WO1994025581A1 - Modified recombinant tissue plasminogen activator - Google Patents

Abstract

The present invention provides a t-PA having substantially all plasminogen catalytic sites inactivated and which is able to bind to a thrombus and/or fibrin. The t-PA may be labelled. It also provides a composition for localising deep venous thrombosis using the compound of this invention, together with a pharmaceutically acceptable carrier, diluent and/or excipient. The invention also provides a process for preparing the compound of this invention, which process comprises reacting t-PA with an agent to inactivate the plasminogen catalytic site followed by cleaving disulfide bridges in the t-PA and labelling the thus-reduced t-PA. The invention specifically provides for the use of recombinant t-PA which is labelled with Tc-99m. Thus the compound of this invention is useful as a radiopharmaceutical for the non-invasive detection of deep vein thrombosis (DVT) in mammals. This radiopharmaceutical has specific thrombus localisation both in vitro and in vivo with an exceptionally fast blood clearance. These properties permit the reliable scintigraphic detection of thrombus in a rabbit model within 60 minutes post injection and in humans within 2hr-6hr post injection. In addition, as rt-PA is a product of recombinant DNA technology, it is unlikely to be antigenic and is thus capable of being used.

Description

Modified recombinant tissue plasminogen activator

Technical Field

This invention relates to compositions which allow in vivo thrombus localisation and methods of using those compositions. In particular, it relates to compositions containing Tc- 99m recombinant tissue plasminogen activator (rt-PA) wherein the t-PA possesses an inactivated plasminogen catalytic site and is able to bind to a fibrin clot. It also relates to methods for using this compound in localisation of thrombi.

Background Art

Deep venous thrombosis (DVT) is a major disease in the community. While DVT does not directly cause death itself, its major complication of pulmonary embolism (PE) is a lethal condition. Approximately, 5 million people each year in the United States have an episode of DVT; approximately 10% of these patients will develop PE and about 10% of these people will die as a direct consequence of PE (Moser 1990). In spite of the clinical awareness of this problem, DVT and PE are under-diagnosed with only 16% to 38% of patients dying from PE correctly diagnosed before death (Palevsky 1991). To this must be added the disappointing observation that our ability to diagnose PE has not significantly improved over the last 40 years (Anderson et al. 1989).

It has been well known for the past two decades that the clinical diagnosis of deep venous thrombosis is unreliable (O'Donnell et al. 1980; Salzman 1987). Consequently, there has been considerable effort over these years to devise the optimal technique for the accurate diagnosis of DVT. At the present time, the reference technique is contrast venography (CV) but this technique has the disadvantages of being^' expensive, invasive, painful (Bettman and Paulin 1977), thrombogenic (Albrechtsson and Olson 1976) and capable of inducing potentially lethal hypersensitivity reactions (Parfrey et al. 1989). In addition, there are some patients in whom venous cannulation is not possible, and technically inadequate studies in up to 10% of patients have been reported (Hull et al. 1985). Interpretation of CV requires considerable expertise and even in experienced hands there is considerable variability with observers disagreeing about the probable presence or absence of thrombosis at some site in the leg, in approximately 10% of examinations (McLaughlin et al. 1979). Ultrasound imaging (US) has recently been advocated as an alternative to CV

(Raghavendra et al. 1986; Cronan et al. 1987a; Appelman et al. 1987; Dauzat et al. 1987; Vogel et al. 1987; Lensing et al. 1989). While this technique has reported sensitivities and specificities in excess of 90% in detecting proximal vein thrombosis, it is significantly less successful in identifying isolated calf vein thrombosis (Cronan and Dorfman 1991 ; Pedersen et al. 1991). As up to 20% of isolated calf vein thrombosis propagate proximally and may further embolise to the lungs, (Moser and LeMoine 1981 ; Philbrick and Becker 1988) repeat US in negative cases has been proposed (Cronan et al. 1987b). Difficulty in identifying pelvic vein thrombosis has also been reported (Pedersen et al. 1991). Similarly to ultra sound imaging, impedance plethysmography (IPG) has sensitivities and specificity in the order of 95% and 96% (Palevsky and Alavi, 1991) for the detection of DVT at or proximal to the popliteal vein. However IPG has a low sensitivity for calf vein thrombi which may be largely due to the fact that the thrombi are smaller and less likely to be occlusive and that they are therefore less likely to be associated with outflow obstruction (Boris et al, 1989).

Various radionuclide techniques have been proposed as possible alternatives to CV. There is considerable interest in developing a radiolabelled monoclonal antibody (MoAb) directed against thrombus using either an anti-fibrin MoAb (Rosebrough et al. 1985; Rosebrough et al. 1987; Rosebrough et al. 1988; Knight et al. 1989; Jung et al. 1989; Lusiani et al. 1989; Alavi et al 1990; DeFaucal et al. 1991; Cerqueira et al. 1992; Lee et al. 1992) or an anti-platelet MoAb (Peters et al. 1986; Oster et al. 1985; Som et al. 1986; Stuttle et al. 1988). The principal limitation in this approach is the slow blood clearance of these radiotracers (in the order of hours to days depending on the MoAb used) as this precludes early imaging and hence severely restricts the clinical application of these radiotracers. As these tracers are murine in origin, there is the additional disadvantage that they may produce anaphylaxis and also induce human anti-mouse antibodies (HAMA) in the patient. This latter complication has been seen in up to 50% of patients (Reynolds et al. 1986) and may preclude the repeated use of this examination. In an attempt to overcome these problems, investigations have been carried out looking at the clot localising properties of radiolabelled recombinant tissue Plasminogen Activator following active site inhibition (Butler et al. 1991). Preliminary results with Indium-I l l labelled recombinant tissue Plasminogen Activator (In-I l l rt-PA) have produced superior results to the currently available radiolabelled anti-thrombus MoAbs. The half clearance time in the circulation of this radiotracer is 4-5 minutes and the absolute uptake into thrombus is comparable to these MoAbs.

Iodine labelled rt-PA, with active site inhibition, has recently been evaluated in a canine model of arterial thrombosis with good localisation of the tracer to thrombus (Ord et al 1992). In that study, thrombus to blood ratios of 8: 1 and 3: 1 were seen in pulmonary artery thrombi and carotid artery thrombi at 2 hours post injection. Planar scintigraphic visualisation of these thrombi, however, was not possible and Single Photon Emission Computed Tomography (SPECT) imaging to localise these thrombi was needed. This investigation also demonstrated that significantly higher thrombus to blood ratios were seen when the tracer was infused while the thrombus was being formed rather than after the thrombus had been formed. A preliminary report of 1-123 labelled rt-PA in 15 patients with DVT has shown that imaging of peripheral vein thrombosis is possible with this agent (Banyai et al. 1992). The main drawback with this procedure is that 1-123 is an expensive isotope that is not freely available.

Preliminary studies (Knight et al. 1992) with Fragment El, a derivative from plasmic digestion of cross-linked fibrin, radiolabelled with Technetium-99m have demonstrated its capacity to adequately bind to thrombus with rapid clearance of unbound radio tracer from the blood and soft-tissue background. Fragment El thrombus-to-blood ratio in dogs was seen to be very high at 15.7: 1 at 4 hours post injection. However, in a protein such as s fragment El, the structure depends on multiple disulfide bridges. It is then important to avoid reduction steps during the labelling procedure that would cleave these bridges and alter the structure and possibly the binding affinity of the protein. Not being able to use conventional Technetium-99m labelling procedures makes the production of this tracer long and tedious. In addition fragment El is not commercially available, and is complex to o produce.

Australian patent application 17889/88 discloses, inter alia, radiolabelled A chain of t-PA for detecting blood clots in vivo. It differs from this invention in that use of the A chain only is contemplated rather than the use of the complete t-PA. In addition, it discloses only the use of the chain labelled with iodine- 125. 5 Australian patent application 26128/88 claims a method for detecting a fibrin-platelet clot in vivo comprising administering to a patient a labelled thrombolytic protein whose clot- dissolving activity is reduced or eliminated and where the label is selectably attached to a portion of the thrombolytic protein other than the fibrin binding domain; followed by detecting the pattern of biodistribution of the labelled thrombolytic protein in the patient. In 0 one embodiment the invention describes a linker molecule (which is the inactivating molecule of this invention) to which is attached a label (which may be the radiolabel, Tc-99m of this invention), the whole then being linked to t-PA. In another embodiment, the specification describes a process for preparing radiolabelled t-PA where the t-PA is firstly inactivated and then labelled. However, this process may be distinguished from the present invention in that 5 labelling of the thrombolytic protein is carried out under different conditions and, based on data disclosed in 26128/88, the results in rabbits are far from satisfactory. No results are disclosed with human subjects.

Further, Australian patent application 62859/90 discloses a diagnostic reagent for the detection in vivo of increased release of a fibrinolytic enzyme or increased fibrinolytic activity 0 in a human or animal patient. The reagent comprises an antibody which is reactive to a fibrinolytic enzyme, or a fragment of the antibody, labelled with the substance which permits the detection in vivo of binding of the antibody to the fibrinolytic enzyme. The application discloses in one embodiment, increased clearance of the antibody provided by subsequent administration of the fibrinolytic enzyme with which the antibody is reactive. The preferred 5 fibrinolytic enzyme is t-PA.

Objects of the Invention It is an object of this invention to provide a composition and a method which will more effectively permit localisation of thrombi. It is a further object of this invention to provide a composition containing tissue plasminogen activator labelled with Tc-99m where the plasminogen catalytic site has been inactivated, and which is still able to bind to a fibrin clot

It is a further object to provide an improved method for visualising e boli and carcinomas, and for determining thrombus age.

Disclosure of the Invention It has been found that t-PA which has been labelled with Tc-99m and which has had the active catalytic site for conversion of plasminogen to plasmin, inhibited, has demonstrated promising results in localising deep venous thrombosis. The Tc-99m-rt-PA of this invention may also be used for localisation of DVT in other areas of the body eg upper limbs. It may also be used for the localisation and determination of severity of pulmonary embolism, determination of thrombus age, and localisation of tumours.

This invention thus describes the preparation and use of Tc-99m radiolabelled recombinant tissue Plasminogen Activator (rt-PA) for an improved method of non-invasive diagnosis of DVT. Thus, this invention provides a radio pharmaceutical which has specific thrombus localisation both in vitro and in vivo with an exceptionally fast blood clearance. Rt-PA is an agent expressing no antigenic properties, which once radiolabelled with Tc-99m promptly and avidly localises thrombi with rapid blood clearance.

A first embodiment of this invention provides a t-PA having substantially all plasminogen catalytic sites inactivated and which is able to bind to a thrombus and/or fibrin. A second embodiment of this invention provides a process for preparing t-PA having substantially all plasminogen catalytic sites inactivated and which is able to bind to a thrombus and/or fibrin, which process comprises reacting t-PA with an agent to inactivate substantially all the plasminogen catalytic sites and which does not effect the ability of the t- PA to bind to a thrombus or fibrin.

This invention also provides compositions containing the partially inactivated t-PA of this invention and methods for localising thrombi and methods for determining the severity of thrombi using said t-PA.

The t-PA may be labelled for example with Tc-99m and may be recombinant in origin. The t-PA may be dissolved in water which is preferably sterile. The concentration of t- PA may be from about 0.25% (w/v) to about 1% (w/v) and is preferably about 0.5% (w/v). This solution of t-PA may be dispensed into aliquots which may be frozen, preferably at - 20°C prior to further processing.

The plasminogen catalytic site is inactivated by suitable inactivating agents. Suitably the inactivating agent is a halomethylketone peptide. Examples of chloromethyleketone peptides are Ala-Ala-Phe-chloromethylketone, Ala-Ala-Pro-Val-chloromethylketone, Leu-chloromethylketone, MeOSuc-Ala-Ala-Pro-Val-chloromethylketone, MeOSuc-Ala-Ala- Pro-Ala-chloromethylketone, Phe-chloromethylketone, D-Phe-Phe-Arg-chloromethylketone, D-Phe-Pro-Arg-chloromethylketone(PPACK), D-Pro-Phe-Arg-chloromethylketone, Tos- Lys-chloromethylketone (TLCK), Tos-Phe-chloromethylketone (TPCK), D-Tyr-Pro-Arg-

SUBSTΠUTE SHEET (Rule 26) chloromethylketone, Z-Ala-Pro-Phe-chloromethylketone, Z-p-fluoro-Phe- chloromethylketone, Z-Gly-Gly-Phe-chloromethylketone, Z-Leu-chloromethylketone, Z- Leu-Tyr-chloromethylketone, Z-Leu-Tyr-chloromethylketone, Z-β-(2-naphthyl)-Ala- chloromethylketone, Z-Phe-chloromethylketone (ZPCK), p-chloro-Phe-OH, p-chloro-D- Phe-OH, p-chloro-DL-Phe-OH, p-chloro-D-Phe-OMe, p-chloro-DL-Phe-OMe, p-chloro- DL-Phe-Arg-OH, (R)-chloropropionic acid, (S)-chloropropionic acid, and 2-chloro-Z-OSu, where Z is benzyloxycarbonyl, and OSu is the N-hydroxysuccinimide ester. These compounds may be obtained from Bachem Bioscience Inc., Philadelphia, USA. We have found that inactivation is efficiently achieved by d-phenylalanine-1-proline-l-arginine chloromethylketone (PPACK). The suitable ratio of the t-PA to inactivating agent is 1:5 to 1 :30 and is preferably 1 : 15. Most preferably 275μL of the PPACK at a concentration of 2mg/mL is added to 5mg rt-PA. The t-PA may be reacted with the inactivating agent for a period between about 2 to about 24 hours and preferably about 2 hours.

Prior to radiolabelling the disulfide bridges in the t-PA may be cleaved by a suitable reducing agent such as dithiothreitol or 2-mercaptoethanol. Preferably the reducing agent is 2-mercaptoethanol. The concentration of reducing agent is from 1 :500 to 1 :3000 and is preferably 1:2000 to 1 :3000. More preferably the molar ratio of t-PA:2-mercaptoethanol is 1:2600. This reducing step is carried out for a period of about 30 to about 60min and is preferably carried out for a period of about 30min on a rotor between about 20 to about 25° C.

The t-PA with the inactivated plasminogen catalytic site is then purified by any standard purification technique and is suitably purified by column chromatography on P6-

DG. The eluate from this step may then be further labelled with a detectable marker forthwith; or stored for later labelling. Thus, the eluate may be frozen or lyophilised if the eluate is to be stored for later labelling.

The eluate from the previous column chromatography step may be further purified. For example, the reduced t-PA may be filtered through a low protein binding filter having a pore size, for example, of 0.22μm (Millipore, Bedford USA).

Examples of radiolabels used in diagnostic scintigraphy are Indium-I l l with a half-life of 2.8 days, gallium-67, gallium-68 and technetium-99m. The radiolabel is preferably Tc- 99m.

The thus-purified t-PA is labelled with Tc-99m by reaction with Tc-99m gluconate.

The amount of Tc-99m gluconate complex used is between about 1 to about 800MBq and is preferably about 500 to about 600MBq. This amount is added to reduced t-PA and reacted for a period of between about 20 to about 60min and preferably reacted for a period of about

30min.

Alternatively, a solution of a gluconate and a tin halide in oxygen-depleted water may be added to the eluate from the P6-DG column. This allows elimination of the filtration step through the low protein binding filter, described above and also allows a shorter incubation time with Tc-99m. The gluconate is preferably calcium gluconate and may suitably be present at a concentration of about 10% (w/v). The tin halide is preferably stannous halide and more preferably stannous fluoride which may suitably be present at a concentration of about 0.28% (w/v). The water which is used to dissolve the stannous halide is preferably sterile and is conveniently Water for Irrigation (Baxter, Sydney, Australia). This water should be oxygen-depleted which may be accomplished by any method known in the art, for example, by purging it with nitrogen gas, suitably for about 30 minutes.

When using this alternative method for labelling the t-PA, a kit is made by firstly dispensing the eluate from the P6-DG column chromatography step into aliquots in freeze dry vials. Then, gluconate solution is added to the reduced t-PA aliquot followed by addition of tin halide solution. This mixture may then be snap frozen, for example on dry ice and then lyophilised. The lyophilised aliquots may then be stored for example at 4°C, prior to further processing.

Prior to injection into the patient, Tc-99m in normal saline (i.e. 0.9% (w/v)) may be added to the lyophilised t-PA kit and incubated for about 2 minutes and preferably may be incubated for 10 minutes. The preparation is stable up to about 24 hours and the incubation time therefore may be extended to 24 hours. Incubation preferably takes place at room temperature. This Tc-99m t-PA is then ready for injection into the patient.

Thus, this alternative method provides the advantage of: (a) eliminating the final protein filtering step through a low protein binding filter; and

(b) reducing the reaction time with Tc-99m from approximately 30 minutes to less than 10 minutes.

Furthermore, by using this kit method, it is only necessary to carry out the procedure down to the P6-DG column chromatography in a sterile cabinet. The remaining steps in the procedure, that is labelling the rt-PA is carried out in the freeze dry vial and may therefore be carried out at the bench rather than in a biological safety cabinet.

Tc-99m rt-PA is suitably prepared in accordance with Example 1 A or Example 2.

The specific activity of t-PA is lOMBq/mg to lOOOMBq/mg and is preferably 300MBq/mg t-PA. The t-PA may also be labelled with a detectable marker for use in detection techniques known in this art. For example, the t-PA may be labelled with a marker suitable for use in magnetic resonance imaging, an example of which is gadolinium.

In this specification "mammal" includes humans.

The compositions of this invention may be administered to the mammal as a bolus injection or as an intravenous infusion. Administration by bolus injection is preferred. The dosage of Tc-99m may be in the range of about 200 to about 1000 and preferably in the range of about 500 to about 700 MBq and more preferably about 500MBq. Rabbit radiation dosimetry is used as a basis for deciding on an acceptable radiation dose for use in humans.

As a suitable inert carrier, 0.05M sodium acetate buffer (pH 4.5) may be used. Following injection of the Tc-99m t-PA, images may be obtained from 0 to 24 hours post-injection. Images obtained early (that is less than two hours) show considerable residual tracer in the blood and thus preclude definite localisation. In practice, the images are collected at between about 2 and about 6 hours after injection. The visualisation of the thrombus is achieved by using a gamma camera interfaced to a dedicated Nuclear Medicine computer. The Tc-99m t-PA is injected intravenously with scintigraphic imaging being performed between about 2 and about 6 hours following such injection. Acquisition uses a 256 x 256 word matrix with a high sensitivity collimator interfaced to an ICON Nuclear Medicine computer. Advantages of the composition and methods of this invention are as follows:

1. rt-PA is commercially available.

2. t-PA is naturally occurring in the body, therefore it is not subjected to antigenic reactions and a mammal can have repeated doses of it.

3. rt-PA has a fibrin binding site as well as a plasminogen catalytic site which causes thrombolysis. The catalytic site can be effectively inhibited without decreasing its ability to bind to fibrin.

4. rt-PA can be readily radiolabelled with Technetium-99m.

5. Technetium-99m exhibits properties that makes it the radioisotope of choice in clinical Nuclear Medicine. Technetium-99m is readily available, cheap and has an optimal gamma energy for most Nuclear Medicine cameras.

6. Tc-99m-rt-PA can rapidly localise DVT with a fast blood and soft tissue background clearance.

However, in comparison to contrast venography Tc-99m-rt-PA requires delayed imaging times. (2hr-6hr). This may be overcome with further mammal studies in order to determine the optimal imaging time.

Previous attempts have been made at preparing a t-PA preparation which was both labelled with Tc-99m and which contained an inactivated plasminogen catalytic site. For example, lmg of rt-PA (2mg/mL) was reacted with 1 lOμL of PPACK (30 PPACK: 1 rt-PA) for two hours at room temperature. To this mixture, lOμL of lOmg/mL diethylenetriaminepentaacetic acid (DTPA) dissolved in DMSO (20 DTPA: 1 rt-PA) and was reacted for 30 minutes at room temperature. This mixture was purified on a G50 column eluting with acetate buffer (pH 4.5). To this mixture was added 23μL of SnF2 (3mg/mL) in 0.01N HC1 (1 rt-PA: 1.5 SnF₂) and reacted at room temperature for 10 minutes. lOOMBq of Tc-99m in about 200-1000μL 0.9% (w/v) NaCl was added and reacted for a further 15 minutes at room temperature. The resultant rt-PA was purified on a P6DG column, eluting with acetate buffer (pH 4.5). Several different concentrations of DTPA:rt-PA were tried ranging from 20: 1 to 100: 1 with an optimal concentration of 20:1 being chosen. Also, different concentrations of DTPA to SnF₂ were attempted, ranging from 1.5:1 to 2.5:1 with an optimal of 1.5: 1 being used. This method was abandoned due to inconsistent labelling efficiency ranging from 5- 75% and widely varying clot to blood ratios ranging from 0.5 to 7.6: 1. In addition to this, scintographic localisation was poor and similarly inconsistent.

Several other modifications to the labelling process described herein were also attempted before gluconate was successfully chosen as the transchelating medium.

Pyrophosphate and methylene diphosphate were used instead of the gluconate, but these resulted in poor labelling efficiencies of 20% for the pyrophosphate and 45% for the methylene diphosphonate; gluconate produces a labelling efficiency of over 95%.

In addition, there have also been several attempts to modify the current labelling process.

Brief Description of the Drawings Figure 1. Fibrin binding results for Tc-99m-rt-PA as compared to the control 1-125- HSA.

Figure 2. Amidolytic activity assay results for inhibited Tc-99m-rt-PA as compared to active unlabelled rt-PA.

Figure 3. Gel electrophoresis of Tc-99m rt-PA (A), unlabelled rt-PA (B) and autoradiograph of Tc-99m rt-PA (C).

Figure 4. Planar anterior image of the neck of a high dose rabbit with a right-sided external vein thrombus (arrow) imaged at 60 minutes post injection. Figure 5. Blood clearance data for rabbit model expressed as percent activity over the left ventricle at first time point (n=6).

Figure 6. Scintigraphic image of patient with acute pain in lower right calf. Figure 7. Scintigraphic image of patient with DVT in left calf.

Figure 8. Blood clearance data expressed as percent activity of the first time point (n=3).

Best And Other Modes Of Carrying Out The Invention

This invention is illustrated by the following Examples without restriction to the scope of the invention:

Example 1A Preparation Of Radiolabelled Tracer 50 mg of sterile lyophilised rt-PA powder (Actilyse, Boehringer Ingelheim, Sydney,

Australia) was diluted with lOmL of sterile water. All the following procedures were carried out in a class III sterile biological safety cabinet. lmL aliquots were dispensed and held at - 20°C. 5mg rt-PA in lmL 0.75M arginine hydrochloride was reacted overnight with 275μl d-phenylalanine-1-proline-l-arginine chloromethylketone (2mg/mL, Bachem Bioscience Inc., Philadelphia, USA) at 4°C. 97μl of a 1 :6 dilution of 2-mercaptoethanol(2-ME):water (Sigma, St. Louis, USA) was mixed with the rt-PA solution for 30 minutes on a rotor in order to give a rt-PA:2-ME molar ratio of 1 :2600. While the reaction was proceeding, 2 x 5mL P6-DG(Bio-Rad, Richmond, USA) prepacked spinning columns were prepared for purification. Each column was mounted into lOmL sterile tubes. 20mL of 0.05M acetate buffer (pH 4.5) was allowed to drip through each column. The columns were then washed with l-2mL of 0.05M acetate buffer and packed for the purification step by centrifuging at 1000RPM for 2.5min. Finally the columns were washed with a further l-2mL of 0.05M acetate buffer and centrifuged at 1500RPM for 2.5min. 550μL of protein mixture is loaded onto each column. rt-PA was then purified by spinning the columns at 1500RPM for 2.5min. The purified samples were collected into two lOmL sterile tubes and stored at -20°C until required. lGBq of pertechnetate (ARI, Sydney, Australia) in l.OmL 0.9% (w/v) NaCl was added to a Tc-99m gluconate kidney scanning kit (ARI, Sydney, Australia) and reacted at room temperature for 30 minutes. Between 500-600MBq of the Tc-99m gluconate complex was then added to 2mg of reduced rt-PA which had been previously filtered through a 0.22μ m low protein binding filter (Millipore, Bedford, USA) and allowed to react for a further 30 minutes. Immediately following reaction time, radiolabelling efficiency assay was performed to ensure the consistent quality of the preparation.

Example IB In Vitro Assays

(A) Radiochemical Purity

Instant Thin Layer Paper Chromatography:

1 drop of the radiolabelled preparation of Example 1A was diluted into 3mL of Normal Saline (Astra, Sydney, Australia). On a 8cm ITLC-SG support (Gelman Sciences, Ann Arbor, USA), the origin was marked at 1cm from the bottom of the strip, then marked again at 6 cm. lμl of the diluted sample was placed at the origin. The strip was developed in a flat bottom tube containing approximately 2mL of Acetone. Once the solvent had reached the top of the strip it was cut into two at the 6cm mark and each half was counted separately in a gamma counter. This process was repeated using Normal Saline as the solvent. The results demonstrated that greater than 95% of the activity in each case was present in the lower half of the strip.

(B) Biological Activity Assay Fibrin Binding Assay: l.Oμg to lO.Oμg of Tc-99m rt-PA in its inhibited form were added to a solution of lmg human fibrinogen (Kabi Vitrum, Stockholm, Sweden) and 100IU Aprotonin (Bayer, Sydney, Australia) in l.OmL HEPES 0.05M (pH7.40). After 10 minutes of reaction at 37°C, 5U Thrombin (Parke Davis, Sydney, Australia) was added and reacted at 37°C for 60 minutes. The resulting clot was then carefully wound onto a wooden applicator stick and the clot and remaining solution counted in a Nal well counter. As a control, the identical assay was performed with l.Oμg- to lO.Oμg of 1-125 labelled human serum albumin (HSA) [Amersham International, Amersham, UK].

The mean (± SEM) for Tc-99m rt-PA was 50.5% (± 1.5%) and for 1-125 HSA was 7.0% (± 1.8%) [n=8]. Fibrin binding was significantly improved for Tc-99m rt-PA than with 1-125 HSA (P<0.001). Figure 1 illustrates these results. (C) Plasminogen Activation Site Inhibition Assay Amidolytic Activity Assay:

The plasminogen activating ability of Tc-99m rt-PA following P-PACK inhibition was assessed using an amidolytic assay. 2.5ng to lOng of unlabelled rt-PA were added to 400μg human fibrinogen (Kabi Vitrum, Stockholm, Sweden) and 0.2 IU plasminogen (Kabi Vitrum, Stockholm, Sweden) in lOOOμl 0.05M Tris buffer (pH 7.40) and reacted for 10 minutes. 100 μl of 5mM S-2251 (D-Ile-Pro-Arg-p-nitroaniline) [Kabi Vitrum, Stockholm, Sweden] was then added and the absorbance at 405nm was determined after a further 10 minute reaction, The identical assay was performed on the inhibited Tc-99m rt-PA except that approximately ten fold higher quantities were assayed.

Unlabelled rt-PA demonstrated a linear relationship between plasminogen activation and rt-PA over the studied range. No measurable amidolytic activity was detected with Tc- 99m rt-PA (n=10) in spite of a tenfold increase in Tc-99m rt-PA used (Figure 2). (D) Polymerisation/Fragmentation And Specific Activity Assays Gel Electrophoresis and Autoradiography:

Unlabelled rt-PA and Tc-99m rt-PA were analysed by electrophoresis according to the method of LaemmLi (LaemmLi 1970) using a 8% polyacrylamide and a Mini-Protean System

(Bio-Rad, Richmond, USA). The gel was then processed for autoradiography according to the method of Laskey et al. (Laskey and Mills 1975) using Kodak AR film (Rochester,

USA).

Figure 3 depicts the electrophoretic pattern of Tc-99m rt-PA (Lane A) and unlabelled rt-PA (Lane B). Both the Tc-99m rt-PA and unlabelled rt-PA have identical migration in the 55,000 to 66,000 Dalton range. The autoradiograph of Tc-99m rt-PA is depicted in Lane C, demonstrating that over 90% of the radioactivity is associated with rt-PA.

A small band in the 30,000 Dalton region was seen in the unlabelled rt-PA but not in the Tc-99m rt-PA. No significant polymerisation of Tc-99m rt-PA is seen.

Example 2 Kit Preparation

50mg of sterile lyophilised rt-PA powder, (Actilyse, Boehringer Ingelheim, Sydney, Australia), was reconstituted with lOmL of sterile water and dispensed into 500μL aliquots and stored at -20°C. 2.5mg rt-PA in 0.5mL 0.75M arginine hydrochloride was incubated for 4 hours with 69μL d-phenylalanine-l-proline-1 -arginine chloromethylketone (P-PACK), (2mg/mL, Bachem Bioscience Inc. Philadelphia, USA) at 4°C (P.PACK:rt-PA; 15:1 mole ratio). 48μL of a 1 :6 dilution of 2-mercaptoethanol(2ME):water (Sigma, St. Louis, USA) was mixed with the rt-PA solution for 30 minutes on a rotor in order to give a rt-PA:2-ME molar ratio of 1:2600. The mixture was then purified on a 5 x 1 cm spinning column of P6- DG (Pharmacia, Stockholm, Sweden) as in Example 1 and eluted with 0.05M acetate buffer (pH 4.5) and dispensed into freeze dry vials. 500mL Water for Irrigation (Baxter, Sydney, Australia) was purged with pure nitrogen for 30 minutes. 14mg stannous fluoride (SnF₂) (Sigma, St. Louis, USA) was added to the water. 150μL calcium gluconate, 10% (w/v) (David Bull Laboratories, Adelaide, Australia), then 500μL of the SnF₂ solution was added to the reduced rt-PA. lmL normal saline or lmL mannitol, 20% (w/v) was then added to the mixture. The mixture was then snap frozen on dry ice and freeze dried.

Radioactive Labelling of Kit Preparation

2.5mL Tc-99m (500MBq) in normal saline was added to the lyophilised rt-PA kit and incubated for 10 minutes at room temperature. The radiolabelled Tc-99m rt-PA was then ready for injection into the patient.

Example 3 In Vivo Studies

(A) Animal Model and Biodistribution Male New Zealand White rabbits (University of New South Wales, Sydney, Australia) [n=18] weighing 2kg, were anaesthetised with 35mg/kg ketamine (Apex Laboratories, Sydney, Australia) and 5mg kg xylazine (Bayer, Sydney, Australia) given intramuscularly. Supplemental doses were given as needed. A previously described model was used (Collen et al. 1983) with minor modifications. The external jugular vein was exposed through a 6- 8cm paramedian incision in the neck. The vein was mobilised over a distance of 4cm. A cotton thread was then introduced into the vein with an ordinary needle and the vein clamped proximally and distally. 50U of thrombin (Parke-Davis, Sydney, Australia) were then injected directly into the vein. After thirty minutes, all clamps were released. Three groups of animals were studied: Group I: lOOμg of inhibited Tc-99m rt-PA (mean injected dose [+SEM] was 20MBq [±2.1]) was administered as a bolus through femoral vein; Group II: lOOOμg of inhibited Tc-99m rt-PA (mean injected dose [±SEM] was 108MBq [±10.7]) was administered as a bolus through femoral vein; Group III: lOOμg of Tc-99m HSA (mean injected dose [±SEM] was 20MBq [±2.3) was administered as a bolus through femoral vein to serve as a control group. HSA was prepared according to a previously described method (Wong et al. 1978). At two hours post injection, the rabbits were then euthanised with intravenous pentobarbital (150mg/kg, Arnolds, Sydney, Australia). The segment of jugular vein containing thrombus was excised and cut open. The clot adherent to the thread was carefully removed and blotted dry. Blood and specimens of different organs were taken, blotted dry and counted in a Nal well counter with appropriate standards.

Significantly higher thrombus-to-blood ratios were seen with both low and high dose groups (PO.001) in comparison with the HSA Group, although no significant difference was demonstrated between the low and high dose groups (P=0.6). The thrombus-to-muscle ratios (mean ± SEM) were 53.6 (±11.8) for the low dose group, 61.6 (±4.9) for high dose group, and 13.2 (±4.1) for the HSA group. As with the thrombus-to-blood ratios, significantly higher muscle-to-blood ratios were detected in both low and high dose groups (P<.01) in comparison to HSA group, although no significant difference was seen between the low and high dose groups (P=0.6). Significantly higher accumulation of Tc-99m rt-PA was seen in the liver, spleen and muscle in the low dose group as compared to the high dose group (p<0.01).

(B) In Vivo Imaging in New Zealand White Rabbits.

At 2 hours post injection, low dose animals and HSA animals were euthanised with intravenous pentobarbital (150mg/kg, Arnolds, Sydney, Australia). 10 minute scans were then performed using a small field of view gamma camera interfaced to a dedicated Nuclear Medicine computer (GE STARCAM-300, Milwaukee, USA) fitted with a high resolution parallel hole collimator with the photopeak set at 140keV±10% with a 128 x 128 word matrix. With the high dose animals, dynamic scanning was performed with sequential five minute images (256 x 256, word matrix) collected commencing immediately following injection and continuing for 120 minutes. All images were viewed by two experienced Nuclear Medicine Physicians on a video screen with interactive display. In the low dose group, 245k counts (±43 k) were collected in each 10 minute image. 4 out of 6 thrombi were clearly visualised. The thrombi not visualised were significantly smaller than those visualised (mean: 164mg vs 30 mg, P<0.01). In the high dose group, all 6 thrombi were visualised between 45 and 60 minutes post injection (Figure 4). The weight of thrombi was 148mg (±19mg). The counts in each image in which the thrombus was first seen was 1240k counts (±128k). Blood Clearance Studies:

Blood clearance was calculated in the high dose animals by placing a region of interest (ROT) over the left ventricle and generating a time-activity curve. The counts in the ROI were normalised so that the first time point was 100%. The means were then graphed against time and then analysed using DeltaGraph Statistics program (Datapoint, Monterey, USA) for a diphasic exponential clearance in the form: _A = _Ale-βlt + A₂e-β2'

Analysis of the blood clearance curve demonstrated a satisfactory biexponential curve fit (P> 0.90) in the form: _y = 144<T°- 1⁷2' + 39.2e-° ⁰⁰⁹⁷' where y is the activity per cardiac pixel (expressed as a percentage of the first frame ROI drawn over the heart) and t is the time in minutes. This gives blood clearance half-times for the fast component of 4 minutes and 72 minutes for the slow component. The clearance curve are depicted in Figure 5 (C) In Vivo Imaging in Humans

Patient Selection Criteria:

To date this study presents the findings of 10 consecutive patients (5 male and 5 female). In each case the patient presented with known DVT, demonstrated first by contrast-

SUBSTTTUTE SHEET (Rule 26) venography. Each patient had the procedure explained to them and all signed a consent form.

Imaging Techniques: 500 MBq was administered as a bolus injection. Images were collected 2 and 6 hours post injection. Scintigraphy was performed using a large field of view gamma camera (DIACAM, Siemens Gammasonics, USA) fitted with a high resolution parallel hole collimator. Digital images were acquired using a 128 x 128 word matrix size and displayed on an ICON computer. The images were collected using a symmetrical 15% window at 140keV. Images were collected for 10 min/view of the anterior pelvis, anterior femur and the posterior calves. Review of Image: All images were interpreted by two experienced observers. A scan was considered a positive result for DVT if there was accumulation of abnormal tracer activity at any time point. From this small group of patients all 10DVT have been clearly visualised. These preliminary findings have demonstrated that the ability of Tc-99m-rt-PA to detect DVT was not affected by medication (eg. anti-coagulants) or the age of the thrombosis. Reference is made to Figs. 6 and 7.

Fig. 6 refers to a patient who presented with an acute pain in his lower right calf which was 2 days old. Contrast venogram was positive for DVT in this region. 10 min static views were taken at 2 hours and 6 hours post injection of Tc-99m rt-PA. Anterior femur, anterior pelvis and posterior calves views were taken at 2 hours and anterior femur and posterior calf views were taken at 6 hours. Both the 2 hour and 6 hour images of the posterior calf demonstrated an abnormal accumulation of activity.

Fig. 7 refers to a patient who had a positive contrast venogram 13 days prior to this radionuclide study. 10 min static views were taken at 2 hours and 6 hours post injection of Tc-99m rt-PA. Anterior femur, anterior pelvis and posterior calves views were taken at 2 hours and 6 hours. At 6 hours a DVT in his left calf was clearly demonstrated. Blood Clearance: Blood clearance was calculated in three patients by taking blood samples every 2 min for the first 20 minutes post injection, then samples at 25, 30min, lhr, 2hr and 6 hr post injection. The samples were weighed then counted in a Nal well counter with appropriate standards. The counts were normalised so that the first time point was 100%. The means were then graphed against time and then analysed using DeltaGraph Statistics program (Datapoint, Monterey, USA) for a diphasic exponential clearance in the form:

I A = A₁e-β¹' + A e-β²'

Analysis of the blood clearance curve demonstrated a satisfactory biexponential curve fit in the form: _{y = 17e}-0-004t _{+ 1}04e-0 012t where y is the activity per minute (expressed as a percentage of the first blood sample) and t is the time in minutes. This gives blood clearance half-times for the fast component of 6 minutes and 180 minutes for the slow component. The clearance curve is depicted in Figure 8. Example 4 Results of Phase II Clinical Trial

This trial was performed using Tc-99m-rt-PA that had been prepared including the final purification step described above by filtering through a low protein binding filter having a pore size of 0.22μm (Millipore, Bedford, USA). (In vitro analysis has not revealed any significant difference between the two preparations).

Patients and Methods

Twenty consecutive patients (10 women and 10 men; average age 63 years range 39- 80 years) with venographically demonstrable deep venous thrombosis of a lower limb underwent scintigraphy using Tc-99m-rt-PA. The mean time between contrast venography and scintigraphy was 1.8 days (range: 4 hours to 4 days). In three of these of these patents, sequential blood samples were taken at specified intervals in order to calculate the blood clearance of Tc-99m-rt-PA. Informed consent was obtained from each patient and the study protocol had been approved by the local ethics committee. Seventeen patients could accurately determine the time of onset of their symptoms. The remaining patients were asymptomatic. In symptomatic patients, symptoms occurred on an average of-4.5 days (range 1-11 days) prior to the contrast venogram. All patients were receiving intravenous heparin when the scintigraphy was performed.

Preparation of Tc-99m-rt-PA

Each patient received 1.0-1.5 mg Tc-99m-rt-PA with an average activity of 560MBq (range 490-790MBq). Instant thin layer paper chromatography was performed prior to injection using ITLC-SG support (Gelman Sciences, Ann Arbor, USA) and acetone as the eluent. At all times, greater then 95% of the activity remained protein bound.

Contrast Venography

This was performed according to standard methods. 50-lOOmL of non ionic contrast was injected into a dorsal vein of the foot suspected of DVT following the placement of tourniquets at the ankle and the calf. Anterior, lateral oblique, medial oblique radiographs of the calf were obtained together with anterior views of the thigh. Additional views were taken as needed. Venograms were viewed by a vascular radiologist who was blinded to all other investigations. The criterion for thrombosis was the presence of an intraluminal filling surrounded by contrast medium outlining the vessel wall. Thromboses were categorised as either proximal (involving the popliteal, the femoral vein or both) or distal (involving only the calf veins).

Scintigraphy Imaging

Images were obtained with a Siemens Diacam gamma camera (Siemans Gammasonics, USA) interfaced to a Siemens ICON computer (Siemens Gammasonics, USA). A high resolution collimator was used with the data acquired in 256 by 256 format with a 15% asymmetric energy window centered on the Tc-99m photopeak. Patients were imaged in the supine position. Two images of the lower legs were performed at 2 hours and 6 hours post injection. Both images were for 10 minutes with the first being an anterior view of the thighs and the second being a posterior view of the calves and knees. In the 2 hours images, the mean (± SEM) number of counts in the thigh images was 580k (± 60k), and the calf was 280k (± 30k); in the 6 hour images the counts were 370k (± 50k) and 180k (± 30k). One patient was only scanned at 6 hours due to her clinical condition. All scans were viewed by two Nuclear Medicine physicians blinded to the results on contrast venography. Images were viewed on video scan using an interactive gray scale threshold. The criterion for thrombosis was unequivocal increased tracer accumulation in comparison with the contralateral side. Thromboses were categorised as either proximal (involving the popliteal, the femoral vein or both) or distal (involving only the calf veins). Blood Clearance

Following injection of Tc-99m-rt-PA, 0.5mL of blood was taken from the contralateral arm at 2, 4, 6, 8,10, 12, 14, 16, 18, 20, 25, 30, 60, 120 and 360 minutes. The blood was taken into weighed tubes and the radioactivity measured in a Nal well counter (LKB Wallac, Finland).

Results

Contrast Venography

All venograms were considered to be of diagnostic quality. 36 thromboses were identified in the 20 patients. 19 were categorised as being distal in location and 17 as proximal. Scintigraphic Imaging

In the 2 hour images, 17 of the 19 patients had abnormal scans; in the 6 hour images, all 20 patients had abnormal scans. In the 2 hour scans, 15 abnormalities were seen in the distal veins and 14 in the proximal veins. In the 6 hour scans, 18 abnormalities were seen in the distal veins and 20 in the proximal veins. Comparison was performed on a site by site basis using the contrast venogram as the accepted definitive study. In the 2 hour scan, 15 of 19 distal thromboses were identified and 12 of 17 proximal thromboses (sensitivity 75% for both distal and proximal thromboses). In the 6 hour scan, 18 of 19 distal thromboses were identified and 17 out of 17 proximal thromboses (sensitivity 97% for both distal and proximal thromboses). A further comparison was performed to measure the agreement between venography and scintigraphy as to the presence or absence of thrombosis. In the 2 hour scans, there was agreement at 28 of 38 sites; in the 6 hour scans, there was agreement at 36 of the 40 sites. In the 6 hour scans, one distal vein thrombosis was not identified and 3 proximal vein abnormalities were seen on scintigraphy that were free of thrombosis on contrast venography. Blood Clearance

The counts were normalised so that the first time point was 100%. The means were then graphed against time and then analysed using DeltaGraph Statistics program (Datapoint, 5 Monterey, USA) for a biphasic exponential clearance in the form:

Analysis of the blood clearance curve demonstrated a satisfactory biexponential curve fit in the form: _{y = 17e}-0-004t _{+ 10}4e-0 012t

lo where y is the activity per minute (expressed as a percentage of the first blood sample) and t is the time in minutes (Fig. 8). This gives blood clearance half-times for the fast component of 6 minutes and 180 minutes for the slow component.

Summary of Phase II Clinical Trial.

At two hours post Injection, Tc-99m-rt-PA has a sensitivity of 75% for detecting deep is venous thrombosis. At six hours post injection, this sensitivity is 97% for deep venous thrombosis detection.

Example 5 Results of Phase III Clinical Trial Patients and Method

Fifty seven consecutive patients (mean age 67 years, range 27-88 years; 31 females and 20 26 males) with suspected deep venous thrombosis of a lower limb, underwent contrast venography and scintigraphic scanning with Tc-99m-rt-PA. The mean time between contrast venography and scintigraphy was 0.8 days (range: 1 hour to 3 days). 42 patients could accurately determine the time of onset of their symptoms. Symptoms (leg pain and/or leg swelling) occurred on an average of 6 days (range 0-42 days) prior to the contrast venogram. 25 Informed consent was obtained from each patient and the study protocol had been approved by the local ethics committee. 27 patients were receiving intravenous heparin when the scintigraphy was performed.

Preparation of Radiolabelled Tracer

Each patient received 1.5mg Tc-99m-rt-PA with an average activity of 64OMBq 30 (range 460-700MBq). Instant thin layer paper chromatography was performed prior to injection using ITLC-SG support (Gelman Sciences, Ann Arbor, USA) and acetone as the eluent. At all times, greater then 98% of the activity remained protein bound.

Contrast Venography

This was performed according to standard methods. 50-lOOmL of non ionic contrast

SυβSΗTUTE SHEET (Rule 26) was injected into a dorsal vein of the foot suspected of DVT following the placement of tourniquets at the ankle and the calf. Anterior, lateral oblique, medial oblique radiographs of the calf were obtained together with anterior views of the thigh. Additional views were taken as needed. Venograms were viewed by two radiologists who were blinded to all other investigations. The criterion for thrombosis was the presence of an intraluminal filling defect surrounded by contrast medium outlining the vessel wall. Venograms were further analysed on a patient by patient basis and a segment by segment basis.

(i) Patient by Patient

Patients were assigned to one of four categories: patients with no thrombosis, those with proximal vein thrombosis (thrombosis in the femoral or popliteal vein with or without concurrent calf vein thrombosis), those with isolated calf vein thrombosis and those whose results were inadequate for interpretation.

(ii) Segment by Segment

Venograms were divided into two segments: proximal (involving the popliteal and the femoral vein) or distal (involving only the calf veins). All segments were classified as being normal, positive for thrombosis or uninterpretable. All segments in which there was disagreement between the radiologists were then classified by a third radiologist using the same criteria.

Scintigraphic Imaging Images were obtained with a Siemens Diacam gamma camera (Siemens Gammasonics,

USA) interfaced to a Siemens ICON computer (Siemens Gammasonics, USA). A high resolution collimator was used with the data acquired in 256 by 256 format with a 15% asymmetric energy window centered on the Tc-99m photopeak. Imaging was performed at 4 hours post injection. An anterior view of the thighs and a posterior view of the calves and knees were obtained at 4 hours post injection. Imaging was performed for ten minutes per image. All scans were viewed by two Nuclear Medicine physicians blinded to the results on contrast venography. Images were viewed on video monitor using an interactive gray scale threshold. The positive criteria were as follows:-

1. femoral vein: increased tracer accumulation in comparison with the contralateral side such that the increase was equal to or greater than adjacent bone marrow accumulation.

2. calf veins: increased tracer accumulation in comparison with the contralateral side.

3. popliteal vein: increased tracer accumulation in comparison with the contralateral side in a patient with co-existing calf or femoral thrombosis. Results Contrast Venography (i) Patient by Patient Analysis.

In this analysis, 44 patients had venograms that were diagnostic of proximal vein 5 thrombosis, calf vein thrombosis or completely normal. 10 patients had proximal vein thrombosis, 13 had isolated calf vein thrombosis and 21 patients had completely normal venograms.

(ii) Segment by Segment Analysis

In this analysis, 57 venograms were evaluated with 18 segments seen to be lo uninterpretable (7 in the proximal segment and 11 in the calf segment). 96 segments were able to be analysed. Thromboses were seen in 33 segments (10 proximal and 23 calf)-

Scintigraphic Imaging

All scans were considered to be of diagnostic quality. (i) Patient by Patient Analysis is 9 of the 10 patients with proximal vein thrombosis had positive scans in the femoral vein, popliteal vein or both. 11 of the 13 patients with isolated calf vein thrombosis had positive scans in the calf veins. 21 of the 24 patients with no thrombosis had negative scans. False positive scans were seen in the proximal vein in 1 patient and in the calf vein of 2 patients. 20 Thus in proximal vein thrombosis, Tc-99m-rt-PA has a sensitivity of 90% and a specificity of 95%. In isolated calf vein thrombosis, Tc-99m-rt-PA has a sensitivity of 85% and a specificity of 91%.

(ii) Segment by Segment Analysis

Of the 33 thrombosed segments, 29 had positive scans. In the 4 false negative scans, 25 one was a proximal vein thombosis, two were isolated calf vein thromboses and one a calf vein thrombosis in a patient with proximal vein thrombosis. In the 63 non thrombosed segments, 57 had negative scans. In the 6 false positive scans, 1 was a proximal vein thrombosis, 2 were isolated calf vein thromboses and 3 were positive in the popliteal veins of a patient with calf vein thrombosis. 30 Thus concerning all segmental thromboses, Tc-99m-rt-PA has a sensitivity of 85% and a specificity of 90%.

If the case of the false negative Tc-99m-rt-PA in , the patient with proximal-vein thrombosis is not included and similarly with the 3 false positive Tc-99m-rt-PA scans in the popliteal veins in patients with calf vein thrombosis, then the sensitivity of Tc-99m-rt-PA is 35 90% and the specificity is 95%.

Summary of Phase IHClinical Trial

In proximal vein thrombosis, Tc-99m-rt-PA has a sensitivity of 90% and a specificity of 95%. In isolated calf vein thrombosis, Tc-99m-rt-PA has a sensitivity of 85% and a specificity of 91%.

Example 6 Selected Haematological and Biochemical Parameters - Phase III Clinical Trial

Various haematological and biochemical parameters were available for some patients who participated in the Phase III Clinical Trial and these values are listed below. Standard analysis was performed using Student's 't' test.

Biochemistry and Haematology

Variance between serum levels before injection and 24 hours after injection.

Mean Before Mean Population P-Value Injection 24 Hours PI

White Cell 1 1 .77 1 1 .78 30 0.98

Haemoglobin 120 121 32 0.534

Platelets 277 299 30 0.037

Albumin 39 37 11 0.372

ALP 189 201 1 1 0.282

AST 43 50 11 0.167

GGT 152 176 11 0.085

Bicarbonate 26 26 22 0.24

Creatinine 0.104 0.104 23 0.932

Chloride 99 99 22 0.909

Sodium 138 138 23 0.861

Potassium 4.2 4.1 23 0.181

Urea 7.8 8.2 22 0.252

A significant increase was observed in the serum platelet count. In view of the clinical condition of these patients (i.e. vendus thrombosis, anticoagulation, hospitalisation) this elevation is not definitely attributable to Tc-99m-rt-PA.

Variance between serum levels before injection and 10 days after injection.

Mean Before Mean Population P-Value Injection 10 Days PI

White Cell 13.28 13.52 5 0.844

Haemoglobin 122 120 22 0.461

Platelets 264 307 21 0.053

Albumin 39 36 5 0.651

ALP 143 125 5 0.174

AST 81 34 5 0.476 GGT 57 60 5 0.747

Bicarbonate 27 26 16 0.542

Creatinine 0.107 0.111 16 0.703

Chloride 98 100 16 0.053

Sodium 138 137 16 0.524

Potassium 4.2 4.5 16 0.048

Urea 8.5 8.1 16 0.510

A significant change was observed in the serum potassium levels from 4.2mmol L to 4.5 mmol L. This is not clinically significant.

Summary

From the data available, there is no definite effect of Tc-99m-rt-PA on standard s haematological or biochemical parameters.

Example 7 Toxicological Data

(i) Normal Volunteers

Five normal volunteers were injected with Tc-99m-rt-PA with vital signs, haematological, biochemical and urinalysis parameters monitored over a 24 hour period. The o variance in biochemistry, haematology, vital signs and urinalysis are listed below. Statistical analysis was performed using Student 't' test.

Biochemistry and Haematology

Mean Before Mean 24 Hours P-value Injection PI

White Cell 6.7 6 1 0.431

Haemoglobin 133 139 0.182

Platelets 213 225 0.184

Albumin 43 44 0.512

ALP 71 188 0.38

AST 24 68 0.367

GGT 27 27 1

Bicarbonate 26 27 0.245

Creatinine 0.086 0.09 0.099

Chloride 105 104 0.501

Sodium 140 142 0.004

Potassium 4 5 0.116

Urea 5.6 5 0.141

Total Bilirubin 10.4 10.6 0.883 A significant change was observed in the serum sodium levels. This is not clinically significant.

Vital Signs

P-Value P-Value

Before Injection v's 5 min Before injection v's 30

PI min PI

BP 0.578 0.179

Heart Rate 1 0.141

Resp. Rate 0.374 0.704

Temp. 0.374 0.621

No significant changes were observed.

Urinalysis

Changes Observed 24 hrs PI

Bilirubin No changes observed

Blood One patient only demonstrated a "negative" result changing to a "trace" result.

Glucose No changes observed

Ketone No changes observed

Leucocytes No changes observed

Protein No changes observed

Specific Gravity No significant change (P-value, 0.252)

Total Bilirubin No significant change (P-value, 0.883)

Urobilinogen No changes observed

No significant changes were observed.

References

Alavi A, Palevsky HI, Gupta N et al. Radiolabelled antifibrin antibody in the detection of venous thrombosis:preliminary results. Radiol 1990; 175:79-85. Albrechtsson U, Olson CG. Thrombotic side effects of lower limb venography. Lancet 1976;

1:723-724.

Anderson RE, Hill RB, Key CR. The sensitivity and specificity of clinical diagnostics during five decades. JAMA 1989; 261 : 1610-1617.

Appelman PT, De Jong TE, Lampmann LE. Deep venous thrombosis of the leg US evaluation. Radiol 1987; 163:743-746.

Banyai M, Virgolini I, Rauscha F, Angelberger P, Sinzinger H. Imaging of peripheral venous thrombosis with 1-123 labelled recombinant tissue plasminogen activator. Nuc Med Commun

1992; 13:233.

Boris LC, Christiansen HM, Lassen MR et al. Comparison of real time B mode ultra sonography and bilateral ascending phlebography for detection of post operative deep vein thrombosis following elective hip surgery. Thromb Haemost 1989;61 :363-365.

Butler SP, Kader KL, Owen J, Wang TST, Fawwaz RA, Alderson PO. Rapid localisation of

In-III labelled recombinant tissue plasminogen activator in a rabbit thrombosi model. J Nuc 5 Med 1991; 32:461-467.

Cerqueira MD, Stratton JR, Vracko R, Schaible TF, Ritchie JL. Non invasive arterial thrombus imaging with Tc-99m monoclonal antifibrin antibody. Circulation 1992; 85:298-

304.

Cronan JJ, Dorf an GS, Scola FH, Schepps B, Alexander J. Deep venous thrombosis: US lo assessment using compression Radiol 1987a; 162 : 191 - 194.

Cronan JJ, Dorfman GS, Scola PH. Deep venous thrombosis: US assessment using vein compression. Radiol 1987b; 168: 100-107.

Cronan JJ, Dorfman GS. Advances in ultrasound imaging of venous thrombosis. Semin Nuc

Med 1991; 21 :297-312. is Dauzat MM, Laroche JP, Charras et al. Real-time B-mode ultrasonography for better specificity irfthe non-invasive diagnosis of deep venous thrombosis. J. Ultrasound Med 1987;

6:625-631.

DeFaucal P, Peltier P, Planchon B et al. Evaluation of In-III labelled antifibrin monoclonal antibody for the diagnosis of venous thrombotic disease. J Nuc Med 1991; 32:785-795. 20 Hull RD, Hirsh J, Carter CJ et al. Diagnostic efficacy of impedance plethysmography for clinically suspected deep venous thrombosis. Ann Int Med 1985; 102: 21-28.

Jung M, Kletter K, Dudczak R et al. Deep vein thrombosis:scintigraphic diagnosis with In-III labelled monoclonal antifibrin antibodies. Radiol 1989;173:469-475.

Knight LC, Maurer AH, Ammar IA et al. Tc-99m antifibrin Fab¹ fragments for imaging 25 venous thrombi: evaluation in a canine model. Radiology 1989; 173: 163-169.

Knight LC, Abrams MJ, Schwartz DA et al. Preparation and Preliminary Evaluation of

Technetium-99m-Labelled Fragment El for Thrombus Imaging. J Nuc Med 1992;33:710-

715.

Knight LC, Scintigraphic Methods for Detecting Vascular Thrombus. J Nuc Med 30 1993;34:554-561.

LaemmLi UK. Cleavage of structural proteins during assembly of the head of bacteriophage

T4.Nature 1970; 227:680-685.

Laskey RA, Mills AD. Quantitative film detection of H-3 and C-14 in polyacrylamide gels in fluorography. Eur J Biochem 1975; 56:335-339. 35 Lee FT, Milner LJ, Boniface GR et al. Evaluation of thrombus detection in a rabbit model using a technetium-99m-labelled anti-fibrin monoclonal antibody. Immunology and Cell

Biology 1992;70: 173-170.

Lensing AW, Prandoni P, Brandjes D et al. Detection of deep-vein thrombosis by real-time

B-mode ultrasonography. N Eng J Med 1989; 320:342-345. Lusiani L, Zanco P, Visona A et al. Immunoscintigrapic detection of venous thrombosis of the lower extremities by means of human antifibrin monoclonal antibodies labelled with In- III. Angiology 1989; 40:671-677.

McLaughlin MS, Thomson JG, Taylor DW, Kelly ME, Sackett DL. Observer variation in the interpretation of lower limb venograms. AJR 1979; 13.

Moser, KM. Venous thromboembolism. Am Rev Respir Dis 1990; 141:235-249.

Moser KM, LeMoine JR. Is embolic risk conditioned by location of deep venous thrombosis?

Ann Intern Med 1981; 94:439-444.

O'Donnell TF, Abbott WM, Athanasoulis CA, Millan VG, Callon AD. Diagnosis of deep venous thrombosis in the outpatient by venography. Surg Gynaecol Obstet 1980;150:69-74.

Ord JM, Hasapes J, Daugherty A, Thorpe SR, Bergmann SR, Sobel BE. Imaging of thrombi with tissue-type plasminogen activator rendered enzymatically inactive and conjugated to a residualising label. Circ 1992; 85:288-297.

Oster ZH, Srivasta SC, Som P el al. Thrombus radioimmunoscintigraphy: an approach using monoclonal antiplatelet antibody. Proc Natl Acad Sci USA 1985; 82:3465-68.

Palevsky HI. The problems of the clinical and laboratory diagnosis of pulmonary embolism.

Sem Nuc Med 1991; 21 :276-280.

Parfrey PS, Griffiths SM, Barrett BJ et al. Contrast material induced renal failure in patients with diabetes mellitus, renal insufficiency or both. N Eng J Med 1989;320: 143-149. Pedersen OM, Aslaksen A, Vik-Mo H, Bassoa AM. Compression ultrasonography in hospitalised patients with suspected deep venous thrombosis. Arch Int Med 1991; 151:2217-

20.

Peters AM, Lavender JP, Needham SG et al. Imaging thrombus with radiolabelled monoclonal antibody to platelets. Brit J Med 1986; 293:1525-1527. Philbrick JT, Becker DM. Calf deep venous thrombosis: a wolf in sheep's clothing? Arch

Intern Med 1988; 148:2131-38.

Raghavendra BN, Horii SC, Hilton S, Subramanyan BR, Rosen RJ, Lam S. Deep venous thrombosis: detection by probe compression of veins. J. Ultrasound Med 1986; 5:89-95.

Reynolds JC, Carrasquillo JA, Keenan AM et al. Human anti-murine antibodies following immunoscintigraphy or therapy with radiolabelled monoclonal antibodies J. Nuc Med 1986;

27:1022-23.

Rosebrough SF, Kudryk B, Grossman ZD et al. Radioimmunoimaging of venous thrombi using Iodine-131 monoclonal antibody. Radiol 1985; 156:515-517.

Rosebrough SF, Grossman ZD, McAfee JG et al. Aged venous thrombi: radioimaging with fibrin specific monoclonal antibody. Radiol 1987; 162:575-577.

Rosebrough SF, Grossman ZD, McAfee JG et al. Thrombus imaging with Indium-III and

Iodine-131-labelled fibrin specific monoclonal antibody and its F(ab')2 and Fab fragments. J.

Nuc Med 1988; 29:1212-22.

Salzman EW. Venous thrombosis made easy. N Eng J Med 1987; 314:847-848. 2:227-229.

SUBSTTTUTΈ SHEET (Rule 26) Som P, Oster ZH, Zamora PO et al. Radioimmunoimaging of experimental thrombi in dogs using Tc-99m labelled monoclonal antibody fragments reactive with human platelets. J Nuc Med 1986; 27:1315-1320.

Stuttle AWJ, Ritter JM, Peters AM, et al. In vitro studies with an antiplatelet monoclonal antibody. Nuc Med Commun 1988; 9: 813-815. Vogel P, Laing FC, Jeffrey RB, Wing VW. Deep venous thrombosis of the lower extremity: US evaluation. Radiol 1987; 163:743-746.

Wong DW, Mishkin F, Lee T. A rapid chemical method of labelling human plasma proteins with Tc-99m pertechnetate at pH 7.4. Int J Appl Rad Isot 1978:29:251-253. WO 88/08878. WO 89/02752. WO 91/02546.

Industrial Applicability This invention provides a composition and a method which permits localisation of thrombi in deep venous thrombosis, localisation and determination of severity of pulmonary embolism, and determination of thrombosis age. Thus, it will find wide use in the medical and veterinary fields.

The foregoing describes only some embodiments of the present invention and modifications obvious to those skilled in the art can be made thereto without departing from the scope of the invention.

Claims

1. t-PA having substantially all plasminogen catalytic sites inactivated and which is able to bind to a thrombus and/or fibrin.

2. The t-PA according to claim 1, in which the plasminogen catalytic sites are s inactivated with a halomethyl ketone peptide.

3. The t-PA according to claim 2, wherein the halomethyl ketone peptide is phenylalanine- 1 -proline- 1 -arginine chloromethylketone.

4. The t-PA according to any one of claims 1 to 3, which is labelled with a detectable marker. o

5. The t-PA according to claim 4, wherein the marker is a radiolabel.

6. The t-PA according to claim 5, wherein the radiolabel is technetium-99m.

7. The t-PA according to any one of claims 1 to 6, which is recombinant.

8. A process for preparing t-PA having substantially all plasminogen catalytic sites inactivated and which is able to bind to a thrombus and/or fibrin, which process comprises s reacting t-PA with an agent to inactivate substantially all the plasminogen catalytic sites and which does not effect the ability of the t-PA to bind to a thrombus or fibrin.

9. The process according to claim 8, comprising the further step of purifying the inactivated t-PA by column chromatography.

10. The process according to claim 8 or claim 9, further comprising labelling the 0 reduced t-PA with a detectable marker.

11. The process according to claim 10, wherein the detectable marker is a radiolabel and said inactivated t-PA has its disulfide bridges cleaved prior to labelling.

12. The process according to any one of claim 11, wherein disulfide bridges in the t- PA are cleaved with a reducing agent. 5 13. The process according to claim 12, wherein the reducing agent is dithiothreitol or

2-mercaptoethanol .

14. The process according to claim 13, wherein the ratio of t-PA:reducing agent is about 1 :500 to 1 :3000.

15. The process according to claim 14, wherein the ratio is about 1 :2000 to 1 :3000. 0

16. The process according to claim 15, wherein the molar ratio is about 1 :2600.

17. The process according to any one of claims 11 to 16, wherein the radiolabel is technetium-99m.

18. The process according to claim 17, wherein the agent used to carry out radiolabelling is Tc-99m gluconate. 5

19. The process according to claim 18, wherein the ratio of t-PA:agent is about 1 :5 to about 1:30.

20. The process according to claim 19, wherein 275 μl of the agent at a concentration of 2mg/mL is added to 5mg t-PA.

21. The process according to any one of claims 18 to 20, wherein the t-PA is reacted with the agent for a period of about 2 to about 24 hours.

22. The process according to claim 21, wherein the period is about 2 hours.

23. The process according to any one of claims 18 to 22, wherein the Tc-99m gluconate is present in an amount of between about 1 to about 800MBq.

5 24. The process according to claim 23, wherein the amount is between about 500 to about 600MBq.

25. The process according to any one of claims 18 to 24, wherein the radiolabelling is carried out at a pH of between about 3.5 to about 7.

26. The process according to claim 25, wherein the pH range is about 4 to about 6. o

27. The process according to claim 26, wherein the radiolabelling is carried out at a pH of about 4.5.

28. t-PA having substantially all plasminogen catalytic sites inactivated and which is able to bind to a thrombus and/or fibrin whenever produced by a process according to any one of claims 8 to 27. s

29. A composition comprising t-PA having substantially all plasminogen catalytic sites inactivated and which is able to bind to a thrombus and/or fibrin, together with a pharmaceutically acceptable carrier, diluent and/or excipient.

30. The composition according to claim 29, in which the plasminogen catalytic sites are inactivated with a halomethyl ketone peptide. o

31. The composition according to claim 30, wherein the halomethyl ketone peptide is phenylalanine- 1 -proline- 1 -arginine chloromethylketone.

32. The composition according to any one of claims 29 to 31, wherein the carrier, diluent and/or excipient is buffered with sodium acetate at a concentration of about 0.05M at a pH of about 4.5. 5

33. The composition according to any one of claims 29 to 32, further comprising a tin halide in oxygen depleted water.

34. A composition according to claim 33, wherein the tin halide is a stannous halide.

35. The composition according to claim 34, wherein the stannous halide is stannous fluoride. 0

36. The composition according to any one of claims 33 to 35, further comprising a transchelating agent.

37. The composition according to claim 36, wherein the transchelating agent is a gluconate.

38. The composition according to any one of claims 29 to 37, in which the t-PA is 5 recombinant.

39. The composition according to any one of claims 29 to 38, which is lyophilised.

40. The composition according to any one of claims 29 to 39, wherein the t-PA is labelled with a detectable marker.

41. The composition according to claim 40, wherein the detectable marker is a radiolabel.

42. A composition according to claim 41, wherein the radiolabel is technetium-99m.

43. A kit comprising (a) a composition according to any one of claims 29 to 39 and (b) a detectable marker.

44. The kit according to claim 43, wherein the detectable marker is a radiolabel.

45. The kit according to claim 44, wherein the radiolabel is technetium-99m.

46. A method for localising thrombi in a mammal requiring said localisation, which method comprises administering to said mammal an effective amount of t-PA according to any one of claims 4 to 7 or of a composition according to any one of claims 40 to 42 followed by visualisation of said thrombus by methods known in this art.

47. The method according to claim 46, wherein administration is by bolus injection.

48. The method according to claim 46 or claim 46, wherein images are obtained between about 0 and about 24 hours post injection.

49. The method according to claim 48, wherein the images are obtained between about 2 hours and about 6 hours post injection.

50. A method for localising and determining the severity of thrombus in a mammal suffering from thrombus which method comprises administering to said mammal an effective amount of t-PA according to any one of claims 4 to 7 or of a composition according to any one of claims 40 to 42 followed by visualisation of said thrombus by methods known in this art.