ORIGINAL ARTICLE 
Annals of Nuclear Medicine Vol. 8, No. 3, 193-199, 1994 
In vivo kinetics of 99*Tc labeled recombinant tissue plasminogen activator in rabbits 
Kazuo ITOH,* Masahiro IEKO,** Etsuro HIRAGUCHI*** Hide KITAYAMA** and Eriko TSUKAMOTO* 
Departments of *Nuclear Medicine, **Second Internal Medicine and ***Second Surgery, Hokkaido University School of Medicine 
Our previous studies demonstrated that 99mTc labeled recombinant tissue plasminogen activator (rt-PA) retained high affinity with fibrin in vitro but showed unexpectedly low uptake in fresh thrombi in vivo. The present study was performed to determine the in vivo kinetics of radiolabeled t-PA in the rabbit. 
Sequential images and blood samples after the intravenous administration of 99mTc labeled rt-PA in thrombus-bearing rabbits were taken. The radioactivity and immunological level of t-PA and PAI-1 in the solution eluted to each fraction by gel permeation chromatography were measured by means of a well scintillation counter and enzyme-linked immunosorbent assay (ELISA). Most of the radioactivity was eluted in the fraction (Fr. 7) of larger molecular weight than that (Fr. 9) of intact t-PA. The level of intact rt-PA was increased with a regimen involving the preadministration of cold rt-PA which was followed by the administration of hot rt-PA. The level of PAI-1 in plasma showed an increased rebound 15 minutes after the intravenous injection. These results suggest two possible reasons why rt-PA retains high affinity with fibrin in vitro, once radiolabeled, but was ineffective in delineating fresh thrombi with a gamma camera: 1) some plasma components such as PAI-1 combine with circulating radiolabeled rt-PA and form a larger molecule immediately and/or 2) radiolabeled rt-PA is modulated as a consequence of the radiolabeling and forms a larger molecule than intact rt-PA. 
Key words: recombinant tissue plasminogen activator, radiolabeling, pharmacokinetics, animal study 
INTRODUCTION 
THROMBUS FORMATION has become a major cause of ischemic manifestations in organ function and death in aged persons. Although many methods for the detection of thrombi have been developed, noninvasive imaging diagnosis is still a challenging problem.1 A scintigraphic method with radionuclide labeled agents that concentrate active thrombi has been considered to be a promising method for localizing their active sites.2-4 
Recombinant tissue plasminogen activator (rt-PA) has very strong fibrinolytic activity as well as affinity with fibrin. The latter biological property is promising for the positive delineation of thrombi by means of scintigraphy. We have studied the ex vivo and in vivo affinities of Technetium-99m (99mTc) labeled rt-PA with the fibrin and venous thrombi formed in rats and cats.5-7 99mTc-rt-PA showed a very high affinity with fibrin ex vivo but showed an unexpectedly low concentration at the sites of fresh thrombi formed in animals. Two reasons were considered for the discrepancy between ex vivo and in vivo results: I ) rt-PA may concentrate in active thrombi but leave immediately after resultant fibrinolytic activity8; 2) only a very small amount of extrinsic rt-PA may accumulate in in vivo thrombi due to interaction with the fast-acting inhibitor of the plasminogen activator (PAI-1) which may be systemically and locally augmented by thrombus formation.9-12 We confirmed that the former mechanism significantly prolonged 99mTc-rt-PA accumulation in thrombi in rats when aprotinin was used.6 Combined dose injection of aprotinin and hot rt-PA, however, was not enough to clearly visualize the clots. 
We extended the study to identify in vivo behavior of intravenously administered hot rt-PA in relation to inter-action with PAI-1. The molecular weight of rt-PA and PAI-1 is estimated to be 67,000 and approximately 50,000, respectively. The complex of rt-PA with PAI-1 in plasma may have an even greater molecular weight. It is proposed that these three molecules are able to be separated from the complex by gel permeation chromatography, at least in the case of rt-PA and PAI-1. The final goal of our study is to determine the clinical usefulness of 99mTc-rt-PA as a promising thrombus imaging agent. 
MATERIALS AND METHODS 
Preparation of 99mTc labeled rt-PA 
The rt-PA (SM-9527, Sumitomo Chemical Co., Osaka, Japan) consisted of a double polypeptide chain. The molecular weight was estimated to be 67,000. The rt-PA was dissolved in 0.7% normal saline containing only 0.01% Tween 80. It was adjusted to pH 3 and was frozen for storage. The specific activity was 7.82 x 106 IU/mg. 
After melting frozen rt-PA at room temperature, a few hundred ul of rt-PA was pipetted into a plastic tube, and thereafter a commercially available stannous solution (containing 1 umol/ml of stannous ion at pH 3, Nihon Medi-Physics Co., Osaka, Japan) used in the 99mTc liver image kit was added. After gently stirring for 5 minutes, freshly eluted 99mTc-pertechnetate was added and followed by incubation for 20 minutes at room temperature with occasional gentle agitation. The amounts of rt-PA, stannous ion and 99mTc-pertechnetate were 1.38 mg, 0.001 umol and 0.1 ml, respectively. The process was described in more detail in our previous report.5 
Imaging of thrombus-bearing rabbits 
Female house rabbits of approximately 3 kg in body weight were anesthetized with ketamine hydrochloride (10 mg/kg). A wire-guide for angiography (0.9 mm in diameter), which was coated with fibrin (Beliplast, Behringwerke Ag, Hoechst Japan limited), was set at the inferior vena cava through the femoral vein. A venula needle was also inserted into the inspilateral vein for further blood sampling. 
The anesthetized rabbit was fixed on a board in a supine position and was put under a digital gamma camera. The rt-PA which was diluted to 0.1 mg/ml just prior to the intravenous injection was administered to two different groups of rabbits through the ear vein. The first group (hot group; n=3) received 99mTc labeled rt-PA (hot rt-PA) alone as an i.v. bolus injection of 1 mg/kg given over 2 minutes. The second group (cold-hot group; n=4) received two injections of intact (0.5 mg/kg) and hot (0.5 mg/kg) rt-PA at 5 minute intervals for a total dose of 1 mg/kg. Sequential images following the injection were obtained with a 64 by 64 matrix at 20 second intervals as storage data for the computer, and with a 128 by 128 matrix at 2 minute intervals for a film display. 3 ml venous blood samples were withdrawn at 0, 5, 10, 15 and 20 minutes after completing the injection. After completing an in vivo study, an image of the inserted wire-guide was taken by a gamma camera following sacrificing of the rabbit. 
Measurements of radioactivity and immunological level of t-PA and PAI-1 in plasma 
Plasma was separated from whole blood by sedimentation and was eluted to the fraction corresponding to the molecular weight by high performance liquid chromatography (HPLC): the column was a 1.4 x 31 cm Superose 12 (Pharmacia LKB, Uppsala, Sweden); elusion was at 1 ml/ minute with 0.02 mol Tris buffer solution (pH 7.4); and each fraction collected was 2 ml. The radioactivity in the solution eluted in each fraction was counted in a well-scintillation counter. The radioactivity in each fraction was expressed as a percentage of the cumulative radio-activity eluted in all fractions . The immunological level of t-PA and PAI-1 in the solution of each fraction was determined with an enzyme-1inked immunosorbent assay (ELISA) kit (Biopool, Umea, Sweden). 
RESULTS 
Imaging of thrombus-bearing rabbits 
99mTc-rt-PA concentrated in the liver immediately and the concentration in the kidney and the urinary bladder occurred later (Fig. 1). The fibrin coated wire-guide was not shown as a high uptake area in the sequential images, but was removed after completion of the study did show a minimal concentration of 99mTc-rt-PA (Fig. 1). Positive uptake in the wire-guide was dependent on the formation of fresh thrombi around the artificially made fibrin clot (Table 1). 
The blood disappearance curve was different in each rabbit (Table 2). The half times for the blood clearance curves for the two groups, calculated by a computer program 2 compartment model, were similar.
Elution curves for plasma sample by HPLC 
The radioactive peak was observed in two fractions: the first peak in fraction No. 7 and the second peak in fraction No. 12 (Fig. 2). The latter peak corresponded to free pertechnetate which was separated from labeled rt-PA during the storage of plasma and gel permeation procedure. The elution pattern of radioactivity was similar for both hot (Fig. 2) and cold-hot groups (Fig. 3). In contrast, the elution pattern for the immunological level of t-PA in each fraction was different in hot and cold-hot groups. In the former group, the peak level of t-PA was observed in fraction No. 8 at time 0. This peak level immediately disappeared in plasma obtained 5 minutes after the injection. In the latter group, the peak level of t-PA was observed in fractions No. 8 and 9 at 0 minutes postinjection. The intact rt-PA was eluted in the fraction No. 9. The peak level of t-PA in fraction Nos. 8 and 9 was also immediately disappeared 5 minutes following the intravenous administration. The level of rt-PA in each fraction of each plasma sampled 10 minutes after the injection of 99mTc-rt-PA appeared to be parallel to the radioactive elution curve. The elution curve of the PAI-1 level in the fractions showed no specific peak (Fig. 2). This result was due to a very low level of PAI-1 in fractions once eluted and represented a non-specific immunological reaction to PAI-1 at such a plasma level. When the level of PAI-1 was measured in plasma before HPLC, it showed an increased rebound at 15 minutes in the hot group and a somewhat delayed increase in the cold-hot group following the administration of hot rt-PA (Fig. 4).
DISCUSSION 
In ex vivo experiments which were performed prior to in vivo studies, 99mTc-rt-PA showed a high uptake in the fibrin coated segment of the wire-guide. However, no radioactive distribution was detected in a wire-guide in the in vivo study, although some wire-guides obtained following the sacrifice of the rabbits showed radioactive deposits in the fibrin-coated portion. These results were mainly due to minimal formation of fresh thrombi around the fibrin-coated wire-guide, which were confirmed by inspection after sacrifice of the rabbits. The positive delineation of fibrin-coated wire guide with 99mTc-rt-PA depended on intravascular formation of fresh red thrombi on the surface of the fibrin paste. It required 4 to 5 hours to form red thrombi around the wire-guide. 
The present study provides in vivo kinetics of 99mTc labeled rt-PA. Most of the radiolabeled rt-PA administered collected had a greater molecular weight than that of the intact rt-PA in circulating plasma. This results suggest that extrinsic rt-PA interacts with fast-acting plasma protein, presumably PAI-1, immediately after the intravenous injection. It is well known that PAI-1 plays a pivotal role in preventing the concentration of externally administered rt-PA in active thrombi.9-11 Thrombus formation also augments the introduction of PAI-1 at the site of the damaged vessel.12 The level of free rt-PA, which can only act biologically as an effector of fibrin-binding in vivo, is therefore exhausted. Multiple injections of t-PA are more effective than a single dose injection on thrombolysis.13 We were able to confirm that the second dose injection of the hot rt-PA extended the return to normal serum PAI-1 level. However, it was not demonstrated that pre-administration of the cold rt-PA results in a prolongation of exhaustion of the externally administered hot rt-PA in the circulation. 
One finding in the present study is the different elution curves of radioactivity and the immunological level of rt-PA, particularly at 0 minutes after the injection in two different doses of hot and cold rt-PA. In the hot group, the radioactive peak was eluted in fraction No. 7 and the immunological peak of t-PA in fraction No. 8. In the cold-hot group, the radioactive peak was observed in fraction No. 7 and the immunological peaks in fractions 8 and 9. The radioactivity in the fraction corresponding to the intact rt-PA did not increase. These results suggest that most of the t-PA in plasma represents the pre-administered cold rt-PA rather than the later hot rt-PA. It is assumed that most of the radiolabeled rt-PA is modulated to a molecule with a weight larger than that of intact rt-PA in the circulation. This molecular modulation seems to be related to the interaction with circulating PAI-1 and/or radiolabeling on protein as an agent. The denaturation of radiolabeled protein influences its accumulation in the liver and elimination from the circulation. Modulation of the carbohydrate structure is reported to be important in the in vivo behavior of radiolabeled rt-PA.14 We confirmed only the inherent affinity of 99mTc-labeled rt-PA with fibrin in vitro and did not investigate its influence on the denaturation of rt-PA in the molecular structure. If 99mTc-rt-PA, which still retains the biological activity of fibrin binding in vitro, is modulated to a larger molecular weight as consequence of radiolabeling without interaction with PAI-1 in vivo, then elimination from the circulation may be altered and thereby accumulation in intravascular thrombi may be affected. Further investigation will be required to clarify the in vivo behavior of 99mTc labeled rt-PA, particularly the relationship between the positive delineation of an active thrombus and the influence of PAI-1. 
The wild-type t-PA has three different active domains: a fibrin-binding domain, fibrinolytic domain and PAI-1 binding domain.15 The recombinant technique can produce mutant t-PA which has only a single domain.15,16 From the pharmacological point of view, the rt-PA which retains only a fibrin-binding domain with a lack of properties such as fibrinolytic as well as PAI-1 binding domains is ideal for thrombus imaging agents,16,17 
In conclusion, most of the radiolabeled rt-PA in rabbit plasma was collected by gel permeation chromatography as a complex with a molecular weight greater than that of intact rt-PA by gel permeation chromatography. This complex may be accounted for by the following: 1) interaction of 99mTc-rt-PA with PAI-1 during circulation and/or 2) modulation of rt-PA as a consequence of radiolabeling. This complex seems to be ineffective for the positive delineation of active thrombi. Clinical application of 99mTc-rt-PA is still far from being realized. 
ACKNOWLEDGMENTS 
We thank Miss Y. Iwasa of the Second Internal Medicine Department and Mr. K. Suzuki, Mr. H. Katsuura. Mr. H. Omote and Mr. H. Arai of Nuclear Medicine Section, Hokhaido University Hospital, for their technical support. 
This study was supported by grant #04454290 from the Ministry of Education of Japan. 
REFERENCES 
1. Moser KM. Venous thromboembolism. Am Rev Respir Dis 141: 235-249 1990. 
2. Khon KA, Knight LC. Radiopharmaceuticals for thrombus detection: selection, preparation, and critical evaluation. Semin Nucl Med 12: 219-228, 1985. 
3. Knight LC. Radiopharmaceuticals for thrombus detection. Semin Nucl Med 20: 52-67, 1985. 
4. Chaudhuri TK, Fink S, Farpour A. Physiological considerations in imaging of lower extremity venous thrombosis. Am J Physiol Imaging 6: 90-103 1991. 
5. Itoh K, Tsukamoto E, Nishibe T, Sakurama S, Ieko M, Tanabe T, et al. Tc-99m labeled tissue-type plasminogen activator: Preparation, stability and preliminary imaging of thrombus-bearing rats. Ann Nucl Med 5: 59-64, 1991. 
6. Tsukamoto E. In vivo distribution of Tc-99m labeled recombinant tissue-type plasminogen activator in control and thrombus-bearing rats. Ann Nucl Med 6: 177-184, 1992. 
7. Takahashi A, Itoh K, Tsukamoto E, Furudate M, Kamiyama H, Abe H . Tc-99m-labeled recombinant tissue plasminogen activator for the imaging of emboli in vivo. Neurol Med Chir 33: 415-419, 1993. 
8. Hunt BJ, Cottam S, Segel H, Ginsburg R. Potter D. Inhibition by aprotinin of t-PA mediated fibrinolysis during orthotopic liver transplantation. Lancet 336: 381, 1990. 
9. Verheijen JH, Chang GTG, Kluft C. Evidence for the occurrence of a fast-acting inhibitor for tissue-type plasminogen activator in human plasma. Thromb Haemastas 51: 392-395, 1984, 
10. Wiman B, Chemielewska J, Randy M. Inactivation of tissue type plasminogen activator in plasma. demonstration of a complex with a rapid inhibitor. J Biol Chem 259: 3644-3647 1984. 
11. Kruithof EKO, Tran-Thang C, Ransijn A, Bachmann F. Demonstration of a fast-acting inhibitor of plasminogen activators in human plasma. Blood 64: 907-913, 1984. 
12. Sawa H, Fujii S, Sobel BE. Augmented arterial wall expression of type-1 plasminogen activator inhibitor induced by thrombosis. Arteriosclerand Thromb 12: 1507-1515, 1992. 
13. Klabunde RE, Burke SE, Henkin J. Enhanced lytic efficacy of multiple bolus injection of tissue plasminogen activator in dogs. Thromb Research 58: 511-517, 1990. 
14. Hotchkiss A, Refino CL. Leonard CK. O'Connor JV, Crowley C. McCabe L, et al. The influence of carbohydrate structure on the clearance of recombinant tissue-type plasminogen activator. Thromb Haemost60: 255-261, 1988. 
15. Bennet WF, Poani NF, Keyt BA, Botstein BA. Jones AJS, Presta L, et al. High resolution analysis of functional determination on human tissue-type plasminogen activator. J Biol Chem 266: 5191-5201, 1991. 
16. Monge JC. Lucore CK, Fry ETA, Sobel BE, Billadello JJ. Characterization of interaction of active-site serine mutants of tissue-type plasminogen activator with plasminogen activator inhibitor-1. J Biol Chem 264: 10922-10925, 1989. 
17. Fry ETA, Mack DL, Monge JC, Billadello JJ, Sobel BE. Labeling of human clots in vitro with an active-site mutant of t-PA. J Nucl Med 30: 187-191, 1990.