IIIa

Feb 1, 1996 - ... Dupont Merck Pharmaceutical Company, 331 Treble Cove Road, North ... under both arterial and venous conditions were not significantl...
0 downloads 0 Views 325KB Size
Download PDF
Bioconjugate Chem. 1996, 7, 203−208

203

Biological Evaluation of 99mTc-Labeled Cyclic Glycoprotein IIb/IIIa Receptor Antagonists in the Canine Arteriovenous Shunt and Deep Vein Thrombosis Models: Effects of Chelators on Biological Properties of [99mTc]Chelator-Peptide Conjugates John A. Barrett,* David J. Damphousse, Stuart J. Heminway, Shuang Liu,* D. Scott Edwards, Richard J. Looby, and Timothy R. Carroll Radiopharmaceuticals Division, The Dupont Merck Pharmaceutical Company, 331 Treble Cove Road, North Billerica, Massachusetts 01862. Received September 18, 1995X

A series of 99mTc-labeled cyclic glycoprotein IIb/IIIa receptor antagonists, [99mTcO(L1-III)]-, [99mTcO(L6-III)]-, [99mTcO(L1-V)]-, and [99mTcO(L6-V)]-, were evaluated in a canine arteriovenous (AV) shunt model for their potential use as thrombus imaging agents. The thrombus formed consists of a platelet-rich head and a fibrin-rich tail. All four agents were incorporated into the growing thrombus under both arterial (platelet-rich) and venous (platelet-poor) conditions. The rank order for uptake was [99mTcO(L1-V)]- > [99mTcO(L6-V)]- > [99mTcO(L6-III)]- > [99mTcO(L1-III)]- (arterial range, 5.8-0.47 % id/g; venous range, 0.58-0.04 % id/g). The uptakes of both [99mTcO(L6-III)]- and [99mTcO(L1-III)]- under both arterial and venous conditions were not significantly greater than that of [99mTc]albumin and [125I]fibrinogen. In contrast, the uptakes of both [99mTcO(L1-V)]- and [99mTcO(L6-V)]were significantly greater than those of [99mTc]albumin and [125I]fibrinogen and comparable to that of [111In]platelets under both arterial and venous conditions. All four [99mTc]chelator-peptide conjugates are cleared faster than the controls with the clearance of the conjugates of peptide III faster than that of the conjugates of peptide V. The differences in incorporation are attributable to the effect of both the cyclic peptide and the chelator. The conjugate [99mTcO(L1-V)]- was also studied using a canine DVT (deep vein thrombosis) model. [99mTcO(L1-V)]- was actively incorporated into the growing thrombus with images clearly detectable within 15 min postinjection. At 2 h postinjection, thrombus/ blood and thrombus/muscle ratios [region of interest (ROI)/background] were approximately 7/1 and 10/1, respectively. This clearly demonstrated that the conjugate [99mTcO(L1-V)]- has the potential for rapid diagnosis of thrombolic events occurring under both arterial and venous conditions.

INTRODUCTION

Venous and arterial thrombus formation are common and potentially life-threatening events. However, existing diagnostic modalities are inadequate for diagnosis and determination of the morphology of the evolving thrombus (1). Thus, the development of agents which will not only detect the location but also determine the age of the thrombi is a critical unmet need in diagnostic nuclear medicine. Deep vein thrombus is the result of a hypercoagulatible state coupled with a period of stasis occurring in a lowshear environment. The end result is the formation of a fibrin-rich thrombus which also contains some platelets and erythrocytes. In contrast, an arterial thrombus is the result of the rupture of an atherosclerotic plaque occurring under high-shear conditions, resulting in the formation of a platelet-rich thrombus (2). Our approach utilizes a series of radiolabeled platelet glycoprotein IIb/ IIIa antagonists. The glycoprotein IIb/IIIa complex (GPIIb/IIIa) is expressed on the membrane surface of activated platelets and plays an integral role in platelet aggregation and thrombus formation (3). It has been demonstrated that the tripeptide sequence Arg-Gly-Asp (RGD) is capable of inhibiting the binding of fibrinogen to the platelet GPIIb/IIIa receptor (4). Cyclic peptides (Figure 1) containing the RGD (Arg-Gly-Asp) sequence * To whom correspondence should be addressed. Telephone: 508-671-8696 (S.L.) or 508-671-8341 (J.A.B.). Fax: 508-4367500. X Abstract published in Advance ACS Abstracts, February 1, 1996.

1043-1802/96/2907-0203$12.00/0

Figure 1. Cyclic GPIIb/IIIa receptor antagonists.

have been shown to be high-affinity antagonists for the GPIIb/IIIa receptor (3). Since the GPIIb/IIIa receptor is expressed only on activated platelets (5), the 99mTclabeled cyclic GPIIb/IIIa receptor antagonists should only be bound to platelets intimately involved in the thromboembolic event. 99mTc is the preferred radionuclide for diagnostic nuclear medicine because of its ideal physical properties (e.g. 6 h half-life, 140 keV, γ-emission), easy availability, and low cost. © 1996 American Chemical Society

204 Bioconjugate Chem., Vol. 7, No. 2, 1996

Figure 2. Two chelators used for peptides.

99mTc

Barrett et al.

labeling of cyclic

In our previous contribution, we described the 99mTc labeling of several cyclic GPIIb/IIIa receptor antagonists by the preformed chelate approach using various chelators. As a continuation of that study, we evaluated four [99mTc]chelator-peptide conjugates, [99mTcO(L1-III)]-, [99mTcO(L6-III)]-, [99mTcO(L1-V)]-, and [99mTcO(L6V)]-, in a canine arteriovenous (AV) shunt model (Figure 2). These agents were incorporated into the growing thrombus under both arterial and venous conditions; differences in incorporation are attributable to the effect of both the chelator and the cyclic peptide. We also evaluated the conjugate [99mTcO(L1-V)]- using a canine DVT (deep vein thrombosis) model. It was clearly demonstrated that the conjugate [99mTcO(L1-V)]- has the potential for rapid diagnosis of thrombolic events occurring under both arterial and venous conditions. EXPERIMENTAL SECTION

Na99mTcO4 was obtained from a commercial DuPont Mo/99mTc generator, North Billerica, MA. [99mTc]Albumin kits, 125I-labeled canine fibrinogen, and [111In]oxine kits were purchased from Medi-Physics Inc., Arlington Heights, IL, and were used as directed. Deionized water was obtained from a Millipore MilliQ Water System and was of >18 MΩ quality. The cyclic GPIIb/IIIa receptor antagonists were labeled with 99mTc by the preformed chelate approach as described in our previous contribution (6). The products were purified by high-performance liquid chromatography (HPLC). The product peaks were collected, and the volatiles were removed under reduced pressure. The residue was redissolved in 0.9% saline solution and analyzed by radio-HPLC. The [99mTc]chelator-peptide conjugates were g95% pure. Doses for biological evaluation were prepared by dilution with 0.9% saline to the required concentration: 300 µCi/mL for AV shunt and 1.5 mCi/mL for DVT studies. Canine Arteriovenous Shunt Methodology. Adult beagle dogs of either sex (9-13 kg) were anesthetized with pentobarbital sodium (35 mg/kg, iv) and ventilated with room air via an endotracheal tube (12 strokes/min, 25 mL/kg). For arterial pressure determination, the left carotid artery was cannulated with a saline-filled polyethylene catheter (PE-240) and connected to a Statham pressure transducer (model P23ID, Gould Co., Oxnard, CA). Mean arterial blood pressure was determined via damping the pulsatile pressure signal. Heart rate was monitored using a cardiotachometer (Grass Instrument Inc., Quincy, MA) triggered from a lead II electrocardiogram generated by limb leads. A jugular vein was cannulated (PE-240) for drug administration. Both femoral arteries and femoral veins were cannulated with silicon-treated (Sigmacote, Sigma Chemical Co., St. 99

Louis, MO), saline-filled polyethylene tubing (PE-200) and connected with a 5 cm section of silicon-treated tubing (PE-240) to form extra corporeal arteriovenous shunts. Shunt patency was monitored using a Doppler flow system (model VF-1, Crystal Biotech Inc., Hopkinton, MA) and a flow probe (2-3.5 mm, Crystal Biotech Inc.) placed proximal to the locus of the shunt. All parameters were monitored continuously on a model 7D polygraph recorder (Grass Instrument Inc., Quincy, MA) at a paper speed of 10 mm/min or 10 mm/s. Upon completion of a 15 min postsurgical stabilization period, an occlusive thrombus was formed by the introduction of a thrombogenic surface (4-0 braided silk thread, 5 cm in length, Ethicon Inc., Somerville, NJ) into one shunt with the other serving as a control. A 1 h shunt period was employed with the test agent (150 µCi/ kg, iv) administered as an infusion over 5 min, beginning 5 min before insertion of the first thrombogenic surface. The thrombus formed was comprised of a platelet-rich component on the thrombogenic surface and a fibrin-rich tail. At the end of the 1 h shunt period, the silk was carefully removed, and the portions were separated and weighed. The percent incorporation was determined via well counting. Thrombus weight was calculated by subtraction of the weight of the silk prior to placement from the total weight of the silk upon removal from the shunt. Arterial blood was withdrawn prior to infusion and every 30 min thereafter for determination of blood clearance, whole blood collagen-induced platelet aggregation, collagen prothrombin time, activated partial thromboplastin time, and fibrinogen and platelet count. Template bleeding time was also performed prior to infusion and every 30 min thereafter. [111In]Platelet Preparation. The isolation and 111In labeling of platelets were performed according to the literature method (7) with some modification. Canine arterial blood (40 mL) was withdrawn in 2.5% ACD solution [2.5 g of trisodium citrate, 1.4 g of citric acid, and 2.0 g of dextrose in 100 mL of H2O (pH 4.5)]. An additional 20 mL of blood was drawn in 3.8% sodium citrate solution which serves as a control. The samples were centrifuged at 1400 rpm for 15 min to form plateletrich plasma (PRP) and platelet-poor plasma (PPP). The ACD-PRP was pelleted (2800 rpm for 15 min) and washed twice with ACD. The platelet pellet was resuspended in ACD (2 mL), and 60-100 µCi of [111In]oxine was added. The suspension was incubated at 37 °C for 5 min, followed by the addition of PPP (5 mL). The suspension was centrifuged at 2800 rpm for 10 min. The supernatant and platelet pellet were counted to assess labeling efficiency (>80% in all cases). The platelet pellet was resuspended in PPP (5 mL) and platelet viability determined by comparison of collagen-induced platelet aggregation using nonlabeled and 111In-labeled platelets. Only those 111In-labeled platelet preparations which were within 25% of control were injected into the animal. Canine Deep Vein Thrombosis Methodology. This model incorporates the triad of events (hypercoagulatible state, period of stasis, and low-shear environment) essential for the formation of a venous fibrin-rich actively growing thrombus. Adult beagle dogs of either sex (913 kg) were anesthetized with pentobarbital sodium (35 mg/kg, iv) and ventilated with room air via an endotracheal tube (12 strokes/min, 25 mL/kg). For arterial pressure determination, the right femoral artery was cannulated with a saline-filled polyethylene catheter connected to a pressure transducer (model P23ID, Gould Co.). Heart rate was monitored using a cardiotachometer (Grass Instrument Inc.) triggered from a lead II electrocardiogram generated by limb leads. The right femoral vein was cannulated for drug administration. For the

Effects of Chelators on [99mTc]Chelator−Peptide Conjugates

Bioconjugate Chem., Vol. 7, No. 2, 1996 205

Figure 3. Thrombus uptakes for [125I]fibrinogen, [111In]platelets, [99mTc]albumin, [99mTcO(L1-III)]-, [99mTcO(L6-III)]-, [99mTcO(L1-V)]-, and [99mTcO(L6-V)]- under both arterial (left) and venous (right) conditions. All values are expressed as the mean ( the standard error of the mean. N ) 10 for platelets and fibrinogen (except for venous conditions where N ) 5). N ) 3 for [99mTcO(L1III)]-. N ) 2 for [99mTcO(L6-III)]-, and N ) 4 for [99mTc]albumin, [99mTcO(L1-V)]-, and [99mTcO(L6-V)]-. An * indicates a significant difference between [111In]platelets and treatment groups, while a + indicates a significant difference from [125I]fibrinogen and [99mTc]albumin using the Newman-Keuls test, P e 0.05. Table 1. Summary of the Hematological and Hemodynamic Effects of [111In]Platelets/[125I]Fibrinogen, [99mTc]Albumin, [99mTcO(L1-III)]-, [99mTcO(L6-III)]-, [99mTcO(L1-V)]-, and [99mTcO(L6-V)]-, 150 µCi/kg, iv in the Canine AV Shunt parametera treatment [111In]platelets/[125I]fibrinogen control EOI 1 h postinjection [99mTc]albumin control EOI 1 h postinjection [99mTcO(L1-III)]control EOI 1 h postinjection [99mTcO(L6-III)]control EOI 1 h postinjection [99mTcO(L1-V)]control EOI 1 h postinjection [99mTcO(L6-V)]control EOI 1 h postinjection

heart rate (bpm)

mean arterial pressure (mmHg)

APTT (s)

platelet count (×103)

aggregation (Ω)

135 ( 5 128 ( 5 124 ( 5

113 ( 8 113 ( 7 112 ( 7

17 ( 1 21 ( 2 17 ( 1

272 ( 17 264 ( 17 264 ( 19

27 ( 3 29 ( 2 29 ( 2

150 ( 6 150 ( 5 147 ( 7

124 ( 8 124 ( 8 124 ( 10

15 ( 0 15 ( 1 15 ( 1

281 ( 8 289 ( 11 284 ( 7

24 ( 1 27 ( 3 23 ( 2

170 ( 10 164 ( 9 159 ( 10

134 ( 12 137 ( 10 139 ( 5

13 ( 0 13 ( 0 13 ( 0

302 ( 9 312 ( 10 323 ( 8

31 ( 7 32 ( 6 32 ( 5

145 ( 17 151 ( 13 138 ( 1

112 ( 14 122 ( 14 115 ( 5

14 ( 1 14 ( 1 14 ( 0

242 ( 37 238 ( 27 250 ( 44

20 ( 2 20 ( 8 14 ( 9

166 ( 11 160 ( 12 150 ( 9

130 ( 9 132 ( 12 126 ( 13

15 ( 1 15 ( 1 15 ( 1

276 ( 31 259 ( 34 282 ( 33

26 ( 4 22 ( 5 28 ( 5

150 ( 14 142 ( 8 133 ( 7

116 ( 11 115 ( 6 115 ( 5

14 ( 0 14 ( 1 15 ( 1

263 ( 13 239 ( 24 263 ( 20

25 ( 4 21 ( 5 21 ( 0

a All values are expressed as the mean ( the standard error of the mean. EOI ) end of infusion. N ) 10 for [111In]platelets/[125I]fibrinogen. N ) 3 for [99mTcO(L1-III)]-. N ) 2 for [99mTcO(L6-III)]-, and N ) 4 for [99mTc]albumin, [99mTcO(L1-V)]-, and [99mTcO(L6-V)]-.

induction of a venous thrombus, a 5 cm segment of both jugular veins was isolated and circumscribed with silk suture. A balloon catheter (3-4 F Baxter Co., McGraw Park, IL) was advanced from the facial vein into the jugular vein. A microthermister probe (Physitemp Co., Clifton, NJ) was placed on the vessel which serves as an indirect measure of venous flow. A period of stasis and hypercoagulatibility was induced by inflating the balloon and the local administration of 5 u of thrombin (American Diagnosticia, Greenwich, CT) into the occluded segment. After an additional 15 min, the balloon was deflated and flow reestablished as verified by the microthermister probe.

The test agent (10 mCi/kg, iv) was administered over 5 min beginning at reflow. Serial images were acquired using a γ-camera (Digital Dyna Camera, Picker International, Cleveland, OH) every 5 min for 2 h and region of interest (ROI) and target/background ratios calculated. Arterial blood was withdrawn prior to administration and every 30 min thereafter for determination of blood clearance, hematology, platelet function, and coagulation status. At the end of the protocol, the animal was euthanized with an overdose of pentobarbital and the vessel excised. The thrombus was removed and weighed, and the amount of incorporation was determined via a γ-well counter (LKB 1282, Wallac Inc., Gaithersburg, MD).

206 Bioconjugate Chem., Vol. 7, No. 2, 1996

Barrett et al.

Figure 4. Thrombus/blood ratios for [125I]fibrinogen, [111In]platelets, [99mTc]albumin, [99mTcO(L1-III)]-, [99mTcO(L6-III)]-, [99mTcO(L1-V)]-, and [99mTcO(L6-V)]- under both arterial (left) and venous (right) conditions. All values are expressed as the mean ( the standard error of the mean. N ) 10 for platelets and fibrinogen (except for venous conditions where N ) 5). N ) 3 for [99mTcO(L1III)]-. N ) 2 for [99mTcO(L6-III)]-, and N ) 4 for [99mTc]albumin, [99mTcO(L1-V)]-, and [99mTcO(L6-V)]-. An * indicates a significant difference between [111In]platelets and treatment groups, while a + indicates a significant difference from [125I]fibrinogen and [99mTc]albumin using the Newman-Keuls test, P e 0.05.

Data Analysis. All values are expressed as the mean ( the standard error of the mean. In the DVT studies, the target (thrombus) uptake was calculated by drawing a 4 × 4 pixel area in the region interest and determining the average intensity. Background values were determined in a similar manner. Blood background was calculated using the jugular vein just distal to the locus of the thrombus. Statistical analysis consisted of a oneway analysis of variance and, when appropriate, a Student’s paired t test, a two-tailed probability for assessing differences within treatment, and a NewmanKeuls test for assessing differences between means of treatment groups. Differences were considered significant at P e 0.05. RESULTS AND DISCUSSION [125I]fibrinogen,

[111In]-

Figure 5. Blood clearance curves for platelets, [99mTc]albumin, [99mTcO(L1-III)]-, [99mTcO(L6-III)]-, [99mTcO(L1-V)]-, and [99mTcO(L6-V)]-. All values are expressed as the mean ( the standard error of the mean. N ) 10 for platelets and fibrinogen (except for venous conditions where N ) 5). N ) 3 for [99mTcO(L1-III)]-. N ) 2 for [99mTcO(L6III)]-, and N ) 4 for [99mTc]albumin, [99mTcO(L1-V)]-, and [99mTcO(L6-V)]-.

Hematologic Studies. Platelet, white blood cell (WBC), and red blood cell (RBC) counts and hematocrit determinations were performed on whole blood collected in 2 mg/mL disodium ethylenediaminetetraacetic acid (EDTA) using a Sysmex K1000 (TEA Medical Electronics Co., Los Alamitos, CA). Template bleeding time was assessed via an incision in the lower lip (Surgicutt, Baxter Co.) and the time to formation of a clot monitored. Whole blood platelet aggregation was measured using a lumiaggregometer (Chrono-Log Co., Havertown, PA) by recording the change in impedance (platelet aggregation). Blood samples were collected in 10 mM sodium citrate and diluted 50% with saline supplemented with 0.5 mM Ca. Aggregation was induced with collagen (5 µg/mL, Chrono-Log Co.), and the changes in impedance were recorded over 6 min. Activated partial thromboplastin time (APTT) and prothrombin time (PT) were monitored using a microsample coagulation analyzer (MCA-210, BIO/DATA Co., Horsham, PA).

In this study, we used a canine AV shunt model to evaluate several 99mTc-labeled cyclic GPIIb/IIIa receptor antagonists for their potential use in the detection of rapidly growing arterial and venous thrombi. The thrombus formed in this model was comprised of a plateletrich head (arterial conditions) and a fibrin-rich tail (venous conditions) which allowed the rapid assessment of an agent under both arterial and venous conditions. To validate this model, the amount of incorporation of [111In]platelets (58 µCi), [125I]fibrinogen (106 µCi), and [99mTc]albumin (153 µCi) was assessed via well counting. [111In]Platelets and [125I]fibrinogen were adequately incorporated into the growing thrombus, with [111In]platelets favoring the platelet-rich head (5.63 ( 0.6 % id/g) with lesser amounts of incorporation observed in the fibrin-rich tail (0.17 ( 0.02 % id/g). Similar amounts of [125I]fibrinogen were incorporated under both arterial and venous conditions (0.11 ( 0.02 vs 0.07 ( 0.02 % id/g, respectively), while little incorporation was observed with [99mTc]albumin (arterial, 0.05 ( 0.01 % id/g; venous, 0.04 ( 0.001 % id/g) (Figure 3). The administration of [111In]platelets, [125I]fibrinogen, and [99mTc]albumin did not alter any of the parameters studied (Table 1). Thus, the deposition on the thrombogenic surface mimics that of arterial conditions, i.e. high-shear and platelet-rich thrombus, and the platelet-poor tail mimics that of a venous thrombus, i.e. platelet-poor and low-shear environment.

Effects of Chelators on [99mTc]Chelator−Peptide Conjugates

Bioconjugate Chem., Vol. 7, No. 2, 1996 207

Figure 6. Illustrated is a typical selection of images derived from studies of [99mTcO(L6-V)]-, 1 mCi/kg, iv, in the canine DVT model. Each image represents a 5 min acquisition beginning at the time listed postinjection.

Figure 7. Assessment of [99mTcO(L6-V)]- and [99mTc]albumin, 1 mCi/kg, iv, in the canine DVT model. Illustrated in the left panel is the thrombus/muscle ratio, while in the right panel is the thrombus/blood ratio calculated from the image. Each histogram is the mean ( the standard error of the mean for 4-6 dogs.

Four [99mTc]chelator-peptide conjugates were assessed in the canine femoral arteriovenous shunt model, and all of them were incorporated into the growing thrombus under both arterial and venous conditions (Figure 3). Greater thrombus uptake occurred under arterial conditions, reflecting a more platelet-rich thrombus. The rank order for uptake was [99mTcO(L1-V)]- > [99mTcO(L6V)]- > [99mTcO(L6-III)]- > [99mTcO(L1-III)]- (arterial range, 5.8-0.47 % id/g; venous range, 0.58-0.04 % id/ g). Under arterial conditions, the uptake of neither [99mTcO(L6-III)]- nor [99mTcO(L1-III)]- was significantly higher than that of [99mTc]albumin and [125I]fibrinogen. In contrast, the uptakes of both [99mTcO(L1V)]- and [99mTcO(L6-V)]- were significantly greater than the negative controls and comparable to the positive control, [111In]platelets, under both arterial and venous conditions. This is probably related to the stronger affinity of the cyclic peptide V for GPIIb/IIIa than that of the cyclic peptide III (8) and the slower rate of clearance of cyclic peptide V from the blood, resulting in a greater receptor residence time (Figure 5).

In spite of the differences in thrombus uptake, all four [99mTc]chelator-peptide conjugates show comparable thrombus/blood ratios at 1 h after administration under both arterial and venous conditions (Figure 4). The conjugates with the highest uptake, [99mTcO(L1-V)]- and [99mTcO(L6-V)]-, have the slower blood clearance rate, while the conjugates with the lower uptake, [99mTcO(L6III)]- and [99mTcO(L1-III)]-, have the faster clearance rates. Figure 5 shows the blood clearance curves for the four [99mTc]chelator-peptide conjugates and the three controls. The rank order for blood clearance was [99mTcO(L1-III)]- > [99mTcO(L6-III)]- > [99mTcO(L6-V)]- > [99mTcO(L1-V)]-. All the [99mTc]chelator-peptide conjugates are cleared faster than the controls with the clearance of the conjugates of peptide III increasingly faster than the conjugates of peptide V. It is quite clear that both the peptide and the chelator affect the blood clearance of the [99mTc]chelator-peptide conjugates. The four [99mTc]chelator-peptide conjugates evaluated did not adversely affect any of the parameters studied (Table 1). Thus, the 99mTc-labeled GPIIb/IIIa receptor antagonists reported in this study have the ability to be rapidly

208 Bioconjugate Chem., Vol. 7, No. 2, 1996

incorporated into growing thrombi in vivo under both arterial and venous conditions. The ability of [99mTcO(L1-V)]- (1 mCi/kg, iv) to be actively incorporated into a growing venous thrombus was also studied using a canine DVT reflow model. [99mTcO(L1-V)]- was administered at reflow with its rate of uptake and contrast ratio assessed using γ-scintigraphy. [99mTc]Albumin was not incorporated into the rapidly growing thrombus. [99mTcO(L1-V)]- was actively incorporated into the growing thrombus with images clearly detectable within 15 min postinjection (Figure 6). At 2 h postinjection, thrombus/blood and thrombus/ muscle ratios (ROI/background) were approximately 7/1 and 10/1, respectively (Figures 6 and 7). The present study is in agreement with the work of Oster et al. (9), who demonstrated that the 7E3 antibody, directed against the platelet GPIIb/IIIa receptor, was also capable of detecting new thrombi in the carotid, femoral, and pulmonary arteries induced by the insertion of a copper coil. The target/background values ranged between 2/1 and 5/1 at 1.5 h postadministration. More recently, a venom derived from the viper Bitis arietans was shown to be effective in detection of venous thrombi. This venom, bitistatin, contains an RGD peptide recognition sequence which has been found to be essential for the binding to the platelet GPIIb/IIIa receptor (10). A 99mTclabeled dimer of a cyclic peptide (P280) has also been shown to detect venous thrombi. However, P280 was rapidly cleared (