In Stage II, patients were treated with a 0.25 mg/kg bolus dose followed by a continuous infusion of 10 Pg/min of c7E3 Fab for 6, 12, or 24 hours. A total of 32 patients (8 women and 24 men) were entered into the treatment group of Stage II of the study. The median age of the c7E3 Fab-treated patients was 57 years (range 38-76). Nine control patients (1 woman, 8 men) were entered. The median age of control patients was 56 years (range 37-74).

Control patients were high risk patients as defined above, who did not receive c7E3, but were monitored and followed in the same fashion as treated patients. A demographic profile for all Stage II patients and for patients within the individual dose groups is listed in Tables 5A and SB.

15 Treatment with c7E3 Fab was initiated 30 minutes prior :.to balloon inflation for PTCA. Aspirin and heparin were given as clinically indicated, with the recommendation that following angioplasty heparin be given at the rate of 800 units per hour. Eleven patients each were entered into the 6 and 12 hour groups, and ten patients were entered into the 24 hour group.

Of the 32 c7E3 Fab-treated patients, 21 patients underwent PTCA of one lesion in a single vessel, 7 patients underwent multi-lesion PTCA in a single vessel, and 3 25 patients had multi-vessel PTCA performed (Table The type of PTCA was not specified in one patient. The coronary location of the dilatations for stage II patients is summarized in Table 6. Of the 9 control patients, 8 underwent PTCA of one lesion in a single vessel, and one patient had multi-vessel PTCA. The 32 c7E3 Fab-treated patients and the 9 control patients had clinical or angiographic characteristics that would classify them as high risk for ischemic cardiac complications of PTCA. Two c7E3 Fab-treated patients and one control patient had unspecified risk factors. The remaining 30 c7E3 Fab-treated patients and 8 control patients had at least one identifiable clinical feature or angiographic characteristic placing then at increased risk of ischeuic complication, and most had more than one risk factor.

Table 7 summarizes these risk factors for the control and c7E3 treatment groups, and individual listings of risk factors by patient within each dose group are presented in Tables SA through 8D.

-51- TABLE 7 HIGH RISK CHARACTERISTICS Stage II Risk Factors Control o7E3 Treatment (n 9) 32) One Type B characteristic 5' 4 Two or more Type B characteristics 12 73 One Type C characteristic 0 2 Unstable angina with no lesion characteristics identified 1 1 Unstable angina Type B characteristic 0 15 Unstable angina 2 Type B characteristics 1 7' Unstable angina Type C characteristic 0 1 pe Unspecified risk characteristic 1 2

S

1 Patient 04-006 had diabetes 2 This patient (04-007) had diabetes 3 Patients 03-001 and 02-007 had diabetes S •25 4 This patient (04-004) had the following additional risk factors: female, age and diabetes Patients 03-003 and 05-001 had diabetes 6 Patient 01-018 had the following additional risk factors: female and age 65. Patient 03-002 had diabetes.

-52- TABLE 8A PREDISPOSING HIGH RISK CHARACTERISTICS Control Patients Patient Number 01-023 Ot-024 04-006 Type of Risk(s) 1 Segment 2 1. Unstable angina 1. One Type B charasteristic 3 Diabetes One Type B characteristic (eccentric) Diabetes Two Type B characteristics (eccentric; moderately angulated segment; 45°, 04-007

LAD

LAD

LCX

RCA

LCX

LCX

RCA

OM

04-008 04-009 01-021 01-022 1. One Type B characteristic (eccentric) 1. Our Type B characteristic (eccentric) 1. One Type B characteristic 3 1. Unstable angina 2. Two Type B characteristics (tubular; irregular contour)

S

03-005 1. Unspecified risk characteristic

RCA

SPotential characteristics listed 2RCA right coronary artery LCX left circumflex coronary LAD left anterior descending OM obtuse marginal branches SCharacteristic not designated in Tables 3 and 4 artery coronary artery of LCX -53- TABLE 8B PREDISPOSING HIGH RISK CHARACTERISTICS 6-Hour Continuous Infusions Patient Number Type of Risk(s) 1 Segment 2 04-001 1. Unstable angina

RCA

2. One Type B characteristic (eccentric) 06-001 1. Unstable angina

LCX

2. One Type B charasteristic (thrombus) 06-002 1. One Type B characteristic

LAD

(tubular [10 to 20mm lesion]) 01-014 1. Unstable angina

LAD

2. One Type C characteristic (Diffuse 2 cm length) 01-013 1. Unstable angina

RCA

2. Two Type B characteristics S* (eccentric, some thrombus) 01-015 1. Two or more Type B

LAD,

characteristics

LADD

01-017 1. One Type B characteristic

LCX

(eccentric) 02-005 1. Unstable angina

LAD

2. Three Type B characteristics (tubular, (10 to 20mm length); irregular contour; ostial in location) 1 2 Potential characteristics listed in Tables 3 and 4 RCA right coronary artery LCX left circumflex coronary artery LAD left anterior descending coronary artery 3 LADD diagonal branch of LAD Characteristic not designated TABLE 8B (cont') PREDISPOSING HIGH RISK CHARACTERISTICS 6-Hour Continuous Infusions (cont.) Patient 2 Number Type of Risk(s) Sgqment 03-001 1. Diabetes LAD 2. Four Type B characteristics (eccentric; moderate angulation, S. 45°, irregular contour; moderate to heavy calcification) 01-012 1. One Type C characteristic LAD (Diffuse >2 cm length) 01-016 1. One Type B characteristic RCA (some thrombus) OMn 1 Potential characteristics listed in Tables 3 and 4 SRCA right coronary artery LCX left circumflex coronary artery LAD left anterior descending coronary artery OM obtuse marginal branches of LCX n TABLE 8C PREDISPOSING HIGH RISK CHARACTERISTICS 12-Hour Continuous Infusion Patient Number Type of Risk(s) 1 Segment 2 01-018 1. Unstable angina

OM

2. Female over 3. Two or more Type B Scharacteristics 01-019 1. Unstable angina Circumflex 2. Two or more Type B characteristics 02-006 1. Unstable angina

RCA

2. One Type B characteristic (total occlusion <3 months old) 02-007 1. Diabetes

LCX

2. Five Type B characteristics

OM

1

LADD

(eccentric, moderate tortuosity S:-i of proximal segment; moderately angulated segment, >450, bifurcation lesions requiring double guidewires; total occlusions <3 mo) 03-002 1. Unstable angina

LAD

2. Diabetes 3. Two Type B characteristics (moderate tortuosity segment; irregular contour) 03-003 1. Unstable angina

RCA

2. Diabetes 3. One Type B characteristic (irregular contour) 2 Potential characteristics listed in Tables 3 and 4 RCA =right coronary artery LCX left circumflex coronary artery LAD left anterior descending coronary artery

OM

n obtuse marginal branches of LCX TABLE 8C (cont') PREDISPOSING HIGH RISK CHARACTERISTICS 12-Hour Continuous Infusion (cont.) Patient Number Type of Risk(s) 1 Segment 2 05-001 1. Unstable angina LADD 2. Diabetes RCA 3. One Type B characteristic (eccentric) O' 0-002 1. One Type B characteristic RCA (tubular) 2. One Type C characteristic LADD (total occlusion >3 months) 05-003 1. Two Type B characteristics LAD (moderately angulated segment, >450, <900; some thrombus) 06-003 1. Unstable angina RCA 2. One Type B characteristic (some thrombus) S04-002 1. Unstable angina LCX 2. One Type B characteristic (tubular 10 to 20 mm) 1 1 Potential characteristics listed in Tables 3 and 4 RCA right coronary artery LCX left circumflex coronary artery LAD left anterior descending coronary artery LADD diagonal branch of LAD OM obtuse marginal branches of LCX Characteristic not designated -57- TABLE 8D PREDISPOSING HIGH RISK CHARACTERISTICS 24-Hour Continuous Infusion Patient Number 01-020 Type of Risk(s) 1 Segment 2 9 02-008 1. Two Type B characteristics (irregular contour, some thrombus) 1. Unstable angina 2. Type B characteristics a) 4 characteristics (tubular; eccentric; moderate tortuosity of proximal segment; irregular contour) b) 3 characteristics (eccentric; moderately angulated segment 1. Unstable angina 2. One Type B characteristic 3 1. Unstable angina 2. One Type C characteristic (diffuse cm length)) 1. Two Type B characteristics (eccentric; bifurcation with double guidewires) 1. Unstable angina 2. Diabetes 3. Female over

RCA

LCX

LCX

RCA

RCA

LAD

LAD

05-005 05-006 04-003 04-004 1 Potential characteristics listed RCA right coronary artery LCX left circumflex coronary LAD left anterior descending LADD diagonal branch of LAD 3 OM obtuse marginal branches Characteristic not designated in Tables 3 and 4 artery coronary artery of LCX TABLE 8D (cont') PREDISPOSING HIGH RISK CHARACTERISTICS 24-Hour Continuous Infusion Patient Number Type of Risk(s) 1 Segment 04-005 1. Two Type B characteristics

RCA

(eccentric, some thrombus) 02-009 1. Unstable angina 2. Type B characteristics a) 3 characteristics

LAD

S(irregular contour, some thrombus; bifurcation lesions requiring double guidewires) b) 1 characteristic

LAD

(bifurcation lesions requiring double guidewires) c) 2 characteristics

DB

(eccentric; bifurcation lesions requiring double guidewires) 05-004 1. One Type B characteristic

OM

(bifurcation requiring double guidewires) O2 o*o 2 Potential characteristics listed in Tables 3 and 4 RCA right coronary artery LCX left circumflex coronary artery LAD left anterior descending coronary artery DB OM obtuse marginal branches of LCX SCharacteristic not designated -59- Inhibition of Platelet Function (Staae I Results) To assess activity of the chimeric 7E3 Fab in inhibiting platelet function, GPIIb/IIIa receptor binding site availability (recorded as median percent GPIIb/IIIa blocked), median inhibition of agonist-induced platelet aggregation in response to 20 IMM ADP, and median bleeding times, were serially measured. Receptor blockade and platelet aggregation in response to agonist were determined essentially as described (Gold, H.K. e al., J. Clin.

Invest:, &:651-659 (1990)). For receptor blockade measurements, receptor availability was measured at time 0 and the number of receptors available were taken as 0% S: receptors blocked (baseline). Other time points are relative to the number of receptors available at baseline 15 or the pre-dose measurement. Bleeding times were determined by the simplate method.

Figures 6A-6C summarize the dose response 2 hours following a single bolus dose of chimeric 7E3 Fab, in terms of receptor blockade (Figure 6A), platelet aggregation 20 (Figure 6B), and bleeding time (Figure 6C). The solid *:"lines in Figures 6A-6C indicate the median values of the patients studied at each dose group. With increasing doses of c7E3 Fab there was a progressive increase in receptor blockade as shown in percent of receptors that are blocked (Figure 6A). The median number of receptors blocked at two hours was 53.8% for the 0.15 mg/kg, 80.2% for the 0.20 mg/kg, and 86.6% for the 0.25 mg/kg dose groups. The increase in receptor blockade was paralleled by inhibition of platelet aggregation, depicted as a percent of the pre-dose value (Figure 6B). Median platelet aggregation at 2 hours was 46.1%, 44.6%, and 17.9% of baseline for the 0.15 mg/kg, 0.20 mg/kg, and 0.25 mg/kg dose groups, respectively. Likewise, a dose-related prolongation of bleeding time was seen at 2 hours post-infusion (bleeding time measurements were truncated at 30 minutes; Figure 6C).

The median bleeding times were 26.0 minutes, 27.5 minutes, and 30 minutes for the 0.15 mg/kg, 0.20 mg/kg, and 0.25 mg/kg doses, respectively. Under the conditions used, and as measured by these assays, the optimal dose for anti-platelet activity was determined to be 0.25 ag/kg.

Figures 7A-7C show the duration of action following a single bolus dose of 0.25 mg/kg, the dose at which maximum platelet effects were seen. The lines indicate the median values from time zero (baseline) through 24 hours, as shown on the x-axis, in terms of receptor blockade in the top panel (Figure 7A), platelet aggregation in the middle panel Figure 7B), and bleeding time in the bottom panel (Figure 7t). Peak effects on recepor blockade, platelet S: aggregation, and bleeding time are seen at 2 hours, with 15 gradual recovery over time. Bleeding times return to near normal values by 12 hours. None of the patients experienced thrombocytopenia.

Inhibition of Platelet Function (Staae II Results) In Stage II, GPIIb/IIIa receptor and platelet 20 aggregation data were not obtained in all patients, and only two patients in the 24 hour infusion had these studies performed. Therefore, only the 6 and 12 hour data are summarized. In both the 6-hour and 12-hour infusion groups, median receptor blockade was maintained to greater than 80% of baseline through the duration of infusion.

Median 20 pM ADP-induced platelet aggregation for the 6 and 12-hour infusion groups was 13% and 15% of baseline at 2 hours, respectively, and in the 12 hour group remained below 25% for the duration of the infusion. The 2-hour median bleeding time for all three infusion durations was greater than 30 minutes. Figures 8A-8C review the results observed in patients who received a 10 Ag/minute continuous infusion of chimeric 7E3 Fab for 12 hours following the 0.25 mg/kg loading dose. The 12-hour infusion period is -61shown in the lightly shaded areas, and the lines represent the median values. The degree of receptor blockade, inhibition of platelet aggregation, and prolongation of bleeding time are maintained for the entire duration of the infusion, with recovery starting as soon as the infusion is discontinued.

TABLE 9 MEDIAN DATA FROM FIGURES BA-8C 0.25 mg/kg 10 pg/min for 12 hours 0 Time following Median Median Median On-set of infusion Bleeding Time Aggreaation Bngia *urs Minutes I Base- ine 11aal~ine 0 5.5 100 0.0 2 30 14.7 93.5 6 30 22.4 89.1 12 23.5 24.4 85.6 .s 18 13.9 61.1 72.9 24 8.6 60.9 69.2 S36 14.5 75.0 60.6 9 9 20 Clinical Outcomes of Stage I and Stage II Patients None of the 47 c7E3 Fab-treated patients experienced a thrombotic event during or after PTCA. All but two of the 47 c7E3 treated patients had a successful PTCA as defined angiographically by a reduction of the lesion(s) to less than 50% luminal diameter narrowing. Of the two unsuccessful dilatations, patient 01-012 had a reduction of a 90% narrowing of the left anterior descending coronary artery to 70%, but further dilation was technically not possible. The second patient, (patient 01-019), reviewed below, had an initially successful dilation, but required intracoronary stent placement for a major longitudinal -62dissection (without evident thrombus). One of the 9 control patients (01-022) experienced thrombotic abrupt closure 15 minutes into the procedure, requiring emergency coronary artery bypass surgery (CABG), from which he recovered. The other 8 control patients had successful dilatations to 50% or less residual narrowing.

Patient 01-019 (12-hour infusion group) had a balloon dilatation of a 95% lesion of the left circumflex coronary artery with a 50% residual narrowing. After the procedure, the patient experienced an apparent vasovagal episode, leading to bradycardia, hypotension, and transient asystole. He was returned to the catheterization laboratory and had urgent intracoronary stent placement for a persistent major longitudinal dissection. The stent 15 became dislodged in the left main coronary artery, and the patient was sent for emergency coronary artery bypass surgery. According to the investigator, there was no evidence of intracoronary thrombosis angiographically or intraoperatively. This patient also experienced a 20 peri-operative myocardial infarction. The patient recovered and was discharged 8 days after surgery.

There were 3 c7E3 Fab-treated patients who each experienced an isolated episode of chest pain post-PTCA of uncertain significance. Patient 01-009 (0.25 mg/kg single dose group) experienced chest pain 9 hours post-c7E3, patient 05-003 (12-hour infusion group) experienced angina 21 hours post-c7E3, and patient 06-003 (12-hour infusion group) experienced angina 2 days post-c7E3. The investigators reported that these episodes of chest pain were unrelated to ischemic symptoms signifying reocclusion.

Patient 02-004 (0.25 mg/kg single dose group) experienced prolonged periods of chest pain prior to c7E3 Fab treatment which continued during the PTCA procedure.

The following day ECG changes accompanied by elevated cardiac enzymes (drawn the preceding day) indicated that -63this patient had experienced a peri-procedural non Q-wave myocardial infarction (peak creatinine kinase 462, MB fraction 64).

There was one death in the trial which occurred 52 days after c7E3 Fab admininstration. Patient 06-002 (6-hour infusion group), who had a history of interstitial lung disease, congestive heart failure and unstable angina, underwent successful PTCA of the proximal left anterior descending coronary artery. During the procedure the patient developed sustained ventricular fibrillation, twice requiring electrical defibrillation, but thereafter the iprocedure proceeded uneventfully. After leaving the catheterization laboratory, the patient developed cyanosis, which initially responded to diuretics and oxygen therapy.

However, this patient subsequently developed progressive respiratory impairment and later required ventilatory support. The patient's subsequent hospital course was complicated by sepsis, adult respiratory distress syndrome, anemia (requiring multiple transfusions), and cardiac S 20 ischemia. This patient died 52 days post-c7E3 Fab due to multi-system failure.

Safety: Stage I and Stage II Observations Figure 9 shows the absolute change in hematocrit from baseline to 24 hours following the end of infusion for all patients by dose group. For reference, a line indicating the zero change point is shown. The hematocrit data from one control patient (01-022) and one c7E3 Fab-treated patient (01-019) are not plotted because both patients required blood transfusions following urgent coronary bypass surgery in the first 24 hours (see below). A second lower line at -12 indicates the change in hematocrit needed to be designated as a minor bleed using the Thrombolysis in Myocardial Infarction (TIMI) criteria (Rao et al., J. Am.

Coll. Cardiol. 11:1-11 (1988)). The changes in hematocrit -64are similar between the control patients and all c7E3 Fab dose groups. Table 10 summarizes the median change in platelet count at 24 hours following the end of infusion.

The change in platelet count in the untreated control and c7E3 Fab-treated groups had similar distributions, with no apparent dose-related effect.

TABLE MEDIAN CHANGE IN PLATELET COUNT AT 24 HOURS FOLLOWING THE END OF INFUSION 10 Dose Change Range Controls (n 9) -0.7 +13.2) 0.15 mg/kg (n 5) -17.4 +19.5) 0.20 mg/kg (n 5) -11.6 +4.7) 0.25 mg/kg (n 5) +2.9 +4.4) 15 6 hours' (n 11) -7.7 +19.4) 12 hours' (n 11) 0.0 +24.4) 24 hours' (n 10) -3.9 +50.0) Period of infusion (10 pg/min) of c7E3 Fab following a bolus injection of 0.25 mg/kg.

Discussion of Stage I and Stage II Results Stage I of this study established that c7E3 exhibits the same dose response characteristic in the PTCA population treated with aspirin and heparin as was seen in stable angina patients in a dose escalation trial (Example Chimeric 7E3 produces a dose-dependent blockade of platelet GPIIb/IIIa receptors, and this receptor blockade correlates with inhibition of platelet function. In addition, Stage II results demonstrate that prolonged inhibition of platelet Fab function up to 24 hours can be achieved by a continuous infusion. In all patients, platelet functional recovery begins by 6 to 12 hours after cessation of the infusion, regardless of the duration of infusion.

The clinical outcome of the c7E3-treated patients, using both angiographic and clinical endpoints, was considerably better than expected based on their risk profile. No patient in the c7E3 treatment group experienced a thrombotic event during or after the procedure. In addition, all but 2 patients had angiographically successful procedures. All patients .enrolled into Stage II and 6 of the 15 patients enrolled in Stage I were high risk patients on the basis of clinical or angiographic characteristics. An individual clinical factor (such as unstable angina, diabetes, women over age 15 65 years) or angiographic lesion-specific characteristic (such as type B or C) places a patient at increased risk of complications, and the effect of multiple factors are wcumulative.

In Stage II, 17 treated patients had unstable angina 20 with or without additional clinical or angiographic lesion-specific risk factors. In addition, 6 patients in Stage I were identified as having unstable angina.

Published series have identified unstable angina patients as having a major complication (death, myocardial infarction, urgent coronary bypass surgery, or repeat PTCA) rate of 10 to 15% (De Feyter, Editorialm.Am. Heart J. 118: 860-868 (1989) and Rupprecht, H.J. et al. Eur Heart J. 1l: 964-973, (1990)). Angiographic characteristics similarly are highly predictive of PTCA complications (Ellis, 1990, "Elective coronary angioplasty: technique and complications", In: Textbook of Interventional Cardioloyv, E.J. Topol, Ed., Saunders Co., Philadelphia); De Feyter, P.J. et al., Circulation 83:927-936 (1991); Ellis, S.G. and Topol, E.J, Am. J.

Cardiol. 66:932-937 (1990); and ACC/AHA Task Force Report: 66 Guidelines for percutaneous transluminal coronary angioplasty, J. Am. Coll. Cardiol. 12:529-545 (1988)).

Twenty-nine Stage II c7E3-treated patients met the eligibility criteria by means of lesion-specific characteristics. Of these, 12 patients had one Type B lesion, 14 had 2 or more Type B lesions, and three had Type C lesions. In addition, many of the patients in the trial had multiple lesions dilatated in a single or more than one vessel, which also potentially increases the risk of the procedure (Samson, M. g al., Am. Heart J. 120:1-12 (1990)). On the basis of both the number and severity of jigh risk angiographically defined risk factors in these patients, ischemic complications would have been expected in the range of 10 to 20% (Ellis, Elective coronary 15 angioplasty: technique and complications, In: Textbook of Interventional Cardioloav. (Ed. E.J. Topol), W.B. Saunders Co., Philadelphia (1990); De Feyter, Circulation 8:927-936 (1991); Ellis, S.G. and Topol, E.J, Am. J.

Cardiol. §:932-937 (1990)).

The control group also was comprised of high risk patients. However, in general, the number and severity of risk factors was lower in the control patients. Five of the 9 control patients had single risk factors of either one type B lesion (4 patients) or unstable angina (1 25 patient), whereas 26 of 32 Stage II c7E3-treated patients had either a type C lesion or two or more other high risk characteristics. This difference in risk status between the 2 groups is significantly different (Fisher's exact test p 0.018). Interestingly, the control patient with the abrupt closure (patient 01-022, unstable angina with 2 type B characteristics) was one of 3 patients identified as having more than one risk factor (one control patient had unspecified risk characteristics). Thus, whereas one of three control patients at highest risk had a thrombotic -67event, none of the 26 c7E3 Fab patients in this highest risk category had a thrombotic event.

This study also demonstrates that the potent antiplatelet effects of c7E3 can be achieved safely in patients already being treated with intravenous heparin and oral aspirin. Bleeding events were comparable in the control and c7E3-treated patients with no difference in hematocrit changes from baseline between dosing groups.

Other adverse events were infrequent and typically of mild 10 or moderate severity. There was one death in the trial, and this occurred almost 2 months after c7E3 Fab treatment in a patient with interstitial lung disease and heart failure who had progressive respiratory failure following PTCA, complicated by sepsis, adult respiratory distress syndrome, and eventually multiple organ failure.

Finally, none of the twenty patients in whom results were available experienced a human anti-chimeric antibody immune response.

In conclusion, chimeric 7E3 Fab potently inhibits platelet function safely in patients treated with aspirin and intravenous heparin who are undergoing PTCA. The antiplatelet action can be maintained for as long as 24 hours without a significant increase in bleeding risk and without immune system reactivity. Among patients at high risk of thrombotic complications, no thrombotic events occurred in the group treated with c7E3, suggesting that c7E3 can reduce the risk of thrombotic complications in this patient population.

Example Treatment of Abrupt Closure During Coronary Anaionlastv Abrupt coronary arterial closure during coronary angioplasty is the major determinant of morbidity and mortality in this procedure. It occurs in approximately 68 of elective angioplasty cases (Detre, K.M. et JL., Circulation a2:739-750 (1991)), but has been noted to occur in up to 20%-40% of patients who undergo angioplasty for unstable angina pectoris or after acute myocardial infarction (Ellis, S.G. et Circulation ZZ:372-379 (1988); DeFeyter, P.J. Cft Circulation g3:927-936 (1991)). The mechanism of abrupt closure is acute thrombosis at the arterial site where angioplasty has created or extended an area of endothelial injury. Usually there are disturbed flow patterns due to altered geometry of the vessel, often from disruption of the plaque material, and there is exposure of subendothelial elements including intimal and often medial dissection. Since initiation of the thrombus requires adhesion and 15 aggregation of platelets, chimeric 7E3 Fab antibody fragment was used for the treatment of abrupt coronary arterial closure complicating a coronary angioplasty procedure.

Case Report 20 The patient is a 45 year old male physician who had been in excellent health previously. Beginning one week prior to his angioplasty procedure, he had begun to experience chest and neck discomfort. When these symptoms persisted and worsened over several days, he sought the advice of a colleague. An electrocardiogram (EKG) revealed anterior precordial T wave inversions. The patient was then hospitalized in the coronary care unit of a local hospital and placed on intravenous nitroglycerin and heparin, and oral aspirin. Serial Cardiac isoenzyme determinations over the next 24 hours did not reveal elevation above the normal range. Serial EKG recordings over the next two days revealed persistent flattening of the anterior precordial T waves but no evolutionary changes of myocardial infarction. On the second day after -69hospitalization the patient was taken to the cardiac catheterization laboratory, where left ventriculography revealed overall normal left ventricular function with a very small hypokinetic area in the anterolateral left ventricular wall, and another hypokinetic area in the inferoposterobasilar zone. The left ventricular ejection fraction was 72%. Coronary arteriography demonstrated a left-dominant coronary system with a small and totally occluded right coronary artery. There was a significant 10 stenosis in the mid portion of the left anterior descending (LAD) coronary artery. A small and diffusely diseased diagonal branch originated just distal to the mid LAD stenosis.

The patient was returned to the coronary care unit and remained on intravenous nitroglycerin and heparin for another 48 hours. He was pain free during this time, cardiac isoenzymes did not rise, and daily EKGs revealed only the persistent flattening of the anterior precordial T i waves. He was transferred to Hermann hospital (Houston, 20 TX) for angioplasty.

Prior to the angioplasty procedure the patient continued to receive intravenous nitroglycerin and heparin, oral aspirin, and he was started on an oral calcium channel blocking agent. The partial thromboplastin time (PTT) had remained in the 70-90 seconds range for several days. At the start of his angioplasty procedure the activated clotting time (ACT) was 173 seconds. The patient received 5000 units heparin intravenously. The left coronary ostium was engaged with a number 8 French JL 3.5 guiding catheter.

The LAD coronary artery was visualized in the caudal right anterior oblique and cranial left anterior oblique projections. The LAD was first instrumented with a 0.018 inch Doppler guidewire (Cardiometrics, Inc., Mountain View, CA). This guidewire is used by us routinely for flow monitoring in patients at higher risk for abrupt closure.

Flow-velocity signals from the LAD proximal and distal to the lesion were recorded. A 2.5 mm coronary balloon catheter (Intrepid, Baxter, Inc., Irvine, CA) was advanced over the Doppler guidewire while the wire was held stationary in the coronary artery. The balloon was positioned so that it straddled the LAD lesion. Sequential brief balloon inflations were made to 6 atmospheres pressure. The severity of the stenosis was reduced as visualized by angiography as well as by increase in the flow velocity signal from a peak flow velocity (APV) of 12 cm/sec to 33 cm/sec.

During several minutes of observation following these dilations it was noted that the flow signal began to diminish. A contrast injection revealed renarrowing of the angioplasty site from elastic recoil, plaque disruption, and formation of thrombus. The balloon was reintroduced to the site of the lesion and another balloon inflation was performed. The artery was reexpanded and the flow signal again returned to an APV of 34 cm/sec.

During several more minutes of monitoring the signal again declined. Within 5 minutes the signal was quite low, at average peak velocity of 3 cm/sec. The patient began to experience chest pain. The EKG monitor of an anterior precordial lead revealed ST segment elevation. Angiography revealed that the artery was completely occluded. The activated clotting time obtained just a few minutes before was 344 seconds.

Chimeric 7E3 monoclonal antibody Fab fragment specific for the platelet GP IIb/IIIa receptor (c7E3 Fab, K) was administered. The dose was 0.25 mg per kilogram given intravenously over 1 minute. Within approximately 1 to 2 minutes after administration of c7E3 Fab, the coronary flow velocity began to increase. An injection of contrast revealed restoration of coronary patency with Thrombolysis -71- In Myocardial Infarction Trial Grade-1 (TIMI 1) flow.

Over the subsequent 15 minutes coronary flow continued to increase and stabilized at an APV of 23 cm/sec. Several other injections of contrast demonstrated improved coronary flow. The patient's chest pain subsided and the ST segment observed in the monitor lead returned to baseline.

Fifteen minutes after administration of c7E3 Fab, an angiogram was made according to protocol. This angiogram revealed TIMI3 coronary flow. The flow velocity 10, signal at this time was 20 cm/sec. Continuous monitoring through the subsequent 5 minutes revealed no further improvement in the coronary flow. During that time the video replay of the angiogram confirmed that there was a small amount of thrombus still visible at the angioplasty site. For this reason it was decided to administer intracoronary urokinase 250,000 units. This thrombolytic agent was infused over approximately the next 10 minutes.

During that time there was no further improvement in flow as measured by the Doppler guidewire. After completion of the intracoronary urokinase infusion, at the 33rd minute after administration of c7E3 Fab, another coronary angiogram was made. The artery was patent with TIMI 3 **flow. Some moderate but definite residual stenosis persisted at the lesion site. In addition, it was observed that the thrombus had diminished further in size but had not been completely dissolved. The decision was made to perform another balloon inflation in order to try to reduce the residual stenosis.

The balloon catheter was again advanced over the guidewire to the site of the lesion. A final balloon inflation to 6 atmospheres for 2 minutes was then performed. Then, the balloon catheter was withdrawn while the wire remained in place. The flow signal increased to an APV of 29 cm/sec and remained stable over several minutes. An angiogram demonstrated adequate reduction in the residual stenosis which had been present. The guidewire was then withdrawn proximal to the stenosis and another flow velocity recording was made. The guidewire, balloon catheter and guiding catheter were withdrawn. This completed the procedure.

The patient was then taken to the coronary care unit.

He remained on oral aspirin, nitrates, a calcium channel blocking agent, and intravenous heparin for several days in order to keep the PTT in the 70-90 seconds range. Serial EKGs demonstrated resolution of the anterior precordial T wave inversions and all subsequent EKGs were normal.

erial creatine kinase (CK) isoenzyme values were S. consistently 100 U/L. The platelet count prior to the PTCA procedure was 248,000, and subsequent platelet counts at 2h, 6h, 12h, 24h, and 48h after c7E3 Fab administration were 304,000, 279,000, 246,000, 185,000 and 220,000, respectively. Platelet aggregation induced by 10 #M ADP was 73% by optical densitometry prior to the procedure, and subsequent values at 2h, 6h, 12h, 24h, and 48h were 0%, 13%, 26%, 45%, and 51%, respectively. One week after the angioplasty procedure, the patient had a follow-up catheterization. The LAD coronary artery was found to be widely patent with TIMI 3 flow. He was discharged home later that same day.

0 Discussion of Case Report In this patient, the combination of 0.25 mg/kg c7E3 Fab intravenously, 250,000 U intracoronary urokinase, and repeat dilatation, successfully treated the acute ischemic coronary syndrome of abrupt closure during coronary angioplasty. These results suggest that antiplatelet therapy that inhibits platelet glycoprotein IIb/IIIa receptor binding and platelet cross-bridging may be efficacious in helping to achieve stable reperfusion of -73acutely occluded coronary arteries in similar clinical settings.

Example 6 A Randomized. Double-Blinded. Evaluation Anti- GPIIb/IIIa Chimeric Antibody Fragment for Preventing Ischemic Complications of High Risk AnaioDlastv Overview Percutaneous myocardial revascularization has grown at a dramatic rate since the advent of coronary angioplasty in 1977 (Gruentzig, A.R. etal., N. Enal. J. Med., 21:1127- 1132 (1987)). Although the procedure is associated with improvement in symptoms of ischemia and quality of life, (Parisi, A.F. et al., N. Enal. J. Med., 326: 10-16 (1992)), acute complications remain a major drawback. The treated vessel closes abruptly during or after the procedure in the hospital in 4 to 9% of cases, and reocclusion or abrupt closure is associated with considerable morbidity and an approximately 10-fold increase in mortality (Lincoff, A.M.

et al., J. Am. Coll. Cardiol., 12: 926-938 (1992); 20 Tenaglia, A.N. et al., Am. J. Cardiol, In press (1993); 0" Detre, K.M. et al., Circulation, 82:739-750 (1990); Ellis, S.G. et al., Circulation, 22:372-379 (1988)). Although the mechanism of abrupt vessel closure is often uncertain, thrombus formation and vessel dissection are contributory factors. Characteristics that identify patients at high risk of acute complications include the presence of a clinical syndrome associated with coronary thrombus (unstable angina, acute or recent myocardial infarction), diabetes, female gender and coronary morphologic characteristics (bend points, thrombus, bifurcation) indicating increased complexity of the individual lesion (Lincoff, A.M. et al., J. Am. Coll. Cardiol., 19:926-938 (1992); Ellis, S.G. et al., J. Am. Coll. Cardiol, 17:(Suppl -74- Bj:89B-95B (1991); Moushmoush, B. et al., Cath. Cardiovasc.

iaSn., 97-103 (1992); and Myler, R.K. et al., Circulation, 82 (SuDDI II):II-88-II-95 (1990)).

Although aspirin has been shown to reduce the risk of abrupt vessel closure and acute myocardial infarction in patients undergoing angioplasty (Schwartz, L. et al., IL Enal. J. Med., 1L:1714-1719 (1988); Barnathon, E.S. et Al., Circulation, 27:125-134 (1987)), its effects on platelet function are relatively weak, and ischemic events continue to occur at a rate of 10-20% in high-risk patients retreated with aspirin (Tenaglia, A.M. et al., J. Am.

•0 .Col. Cardiol., (1993, in press)). In contrast, administration of the Fab fragment of chimeric 7E3 antibody 9* to humans achieved substantial blockade of the GPIIb/IIIa 15 receptors and inhibition of platelet aggregation (see Example 4, Inhibition of Platelet Function).

In an initial study of patients undergoing angioplasty, c7E3 Fab reduced the risk of abrupt vessel closure during and after percutaneous intervention (see 20 Example 4 and Ellis, S.G. et al., Cor. Art. Dis., 4:167-75 (1993)). The current randomized study was designed to further evaluate the efficacy of chimeric antibody fragments which bind selectively to the glycoprotein IIb/IIIa receptor in preventing ischemic complications (EPIC trial, Evaluation of c7E3 Fab to Prevent Ischemic Complications). In particular, the clinical efficacy of a chimeric 7E3 Fab fragment (c7E3 Fab) in patients undergoing angioplasty who were at high risk of procedural complications was evaluated in a prospective, double-blind, placebo-controlled, randomized clinical trial. The trial included 2099 patients at 56 sites who were scheduled to undergo coronary angioplasty or directional coronary atherectomy in high-risk clinical situations: severe unstable angina with rest pain and documented electrocardiographic changes, evolving acute myocardial infarction, or clinical and coronary lesion morphological characteristics associated with a high risk of periprocedure complications. Patients received either (a) placebo bolus and infusion, a bolus of c7E3 Fab and placebo infusion, or a bolus and infusion of c7E3 Fab.

The primary endpoint was a composite which included the occurrence of any one of the following components: death, nonfatal myocardial infarction, unplanned surgical revascularization or repeat percutaneous procedure, unplanned coronary stent implantation, or intraaortic balloon pump insertion for refractory ischemia.

The bolus and infusion resulted in a 35% reduction in the primary endpoint (12.8 vs P 0.008), while an 11 S. percent reduction was observed with the bolus alone (12.8 15 vs 11.4%, P 0.43). The reduction in events achieved by administration of bolus plus infusion was consistent within each of the endpoint components, and in addition, in major patient subgroups including age, gender, preexistence of intercoronary thrombus, and acute coronary syndromes 20 (myocardial infarction, unstable angina). Bleeding Sepisodes and transfusions were increased in the bolus and infusion group and intermediate for the bolus alone regimen. This controlled trial of high-risk patients undergoing coronary intervention revealed that administration of an antibody fragment directed against the platelet IIb/IIIa receptor led to sustained clinical benefit through a significant reduction in ischemic complications.

Methods Based on previous studies stratifying risk with conventional percutaneous intervention, patients were eligible if they were at high risk for abrupt vessel closure and had no major contraindication due to high risk of bleeding. Patients were considered at high risk if they -76were in one of three clinical groups: acute evolving myocardial infarction within 12 hours of the onset of symptoms undergoing either direct or "rescue" percutaneous intervention; early postinfarction angina or unstable angina with at least 2 episodes of rest angina associated with resting electrocardiographic changes in the previous 24 hours despite medical therapy; or high-risk clinical and/or angiographic criteria using the American Heart Association/American College of Cardiology criteria (Ryan, T.J. et al., J. Am. Coll. Cardiol., 12:529-45 (1988)) as modified by (Ellis, S.G. et al., J. Am. Coll. Cardiol., j7 S.(Sup l B):89B-95B (1991)). These high-risk clinical and 0. angiographic criteria included either two Type B characteristics or 1 Type C characteristic in the target 15 lesion, or one Type B characteristic in women over age or in patients with diabetes.

The specific inclusion criteria are presented in more detail below: Referred for elective or urgent coronary balloon 20 angioplasty or atherectomy with an FDA approved device in one of the following settings: Unstable Angina and/or Non-Q Wave Myocardial Infarction defined as: 1) Angina at rest: two or more episodes of angina at rest with ischemic ST segment or T wave abnormalities; or 2) Recurrent angina: recurrent angina with ischemic ST segment or T-wave abnormalities while hospitalized not responsive to standard pharmacologic intervention; or 3) early post-infarction angina: angina within 7 days of documented myocardia infarction, with angina at rest accompanied by ischemic ST segment or T wave changes; or angina provoked by minimum exertion (2 mets), -77wherein transient ischemic ST segment or T wave abnormalities were defined as: a) a 1 mm ST segment depression (80 msec after the J point) or elevation (20 msec after the J point), and/or b) T wave changes (usually inversion), and wherein creatine kinase (CK) must have been less than 2 times normal, for all patients at the time of enrollment.

Acute Q-wave Myocardial Infarction 1) Direct intervention during myocardial infarctioh without antecedent thrombolytic .therapy, or 2) Rescue angioplasty for failed thrombolytic 15 therapy during myocardial infarction, wherein an enrolling myocardial infarction was defined as the presence of at least two of the following three criteria: prolonged angina (greater than 30 minutes); total creatine kinase elevation to greater than 2 times the upper limit of normal (confirmed by CK-MB isoenzyme elevation) ECG evidence of infarction defined as: a) ST segment elevation of at least 0.1 mV 25 (measured 0.2 seconds after the J point) in at least one of three locations: i) at least 2 of 3 inferior leads (II, III, aVF); or ii) at least 2 of 6 precordial leads or iii) leads I and aVL; or iv) ST segment depression of the precordial leads V,-V 4 consistent with posterior current of injury (mirror rule); or -78-

W

ri u v) in the presence of left bundle branch block, primary ST changes in the inferior or anterior leads; b) new significant Q wave of t 0.04 seconds duration or having a depth k one-fourth of the corresponding R wave amplitude, or both.

High Risk Clinical/Morphological Characteristics 1) Stenosis with two or more type B lesionspecific characteristics in the artery to be dilated. Lesion specific-characteristics were based on ACC/AHA criteria; 2) Stenosis with one or more type C characteristics in the artery to be dilated; 3) Age k 65 years combined with female gender and stenosis with at least one type B characteristic; 4) Diabetes mellitus and stenosis with at least one type B characteristic in the artery to be dilated; or Angioplasty of an infarct-related lesion within 7 days of myocardial infarction documented by characteristic CK-MB isoenzyme elevation.

Men between the ages of 18 and 80, and women between the ages of 18 and 80, not of childbearing potential surgically sterilized or post-menopausal, defined as not having a menstrual period for at least one year.

II.

III. Written informed consent provided before initiation of protocol-specific procedures and study agent administration.

-79p Patients who otherwise qualified for the study were excluded from participation for any of the following reasons: History of hemorrhagic diathesis; Major surgery within 6 weeks of study enrollment; Recent (within 6 weeks of enrollment) gastrointestinal or genitourinary bleeding of clinical significance; Stroke within 2 years prior to enrollment or any stroke with significant residual neurological deficit; A greater than 50% occlusion of the left main coronary artery; Presumed or documented history of vasculitis; Participation in other clinical research studies involving the evaluation of an investigational drug or device within 7 days prior to proposed study agent infusion; Administration of oral anticoagulants within 7 days prior to randomization to study agent, unless the pre-randomization prothrombin time is S 1.2 times control; Use of intravenous dextran before or planned for during the treatment angioplasty History of prior administration of murine monoclonal antibodies or known allergies to murine proteins; or (11) Inability to give informed consent.

Institutional review board approval was obtained at all institutions and informed consent was obtained from all patients. Recruitment into the trial occurred between December 1991 and November 1992; 2099 patients were enrolled at 56 institutions in the United States.

Study Protocol All patients were treated with aspirin and heparin.

Aspirin was administered orally at a dose of 325 mg at least 2 hours before the procedure, and were maintained a dose of 325 mg once a day thereafter. Heparin (porcine) was administered as an initial bolus of 10,000 to 12,000 units intravenously followed by incremental boluses of up to 3,000 units at 15 minute intervals, but not exceeding a total of 20,000 units; the goal was to keep the activated clotting time at a "therapeutic" range, generally .considered to be 300 to 350 seconds during the procedure (Dougherty, K.G. et al., Abstracts of the 63rd Scientific Sessions, III-189 (1991); Rath, B. et al., Br. Heart. J., 63:18-21 (1990); Ogilby J.D. et al., Cath. Cardiovasc.

iDjf., 18:206-9 (1989)). Heparin was continued by constant infusion at a rate of 1,000 units per hour for at least 12 hours. Intravenous and intracoronary nitrates could be used as clinically indicated. Chimeric 7E3 Fab K) was supplied as a sterile, nonpyrogenic solution containing 2 mg of monoclonal Fab per mL of 0.15 M sodium chloride, 0.01 M sodium phosphate and 0.001% polysorbate 80, pH 7.2. The only medication required at discharge was aspirin at a dose of 325 mg per day.

Patients were randomized equally into one of three 25 treatment arms with a double-blind study design. One group of patients was to receive a bolus of c7E3 Fab at a dose of 0.25 mg/kg followed by a 12-hour continuous infusion of c7E3 Fab at a dose of 10 pg/min. A second group was to receive a bolus dose of 0.25 mg/kg c7E3 Fab and 12-hour continuous infusion of placebo solution. A third group was to receive a placebo bolus and 12-hour continuous infusion of placebo solution. The bolus was started at least minutes before the procedure and given over 5 minutes, and -81the infusion was continued for 12 hours unless a clinical contraindication developed.

Blood samples for platelet counts were drawn minutes, 2, 12 and 24 hours following initiation of the drug and then daily until hospital discharge to carefully examine for evidence of thrombocytopenia. A predesigned algorithm was used to evaluate and treat life-threatening bleeding and thrombocytopenia (Sane, D.C. et al., Ann.

Intern. Med., :1:1010-22 (1989)). The protocol contained no specific indications governing red blood cell -transfusion; instead, transfusions were prescribed according to local practice patterns existing at each site.

Angioplasty was performed according to standard protocols.

Pre- and post-procedure angiography were performed after 15 coronary vasodilation with 150-300 ugrams of intracoronary nitroglycerin. After the procedure, vascular sheaths were maintained for at least 6 hours after the end of study agent infusion. Additionally, sheaths were left in place until at least 4 hours after the end of the heparin infusion and until an acceptable activated partial thromboplastin time was achieved to maintain hemostasis.

Study Endoints An independent Clinical Endpoint Committee reviewed all episodes that might have represented a study endpoint or a major adverse event. This committee, which remained blinded to treatment throughout the study, reviewed the case report forms, electrocardiograms and pertinent medical records when needed. A consensus of two reviewers was required to classify an event.

The primary endpoint of the trial was a composite clinical endpoint including the occurrence of any one of the following events in the first 30 days after randomization: -82- Death from any cause; Nonfatal myocardial infarction; or Urgent intervention: Second angioplasty. Repeat percutaneous intervention for recurrent acute ischemia (balloon angioplasty or coronary atherectomy). Scheduled staged procedures) were not considered endpoint events; Coronary artery bypass graft. Urgent (nonelective) surgical intervention to treat recurrent acute ischemia; Insertion of a coronary endovascular stent.

Intracoronary stent placement in order to 15 maintain the immediate patency of the dilated vessel; or Insertion of an intraaortic counterpulsation balloon pump. Balloon pump placed for recurrent ischemia in patients not considered candidates for repeat angioplasty or surgical intervention.

For the purposes of this study, "patency" was defined as TIMI Grade 2-3 flow with a less than or equal to isual stenosis as determined by the operator and no ECG 25 evidence of ischemia.

Endpoint myocardial infarctions were defined as follows: 1. In patients randomized within 24 hours of an acute evolving myocardial infarction, one of two enzymatic criteria was required for the diagnosis of a subsequent nonfatal infarction: a creatine kinase (CK) or creatine kinase MB at least 3 times the upper limit of normal, representing an increase of at least 33% from the previous "valley" (defined as a 25% decrease from a prior -83peak value but remaining at least twice the upper limit of normal); or an increase in CK or CK-MB of at least 100% and remaining three times the upper limit of normal following a 50% decrease from a prior peak level and a "valley" level less than twice the upper limit of normal.

A documented new episode of prolonged angina (greater than minutes) accompanying the re-elevation of cardiac enzymes was used to establish the time of onset of reinfarction. In the absence of documented angina, the onset of reinfarction was set as the time of measurement of the valley enzyme level immediately preceding new elevation. Definition 1 was used only for those patients in whom the onset of recurrent angina and/or the valley enzyme level occurred within 24 hours of the onset of the 15 myocardial infarction present at study entry. In all cases the CK-MB level was used unless it was not available, in which case the total creatine kinase values were used.

2. In patients entered into the trial more than 24 hours after an acute infarction or without a recent infarction, one of two criteria had to be met for a diagnosis of inhospital myocardial infarction: a new Q wave a 0.04 i sec in duration or with a depth 1/4 the corresponding R wave amplitude in two or more contiguous leads; or a CK-MB level at least 3 times the upper limit of normal and 25 an increase in that level 50% over the previous "valley" level. For this definition, the time of onset of reinfarction in patients with acute myocardial infarction at study entry was either the time of occurrence of a new episode of prolonged angina 20 minutes) or the time of measurement of a valley enzyme level preceding the new enzyme elevation. For this definition to be applicable, either time must have been greater than 24 hours following the initial infarction.

3. After hospital discharge, one of the two following criteria was required to diagnose a myocardial infarction: -84a new significant Q wave 2 0.04 sec in duration or with a depth a 1/4 the corresponding R wave amplitude, or both, in two or more contiguous leads; or a CK or CK-MB level greater than twice the upper limit of normal.

Another component of the primary endpoint was the need for urgent repeat intervention, defined as an unplanned return to the angioplasty suite for intervention; planned staged procedures were not included in the primary endpoint. Similarly, only urgent coronary surgery to treat recurrent ischemia or a failed angioplasty procedure was ,counted as a primary endpoint. Intracoronary stent placement was considered a primary endpoint when the stent was placed to treat threatened or actual abrupt closure of the vessel undergoing angioplasty. Intraaortic balloon S 15 pump placement was considered to be a primary endpoint when the pump was placed for recurrent ischemia in a patient not undergoing a repeat revascularization procedure.

Bleeding events were classified as major, minor or insignificant using the Thrombolysis in Myocardial Infarction Study Group criteria (Rao, A.K. et al., J. Am.

Coll. Cardiol., 11:1-11 (1988)). Major hemorrhages were defined as intracranial bleeding or bleeding associated with a decrease in hemoglobin greater than 5 g/dl (or, when hemoglobin was not available, a hematocrit decrease of at 25 least Minor bleeding was either spontaneous and observed as gross hematuria or hematemesis, associated with a drop in hemoglobin greater than 3 g/dl (or, when hemoglobin was not available, a decrease in hematocrit of at least or when no bleeding site was identified, a decrease in hemoglobin greater than 4 g/dl (or, when hemoglobin was not available, a hematocrit decrease of at least In patients who underwent blood transfusion, the number of units transfused was added to the observed drop in hematocrit divided by 3 in order to obtain the total decrease in hemoglobin used for deciding whether major or minor bleeding occurred (Landefield, C.S. et al., Am. J. Med., .2:703-13 (1987)).

Data Manaaement and Statistics Patients were randomized via telephone communication with the Data Coordinating Center at Duke University.

Randomization was stratified by study site and according to whether the patient was having an acute evolving myocardial infarction. Based on previous data, a sample size of 2,100 patients was planned in order to detect a 33% reduction in the primary endpoint (predicted to be 15% in the placebo group) with a power of 0.8 and a 0.05.

*0 Data were collected by study coordinators on case report forms and monitored by blinded study monitors prior to data entry. The sponsor remained blinded to the 15 randomization code and study results until all patients were enrolled and all endpoints had been adjudicated by the clinical endpoints committee.

Baseline characteristics are displayed as medians and and 75th percentiles for continuous variables and as 20 percentages for discrete variables in the tables below.

The primary endpoint of the trial was analyzed by considering the time until the first occurrence of any one ftf**: of the components of the composite endpoint within the first 30 days after enrollment. If no event occurred within the 30-day interval, the patient's follow-up was censored after 30 days. Kaplan-Meier survival curves for each treatment were used to graphically display the results (Kaplan, E.L. et al., J. Am. Stat. Assn., 53:457-81 (1958)). All treatment comparisons were performed using the intention-to-treat principle. For the primary endpoint, a log-rank test for trend (dose-response) was performed, considering the bolus only patients intermediate to the bolus and infusion group (Kalbfleisch, J.D. and R.L.

Prentice, The Statistical Analysis of Failure Time Data, -86- John Wiley and Sons, New York (1980)). The analysis plan then called for pairwise log-rank comparisons between the control group and each of the two c7E3 Fab groups if the test for trend was significant. Interim analyses of safety were performed when data were available on one-third and approximately two-thirds of the patients. The nominal alpha level used for judging significance of the test for dose-response at each interim analysis was prespecified to maintain an overall type I error rate 5 0.05. At the final analysis, the significance level used for this comparison was 0.036. A similar strategy (test for trend followed by pairwise treatment comparisons as appropriate) was employed in the final analysis to explore the relationship of treatment to each component of the composite endpoint, 15 although these comparisons were primarily for explanatory purposes. The final analysis also employed this strategy to compare treatments with respect to measures of bleeding complications, using conventional chi-square testing. Odds ratios and confidence intervals for the treatment effect in major subgroups (age, sex, weight, clinical subgroup) were calculated and graphically displayed.

Results The baseline clinical characteristics of the patients groups show that the population was at increased risk for acute complications of angioplasty because of high proportions of diabetes, recent myocardial infarctions, advanced age and women in combination with the lesion characteristics comprising the entry criteria (see Table 11). The majority of patients had one or two vessel disease with good left ventricular function. Details of the interventional procedures are listed in Table 12. No substantial differences according to treatment assignment are evident.

-87- The overall primary endpoint and its components are shown in Table 13. Compared with placebo, a graded effect of c7E3 Fab is demonstrated (p 0.009), with an 11% reduction in the composite event rates in the bolus alone group (p 0.43) and a 35% reduction in event rates in the bolus plus infusion group (p 0.008). The same graded effect was observed for each of the most important ischemic endpoints shown in Table 13. Thus, the sustained glycoprotein IIb/IIIa blockade reduced nonfatal infarction, emergency coronary bypass surgery and emergency percutaneous revascularization, while an insignificant trend in the same direction was present with the short term blockade produced by the bolus alone. Three of the deaths in the bolus and infusion group occurred in patients who died after randomization but before receiving the drug; despite this, these deaths are included in the analysis according to the intention-to-treat principle.

Since nonfatal ischemic events were prevented by c7E3 Fab, the severity of the nonfatal myocardial infarctions prevented is of interest. As shown in Table 14, both Q wave infarctions and infarctions associated with large enzyme elevations were prevented, and a dose-response effect was present.

•s -as- Table 21 Baseline Characteristics Placebo Bolus Bolus Infusion (K=696) (Nin695) Age (yr)4 62 (53, 69) 61 (52, 68) 63 (53, 69) Male 73 72 71 Weight (kg) 84 82 82 Risk Factors "'Diabetes 26 23 23 Hypertension 55 55 54 *Elevated cholesterol 57 59 Smoking 65 71 68 *9 Vascular Disease() Peripheral 9 9 ***Cerebral 3 4 4 Previous Hi() *None 34 30 days 25 28 27 8-30 days 13 13 14 <8 <days 28 29 29 Prior Procedure (t) *Angioplasty 25 20 22 Bypass surgery 15 14 16 Coronary Anatomy() 1 vessel disease 54 51 2 vessel disease 29 34 31 3 vessel disease 16 15 13 MI1 myocardial infarction.

+Median (25th, 75th percentiles) to I.: to 900

S

S S S S SS S S S S S S *5 S *S

S

Table 12 Interventional Procedural Details Placebo Blus (N-696) (N=.695) Blus infusion (N=708S) Procedure Balloon angioplasty Atherectomy Both Time in lab (min)+ Contrast used (ml)+ Thrombolytics Lowest ACT (sec).

134 (71,394) 200 (150,286) 3.2 271 (190,345) 129 (75,401) 200 (150,277) 3.7 284 (190,371) 141 (69,513) 200 (150,285) 2.3 285 (185,388) ACT activated clotting time +Median (25th, 75th percentiles) 9

S.

S

0 9** *0* 9 9 9* 9 9.

9. *9 S S S S S S S *5 S Placebo (M-.696) Table 13 Primary outcome Events, Bolus (N=695) Bolus and Infusion Dose-response P-values (N0708) Primary Endpoint* 89(12.8%) Composites of Primary Endpoint: Death 12(l.7%) Nonfatal MI 60(8.6%) Emergency PTCA 31(4.5%) Emergency CABG 25(3.6%) Stent 4(0.6%) Balloon pump 1(0.1%) 79(11.41) 9(l.3%) 43 2%) 25(3.6%) 16(2.3%) 12(l.7%) 1(0.1%) 59(8.3%) 1 2 (l.7%)t 37(5.2%) 6(0.8%) 17(2.4%) 1(0.1%) 0.009 0.*96 0.*013 'C0.0001 0.177 0.97 *p 0.009 for overall test for trend; p a 0.43 for comparison of placebo with bolus; p 0.008 for comparison of placebo with bolus and infusion.

t Three patients who died were assigned to this treatment limb but never actually received therapy.

MI a myocardial infarction; PTCA percutaneous angioplasty or atherectomy; CABO a coronary artery bypass grafting.

9., p a 9* *9 *9 9 to 0 'o 00: of *t to Table 14 Effect of c783 Fab on M~yocardial Placebo Bolus Q wave M1 M% Large non-Q wave MI 1 Small non-Q wave MI 1 Total (N=6 96) 2.3 4.0 2.3 8.6 (N-695) 0.9 2.4 2.9 6.2 Infarction Bolus and Infusion (3=708) 0.8 2.8 1.6 5.2 Dose-response P-values 0.018 0.197 0.342 0.013 MI myocardial infarction Large non-Q wave MI is defined by a peak CK-MB or total CK S times upper limit of normal; Small non-Q wave M41 is defined by a peak CK-MB or total CK 3 to 5 times upper limit of normal.

-92- The timing of nonfatal ischemic events was different in the 3 groups for urgent repeat angioplasty, an event that could be timed accurately (See Figure 10). The majority of events in the placebo group occurred in the first hours after the index procedure, while a delay of several hours (-6-12 hours) until the occurrence of events was evident among the bolus group, corresponding with the time of maximal receptor blockade. There was a marked delay in the onset of ischemic events in the bolus plus infusion group as well as a marked reduction in their absolute frequency.

The profile of bleeding complications during hospitalization is shown in Table 15. As with the primary endpoint for efficacy, a graded effect of treatment on 15 bleeding is evident. Patients in the bolus plus infusion group displayed substantial increases in both major ,bleeding rate and transfusion rate, while patients receiving bolus alone displayed only a moderate increase.

The majority of bleeding episodes occurred during coronary 20 artery bypass grafting or at the site of vascular puncture in the groin, although the rate of surgical vascular repair was evenly distributed in the placebo and bolus plus infusion groups and 2% in the bolus only group).

Similarly, six patients had an intracranial hemorrhage, 25 with two events in the placebo group, one in the bolus only group and three in the bolus plus infusion group, one of whom did not receive the drug because the event occurred after randomization but prior to the angioplasty.

al 1 4 4 Bledn Coplctin an Heatloi Measures*4 4 Placebo (N=6 96) Bolus (N=69 5) Bolus and infusion (N-7 08) Major bleed 4 6 (7:t) 76 (11%) 97 (14%) Transfusions Red cells Platelets Blood count Nadir hematocrit t A hematocrit t Index t Nadir platelet t Platelet count 1000000 49 18 (3t) 35 (32,38) 5.3 (3.2,7.9) 1.8 (1.1t2.7) 196 (159,240) 24 92 (13%) 29 34 (29,38) 6.5 (4.2t9.8) 2.2 (1.4t3.5) 194 (153,236) 29 109 39 33 (29,37) 6.8 (4.1,10.5) 2.3 (1.4,3.8) 193 (154,231) 42 p= 0. 001 0.001 t Median (25th, 75th percentiles).

The frequency of secondary clinical events was low and no major differences in these outcomes were observed according to treatment. The 30-day rates of heart failure sustained hypotension 3.6%, ventricular fibrillation and clinical occurrence of ischemia 17%, 18%) were similar in the placebo, bolus and bolus and infusion groups, respectively.

When the treatment effect was evaluated in subgroups (see Figure 11) defined according to whether patients were enrolled with an acute infarction, unstable angina, or high risk anatomy, the benefit of c7E3 Fab was present in all three enrollment categories. Similarly, the treatment effect was homogeneous across subgroups defined by age and 15 by gender. As a function of weight, the efficacy benefit of c7E3 Fab bolus plus infusion was present across the entire spectrum of patients, although the treatment effect was more pronounced in heavier patients.

The risk of major bleeding was increased in lighter 20 weight patients as compared with heavier patients for both the bolus only and the bolus plus infusion regimens. In the lightest tertile of patients, major bleeding occurred S* in 21 versus 15 versus 7 percent, respectively, for bolus plus infusion, bolus only and placebo-treated patients, 25 whereas in the heaviest tertile, the corresponding bleeding rates were 8, 7 and 7 percent. Transfusion of packed red blood.cells in the lightest tertile of patients occurred in 24, 20 and 11 percent, respectively, for bolus plus infusion, bolus only and placebo-treated patients, while in the heaviest tertile, the corresponding rates were 11, 7 and 4 percent.

Discussion These results confirm the importance of platelets, presumably platelet-derived mediators, and platelet function in the occurrence of acute ischemic events in patients undergoing percutaneous transluminal coronary angioplasty. The EPIC trial was designed specifically to enroll a cohort of patients at high risk for abrupt vessel closure (reocclusion) and its attendant complications.

Analysis of previous databases indicated that some patients could be classified as high risk based on clinical indicators of thrombus in the vessel, such as acute infarction (Stack, R.S. et al., J. Am. Coll. Cardiol., 11:1141-49 (1988)), severe unstable angina (Myler, R.K. et ai., Circulation, 82 (Sunl II):II-88-II-95 (1990)) or the 15 angiographic appearance of thrombus (Sugrue, D.D. et al., Br. Heart 56:62-66 (1986); Hettleman, B.D. et al., J Am. Coll. Cardiol., 15:154A (1990)). Other patients were at high risk because of mechanical factors, such as small S:vessel caliber, diffuse disease or unfavorable vessel 20 morphology (Sinclair, I.N. et al., Am. J. Cardiol., 61:61G- 66G (1988); Ruocco, N.A. et al., Am. J. Cardiol., 69:69-76 (1992); Ellis, S.G. et al., Am. J. Cardiol., 63:30-4 :"i (1989)). A substantial number of both types of patients were enrolled in the trial, allowing us to provide some 25 insight into the relevance of the inhibition of platelet aggregation for each type of patient. The inclusion of these high-risk patients was expected to result in an ischemic event rate of 15% in the placebo-treated patients, despite the use of aspirin and high-dose heparin in every patient; this expected event rate was nearly realized.

The administration of c7E3 Fab resulted in a reduction in the composite event rate, with the primary effect observed in the reduction in nonfatal myocardial infarction, need for emergency angioplasty and emergency coronary bypass surgery. The bolus of c7E3 Fab produced a -96delay in the onset of these events, corresponding to length of time that platelet aggregation is affected. Four to six hours after the bolus, however, ischemic events began to occur. This interval corresponds to the period following in which platelet aggregation has been shown to return to approximately 50% of its basal value following a bolus of murine 7E3 Fab.

In addition to delaying events, the combined bolus and infusion treatment, which produces profound and prolonged platelet inhibition (see Example also prevented acute ischemic events from occurring. The composite endpoint provides an overall estimate of the impact of this therapeutic approach on ischemic events in the .periangioplasty period.

15 One of the most important findings from this trial is the consistency of the event reduction profile across the various endpoints. The reduction in myocardial infarction was substantial and consistent with the simultaneous reduction in the clinical need for subsequent emergency 20 procedures. The classification of nonfatal infarction has become a major issue in the evaluation of percutaneous coronary interventions. Elevations of creatine kinase MB isoenzyme above the upper limit of normal are common, ranging from 4 to 21% of patients in reported series 25 (Klein, L.W. et al., J. Am. Coll. Cardiol., 12:621-6 (1991); Hunt, A.C. et al., Eur. Heart 12:690-3 (1991; Pauletto, P. et al., Am. J. Cardiol., A9:999-1000 (1987); Spadaro, J.J. et al., Cath. Cardiovasc. Diagn., 12:230-4 (1986)). When these enzyme elevations are not uniformly associated with discrete clinical events, no association with long term adverse outcomes has yet been documented.

Thus, the prevention of isolated enzyme elevations without associated clinical ischemic events may not be prognostically meaningful. In order to assure objectivity in this subtle area, enzymes and electrocardiograms were -97systematically collected, a blinded endpoints committee was used, and at least a three-fold increase in myocardial specific enzymes to classify an event as a myocardial infarction was required. The finding that c7E3 Fab reduced the full spectrum of myocardial infarctions, including those associated with moderate enzyme elevations, large enzyme elevations and Q wave development, provides reassurance of the clinical importance of the events that were prevented, especially since the need for emergency coronary revascularization procedures were also reduced.

S Although no effect on mortality was expected or observed, the occurrence of 3 deaths in the bolus and infusion group in patients who never received the drug should be noted. These deaths were counted in the primary 15 analysis according to the intention-to-treat principle.

All other patients who died actually received their assigned therapy. Given the low mortality associated with angioplasty, more than 20,000 patients would be required to detect a 25% beneficial or detrimental treatment effect on mortality.

The beneficial effect of glycoprotein IIb/IIIa receptor blockade on clinical endpoints in the setting of high-risk angioplasty is convincing, and is consistent with the positive results of a recent initial trial using the 0. 0 25 same antibody in patients undergoing angioplasty in the setting of refractory unstable angina (Simoons, M.L. et al., J. Am. Coll. Cardiol., 21:269A (1993)), With this being the first large scale trial of IIb/IIIa receptor blockade, there was concern about inducing thrombocytopenia. However, there was only a small, clinically unimportant increase in thrombocytopenia with c7E3 Fab. In particular, analysis revealed that more patients in the bolus plus infusion treatment group experienced thrombocytopenia (platelet count 100,000/gL) than in the bolus or placebo treatment -98groups. Thus, there was an increase in incidence of thrombocytopenia (platelet count 1 0 0,000/ML) in the bolus plus infusion treatment group compared with the placebo group (pairwise p 0.062). Severe thrombocytopenia (platelet count 50,000/#L) occurred in 11 of the patients in the bolus plus infusion treatment group and of the patients in the placebo treatment group.

Only 4 patients in each of the bolus plus infusion and placebo treatment groups had both severe thrombocytopenia and serious, life-threatening, or fatal adverse events. All episodes of thrombocytopenia were transient and typically occurred during the first few days.

1 A significant increase in bleeding complications and transfusion was observed in treated patients. This i 15 increase occurred primarily as a result of bleeding at the femoral puncture site and did not result in a marked S- difference in nadir hematocrits or life-threatening complications among the three groups. The trends remained the same whether or not surgical patients were included in 20 the analysis. In this blinded study, the uncertainty about which treatment was being given may have led to a lower threshold for transfusion at some sites due to concern about the ability to stop the bleeding. Our previous experience has demonstrated that a refined protocol for 25 management of bleeding and administration of transfusions in patients treated with thrombolytic therapy can effectively reduce the administration of blood products (Wall, T.C. et al., J. Am. Coll. Cardiol., 21:597-603 (1993)).

The interplay between treatment benefit versus bleeding risk as a function of weight is more complex than was expected. Although the primary event rate and the risk of major bleeding did not vary appreciably according to weight in placebo-treated patients, there was a clear trend with decreasing weight toward a higher proportion of -99primary outcome events and a higher proportion of major bleeds in both the c7E3 Fab bolus only patients and the bolus plus infusion patients.

Judgment about the clinical utility of this approach to blocking the glycoprotein IIb/IIIa receptor in patients before, during or after percutaneous revascularization depends on the relative value of avoiding ischemic events versus that of giving blood products. In the high-risk patients enrolled in this trial, the balance appears to be favorable. The prognostic implications of acute myocardial infarction or emergency repeat revascularization are serious, and fortunately the risk of transfusion continues to decline (Donahue, J.G. et al., N. Enol. J. Med., 212=:369-73 (1992); Dodd, R.Y. N. Enql. J. Med., 327:419-21 S 15 (1992); Nelson, K.E. et al., Ann. Intern. Med., 11:554-9 (1992)). Efforts to employ practice algorithms to reduce Sunnecessary bleeding and transfusions and to more effectively dose antithrombotic therapy, including antithrombin and antiplatelet effects, should further enhance the clinical benefits observed in this trial. In practice, potent, parenterally administered antithrombotic agents in instrumented patients must focus on weight adjusted dosing of antithrombotic therapy heparin doses were not weight adjusted here). Use of more detailed 25 mechanistic evaluation and protocols that define approaches to reduce bleeding to be applied consistently across participating centers can yield more information about the occurrence of bleeding complications.

The broad consistency of treatment effect across the patient population is strong evidence that although in some cases abrupt closure may be primarily thrombotic or primarily mechanical, thrombus formation has a more important role in many patients. The delay in events engendered by the bolus therapy and the prevention of events by the bolus and infusion implies that in most -100situations the surface of the disrupted artery has lost much of its thrombogenicity by 18 to 24 hours after the procedure. Therapeutic approaches should take into account the need for sustained antithrombotic effect in patients with a significant risk of abrupt vessel closure.

In summary, this trial demonstrates a beneficial effect of sustained blockade of the glycoprotein IIb/IIIa receptor in patients undergoing high-risk percutaneous revascularization procedures in reducing and/or preventing reocclusion or abrupt closure. Although this benefit was achieved at the risk of increased bleeding, consideration of the overall clinical outcomes favors this approach in S. the treatment of patients known to be at high risk of acute ischemic complications, based on clinical and angiographic 15 predictors prior to the procedure. The trial provides the first validation of a meaningful therapeutic approach to inhibition of the function of cell integrins, paving the way for other selectin and integrin targets for biotechnology in the future, as well as non-antibody or 20 peptide approaches to this specific IIb/IIIa glycoprotein.

Example 7 Reduction of Clinical Restenosis Following Coronary Intervention with Early Administration of Anti-GPIIb/IIIa Chimeric Antibody Fragment The process of restenosis following balloon angioplasty and percutaneous coronary interventions is extremely common, leading to recurrence of anginal symptoms and the need for repeat revascularization procedures in more than 25% of cases within 6 months, at an overall cost of more than $2 billion per year in the United States (Popma, J.J. et al., Circulation, 84:1426-1436 (1991); Topol, E.J. et al., Circulation, 87:1489-1497 (1993); Herrman, et al., Drugs, 46:249-262 (1993)). The -101major biologic trigger of restenosis is vascular injury, induced by the inflated balloon or alternative device at the site of intervention, and accompanied by plateletthrombus formation and change of phenotype of medial smooth muscle cells from their resting, contractile state to one capable of migratory, proliferative and secretory function (Forrester, J.s. et al., J. Am. Coll. Cardiol., U:758-769 (1991); Ip, J.H. et al., J. Am. Coll. Cardiol., 12:77B-88B (1991); Casscells, Circulation, f8:723-729 (1993)).

Although various pharmacologic agents have been successful .in experimental models in modulating the characteristic myointinal growth that occurs after vascular injury and '0 small studies have suggested an angiographic benefit, there has not yet been a large-scale clinical trial in patients 15 that demonstrates an effective agent and, until now, there 0*0' was no known pharmacologic treatment to reduce the likelihood of this event (Popma, J.J. et al., Circulation, 84:1426-1436 (1991); Herrman, et al., Drugs, 46A:249-262 (1993); Mercator Study Group, Circulation, 20 86:100-110 (1992)).

Coronary angioplasty is routinely performed with adjunctive oral aspirin and intravenous heparin. This 00000: a anti-thrombotic approach, however, is only weakly inhibitory of platelet aggregation. A variety of agonists 25 including thrombin, collagen, and adenosine diphosphate can stimulate platelets even in the face of aspirin therapy.

Molecular biology of the platelet has elucidated the glycoprotein IIb/IIIa integrin as the receptor responsible for platelet aggregation (Plow, E.F. et al., Proa.

Hemostas. Thromb., 296:320-331 (1988); Coller,

J.

Clin. Invest., 76:101-108, (1985)). The chimeric 7E3 antibody Fab fragment binds selectively to the platelet IIb/IIIa integrin. After initial studies confirmed preliminary safety and efficacy of the chimeric monoclonal antibody Fab fragment, a multicenter, double-blind, -102placebo-controlled trial in 2,099 patients was performed (see Example In addition to the primary efficacy endpoint of a reduction of major ischemic events in the acute phase, representing suppression of abrupt closure (see Example it was determined that c7E3 is capable of reducing the incidence of clinical restenosis, as defined by ischemic events or the need for repeat revascularization during the subsequent 6-month follow-up phase.

Methods Details of the study population and protocol are described in Example 6. To recapitulate, patients were eligible if they were undergoing coronary angioplasty or directional atherectomy and had an evolving or recent myocardial infarction, unstable angina, or high risk S. 15 angiographic lesion morphology as defined by the American Heart Association/American College of Cardiology criteria (ACC/AHA Task Force Report, J. Am. Coll. Cardiol., 12:529- 545 (1988)). Exclusion criteria were bleeding diathesis, age a 80 years, stroke within 2 years, or major surgery within 6 weeks. The protocol was approved by the Institutional Review Board at all 56 participating sites and informed consent was obtained in all patients.

Patients received oral aspirin (325 mg per day), with the first dose administered at least 2 hours before the 25 procedure. Intravenous heparin was given during the procedure to achieve an activating clotting time of at least 300 seconds. In addition to aspirin and heparin, patients were randomly assigned to one of three alternative regimens: placebo bolus and placebo 12 hour infusion; active c7E3 (Centocor, Malvern, PA) bolus at 0.25 mg/kg and placebo 12 hour infusion; or active c7E3 bolus at the same dose followed immediately by c7E3 infusion at gg/min for 12 hours. The bolus was given at least minutes prior to the coronary intervention procedure.

-103- The primary endpoint was the 30-day composite incidence of death from any cause, myocardial infarction, coronary artery bypass surgery for acute ischemia, repeat percutaneous coronary intervention for acute ischemia, need for an endoluminal stent or insertion of an intra-aortic balloon pump to treat ischemia. All of these events were reviewed by an independent Clinical Endpoints Committee which remained blinded to treatment throughout the study and required consensus of at least two reviewers for classification.

During 6 month follow-up, the double-blind was preserved. In addition to subsequent ischemic events of death or nonfatal myocardial infarction, patients were followed to determine the need for a repeat 15 revascularization procedure consisting of percutaneous coronary interventions, or coronary artery bypass surgery, or both. Unlike the acute phase endpoint, stenting or the use of an intra-aortic balloon pump was not included as an outcome since the focus was the need for revascularization 20 procedures rather than surrogates for a sudden ischemic event. The criteria for diagnosis after hospital discharge S. of myocardial infarction required either a new significant Q-wave 0.04 sec in duration or with a depth 2 1/4 the corresponding R wave amplitude in two or more contiguous 25 leads; or creatine kinase or creatine kinase myocardial band greater than twice the upper limit of normal.

Resvascularization data were collected as well as whether the original target vessel was subjected to repeat surgical or percutaneous revascularization. Follow-up was 97.2% complete.

To assess whether the 6 month outcomes were separate from the acute phase results, the analysis included all events from baseline to 6 months, events occurring after the 30 day endpoint in patients with an initial successful intervention (defined as achievement of a final stenosis -104less than 50% according to the reading of the clinical investigator and without an ischemic complication), and events occurring after 48 hours in patients with an initial successful intervention. The 30-day endpoint was prospectively selected because of precedence in many cardiovascular intervention trials. The 48 hour cutoff was used because it is known that, by definition, nearly all abrupt closure events that occur following coronary intervention take place within this time window (Detre, K.M. et al., J. Am. Coll. Cardiol., 1.:230A (1989); Lincoff, A.M. et al., J. Am. Coll. Cardiol., 19:926-938 (1992); de Feyter, P.J. et al., Circulation, 83:927-936 S(1991)).

Randomization was performed via a telephone call to 15 the Duke Coordinating Center and stratified by study site and whether the patient was having an acute myocardial infarction. Data were collected by study coordinators on a separate 6 month case report form that was quality assured by source documentation by blinded study monitors prior to 20 data entry. The sponsor remained blinded to the follow-up results until all patients had completed follow-up, the events were adjudicated by the Endpoints Committee, and the i database was complete.

Statistical Analysis 25 All treatment comparisons were performed with the intention-to-treat principle. Event rates were estimated using the Kaplan-Meier method (Kaplan, E.L. and P. Meier, J. Am. Stat. Assn., 53: 457-481 (1958)) and survival curves were used to graphically display the results. A test for dose-response in event rates from placebo to bolus to bolus and infusion (with scores 0, 1, and 2, respectively) was performed using the generalized log-rank statistic.

Pairwise comparisons between placebo and each of the c7E3 arms were also performed using the log-rank statistic.

-105- Proportional hazards (Cox) models were fit to examine possible associations between baseline character-istics and outcome. These were done with all treatment groups combined with the treatment differences modelled, and separately by treatment group to examine for differences between treatment arms. In addition, proportion hazards regression (Cox) models were fit for the all event composite endpoint after the 48 hour follow-up to examine factors which might be associated with late events or treatment effects. The factors included in this analysis were treatment, single lesion or multiple lesions revascularized, duration of the procedure, myocardial infarction or unstable angina at baseline or other highrisk entry criteria, gender, age 2 65 years or 65 years, *15 weight, and diabetes mellitus.

Results Enrollment began December 1, 1991 and ended November 18, 1992 with 2099 patients. Features of the entire study cohort are presented above (see Example 6, Table 10). The 20 baseline characteristics of the patients who had a successful initial angioplasty or atherectomy procedure, thus being eligible for subsequent clinical restenosis, are presented in Table 16. There were no significant differences in baseline features of the patients who had a successful initial procedure by treatment assignment.

-106- Table 16 Demographic Features of patients with a Successful Initial Procedur~e Age (years) Male Weight (kg) Risk Factors (t) Diabetes HypAirtension Elevated Cholesterol Smoking :Vascular Disease(% Peripheral Cerebral Placebo (N 609) 60.0 ±10. 3 430 (71.7%) 84.9 16. 0 155 (25. 8%) 321 (53.7% 316 (52.7%) 377 (64.7%) 47 21 Bolus (N =605) 59.7 ±10.4 434 (71.7%) 83.5 16.5 146 (24.1%) 329 (54.8%) 337 (55.7%) 428 (71.8%) 52 19 Boliu and Infusion (N =620) 60.2 ±10.6 445 (71.8%) 83.3 15.7 139 (22.4%) 314 (50.8%) 324 (52.3%) 423 (69.5%) 52 27 (4.4t) Previous MI(% None 30 days 8-30 days 8 days 275 109 45 171 (45.8%) (18.2%) 5%) (28.5%) 239 126 55 185 (39.5%) (20.8%) (30.6%) 253 128 57 182 (40.8%) (20.6%) (29.4%) Prior Procedure(% Angioplasty Bypass Surgery Coronary Anatomy(% 1 Vessel Disease 2 Vessel Disease 3 Vessel Disease Type of Procedure Balloon Atherectomy Both Target Vessel

LAD

LCX

RCA

Left Main Graft 145 121 (20.0t) 88 (14.7t) 85 (14.0%) 139 (22.6%) 95 (15.3%) 337 170 93 540 37 23 241 144 234 4 35 (56.2%) (28.3%) (15.0%) (90.0%) (40.2%) (24.0%) (39.0%) 324 198 83 549 27 29 229 159 253 1 35 (53.6%) (32.7%) (13.7%) (90.7%) (37.9%) (26.3%) (41.8%) 354 191 75 561 34 25 262 165 219 3 43 (57. 1%) (30.8%) (12.1%) (90.5%) (42.3%) (26.696) (35.3%) -107- Patients receiving c7E3 bolus or the bolus and infusion had a significant increase in bleeding complications, predominantly in the first 48 hours, with approximate doubling of transfusion rate (placebo bolus only 13%, bolus and infusion 15%, P 0.001). The 12 hour infusion was not fully completed in 48 patients receiving placebo, 85 patients assigned to bolus only, and 107 patients in the bolus and infusion group.

There were no significant increases in thrombocytopenia with c7E3 and no hypersensitivity or allergic effects were manifest. Positive human antichimeric antibody (HACA) responses occurred in 5.2% of the bolus-treated patients and 6.5% of the bolus plus infusion- 15 treated patients. Most of the patients with a positive HACA response had low-titer responses. All of the 32 patients who had a positive HACA response in the bolus treatment group had a titer 5 1:1600. Thirty-four of the 40 patients in the bolus plus infusion treatment group who had a positive HACA response had a titer S 1:1600. There were 6 patients in the bolus plus infusion treatment gorup, but none in the bolus treatment goup, with HACA titers ranging between 1:6400 and 1:51200.

At 30-days, there was a 35% reduction of major 25 ischemic events (death, myocardial infarction, urgent revascularization) for patients treated with c7E3 bolus and infusion compared with placebo treatment P 0.009) (see Example 6, Table 13). The 6-month data is presented in Table 17, which shows the outcomes of death, myocardial infarction, and the need for coronary artery bypass surgery or repeat coronary intervention, with target vessel revascularization for all patients enrolled, (b) those patients with a successful procedure, without an ischemic complication in the first 48 hours after enrollment, and for the cohort of patients who had an -108initially successful procedure and did not have an event in the first 30 days. At 6 months, there was a 23% reduction of ischemic events and revascularization (27% versus 35%, p 0.001; see Table 17, All Patients Enrolled, Composite Death, MI, CABG, PTCA). The favorable long-term effect was chiefly due to less need for bypass surgery or repeat angioplasty in patients with an initial successful procedure, as repeat target vessel revascularization was lessened by 26% for c7E3 bolus and infusion versus placebo treatment P 0.007; see Table 17). The bplus only patient group had an intermediate outcome which was not significantly better than placebo by the criteria used in this trial.

In a separate analysis, 6 month data of all patients not experiencing a primary efficacy event in the first days were studied. The results of this analysis are shown in Table 18. These data suggest a 21% reduction of the need for repeat revascularization procedures upon the acute phase administration of c7E3 bolus plus infusion.

S

-109- Table 17 Outcomes at Six Months Placebo Bolus Bolus infusion P-Value (K=696) (N=695) (M-708) a a a a a a a ALI Patients Unrolled Death 3.4 2.6 KI 10.5 8.0 CAWG M% 11.0 9.7 PTCA 20.8 19.8 Composite Death, MI, CAUG, PTCA 35.0 32.4 Any $evaseularization (CABG/PTCA) 29.4 27.2 Target Vessel Repeat Revascularization 22.3 20.8 All Patients With a Successful Procedure at 48 Hours (N-606) (N-618) Death 2.3 KI 2.6 2.4 CAUG 7.4 PTCA ()16.4 17.23 Composite Death, HI, CABG, PTCA 25.3 24.1 1 Any Revascularizat ion (CABG/PTCA) 23.0 22.5 1 (All patients) (N=636) (14-655)

(N~

Target Vessel Repeat Vascularizatj~on 18.9% 18.4% 1 All Patients With a Successful Procedure and No Event (N-549) (N=576) Death 1.7 1.3 HI 1.9 CABG 5.6 5.7 PTCA 12.5 14.4 1 Composite Death, MI, CABG, PTCA 19.2 20.1 1 Any Revascularization (CABG/PTCA) 18.4 18.3 1 Target Vessel Repeat Vascularization 16.8 16.4 1 3.0 6.9 9.4 14.*3 26.9 23.1 16.4 0.827 0.016 0.339 0.001 0.001 0.008 0.007 (N4-639) 3.0 2.5 7.0 1.5 .9.1 -=666) 5.6% at 30 Days N-=598) 1.5 1.7 4.7 0.0 5.2 4.6 4.3 0.664 0.860 0.701 0.010 0.007 0.025 0.134 0.789 0.716 0.438 0.180 0.072 0.077 0.265 Table 18 Outconex at 6 Monthx in All Patients Having No event at 30 Days Placebo Bolus Only Bolus P-Value Infusion (Nin601) 611) (N-643) Death 2.8 1.3 1.0 0.38 M1 2.4 2.5 2.2 0.85 CABG 5.9 5.5 4.4 0.24 PTCA 13.8 14.8 10.7 0.12 CABG/PTC& 18.4 18.3 14.6 0.077 Composite: Death, MI MI, CAWO, PTCA 19.5 19.6 15.4 0.059 Composite from Baseline (including 0-30 days) 34.8 32.2 26.9 0.001

M%

-111- The data for all events (death, nonfatal infarction or need for coronary revascularization) for all patients who were entered into the trial are also presented in Figure 12. The data for patients who had a successful intervention and no events until 30 days are also shown in Figure 13.

Of the acute phase endpoint, 81% of events had occurred by 48 hours. This was similar across treatment groups (82.0% placebo, 79.7% bolus, 81.4% bolus and infusion). By considering events after the first 48 hours in patients with initially successful intervention, the elective target vessel revascularization in the first days is identified. As shown in Figure 14, there was little difference for subacute ischemic events between c7E3 bolus and the bolus plus infusion groups until after the day endpoint.

SInstead of the composite of death, myocardial revascularization, and any revascularization, which would 0 include vessels not initially approached, it is helpful to 20 focus on target vessel revascularization only. For the entire patient cohort during the 6 month period, there was Sa significant 26% reduction in target vessel revascularization for the bolus and infusion group patients compared with the other treatment groups (see Figure S" 25 Of note, little effect of the bolus alone on target vessel revascularization during follow-up was observed under the conditions of the trial.

Subgroup analysis compared patients who at baseline were diagnosed to have acute coronary syndromes (unstable angina, recent or acute myocardial infarction) with the remaining patients who had stable angina but high-risk angiographic morphology (Table 19). This revealed a significant effect in reduction of composite events for both subgroups, but the reduction in need for repeat coronary interventions was only significant in the stable -112angina patients (Table 19). This finding was consistent, whether events were analyzed from baseline or after 48 hours in the patients with a successful procedure.

Table 19 Subgroup Analysis of Nvents for Acute Coronary Syndrome Versus Stable Angina Patients Events from Baseline to Six Month.

Placebo Bolus Bolus

P

Infusion Acute Coronary Syndrome N 288 306 299 Composite Event 33.1% 28.8% 25.5% 0.039 *Re-PTCA 20.2% 16.9% 15.4% 0.129 .Stable Angina N 408 389 409 Composite 36.3% 35.3% 28.0% 0.022 Re-PTCA 21.3% 22.1% 13.5% 0.004 Events from 48 Hours to Six Months P~Acute Coronary Syndrome N 252 277 271 Composite Event 23.5% 21.4% 17.8% 0.094 Re-PTCA 14.8% 15.2% 11.9% 0.325 Stable Angina N 354 341 368 Composite Event 26.6% 26.4% 20.0% 0.034 Re-PTCA 17.5% 18.7% 11.1% 0.013 -113- Discussion The current findings from a large-scale, multicenter, randomized trial support a reduction of clinical events reflecting less restenosis for patients undergoing coronary intervention who received a bolus and infusion of platelet IIb/IIIa integrin blockade. The extent of the benefit at 6 months was approximately 23% reduction of overall ischemic events, including death, nonfatal myocardial infarction, and need for revascularization, and a 26% decrease in target vessel revascularization. These results extend the benefit of c7E3 bolus and 12 hour infusion from reducing abrupt closure and acute phase adverse outcomes to a diminished need for subsequent coronary revascularization procedures.

15 The monoclonal Fab fragment used in the trial has potent affinity for binding to the platelet IIb/IIIa surface integrin with minimal dissociation. Previous studies in patients undergoing angioplasty with c7E3 have shown that even after the infusion of the antibody is terminated, there is persistent occupancy of the IIb/IIIa binding sites for at least 36 to 48 hours, and evidence of inhibition of platelet aggregation for at least 72 hours (see Example 4; see also, Ellis, S.G. et al., Cor, Art.

4:1675-175 (1993); Tcheng, J.E. et al., Circulation, 25 j8: (1993)). While both of these effects diminish and revert to baseline over time, the duration of the c7E3 inhibitory effect on platelet aggregation is interesting in view of the observation that the bolus plus placebo infusion did not have a clinically meaningful effect on either acute or 6 month outcomes under the criteria and conditions used in this trial. This observation suggests that reduction of acute ischemic events or clinical restenosis via administration of anti-GPIIb-IIIa antibody may require a more prolonged exposure to agent as achieved by bolus plus infusion of drug), and that more -114prolonged suppression of the GPIIb/IIIa integrin could be accompanied by further improvement of outcomes.

It is also noteworthy that c7E3 has been reported to bind to the vitronectin receptor (Hynes, Cell, 1S:11- 25 (1992)), probably because this receptor contains the #3 component of GPIIb/IIIa. This integrin (vitronectin) may have a role in modulating stenosis or restenosis, and binding of anti-GPIIb/IIIa to the vitronectin receptor may contribute to the effect observed. Other GPIIb/IIIa receptor inhibitors have varying degrees of specificity for the target and for homologous integrins (Sutton, J. et al., Clinical Research AFCR, A1:118A (1993)). Comparative studies can be performed to dissect the contribution of these molecular interactions to clinical outcome.

15 In the current trial, systematic six month repeat angiography to quantitatively determine the extent of renarrowing in the treatment groups was not carried out.

Although repeat angiography has been conducted in many restenosis trials (Forrester, J.S. et al., J. An. CoIl.

Cardiol., 12:758-769 (1991); Ip, J.H. et al., J. Am. Coll.

Cardi., 17:77B-88B (1991); Casscells, Circulation, :723-729, (1993); Topol, E.J. et al., N. Enol. J. Med., 2129:228-233 (1993); Adelman, A.G. et al., N. Enal. J. Med., 329:228-233 (1993); Serruys, P.W. et al., Circulation, 25 84:1568-1580 (1991)), it has a major drawback because diagnosis of target vessel stenoses in asymptomatic patients often leads to repeat procedures that would not have occurred in practice. In contrast, the present design offers a simulation of clinical practice in a large population of patients. It is the net objective of restenosis trials to demonstrate a significant reduction in the need for repeat revascularization procedures, because angiographic benefit per se, which has been documented in some recent restenosis trials, is not singularly adequate -115or completely clinically relevant. Furthermore, since death and myocardial infarction are unusual in patients following percutaneous coronary intervention, the main outcome that should be modulated by an effective pharmacologic intervention is repeat target vessel revascularization, as observed here. Because the patients in the study were treated similarly except for the study drug bolus and infusion, including careful preservation of the double-blind until the follow-up was complete, it reasonable to conclude that a reduction in renarrowing explains the observed clinically meaningful benefit.

This represents the first large-scale randomized trial to demonstrate a clinically meaningful reduction in the need for subsequent revascularization procedures that can 15 be interpreted as less clinical restenosis. This was achieved with a bolus and infusion of IIb/IIIa blockade and points towards true passivation of the injured vessel wall.

Even through the infusion of c7E3 was only maintained for 12 hours, the agent has an extended anti-platelet effect for several days and there was no evidence of rebound in ischemic events in the acute phase. Importantly, the independent benefit of reduced target vessel revascularization at 6 months suggests a lasting effect of the acute c7E3 pharmacologic intervention and can be 25 offered as clinical proof of vessel wall passivation.

The finding of less clinical evidence of restenosis with platelet IIb/IIIa blockade further emphasizes the role of the platelet-thrombus in restenosis, which has been raised as a potentially key pathway in the development of the post-angioplasty or endothelial injury neointimal lesion (Schwartz, R.S. et al., J. Am. Coll. Cardiol., 2Q:1284-1293 (1992); Topol, E.J. Mayo Clin. Proc., 68:88-90 (1993); Willerson, J.T. et al., Proc. Natl. Acad. Sci. USA, 88:10624-10628 (1991)). While the proliferation of medial smooth muscle cells may also play a prominent role in -116restenosis (Forrester, J.S. et al., J. Am. Coll. Cardiol., 12:758-769 (1991); Ip, J.H. et al., J. Am. Coll. Cardiol., 1L:77B-88B (1991); Casscells, Circulation, 8&:723-729, (1993)), the results here suggest that potent anti-platelet and anti-throabotic approaches to this important clinical phenomenon may be particularly fruitful. The trial supports the notion that the current approach of using aspirin during coronary intervention (Schwartz, L. et al., N. Enal. J. Med., 218:1714-1719 (1988)) as the sole antiplatelet agent is insufficient to antagonize the platelet response to vascular injury.

Example 8 Additional Findings From the Randomized. Double-Blind.

Placebo-Controlled Trial 15 Sheath Size and Bleeding Complications Bleeding complications were assessed in the EPIC trial (see Examples 6 and 7) to determine if sheath size during PTCA/DCA correlates with bleeding complications Sheath and guiding catheter size were determined clinically by the interventionalist. Major bleeding episodes, groin bleeding, transfusion, vascular repair, nadir Hct were prospectively assessed.

Sheath size predicted groin bleeding even after adjustment for known bleeding predictors, including treatment assignment and heparin use during catheterization (p 0.0004). The trend toward more vascular repair with larger sheath sizes was not significant (p 0.0004).

Major bleeds transfusions and nadir Hct (34) did not vary with sheath size. c7E3 Fab patients had more groin bleeds than non-7E3 patients (55% vs. 30%, p 0.0001).

-117- 6F-7.5F 8F-8.5F 9F-11F n 375 n 1416 a 291 Groin Bleed 140 674 147 (51%) Vase Repair 17(1.2%) 7 These results indicate that larger sheath sizes are associated with more groin bleeding but do not contribute to major bleeding complications of PTCA/DCA. c7E3 Fab is associated with increased groin bleeding, but this may be minimized by using smaller sheaths and guiding catheters.

o-Incidence of Ischemic and Bleedina Events After Platelet GPIIb/IIIa Receptor Inhibition for Coronary Intervention The EPIC (Evaluation of c7E3 in the Prevention of Ischemic Complications) trial (see Examples 6 and 7) demonstrated that therapy with c7E3 Fab, a potent platelet 15 receptor GPIIb/IIIa antagonist, prevents ischemic complications in the setting of high risk coronary angioplasty (PTCA), but bleeding episodes requiring transfusion increased from 7% in the placebo group to 14% in c7E3 Fab group. To investigate this further, the 20 relationship between bleeding indices (nadir hematocrit, bleeding index, change in hematocrit, units of packed red cells transfused) and the primary endpoint of the study (death, myocardial infarction, coronary artery bypass grafting (CABG) or PTCA for acute ischemia or insertion of a coronary stent for procedural failure) was investigated.

A strong association between bleeding and the primary endpoint was noted (p 0.0001 for all bleeding indices).

This association was also present for each of the therapies: placebo, bolus c7E3 Fab, and bolus plus infusion c7E3 Fab. Thus, patients with significant bleeding were more likely to have ischemic complications.

This strong relationship may be due to the increased bleeding associated with primary endpoint events (e.g.

-118- CABG). Alternatively, bleeding with associated hypotension may be a major contributor to post-procedural ischemic complications. In support of this, patients who developed hypotension (excluding hypotension after a primary outcome event) after successful PTCA were significantly more likely to have a primary outcome event and were also more likely to experience major bleeding and a primary outcome event.

Hypotension No Hypotension

P

Number 239 1597 Primary Endpoint 46(16.1) 103(6.5) <.001 Endpoint Major Bleed 18(39.9) 10(9.7) <.001 In conclusion, bleeding appears to induce ischemic complications in some patients, and measures to reduce bleeding (for example, modifications of heparin dosage) may further enhance the anti-ischemic efficacy of GPIIb/IIIa inhibition for coronary intervention.

Economic Benefits and Detriments of Aggressive Platelet Inhibition During High Risk Angioplasty 20 In the 2,100 patient EPIC randomized trial (see Examples 6 and aggressive platelet inhibition with high dose c7E3 Fab was associated with a 35% decrease in subsequent death, reinfarction and recurrent ischemia during high-risk coronary angioplasty (PTCA), but twice as many post-PTCA major bleeds. To evaluate the economic consequences of this combination of clinical effects, a prospective economic substudy was conducted. Hospital costs (not charges) and resource use data were collected prospectively for each participant for six months following enrollment. Average baseline hospital costs for patients with an uncomplicated hospital course was $9300. The effects of major complications on baseline hospital costs -119were examined using a multivariable linear regression model: Mean hospital costs $9065 $5923 urgent PTCA $28,219 urgent CABG $3645 (re) infarction $3462 major bleed This model shows that by reducing urgent PTCA from 4.5% to urgent CABG from 3.6% to and (re) infarction from 8.6% to therapy with chimeric 7E3 fragment saved an average of $682 per patient over placebo therapy.

However, by doubling the rate of major bleeding from 7% to 14%, the therapy lost $242 of its potential cost advantage, resulting in a projected net cost savings per person of $440. The observed cost average difference between high dose c7E3 Fab (X $10,970 7,284) and placebo (X 15 $11,376 12,555) was $406, agreeing closely with the model prediction.

Thus, by significantly reducing the ischemic complications of high risk PTCA, aggressive platelet inhibition with a chimeric anti-GPIIb/IIIa fragment offers both improved medical outcomes and a net cost savings.

Measures to lower major bleeding rates upon administration of c7E3, while preserving treatment benefits, could yield a projected net cost savings of up to $700 per patient Activated Clotting Time is Increased During Coronary Interventions with Platelet GPIIb/IIIa Antagonism The activated clotting time (ACT) has been used during percutaneous transluminal coronary angioplasty (PTCA) to monitor the extent of thrombin inhibition and anticoagulation in an attempt to minimize untoward thrombotic events. With the introduction of potent platelet inhibitors, such as chimeric monoclonal antibody c7E3 Fab, the utility of measuring and regulating

ACT

during PTCA has not been examined. To this point, an effect of c7E3 on ACT was not known or suspected. The -120possible influence of platelet GPIIb/IIIa antagonism on procedural ACT was investigated. In the trial, 2099 subjects undergoing PTCA were randomized to receive placebo (n 696) or GPIIb/IIIa antagonist, c7E3 Fab (n=1403).

Despite receiving a similar amount of heparin and fewer patients receiving very high heparin doses (>14,000 units) compared to the placebo group, those receiving c7E3 Fab had a significantly higher (p 0.001) ACT when corrected for bodyweight.

Placebo Procedural Reparin Subjects c7E3 Fab Subjects <10,000 units or infusion only 112 (16) 265 (19) 15 10,000 units 220 (32) 497 (36) 10,000-14,000 units 141 (21) 298 (22) >14,000 units 209 (31) 316 (23) In conclusion, the activated clotting time is increased 40 seconds by the platelet GPIIb/IIIa antagonist c7E3 Fab.

20 This has important implications for dosing conjunctive heparin therapy and performing coronary artery interventions in the setting of GPIIb/IIIa-directed therapy.

Gender Differences in a Trial Evaluating the Prevention of 25 Ischemic Complications of Angioplastv with a Monoclonal Antibody to the Platelet GPIIb/IIIa Receptor Gender differences were explored in a trial evaluating the prevention of ischemic complications of angioplasty with the use of c7E3 Fab, a monoclonal antibody to the platelet GPIIb/IIIa receptor (see Examples 6 and 7).

Patients underwent percutaneous intervention (PTCA) and received one of three blinded treatments shortly preceding the intervention: c7E3 bolus followed by a 12 hour c7E3 infusion, c7E3 bolus only, or placebo.

Despite the fact that women tended to be older, lighter in weight, and had more cardiovascular risk factors -121than men, women and men did not differ in frequency of death vs. MI vs. emergency PTCA vs. urgent bypass surgery vs. stent placement, intraaortic balloon pump placement or a composite measure of those adverse ischemic events to days (10.5% vs. 11.0%, p=0.74). Adjustment for known predictors of the composite measure (treatment assignment, weight, hypertension, peripheral vascular disease), as well as for any potential statistical interactions with gender did not alter these results. Treatment with a c7E3 Fab 'bolus and infusion decreased the frequency of ischemic events similarly in both sexes.

Women displayed more serious bleeding (12.6% vs.

required transfusion (PRBC) more frequently (19.5% 15 vs. and had a higher bleeding index (A hematocrit/3 units PRBC, 2.4 vs. 1.9) than men. In a regression model to predict the bleeding index, despite adjustment for known predictors of bleeding (treatment allocations, age, weight, baseline hematocrit, 20 hypertension) gender remained a statistically independent predictor (p .0041).

In conclusion, women experienced more bleeding than e men, but did not exhibit an excess of other adverse outcomes following high risk angioplasty and treatment with c7E3 Fab.

Examole 9 Affinity of Chimeric 7E3 Fab for GPIIb/IIa on Human Platelets The binding of chimeric 7E3 Fab and murine 7E3 IgG to platelets from three different normal human donors were studied at 376C, and these were analyzed both by direct binding isotherm plots and by linear representations of mass action law.

-122- Labelled antibodies '"I-murine 7E3 IgG (m7E3 IgG) was prepared by iodogen labelling. Specific activity and protein concentrations were determined to be 4.1 pCi/pg and 45 pg/ml, respectively. A 1:20 dilution was made with HSA-saline diluent human serum albumin in 0.9% NaC1 solution) to yield 100,000 cpm/10 pl. '"I-chimeric 7E3 Fab (c7E3 Fab) was also prepared by iodogen labelling. The specific activity and protein concentrations were 0.995 CiJg and 0.29 mg/ml, respectively. A 1:62.5 dilution was made with HSA-saline diluent to yield 100,000 cpm/10 pl.

Antibodies Working stock solutions at a concentration of 250 Ag/ml were prepared for murine 7E3 (m7E3) IgG and for c7E3 Fab. Several dilutions were made with HSA-saline diluent to desired concentrations. Fixed amounts of corresponding labelled antibodies, 10 p1, were added to 40 1l of cold antibody preparations and this mixture was used in antibody binding assays.

Platelets Platelet Rich Plasma (PRP) was prepared according to standard procedures. Blood was collected into sodium citrate anticoagulant and PRP was adjusted to 200-300,000 platelets/pl and used within an hour.

Antibody binding assay Antibody binding was performed by a modification of the method described by Coller (Coller, J. Clin, Invest. 76:101-108 (1985)). A 50 p1 aliquot of antibody solution was mixed with 450 1l of PRP, both pre-warmed to 37 0 C and the mixture was incubated for 30 minutes at 37°C.

One hundred microliter aliquots from the reaction mixture -123were immediately layered on 200 pl of 30% sucrose, in triplicate, and were centrifuged for 5 minutes at 10,000 x g. The pellet containing the sedimented platelets with bound antibodies (the tip of the centrifuge tube) and the supernatant containing the unbound antibody were separated by cutting of the microfuge tube, and the associated radioactivity in each compartment was determined and recorded.

Three separate experiments were performed with platelets from three different donors.

10 Data analysis For each antibody concentration, the fraction of bound antibody was calculated by using the radioactive counts measured in the pellet and supernatant. It was assumed that the binding of radioactively labelled antibody was identical to that of cold antibody and that the value of the bound antibody fraction calculated from labelled antibody was representative of the entire population of antibody molecules. This bound antibody (Ab) fraction, together with values for the total concentration of 20 antibody were used to calculate the molar concentrations of bound and free antibodies for each data point. First a binding isotherm plot was drawn by plotting [Bound Ab] versus [Free Ab]. A four parameter curve was fitted by KaleidaGraph software (Synergy Software, Reading, PA) in a Macintosh computer. The equation and the four parameters of the curve fit were: y ml ((m2 ml)/(l where y [Bound Ab], x [Free Ab], and the four parameters were: ml value of upper plateau (total antigen/epitope concentration, [Total Ag], at saturation binding); m2 value of lower plateau (value of [Bound Ab] approaching zero at very low concentrations of added Ab; -124m3 value of x at mid-value of y (value of [Free Ab] at binding of antigenic epitope); and m4 an exponent related to the slope of the linear portion of the curve.

The value of [Free Ab] at 50% of binding (third parameter calculated by the program) is the Kd of the reaction between Ab and antigen according to the mass action law, assuming independent non-interacting sites. Additionally, the value of the [Bound Ab] at the upper plateau (first parameter) which corresponds to complete saturation of antigenic sites, gives the concentration of the antigenic sites. With the latter value, it is possible to calculate the number of GPIIb/IIIa receptors per platelet. In addition to direct analyses of the binding isotherm, the data were also analyzed according to linear transformations of the mass action law formula. There were six possible linear transformations, one of which is the commonlydesignated Scatchard plot (plot a below) (Fazekas de St.

Groth, "The Quality of Antibodies and Cellular 20 Receptors", In: Immunoloaical Methods. Vol. I, Lefkovits, I. and B. Pernis, Editors, Academic Press, Inc. New York, 1-42, 1979).

Formulae for plots All formulae are derivations from the mass action formula: Kd [Ab] x [Ag] [Ab:Ag] where [Ag] and [Ab:Ag] are equilibrium molar concentrations of free Ab, free Ag and Ab:Ag complex.

Binding isotherm: [Bound Ab]/[Total Ag] (1/(Kd [Free Ab]) x [Free Ab] Plot a: [Bound Ab]/(Free Ab] ([Total Ag]/Kd) 1/Kd x [Bound Ab] Plot b: 1/[Free Ab] ([Total Ag]/Kd) X 1/(Bound Ab] -125- Plot c: 1/(Bound Ab) (1/[Total Ag]) (Kd/[Total Ag)) x 1/[Free Ab) Plot d: [Free Ab]/[Bound Ab] (Kd/(Total Ag)) 1/(Total Ag] x [Free Ab) Plot e: [Free Ab) Kd [Total Ag] x [Free Ab]/[Bound Ab] Plot f: (Bound Ab) (Total Ag] Kd [Bound Ab]/[Free Ab] Bindabilitv of labelled antibodies When antibodies are labelled with radioactive isotopes, a portion of the antibodies may lose activity and the ability to bind to the target antigen. The ratio of unactive labelled antibody must be determined Experimentally and subtracted from the total antibody concentration before the calculations for the graphs (Trucco, M. and S. de Petris, "Determination of Equilibrium Binding Parameters of Monoclonal Antibodies Specific for Cell Surface Antigens", In: Immunological Methods. Vol. II, Lefkovits, I. and B. Pernis, Editors, Academic Press, N.Y.

1-26 (1981)). It was assumed that for the unactive antibody population, loss of activity due to labelling was complete and the rest of antibody population, 20 bindable antibodies) have the same affinity constant.

'I-labelled 7E3 Fab and 7E3 IgG, at 0.14-0.2 .g/ml, were incubated with platelets at approximately 300,000 platelets/Al. The bound Ab ratios at equilibrium were determined to be 89% versus 78% for 7E3 Fab and 7E3 IgG, respectively. The expected ratios for fully active antibodies as calculated from the mass action formula were 93% for both Ab's at these concentrations. Hence, the results suggest that 4% of the 7E3 Fab and 15% of the 7E3 IgG were inactive, likely as a result of the radiolabelling procedure.

Results The experimental data was plotted as described above (not shown); binding isotherms and linear plots with the -126equation of the fitted curve for each experiment were prepared. The Ka values and epitope densities (number of GPIIb/IIIa molecules per platelet) determined from each plot using the equations of the fitted curves are presented in Table Binding of c7E3 Fab to human platelets follows a conventional binding pattern. The binding isotherm revealed a smooth predicted curve expected from binding of a homogenous monoclonal Ab Fab fragment to homogenous cell surface antigen. The dissociation and association constants of the reaction were calculated to be 5.15 nanomolar and 1.94E08 respectively. The number of GPIIb/IIIa antigens present on the platelet surface was determined to be 69590 (-70,000). By comparison, m7E3 IgG 15 binds to platelets with dissociation and association constants of 3.56 nanomolar and 2.81E08 respectively.

The GPIIb/IIIa density calculated for 7E3 IgG was found to be 73,355 epitopes/platelet. These numbers indicate that intrinsic reaction constants of the antigen binding sites of 7E3 Fab and 7E3 IgG are similar. It was concluded that m7E3 IgG binds to platelets with both arms based on an analysis of this data and consideration of the off rate constants.

9 9., *9* 4 9 *9 .9 99 9 9 *e9 9 9..

*9# 9. 9.

9 .9 9* 9 99. 9 9 9* 99 9 9 99 .9 9 *9 9 9 TABLE 7E3 Binding to Human Platelelt.

Affinity constant and Spitope Density on Platelet.

Donor 11091 #1253 Ant ibody c7E3 Fab m7E3 IgG m7E3 IgG c7E3 Fab m7E3 IgG m7E3 IgG c7E3 Fab m7E3 IgG m7E3 IgG c7E3 Fab m7E3 IgG m7E3 IgG b iva lent monovalent bivalent monovalent biva lent monovalent 1. 26E08 1. 71E08 2.72E08 1.33E08 2.95E08 3. 53E08 3.23E08 3.77E08 3. 62E08 1.94E08± 2.81E08± 3.29E08± 71,506 68,.841 29,408 78,136 78,685 33,780 Ebitopesl1atclet 59, 129 72,540 31,898 #2010B Averages biva lent monovalent 1.12E08 1. 04E08 4.96E07 69590 k9647 73355 ±4972 31695 ±2193 -128- Example Bindina of Chimeric 7E3 Fab to the Vitronectin Receptor on Endothelial Cells Platelet glycoprotein GPIIb/IIIa belongs to a family of integrin receptors which share structural and immunological characteristics. An integrin closely related to GPIIb/IIIa is the vitronectin receptor (a,8 3 which utilizes the same P subunit as GPIIb/IIIa but has a different a subunit. The vitronectin receptor is expressed on endothelial cells and mediates adhesion to a variety of extracellular matrix proteins vitronectin, *fibronectin, von Willebrand Factor, fibrinogen, osteopontin; thrombospondin, collagen, perlecan). The .homology between GPIIb/IIIa and the vitronectin receptor is 15 sufficient so that 7E3, an antibody directed against GPIIb/IIIa, also binds to the vitronectin receptor '.e expressed on endothelial cells. Therefore, this study was undertaken to characterize the interaction of chimeric 7E3 Fab (c7E3 Fab) with endothelial cells, and to determine the 20 functional consequences, if any, of the interaction.

Antibodies The antibodies used in these studies included the following: Anti-GPIIb/IIIa chimeric 7E3 Fab (C-116E; an IgG1 Fab produced by papain digestion); anti-CD4 chimeric MT412 Fab used as an isotype-matched chimeric Fab fragment control (produced by cell line C128A; published International application number WO 91/10722); anti-E-selectin H18/7 F(ab') 2 (gift of M. Bevilaqua); anti-ICAM-1 #19 (gift of G. Riethmuller); anti-CD51 (AMAC), a monoclonal antibody which recognizes the a chain of the vitronectin receptor; anti-IIIa (AMAC), a monoclonal antibody which reacts with GPIIIa; murine 7E3 IgG; anti-7E3 -129is a rabbit, variable region specific anti-7E3 polyclonal antibody preparation; monoclonal antibody LM609 (gift of D.

A. Cheresh, Scripps Research Institute, La Jolla, CA), which binds to complexed a83 (vitronectin receptor) but does not bind to GPIIb/IIIa; monoclonal antibody 10E5, which reacts with GPIIb/IIIa but does not recognize endothelial cell a,83 (Centocor).

lodination of antibodies Chimeric 7E3 Fab and chimeric anti-CD4 MT412 Fab (cMT412 Fab) were filtered through a 0.22 micron, 13 mm filter unit (Millipore, Millex-GV #SLGV01305) prior to radiolabeling. Antibodies were radiolabeled with Na 1 2 (Amersham) using lodobeads (Pierce Chemicals, Rockford, IL) and passed through a Sephadex G25 column (Pharmacia 15 Sephadex G-25M), to remove unreacted '"I-iodide. The column had been blocked previously with 0.1% human serum albumin (Albuminar-25, Armour Pharmaceutical Co., Kankakee, IL) in phosphate buffered saline and equilibrated with 0.01% Tween 80-PBSS elution buffer. Antibodies were 0.22 20 micron filtered after iodinaton and antibody concentration determined by measuring the absorbance at 280 nm and using OD/mg mL 1 as the absorptivity coefficient.

Culture of HUVEC HUVEC at passage 1 were purchased from Cell Systems (Kirkland, WA) and cultured with serum-containing media (HUVEC media, Cell Systems) in 2% gelatin-coated tissue culture flasks until passage 4 at which time they were frozen at 5 x 106 cells/ml. For HUVEC binding and activation experiments, cells were thawed and seeded directly at -1 x 104 cells/well into 2% gelatin-coated 96-well tissue culture plates and grown for 3 to 5 days to confluence at whichtime they were assayed. For experiments -130examining HUVEC spreading and adhesion, cells were thawed into gelatin-coated T-150 tissue culture flasks and grown to -85% confluence. Cells were then trypsinized and seeded onto coated matrices as described below.

'~I-c7E3 Bindina HUVEC were seeded into 96-well removacell tissue culture plates (Dynatech) and grown to confluence. For saturation binding, 2 I-c7E3 Fab was diluted in HUVEC media containing 10% FCS (or serum free media if indicated). One ,set of cells was incubated with tracer in the presence of 0.02% sodium axide to prevent capping and internalization of tracer antibody. A 100-fold excess of cold c7E3 Fab was used to define non-specific binding. The cells were incubated with tracer for 4 hours at 370C, washed with 2 x 200 pl of media, and the wells were removed and radioactivity bound was quantitated using a gamma counter.

The number of cells per well was quantitated by trypsinization of sample wells and counting of cells using a hemacytometer. Assays were performed in triplicate 20 determinations. For Scatchard data analysis, bound '"I-c7E3 was plotted on the abscissa and the amount bound divided by the concentration of free antibody was plotted on the ordinate. A linear regression through the curve yielded a slope which was defined as the K. value. The intercept with the X-axis was defined as the or maximal amount of antibody bound. The B. value was converted to sites per cell using the following formula: modecaes bouncell B, (in fmo x (mot 10' 1 fnmo x (Avogadrds numbemonaleloell w For competition binding, methods were the same as those for the Scatchard analysis, but a constant, 1 pg/ml concentration of 25 I-c7E3 Fab was used with increasing -131concentrations of unlabeled competitor. The assays were done in HUVEC complete media containing 10% FCS.

Measurement of HUVEC activation A. E-selectin and ICAM-1 expression HUVEC were seeded into 96-well pop-out well tissue culture plates and grown to confluence. Cells were treated in 100 p1 of HUVEC complete media containing the indicated concentration of antibody for either 4 or 24 hours. TNFa ,(Genzyme) at 50 units/ml was used as a positive control to increase E-selectin and ICAM-1 expression. After incubation, media was removed and replaced with 50 pl of HUVEC complete media containing 1 pg/ml 2I-anti-E-selectin antibody (for the cells stimulated for 4 hours) or 1 pg/ml '"I-anti-ICAM-1 antibody (for the cells stimulated for 24 15 hours). The cells were incubated for 1 hour at 370C, washed with 2 x 200 pl of media, the wells were removed and radioactivity bound was quantitated with a gamma counter.

B. PMN adhesion to HUVEC HUVEC were seeded into 96-well pop-out tissue culture 20 plates and grown to confluence. Cells were treated in 100 li of HUVEC complete media containing the indicated concentration of antibody for either 4 or 24 hours. TNFa (Genzyme) at 50 units/mi was used as a positive control to increase E-selectin and ICAM-1 expression and therefore adhesiveness for polymorphonuclear leukocytes (PMNs). PMN were isolated from heparinized human blood using Monopoly Resolving Medium (Flow Labs). PMN were resuspended in 5 ml of RPMI and '"Indium labeled for 15 minutes at room temperature with 100 p1 "'Indium (Amersham). The cells were washed twice with 50 ml of RPMI and resuspended to 4 x 106/ml in RPMI containing 10% FCS. Media was removed from -132the HUVEC monolayer and 100 Ml of PMN were added to each well and incubated for 30 minutes at 370C. Unbound PMNs were removed by 2 x 200 pl washes with media. Bound PMNs were quantitated by counting of the wells in a gamma counter. The experiment was done with triplicate determinations.

C. Spreading and adhesion of HUVEC to coated surfaces HUVEC were seeded into 2% gelatin coated T-150 flasks and used when -85% confluent. Cells were briefly trypsinized, washed, and resuspended in HUVEC complete media at 3 x 105 cells/ml. The cells were treated with either c7E3 Fab or cMT412 Fab at 10 pg/ml and immediately seeded at 300 ll/well into either glass 8-chamber slides (NUNC #177402) or Permanox plastic 8-chamber slides (NUNC #177445) or 48-well tissue culture plastic plates (Corning) and placed in a 37"C CO, humidified incubator. Each type of surface had been precoated for 4 hours at room temperature with either fibronectin at 20 Ag/ml (Sigma F2006), fibrinogen at 40 pg/ml (Sigma F4883), or S 20 vitronectin at 20 Ag/ml (Sigma V8379) each diluted in PBS.

After coating, the wells were briefly rinsed twice with PBS. Photographs of the cells were taken at 6 and 24 hours o after seeding with a camera attached to an inverted phase contrast microscope.

Results Binding of c7E3 Fab to Human Endothelial Cells Affinity and number of antibodies bound per cell Scatchard analysis of saturation binding data (Figures 16 and 17A-17E) using 2 SI-c7E3 Fab revealed that the affinity of 125I-c7E3 for the human umbilical vein -133endothelial cells (HUVEC) was approximately K. 1 x 10' Mp (Table 21). The affinity did not change after stimulation of the cells with TNFa, a treatment which upregulates inflammation-mediating proteins such as E-selectin and ICAM-1 on the endothelial cells. The affinity also did not change when the assay was performed with serum-free media or in the presence of sodium axide to prevent capping and internalization (Table 21).

*Table 21 10 Summary of Scatchard Analysis of 25 I-c7E3 Fab Binding to Endothelial Cells S7E3 Fab bound/cell K (M-I) Unstimulated HUVEC 677,000 1.34 x Unstimulated HUVEC/serum-free 677,000 1.19 x 15 Unstimulated HUVEC azide 677,000 1.23 x HUVEC stimulated for 4 hrs with TNF 677,000 1.38 x HUVEC stimulated for 24 hrs with TNF 602,000 1.28 x HUVEC were incubated with increasing concentrations of 12I-labeled c7E3 Fab in the presence or absence of a 20 100-fold escess of cold c7E3 Fab to define non-specific binding. Data were analyzed using Scatchard analysis as described above.

Scatchard analysis also revealed that approximately 650,000 5 I-c7E3 Fab molecules were bound per endothelial cell.

This number did not change after stimulation of the cells with TNFa or in the presence of serum-free media or sodium azide. If 5 I-c7E3 Fab binds to a83 on endothelial cells, then there should be a similar number of sites per cell found with saturation binding analysis using LM609, an antibody which recognizes the a,03 complex. Scatchard analysis indicated that 300,000 LM609 antibodies were bound per cell (Figure 18).

-134- SDecificitv of 'sI-c7E3 binding to endothelial cells A competition binding experiment was performed to determine whether any other anti-GPIIb/IIIa, anti-aB3, or other antibodies could inhibit the binding of 1 2 I-c7E3 Fab (Figure 19). LM609, an antibody which binds to complexed a.

3 (vitronectin receptor) but does not bind to GPIIb/IIa, competed effectively for '5I-c7E3 Fab binding with an IC, of -0.03 ,g/ml. The murine IgG form of 7E3 and the chimeric Fab form of 7E3 competed for 1 I-c7E3 Fab binding with IC, values of approximately 0.2 and 1.0 pg/ml, respectively. A rabbit, variable region specific anti-7E3 antibody also blocked 'UI-c7E3 Fab binding with an IC, value of approximately 1.0 Ag/ml. An isotype-matched control Fab fragment MT412 (anti-CD4) did not compete for 15 '"I-c7E3 Fab binding. The 10E5 antibody which reacts with GPIIb/IIIa but does not recognize endothelial cell "GPIIb/IIIa" also did not compete. An anti-a, antibody (clone AMF7 purchased from AMAC) and an anti-IIIa antibody (clone SZ.21 purchased from AMAC) and the vitronectin 20 protein did not compete for 2 I-c7E3 Fab binding.

Effect of chimeric 7E3 Fab on HUVEC activation Endothelial cells can be activated by LPS, IL-1, and TNFa to express adhesion proteins such as E-selectin and ICAM-1. These adhesion proteins mediate the adhesion of leukocytes to the endothelium and allow their transmigration to sites of inflammation. In vitro, 4 hours of exposure to an activating agent provides a stimulus for optimal expression of E-selectin, and 24 hours of incubation is optimal for ICAM-1 expression. Incubation of HUVEC for either 4 or 24 hours with c7E3 Fab did not change either E-selectin or ICAM-1 expression (Figures 20A-20B) as measured by '"I-antibody binding. Chimeric 7E3 Fab treatment (0.01, 0.1, 1.0, 10, or 100 g/ml) or control Fab -135treatment 10, or 100 pg/ml chimeric MT412 Fab) of HUVEC also did not significantly increase adhesiveness of the HUVEC for PMN (Figures 21A-21B; 100 pg/ml shown).

Microscopically, the cells treated for 4 or 24 hours with up to 100 pg/al c7E3 Fab did not look any different from untreated cells, either before or after washing the monolayer.

Effect of c7E3 Fab on spreadina and adhesion of HUVEC to substrate-coated surfaces S 10 HUVEC were treated with either media alone and seeded on vitronectin-coated glass or plastic, or were treated with chimeric 7E3 Fab (10 Ag/ml) and seeded on (a) vitronectin-coated glass or plastic, fibrinogen-coated Sglass or plastic, or fibronectin-coated glass or plastic. In particular, HUVEC were trypsinized and treated at time 0 with 10 pg/ml c7E3 Fab. Immediately afterwards, cells were seeded onto chamber slides or 48-well tissue culture plates, and incubated at 370C.

Phase contrast photomicrographs were taken at 6 hours after 20 seeding with an inverted phase contrast microscope.

Untreated cells which were seeded on fibrinogen or fibronectin-coated plastic looked similar to untreated cells grown on vitronectin-coated plastic. Chimeric MT412-treated cells looked like untreated cells. Compared to untreated and cMT412 Fab treated cells, there was no effect of c7E3 Fab addition on HUVEC adhesion and spreading after 6 hours to fibronectin, vitronectin, or fibrinogen coated Permanox plastic, glass, or tissue culture plastic.

There was also no effect seen at 24 hours after cell seeding.

Chimeric 7E3 Fab bound to human endothelial cells with an affinity comparable to that of chimeric 7E3 Fab for GPIIb/IIIa on platelets (Ka 1.94 x 10' The affinity -136of c7E3 did not change in the presence of azide, indicating that internalization of the antibody did not occur. The affinity was also the same in serum-free media, suggesting that proteins in fetal calf serum did not alter the binding of c7E3 Fab to endothelial cells. Approximately 650,000 c7E3 Fab molecules bind to each endothelial cell and 80-100,000 c7E3 Fab molecules bind each platelet. Assuming that the surface area of an endothelial cell is 500 pm2 and that of a platelet is 22 Am 2 then the density of c7E3 Fab binding sites is approximately 4500/Am 2 for platelets and 000/Am 2 for endothelial cells. Therefore, although there appear to be many 7E3 binding sites on endothelial cells, the density of sites is less than one-quarter that of platelets.

15 Competition binding experiments suggested that c7E3 Fab bound specifically to endothelial cells through the vitronectin receptor, since a vitronectin receptor-specific antibody LM609 completely inhibited c7E3 Fab binding to endothelial cells. The number of vitronectin receptors on 20 endothelial cells measured by saturation binding with a vitronectin receptor-specific antibody was 300,000. This was half the number of sites bound by c7E3 Fab. This discrepancy could be attributed to the increased ability of the smaller c7E3 Fab to reach receptor sites unavailable to the LM609 IgG or bivalent binding of the LM609 IgG antibody.

Binding of c7E3 Fab did not appear to activate endothelial cells as measured by the upregulation of E-selectin or ICAM-1 adhesion proteins or by the ability of endothelial cells to bind PMNs. Binding of c7E3 Fab to endothelial cells immediately before seeding them onto vitronectin, fibronectin, or fibrinogen coated glass or plastic did not alter the cells' ability to spread and adhere to the surfaces.

-137- This finding is different from a previous report of inhibition by m7E3 IgG (20 ig/ml) of HUVEC adherence and spreading on fibrinogen- and vitronectin- coated glass (Charo et al., J. Biol. Chem. 262, 9935-9938, (1987)).

Possible explanations for these differences include: (1) the assay described here was done in serum (10% FCS) containing media, while the previous assay was done in serum-free media; c7E3 Fab was used in these assays, while m7E3 IgG was used in the other assay; 10 Ig/ml c7E3 Fab was used here, while 20 pg/ml m7E3 IgG was used "previously. However, in the HUVEC activation experiments described here, up to 100 pg/ml c7E3 Fab did not disrupt the already established HUVEC monolayer which had been seeded onto gelatin-coated tissue culture plastic.

15 In summary, chimeric 7E3 Fab binds to endothelial cell vitronectin receptors in vitro. Chimeric 7E3 Fab binds to approximately 650,000 sites per endothelial cell with an affinity of approximately K, 1 x 10 8 The antibody appears to bind specifically to endothelial cells through 20 the vitronectin receptor (aB3), since an aj-specific .antibody LM609 inhibited c7E3 Fab binding completely.

The binding of c7E3 Fab to endothelial cells did not appear to activate the cells, as assessed by the expression of activation-specific markers. In particular, E-selectin and intercellular adhesion molecule-1 (ICAM-1) expression was not increased upon treatment with c7E3 Fab.

Binding of the antibody does not appear to disrupt endothelial cell monolayers or prevent their establishment on matrix protein-coated surfaces. Exposure of HUVEC to c7E3 Fab did not alter the ability of HUVEC to spread and adhere to vitronectin, fibronectin, or fibrinogen coated surfaces. In addition, adhesiveness of human umbilical vein endothelial cells (HUVEC) for polymorphonuclear leukocytes (PMN) was not enhanced by antibody treatment.

-138- In summary, c7E3 binds to the vitronectin receptor on endothelial cells in vitro, and this binding appears not to activate the cells or affect their ability to adhere and spread on matrix proteins.

Euivalents Those skilled ascertain using no equivalents to the described herein.

10 encompassed by the in the art will recognize, or be able to more than routine experimentation, many specific embodiments of the invention Such equivalents are intended to be following claims.

rs

Claims (36)

1. A method of inhibiting stenosis and/or restenosis, comprising administering to a patient an effective amount of a compound which binds selectively to glycoprotein IIb/IIIa and the av43 vitronectin receptor.

2. The method of Claim 1 wherein the compound is an immunoglobulin or immunoglobulin fragment having 10 specificity for glycoprotein IIb/IIIa and the av#3 vitronectin receptor.

3. The method of Claim 2 wherein the immunoglobulin or immunoglobulin fragment having specificity for glycoprotein IIb/IIIa and the aV~3 vitronectin 15 receptor is the 7E3 monoclonal antibody or a portion thereof. A method of Claim 2 wherein the immunoglobulin or immunoglobulin fragment is a chimeric immunoglobulin or chimeric immunoglobulin fragment comprising an 20 antigen binding region of nonhuman origin and at least a portion of a human constant region. The method of Claim 4 wherein the antigen binding region is derived from an antibody having specificity for glycoprotein IIb/IIIa and the ayv3 vitronectin receptor the 7E3 monoclonal antibody).

6. The method of Claim 4, wherein the inmunoglobulin fragment is an Fab, Fab' or F(ab') 2 fragment. -140-

7. A method of inhibiting stenosis and/or restenosis, comprising administering to a patient having coronary artery disease an effective amount of a compound which binds selectively to glycoprotein IIb/IIIa and the av4 vitronectin receptor.

8. The method of Claim 7 wherein the compound is an immunoglobulin or inmunoglobulin fragment having specificity for glycoprotein IIb/IIla and the avy3 vitronectin receptor. C. *b*.a 10 9.

10.

11. The method of Claim 8 wherein the immunoglobulin or immunoglobulin fragment having specificity for glycoprotein IIb/IIIa and the av0 3 vitronectin receptor is the 7E3 monoclonal antibody or a portion thereof. A method of Claim 8 wherein the immunoglobulin or immunoglobulin fragment is a chimeric immunoglobulin or chimeric immunoglobulin fragment comprising an antigen binding region of nonhuman origin, and at least a portion of a human constant region. The method of Claim 10 wherein the antigen binding region is derived from an antibody having specificity for glycoprotein IIb/IIIa and the av#3 vitronectin receptor the 7E3 monoclonal antibody).

12. A method of inhibiting stenosis and/or restenosis following a coronary artery intervention procedure in a human, comprising administering to said human an effective amount of a chimeric immunoglobulin or immunoglobulin fragment comprising an antigen binding -141- region of nonhuman origin specific for platelets and a human constant region.

13. The method of Claim 12 wherein the procedure is angioplasty.

14. The method of Claim 12 wherein the procedure is placement of a stent. The method of Claim 12 wherein in the procedure comprises angioplasty and placement of a stent. 9. 10 16. The method of Claim 13 wherein the antigen binding region is specific for glycoprotein IIb/IIIa receptor.

17. The method of Claim 16 wherein the antigen binding region is derived from the monoclonal antibody 7E3.

18. The method of Claim 17 wherein the immunoglobulin 15 fragment is selected from the group consisting of Fab, Fab' and F(ab'),fragments.

19. A method of inhibiting stenosis and/or restenosis following a vascular intervention procedure in a human, comprising administering to said human an effective amount of a compound which binds selectively to glycoprotein IIb/IIIa. The method of Claim 19 wherein the procedure is placement of a stent.

21. The method of Claim 19 wherein the vascular intervention procedure is a coronary artery intervention procedure. -142-

22. The method of Claim 21 wherein the procedure is placement of a stent.

23. The method of Claim 21 wherein the procedure is angioplasty.

24. The method of Claim 21 wherein in the procedure comprises angioplasty and placement of a stent. The method of Claim 19 wherein the compound is an immunoglobulin or immunoglobulin fragment having specificity for glycoprotein IIb/IIIa. 10 26. The method of Claim 25 wherein the immunoglobulin or immunoglobulin fragment having specificity for glycoprotein IIb/IIIa is the 7E3 monoclonal antibody or a portion thereof.

27. A method of Claim 25 wherein the immunoglobulin or immunoglobulin fragment is a chimeric immunoglobulin or chimeric immunoglobulin fragment having specificity for glycoprotein IIb/IIIa comprising an antigen binding region of nonhuman origin and at least a portion of a human constant region.

28. The method of Claim 27 wherein the antigen binding region is derived from an antibody having specificity for glycoprotein IIb/IIIa the 7E3 monoclonal antibody).

29. The method of Claim 27 wherein the immunoglobulin fragment is an Fab, Fab' or F(ab') 2 fragment. -143- A method of inhibiting stenosis and/or restenosis following a vascular intervention procedure in a human, comprising administering to said human an effective amount of a compound which binds selectively to glycoprotein IIb/IIIa and the ayv 3 vitronectin receptor.

31. The method of Claim 30 wherein the procedure is placement of a stent. S2. The method of Claim 30 wherein the vascular 10 intervention procedure is a coronary artery intervention procedure.

33. The method of Claim 32 wherein the procedure is placement of a stent.

34. The method of Claim 32 wherein the procedure is angioplasty.

35. The method of Claim 32 wherein in the procedure comprises angioplasty and placement of a stent. S 36. The method of Claim 30 wherein the compound is an immunoglobulin or immunoglobulin fragment having specificity for glycoprotein IIb/IIIa and the aV8 3 vitronectin receptor.

37. The method of Claim 36 wherein the immunoglobulin or immunoglobulin fragment having specificity for glycoprotein IIb/IIIa and the a03 3 vitronectin receptor is the 7E3 monoclonal antibody or a portion thereof. -144- a. r a. a. r a

38. A method of Claim 36 wherein the immunoglobulin or immunoglobulin fragment is a chimeric immunoglobulin or chimeric immunoglobulin fragment comprising an antigen binding region of nonhuman origin, and at least a portion of a human constant region.

39. The method of Claim 38 wherein the antigen binding region is derived from an antibody having specificity for glycoprotein IIb/IIIa and the ovy3 vitronectin receptor the 7E3 monoclonal antibody).

40. The method of Claim 38 wherein the immunoglobulin fragment is an Fab, Fab' or F(ab') 2 fragment.

41. A method of reducing or preventing ischemic complications of angioplasty and/or stent placement in a human, comprising administering to said human an effective amount of a compound which binds selectively to glycoprotein IIb/IIIa.

42. A method of Claim 41 wherein the compound is an immunoglobulin or immunoglobulin fragment having specificity for glycoprotein IIb/IIIa.

43. A method of Claim 42 wherein the immunoglobulin or immunoglobulin fragment is a chimeric immunoglobulin or chimeric immunoglobulin fragment comprising an antigen binding region of nonhuman origin derived from an antibody having specificity for glycoprotein IIb/IIIa and at least a portion of a human constant region. -145-

44. A method of Claim 41 wherein the compound is a compound which binds selectively to glycoprotein IIb/IIIa and the av 3 vitronectin receptor. A method of Claim 44 wherein the compound is an immunoglobulin or immunoglobulin fragment having specificity for glycoprotein IIb/IIIa and the aoB, vitronectin receptor. -46. The method of Claim 45 wherein the immunoglobulin or immunoglobulin fragment having specificity for glycoprotein IIb/IIIa and the avS, vitronectin receptor is the 7E3 monoclonal antibody or a portion thereof. a*

47. A method of Claim 45 wherein the immunoglobulin or immunoglobulin fragment is a chimeric immunoglobulin 15 or chimeric immunoglobulin fragment comprising an antigen binding region of nonhuman origin and at least a portion of a human constant region.

48. The method of Claim 47 wherein the antigen binding region is derived from an antibody having specificity 20 for glycoprotein IIb/IIIa and the av03 vitronectin receptor the 7E3 monoclonal antibody).

49. The method of Claim 48 wherein the immunoglobulin fragment is an Fab, Fab' or F(ab') 2 fragment. DATED this 9th day of November, 1999 Centocor, Inc. AND The Research Foundation of State University of New York By DAVIES COLLISON CAVE Patent Attorneys for the Applicants