Journal Logo


Antithrombotic Activity of NSP-513, a Novel Selective Phosphodiesterase 3 Inhibitor, on Femoral Arterial Thrombosis Induced by Physical Stenosis and Electrical Current: Comparison of Antithrombotic and Hemodynamic Effects

Hirose, Hiroyasu; Mashiko, Satoshi; Kimura, Toshifumi; Ishida, Fumiaki; Mochizuki, Nobuo*; Nishibe, Tadayuki*; Nishikibe, Masaru

Author Information
Journal of Cardiovascular Pharmacology: April 2000 - Volume 35 - Issue 4 - p 586-594
  • Free


Cyclic nucleotide phosphodiesterases (PDEs) are classified largely into seven isozyme families (PDE 1-7), and are distributed in many cells and tissues, including platelets (1-3). In platelets, the major PDE isozymes are PDE 3 and 5 (4,5). Because the cyclic adenosine monophosphate (cAMP) specificity of PDE 3 is higher than that of PDE 5, it is believed that the activity of PDE 3 dominates the hydrolysis of cAMP in platelets. In fact, selective PDE 3 inhibitors raise intracellular cAMP levels, decrease calcium levels in platelets, and prevent platelet aggregation and adhesion induced by most physiologic platelet activators, such as collagen, thrombin, and adenosine diphosphate (ADP) (4,6). Thus PDE 3 inhibitors display a broad spectrum of antiplatelet activities, as well as potential for a new family of antiplatelet drugs. It is well known that PDE inhibition causes depressor effects due to vascular relaxation and cardiotonic effects due to positive inotropic action (7-9). In the clinical development of PDE 3 inhibitors as antiplatelet drugs, hemodynamic adverse effects are a very important issue. However, several reports suggested that two cyclic nucleotide-mediated processes, vascular relaxation and inhibition of platelet aggregation, can be dissociated by selective inhibitors of PDEs (10,11). Furthermore, in rat cardiac muscle, it has been reported that selective inhibitors of neither PDE 3 nor PDE 4 increase the force of contraction when given alone, but these substances do increase contractility when administered in combination (12). Thus the inhibition of a sufficient proportion of total PDE activity, but not solely PDE 3 activity, in relevant intracellular sites may be generally important for the modulation of hemodynamic functions. The differential modulation of tissue functions may lead to in vivo selectivity between the antiplatelet and hemodynamic activities of these enzymes. However, little evidence is available on whether selective PDE 3 inhibitors provide in vivo selectivity between antiplatelet and hemodynamic activity. Cilostazol (Pletaal) is a selective PDE 3 inhibitor that is clinically available as an antiplatelet drug (13,14). Although it has been reported that the antithrombotic efficacy of cilostazol in arterial occlusive diseases is clinically achieved with a minimum of cardiovascular adverse effects, cilostazol's selectivity for antithrombotic activity over hemodynamic activity is not clear.

NSP-513, (R)-4,5-dihydro-5-methyl-6-[4-(2-propyl-3-oxo-1-cyclohexenyl)amino] phenyl-3(2H)-pyridazinone, is a potent and selective PDE 3 inhibitor (15,16). PDE 3 inhibition by NSP-513 is more potent than that achieved by previously described PDE 3-selective inhibitors such as cilostazol and milrinone, and the selectivity of NSP-513 for PDE 3 versus the other PDE isozymes (1,2,4,5) is much greater than those of cilostazol and milrinone (15,16). The aims of our investigation were to (a) estimate the antiplatelet and antithrombotic potencies of NSP-513 by using a canine model, (b) determine whether NSP-513 displays in vivo selectivity for antiplatelet or antithrombotic activity over hemodynamic activity, and (c) compare this selectivity with that of cilostazol.


Antiplatelet studies

In vitro platelet-aggregation assay. Venous blood samples (6 ml) were obtained from conscious beagle dogs (8.0-12.0 kg) by using a plastic syringe containing one part (0.6 ml) of 3.8% trisodium citrate. Then platelet-rich plasma (PRP) was prepared by centrifugation of whole blood at 50 g for 5 min at room temperature. If necessary, the centrifugation was repeated once more. Platelet counts in PRP were determined by using an automated particle counter (Coulter ZI; Coulter, Luton, U.K.). Platelet concentration was adjusted to 2 × 108 platelets/ml with time-matched platelet-poor plasma (PPP), which was obtained from the precipitated fraction of PRP by centrifugation at 1,700 g for 15 min at room temperature. All aggregation tests were performed according to Born's turbidimetric method (17). Aliquots (0.2 ml) of platelet suspension (4 × 107 platelets) were gently stirred at 1,000 rpm in siliconized cuvettes. Then 1 μl of drug solution [NSP-513 or cilostazol dissolved in dimethyl sulfoxide (DMSO)] was added to the PRP. After 3 min of preincubation at 37°C, collagen (final concentration, 20 μg/ml) and ADP (final concentration, 10 μM) were added to the cuvettes. Changes in light transmission were measured by using an aggregometer (Mebanix, PAM-8C, Tokyo, Japan) until each trace reached a plateau. Platelet aggregation was expressed as the percentage of light transmittance, with PPP representing 100% transmittance. The extent of inhibition of aggregation was calculated as the degree of aggregation observed in the presence of increasing concentrations of NSP-513 or cilostazol as a percentage of that observed in the absence of NSP-513 or cilostazol.

Ex vivo platelet-aggregation assay. Conscious beagle dogs (8.0-12.0 kg) were used after an overnight fast. NSP-513 (0.03-1 mg/kg) and cilostazol (50 mg/kg) were suspended in 0.5% methylcellulose and administered by oral gavage in a volume of 2 ml/kg. The control animals received only vehicle. Blood samples (6 ml) were drawn before and 0.5, 1, 2, 4, 6, and 8 h after drug dosing from the cephalic vein of each animal by using a plastic syringe containing one part (0.6 ml) of 3.8% trisodium citrate. Then PRP was prepared, and collagen-induced (20 μg/ml) platelet aggregation was measured by the method described earlier.

In this antiplatelet study, NSP-513 and cilostazol levels in PPP at 0.5, 1, 2, 4, 6, and 8 h after drug dosing were measured by high-pressure liquid chromatography (HPLC) analysis.

Antithrombotic studies

Induction of femoral arterial thrombus formation. After overnight fasting, beagle dogs (7.5-12.8 kg) were anesthetized with sodium pentobarbital (30 mg/kg, i.v. bolus, followed by 5 mg/kg/h i.v. infusion). Each animal was intubated, and arterial blood gases were maintained in the ranges of 100-150 mm Hg (PO2), 30-40 mm Hg (PCO2), and 7.35-7.45 (pH) under artificial ventilation (model 613; Harvard, South Natick, MA, U.S.A.). Arterial blood pressure was measured by a pressure transducer (DX-360; Nihon Kohden, Tokyo, Japan) connected to the right carotid artery via a polyethylene catheter. Heart rate was determined as arterial pulse rate by a cardiotachometer (AT-601G; Nihon Kohden). A polyethylene cannula was inserted into the duodenum for drug administration. A stainless steel L-shaped anodal electrode, constructed from a 23-gauge needle tip (5 mm) attached to an enamel-insulated stainless wire (φ, 0.25 mm), was placed in contact with the intimal surface of the right femoral artery. The cathode was a silver hook-shaped wire that was attached to the right saphenous artery. An electromagnetic blood flow probe (φ, 3 mm, MFV-3100; Nihon Kohden) was placed around the femoral artery of the vessel upstream from an anodal electrode. A vessel occluder (VO-3; Star Medical, Tokyo, Japan) was placed around the vessel downstream from an anodal electrode.

Basal femoral blood flow was measured with an electromagnetic blood flow-meter. The electrical stimulus was applied to the femoral artery after 50 min of vessel occlusion, which was adjusted to produce a 50% reduction in the height of the waveform of basal pulsatile blood flow, and 30 min after intraduodenal administration of drug, either NSP-513 (0.01 and 0.03 mg/kg) or cilostazol (30 and 100 mg/kg). The experimental protocol is diagrammed in Fig. 1. The control animals received only vehicle before electrical stimulus, and the sham-operation group also received only vehicle, but no electrical stimulus. In the electrically stimulated groups, a 200-μA anodal current was delivered for 40 min by using an electronic stimulator (SEN-3201; Nihon Kohden) and an isolator (SS-102J; Nihon Kohden). Fifty minutes after the cessation of electrical stimulus, the femoral arteries of all animals were opened lengthwise, and the observed thrombi were removed and weighed immediately. The percentage inhibition of thrombus formation was calculated by the following formula: EQUATION where TWs, TWc, and TWd are the mean thrombus weights of the sham-operated, control, and drug-treated groups, respectively. The drug dose causing a 50% inhibition of thrombus formation (ED50) was calculated from the dose-inhibition curve. The change in blood flow was expressed as a percentage of the basal blood flow observed at 50 min before electrical stimulus.

FIG. 1
FIG. 1:
Experimental protocol for the canine femoral arterial thrombosis model with intraduodenal (i.d.) administration of NSP-513 and cilostazol. Stenosis, 50% reduction in the height of the waveform of basal pulsatile femoral artery blood flow; electrical current, stimulation with of 200-μA anodal current.

In this antithrombotic study, plasma levels of drugs and bleeding time were measured. Blood samples were collected from the right carotid artery into plastic syringes containing heparin (final concentration, 10 U/ml). Plasma was prepared by centrifugation at 1,700 g for 10 min at 4°C. NSP-513 and cilostazol levels in plasma were measured by HPLC analysis 60 min after drug dosing. Bleeding time was measured 30 min before, and 60 and 120 min after drug dosing (Fig. 1), as described later.

Measurement of template bleeding time. The mucous membrane on the buccal cavity side of each animal's upper lip was incised by using a Simplate R (Organon Teknika Corp., Durham, NC, U.S.A.). The blood shed by this procedure was blotted every 10 s with fresh filter paper without touching the edge of the wound. When the filter paper no longer became blood-stained, the time corresponding to the nearest 10 s was recorded.

Hemodynamic studies

Surgical preparation. Beagle dogs (9.0-13.1 kg) were anesthetized with sodium pentobarbital (35 mg/kg, i.v.). An endotracheal tube was introduced. All surgery was performed aseptically. Polyethylene catheters filled with heparin-saline (500 U/ml) were inserted into the left femoral artery to monitor blood pressure and heart rate, and into the left ventricle via the left carotid artery to monitor left ventricular pressure. After cannulation, one end of each catheter was subcutaneously fixed at the cervix, and the ambient skin was sutured. Animals were then allowed to recover from anesthesia. After the operation, the dogs were housed in individual cages, and allowed ≥7 days to recover from surgery.

Measurement of hemodynamic parameters. After overnight fasting, the cannulated, conscious, nonrestrained animals were used in the experiments. Arterial blood pressure was measured with a pressure transducer (DX-360; Nihon Koden). Heart rate triggered by the arterial pulse was counted with a cardiotachometer (AT-601G; Nihon Koden). The left ventricular pressure (LVP) was measured with a pressure transducer. As an index of cardiac contractility, left ventricular dP/dtmax (LVdP/dtmax) was obtained by differentiating the LVP signal with an electronic differentiator (EQ-601G; Nihon Koden). After determination of the baseline parameters, NSP-513 (0.03-1 mg/kg) and cilostazol (50 mg/kg) were orally administered by gavage in a volume of 2 ml/kg. The control animals received only vehicle.

Mean blood pressure, heart rate, and LVdP/dtmax were recorded before and 0.5, 1, 2, 3, 4, 5, 6, 7, and 8 h after drug dosing. The changes in mean blood pressure, heart rate, and LVdP/dtmax were expressed as changes (Δ) from the basal values determined just before drug dosing.


NSP-513 was synthesized in Nippon Soda Co., Ltd., Odawara, Japan. Cilostazol was extracted and purified from Pletaal tablets (Otsuka Pharmaceutical Co., Japan). NSP-513 and cilostazol were suspended in 0.5% methylcellulose just before the experiments were started. Collagen was purchased from NYCOMED (Germany). ADP and prostaglandin E1 (PGE1) were purchased from Sigma Chemical Co. (St. Louis, MO, U.S.A.). All other agents were of the highest grades commercially available.

Statistical analyses

Data are expressed as the mean ± SEM. The data were analyzed by one-way analysis of variance for group comparisons and for repeated measures followed by Dunnett's test to determine the level of significance.


In vitro antiplatelet effects

NSP-513 showed potent and concentration-dependent inhibition of canine platelet aggregation induced by collagen and ADP with IC50 values of 0.093 and 0.15 μM, respectively (Table 1). These IC50 values revealed that NSP-513 was ∼30-80 times more potent than cilostazol for inhibiting platelet aggregation. In addition, the IC50 values of NSP-513 for ADP-induced platelet aggregation were decreased to 0.11 and 0.032 μM by the addition of 3 and 10 nM PGE1 to the reaction mixture, respectively. PGE1 alone did not show any significant inhibition of ADP-induced canine platelet aggregation in vitro (data not shown).

Inhibitory effects of NSP-513 and cilostazol on ADP- or collagen-induced canine platelet aggregation

Ex vivo antiplatelet of orally administered NSP-513 in conscious dogs

In the group that received only vehicle, there were no differences in the platelet-aggregation responses to collagen during the experimental periods (n = 6). Orally administered NSP-513 (0.03-1 mg/kg) showed dose-dependent inhibition of ex vivo platelet-aggregation responses to collagen; maximal inhibition occurred at 1-2 h after drug administration (Fig. 2). The dose of NSP-513 that produced 50% inhibition (ID50), calculated from the maximal percentage inhibition of platelet aggregation, was estimated as 0.056 mg/kg. In the group treated with cilostazol at 50 mg/kg, maximal inhibition at 1 h after drug administration was 49.7 ± 14.2% (n = 6, p < 0.05).

FIG. 2
FIG. 2:
Effects of orally administered NSP-513 (top) and cilostazol (bottom) on ex vivo collagen-induced (20 μg/ml) platelet aggregation in conscious dogs. Changes in ex vivo platelet aggregation are expressed as percentage inhibition of the basal platelet aggregation measured just before drug dosing in each dog. Data represent mean ± SEM of five to six animals per group. *p < 0.05 versus vehicle.

NSP-513 at 0.03-1 mg/kg achieved a maximal plasma concentration (Cmax) at 0.5-2 h after oral dosing. The plasma level of NSP-513 increased in an approximately dose-proportional fashion with a half-life (T1/2) of 5-6 h. The plasma Cmax of NSP-513 at doses of 0.03, 0.1, 0.3, and 1 mg/kg was 9.2 ± 1.2 ng/ml (0.027 ± 0.003 μM), 29.5 ± 4.8 ng/ml (0.087 ± 0.014 μM), 110.0 ± 6.3 ng/ml (0.32 ± 0.02 μM), and 367.5 ± 80.0 ng/ml (1.08 ± 0.23 μM), respectively. In contrast, cilostazol at a 50 mg/kg achieved a Cmax of 0.39 ± 0.21 μg/ml (1.06 ± 0.57 μM) at 0.5-1 h after oral dosing and had a T1/2 value of 1.1 h.

In vivo antithrombotic effects of intraduodenally administered NSP-513 in anesthetized dogs

In the vehicle-treated control group, occlusive thrombus in the right femoral artery was induced by the electrical stimulus. As shown in Fig. 3, femoral blood flow was severely attenuated, and the thrombus weighed 48.1 ± 4.7 mg (n = 6) in the vehicle-treated group versus only 2.7 ± 2.0 mg in the sham-operated group (n = 5; Fig. 3). Intraduodenal administration of NSP-513 (0.01 and 0.03 mg/kg) appeared to attenuate the decreases in these circulation indices. The estimated ID50 for thrombus formation was 0.016 mg/kg for NSP-513. In the 0.03 mg/kg NSP-513-treated group, the average thrombus weighed only 12.9 ± 4.9 mg (n = 5; p < 0.05), and there was no significant change in femoral blood flow throughout the experiment (Fig. 3). Intraduodenal administration of NSP-513 (0.01 and 0.03 mg/kg) did not affect either mean blood pressure, heart rate, or buccal mucosal bleeding time (Fig. 4). The higher dose (0.03 mg/kg) of NSP-513 produced a plasma concentration of 11.6 ± 0.8 ng/ml (0.034 ± 0.002 μM) at 60 min after dosing. Parallel studies with cilostazol showed this drug to be less effective in its antithrombotic action. At a dose of 30 mg/kg, thrombus weight was 35.8 ± 8.7 mg, indicating only 27.1% inhibition (n = 4, NS). Even a higher dose of 100 mg/kg did not increase the antithrombotic effect (thrombus weight, 37.3 mg; 23.8% inhibition, n = 2). The plasma concentration of cilostazol was 0.24 ± 0.07 μg/ml (0.65 ± 0.19 μM) for a 30-mg/kg dose and, in two animals that received 100 mg/kg, the plasma level at 60 min after dosing was 0.27 μg/ml (0.73 μM) and nearly zero, respectively.

FIG. 3
FIG. 3:
Effects of intraduodenal (i.d.) administration of NSP-513 and cilostazol on femoral arterial blood flow (FBF) (top) and thrombus weight (bottom) in the canine femoral arterial thrombosis model. NSP-513, cilostazol, or vehicle was administered 30 min before the application of anodal current (200 μA, 40 min). The sham-operation group (sham ope.) received vehicle without the electrical stimulus. Changes in FBF are expressed as a percentage of the basal FBF measured just before the application of stenosis in each dog. The basal FBF values were 80.6 ± 14.5 ml/min (sham ope.), 78.8 ± 8.8 ml/min (vehicle), 76.5 ± 15.0 ml/min (NSP-513, 0.01 mg/kg), 70.6 ± 12.7 ml/min (NSP-513, 0.03 mg/kg), and 71.3 ± 13.0 ml/min (cilostazol, 30 mg/kg). Data represent mean ± SEM of four to six animals per group. *p < 0.05 versus vehicle.
FIG. 4
FIG. 4:
Effects of intraduodenal (i.d.) administration of NSP-513 and cilostazol on mean blood pressure (MBP) (top), heart rate (HR) (middle), and bleeding time (bottom) in the canine femoral arterial thrombosis model. NSP-513, cilostazol, or vehicle was administered 30 min before the application of anodal current (200 μA, 40 min). The sham-operation group (sham ope.) received vehicle without the electrical stimulus. Data represent mean ± SEM of four to six animals per group.

Hemodynamic effects of orally administered NSP-513 in conscious dogs

As shown in Fig. 5, oral administration of NSP-513 at doses of 0.03 and 0.1 mg/kg did not affect either mean blood pressure (MBP), heart rate (HR), or LVdP/dtmax. However, higher doses (0.3 and 1 mg/kg) of NSP-513 decreased MBP and increased HR and LVdP/dtmax. A 0.3-mg/kg dose of NSP-513 produced maximal effects on MBP (17.8 ± 2.4 mm Hg decrease at 1 h), HR (26.2 ± 8.0 beats/min increase at 5 h), and LVdP/dtmax (518.7 ± 195.9 mm Hg/s increase at 4 h).

FIG. 5
FIG. 5:
Effects of orally administered NSP-513 (left) and cilostazol (right) on mean blood pressure (MBP), heart rate (HR), and left ventricular dP/dt max (LVdP/dt max) in conscious dogs. Changes in MBP, HR, and LVdP/dt max are expressed as change (Δ) from the basal value just before oral dosing in each dog. The basal MBP values (in mm Hg) were 104 ± 8 (vehicle), 103 ± 4 (NSP-513, 0.03 mg/kg), 106 ± 4 (NSP-513, 0.1 mg/kg), 100 ± 6 (NSP-513, 0.3 mg/kg), 95 ± 5 (NSP-513, 1 mg/kg), and 104 ± 4 (cilostazol, 50 mg/kg). The corresponding basal HR values (in beats/min) were 106 ± 10, 106 ± 4, 108 ± 7, 114 ± 5, 109 ± 4, and 102 ± 7. The corresponding basal LVdP/dt max values (in mm Hg/s) were 2,870 ± 171, 2,896 ± 86, 2,793 ± 154, 2,933 ± 188, 2,851 ± 73, and 3,035 ± 132. Data represent mean ± SEM of six to seven animals per group. *p < 0.05 versus vehicle.

In another set of experiments, cilostazol was orally administered at a dose of 50 mg/kg. This drug produced similar effects over a similar time frame, with a decrease in MBP and increases in HR and LVdP/dtmax(Fig. 5). The hemodynamic effects of 50 mg/kg of cilostazol were comparable to those of 0.3 mg/kg of NSP-513.


In the ex vivo antiplatelet study, we confirmed that oral administration of NSP-513 (0.03-1 mg/kg, p.o.) inhibited collagen-induced platelet aggregation with an ID50 value of 0.056 mg/kg. The plasma Cmax values of NSP-513 after doses of 0.03 and 0.1 mg/kg were 9.2 ± 1.2 ng/ml (0.027 ± 0.003 μM) and 29.5 ± 4.8 ng/ml (0.087 ± 0.014 μM), respectively. These results suggest that the ex vivo antiplatelet potency of NSP-513 is nearly consistent with that observed in vitro. We also evaluated the ex vivo antiplatelet effects of cilostazol, another antiplatelet PDE 3 inhibitor. Cilostazol at a dose of 50 mg/kg (p.o.) produced ∼50% inhibition of collagen-induced platelet aggregation (Fig. 2). A comparison of ex vivo antiplatelet potency between the two drugs suggests that the inhibitory potency of NSP-513 is extremely higher than that of cilostazol. We have already confirmed the presence of PDE 2, 3, and 5 in human platelets. The IC50 values for NSP-513 on human PDE 2, 3, and 5 were 45, 0.039, and 33 μM, respectively. In contrast, the corresponding IC50 values for cilostazol were 5.6, 0.50, and 23 μM, respectively (15). PDE 2 hydrolyzes both cAMP and cyclic guanosine monophosphate (cGMP); it is thought that this enzyme participates in the catabolism of cAMP, particularly when it is activated by cGMP (18). However, the inhibition potency of NSP-513 for PDE 2 is rather less than that of cilostazol. Thus in view of the inhibition potency on PDE 2, it seems to be impossible to explain the difference in ex vivo potency between NSP-513 and cilostazol. Compared with NSP-513 at a dose of 0.3 mg/kg, cilostazol at 50 mg/kg has a similar Cmax (0.39 vs. 0.32 μg/ml), but a shorter T1/2 (1.1 vs. 5-6 h), indicating less favorable pharmacokinetic characteristics for cilostazol with respect to duration of action and (probably) bioavailability in the circulatory system of dogs. It has been reported that cilostazol exhibits dose-dependent inhibition of ex vivo collagen-induced platelet aggregation, and almost linear kinetics in plasma levels at the oral dose levels of 3-30 mg/kg (6, 19). Thus even without any additional doses, it is evident that cilostazol is less potent. Therefore, these pharmacokinetic differences may explain the difference in potency between the in vitro and ex vivo effects of cilostazol.

In the antithrombotic study, we used both physical stenosis and electrical injury as stimuli for experimental thrombus formation. Because both stress on platelets and injury to endothelial cells can stimulate platelet aggregation, the experimental arterial thrombosis model using artificial stenosis of a vessel and/or the injury to endothelial cells has been developed successfully by other investigators (20). In our canine femoral arterial thrombosis model, intraduodenal administration of NSP-513 exerted a potent antithrombotic effect (ID50, 0.016 mg/kg, i.d.) without influencing other hemodynamic indices such as blood pressure, HR, and buccal mucosal bleeding. In this antithrombotic study, a 0.03-mg/kg dose of NSP-513 produced 77.5% inhibition of thrombus formation (thrombus weight), and yielded a plasma level of 11.6 ± 0.8 ng/ml (0.034 ± 0.002 μM) at 60 min after dosing. This plasma level obtained with an i.d. dose of 0.03 mg/kg is consistent with that achieved at the same dose given p.o. in our ex vivo antiplatelet study. These findings suggest that the in vivo antithrombotic effectiveness of NSP-513 in our canine femoral arterial thrombosis model is more potent than its ex vivo antiplatelet effectiveness, despite the similar plasma levels under both in vivo and ex vivo conditions. Other investigators, using ibuprofen (a nonsteroidal antiinflammatory agent) in a canine coronary thrombosis model, found a similar discrepancy between inhibition of thrombosis formation in vivo and inhibition of platelet aggregation ex vivo (21). Generally, in vivo platelet aggregation in experimental thrombosis models is affected by more platelet activators than that observed in ex vivo models. Furthermore, platelet function is modulated by several endogenous platelet activators and inhibitors. It is well known that endothelial cells produce platelet inhibitors, such as prostacyclin (PGI2) and nitric oxide. cAMP level in platelets is a primary factor in the cell-aggregation process, and PGI2 is known to upregulate adenylate cyclase activity (22). To mimic the intact living system, PGE1 at a concentration that does not affect in vitro ADP-induced platelet aggregation was added to the reaction mixture in our in vitro experiments, because PGE1 shares the same receptor on the platelet cell surface with PGI2(23-25). Our results demonstrated that NSP-513 effectively suppressed ADP-induced platelet aggregation, depending on the PGE1 concentration. A related observation showed that the in vitro antiplatelet activity of cilostazol was potentiated in the presence of endothelial cells, which are thought to release PGI2(26). These in vitro studies may provide a clue to understanding the discrepancy between the in vivo efficacy of NSP-513 (ID50, 0.016 mg/kg) and its ex vivo efficacy (ID50, 0.056 mg/kg). Consequently, NSP-513 may be more potent with respect to its in vivo antithrombotic activity, because of the assistance of endothelially derived PGI2, than in either in vitro or ex vivo antiplatelet activity, although the interactions between platelets and the vascular endothelium are critical in determining the extent of in vivo aggregation (27,28). In addition, the experimental conditions differ between in vivo and in vitro platelet aggregation. Injury to the arterial endothelium in vivo reveals subintimal structures, such as collagen, which stimulate circulating platelets to adhere to the site of injury and release ADP or thromboxane A2, promoting further aggregation of platelets. In contrast to platelet aggregation in vivo, when collagen or ADP is added in vitro or ex vivo to platelets concentrated in PRP, large numbers of platelets simultaneously undergo metabolic changes, culminating in the release of proaggregating substances and massive platelet aggregation. The accumulation in vitro of these substances that promote platelet aggregation influences the sensitivity and reactivity of those platelets that were not stimulated by initial exposure to collagen. Hence, in vitro or ex vivo assessment of platelet function may be inherently insensitive to changes in platelet dynamics that may be of significance in vivo.

The parallel studies with cilostazol showed this drug to be less effective in its antithrombotic action. It has been shown that mainly PDE 2 and 4 are present in cultured bovine or pig aortic endothelial cells, and that PDE 2 inhibition markedly increases cAMP levels in endothelial cells when it is activated by an increase in cGMP resulting from an increase in nitric oxide (NO) (18,29). Because the increase in cAMP levels reduces PGI2 generation in endothelial cells (4), the possibility is considered that PDE 2 inhibition reduces PGI2 generation in endothelial cells. The potency of cilostazol to inhibit PDE 2 is 8 times higher than that of NSP-513, although the inhibition potency of cilostazol for PDE 4 is similar to that of NSP-513. Cilostazol was only 11 times more potent against human platelet PDE 3 than against PDE 2, whereas NSP-513 was 1,200 times more potent. Therefore one reason that cilostazol was less effective in an in vivo thrombosis model may be its inhibition potency for PDE 2 in endothelial cells. In addition, under the conditions of pentobarbital anesthesia, unfortunately, no increase in plasma level was observed when the dose was increased from 30 to 100 mg/kg, suggesting limited absorption or malabsorption of cilostazol into the circulation, which in turn may explain the reduced potency of this drug. In our unpublished study with dogs, the absorption of orally administered cilostazol seems to be completely saturated at 100 and 300 mg/kg. Thus we considered that the in vivo antithrombotic efficacy of cilostazol, even with higher doses, would be less potent than that of NSP-513.

More recently, glycoprotein (GP) IIb/IIIa antagonists have been administered during and/or shortly after angioplasty (30,31) to prevent ischemic episodes. Although the results of using GP IIb/IIIa antagonists in percutaneous transluminal coronary angioplasty (PTCA) studies are encouraging, bleeding is a major side effect of these drugs. Despite the antithrombotic efficacy of NSP-513, significant prolongation in buccal mucosal bleeding time was not found in our thrombosis model. These findings may indicate that the prolongation of hemostasis in mucosal small blood vessels, but not large arteries, is less sensitive to selective PDE 3 inhibition. However, because the mucosal bleeding time in our template model dose not necessarily reflect the potential bleeding risk in humans (32), NSP-513 should be used with caution.

NSP-513 at doses of 0.3 mg/kg and higher caused decreases in MBP and increases in HR and LVdP/dtmax. As reported previously (33,34), cilostazol, another PDE 3-selective inhibitor, produced similar hemodynamic effects. It has been suggested that inhibition of PDE activity results in an increase in cAMP level, leading to both arterial vasodilation (decrease in MBP) and cardiotonic action (increase in dP/dtmax) (35,36). In addition, because it is generally accepted that vasodilation causes reflex tachycardia (37-39), the NSP-513-induced increase in HR observed in our study may at least partly reflect the decreased blood pressure mediated by this drug's vasodilatory effect.

As mentioned, the hemodynamically effective dose of NSP-513 was ≥0.3 mg/kg in our study. In contrast, in antiplatelet and antithrombotic studies, the ID50 of NSP-513 was 0.056 mg/kg for platelet aggregation and 0.016 mg/kg for thrombus formation. Thus the hemodynamically effective dose (≥0.3 mg/kg) of this drug is 5 and 19 times higher than its ID50 values for platelet aggregation and thrombus formation, respectively. It has been reported that distinct PDE isoforms are distributed in different tissues (2,3), and that platelet PDE activity is more sensitive to cilostazol, a selective PDE 3 inhibitor, than is the PDE activity in vascular endothelial cells (4). Interestingly, Mochizuki et al. (35) reported that the selective PDE 3 inhibitors, which are structurally related to NSP-513 (dihydropyridazinone derivatives), showed biphasic positive inotropic effects. We also observed that NSP-513 exerted biphasic positive inotropic effects in isolated guinea-pig right atrium, despite yielding a monophasic concentration-inhibition curve for platelet aggregation (unpublished data). Thus the potential in vivo selectivity of NSP-513 for antithrombotic versus hemodynamic activity may be related to the different PDE isoforms present in platelets and hemodynamic tissues (vascular smooth muscle, ventricular muscle, etc.). It is widely assumed that basal vascular tone is maintained at a constant level by the systemic vascular autoregulation. Therefore we cannot rule out the possibility that the hemodynamic activity of low-dose NSP-513 may be masked by autoregulation in the in vivo vascular system.

Because the maximal hemodynamic change with cilostazol occurred at 50 mg/kg, a dose that exhibits ∼50% inhibition of platelet aggregation, and these changes were comparable to those seen with NSP-513 at a dose of 0.3 mg/kg, we could not determine the selectivity of cilostazol in this study. Thus the difference in selectivity between NSP-513 and cilostazol is still in question. However, we have obtained data showing that NSP-513 and cilostazol are 850-fold and 11-fold more selective for PDE 3 than for the other PDE isozymes (canine PDE 1, 2, and 4, and human PDE 5), respectively (15). Because NSP-513 is a much more selective PDE3 inhibitor than cilostazol, NSP-513's selectivity for antithrombotic over hemodynamic effects may be related in part to its extremely high selectivity for PDE 3. In addition, different tissue concentrations of cilostazol after oral administration have been observed in several cardiovascular tissues, including blood, aorta, vena cava, heart, etc. (19). Furthermore, it is possible that NSP-513 and cilostazol have different tissue distributions, although data for NSP-513 are not available. Further study is needed to elucidate the mechanism of NSP-513's selectivity for antithrombotic effects.

In conclusion, our results indicate that NSP-513 is a potent antithrombotic agent, effective for preventing thrombus formation. The ID50 of NSP-513 was 0.056 mg/kg (p.o.) for platelet aggregation and 0.016 mg/kg (i.d.) for thrombus formation. In addition, NSP-513 at oral doses of 0.03 and 0.1 mg/kg produced little or no hemodynamic change in MBP, HR, and LVdP/dtmax. However, higher doses (0.3 and 1 mg/kg) of the drug decreased MBP and increased HR and LVdP/dtmax. These results suggest that NSP-513 has in vivo selectivity for antiplatelet and antithrombotic activities over hemodynamic activity, and that this selectivity is higher than that of cilostazol in dogs. Thus NSP-513 may have an important potential for treating arterial thrombosis disorders with reduced hemodynamic adverse effects.


1. Beavo JA, Conti M, Heaslip RJ. Multiple cyclic nucleotide phosphodiesterases. Mol Pharmacol 1994;46:399-405.
2. Beavo JA, Reifsnyder DH. Primary sequence of cyclic nucleotide phosphodiesterase isozymes and the design of selective inhibitors. Trends Pharmacol Sci 1990;11:150-5.
3. Nicholson CD, Challiss RAJ, Shahid M. Differential modulation of tissue function and therapeutic potential of selective inhibitors of cyclic nucleotide phosphodiesterase isoenzymes. Trends Pharmacol Sci 1991;12:19-27.
4. Tani T, Sakurai K, Kimura Y, Ishikawa T, Hidaka H. Pharmacological manipulation of tissue cyclic AMP by inhibitors: effects of phosphodiesterase inhibitors on the functions of platelets and vascular endothelial cells. Adv Second Messenger Phosphoprol Res 1992;25:215-27.
5. Hidaka H, Hayashi H, Kohri H, et al. Selective inhibitor of platelet cyclic adenosine monophosphate phosphodiesterase, cilostamide, inhibits platelet aggregation. J Pharmacol Exp Ther 1979;211:26-30.
6. Kimura Y, Tani T, Kanbe T, Watanabe T. Effect of cilostazol on platelet aggregation and experimental thrombosis. Arzneimittelforschung/Drug Res 1985;35:1144-9.
7. Alvarez R, Banerjee GL, Bruno JJ, et al. A potent and selective inhibitor of cyclic AMP phosphodiesterase with potential cardiotonic and antithrombotic properties. Mol Pharmacol 1986;29:554-60.
8. Lanza F, Beretz A, Stierle A, Corre G, Cazenave JP. Cyclic nucleotide phosphodiesterase inhibitors prevent aggregation of human platelets by raising cyclic AMP and reducing cytoplasmic free calcium mobilization. Thromb Res 1987;45:477-84.
9. Simpson AWM, Reeves ML, Rink TJ. Effects of SK&F94120, an inhibitor of cyclic nucleotide phosphodiesterase type III, on human platelets. Biochem Pharmacol 1988;37:2315-20.
10. Lugnier C, Stierle A, Beretz A, et al. Tissue and substrate specificity of inhibition by alkoxy-aryl-lactams of platelet and arterial smooth muscle cyclic nucleotide phosphodiesterases: relationship to pharmacological activity. Biochem Biophys Res Commun 1983;113:954-9.
11. Hidaka H, Tanaka T, Itoh H. Selective inhibitors of three forms of cyclic nucleotide phosphodiesterases. Trends Pharmacol Sci 1984;5:237-9.
12. Shaid M, Nicholson CD. Comparison of cyclic nucleotide phosphodiesterase isoenzymes in rat and rabbit ventricular myocardium: positive inotropic and phosphodiesterase inhibitory effects of Org 30029, milrinone and rolipram. Naunyn Schmiedebergs Arch Pharmacol 1990;342:698-705.
13. Katsumura T, Masaki H. Antiaggregatory effect of cilostazol (OPC-13013) in arteriosclerotic disease. Jpn Pharmacol Ther 1986;14:1531-6 (Abstr in English).
14. Okuda Y, Kimura Y, Yamashita K. Cilostazol. Cardiovasc Drug Rev 1993;11:451-65.
15. Hirose H, Kimura T, Okada M, et al. Antiplatelet and antithrombotic effects of a novel selective phosphodiesterase 3 inhibitor, NSP-513, in mice and rats. Jpn J Pharmacol 2000;82:182-99.
16. Mochizuki N, Uchida S, Kanaguchi Y, Ichihara H. Nishibe T. Selective inhibiting action against cAMP phosphodiesterase, dog platelet and influence on cardiovascular system. 114th Ann Meet Pharmaceut Soc Jpn 1994;Abst 29(P1):13-48 (Japanese).
17. Born GVR. Aggregation of blood platelets by adenosine diphosphate and its reversal. Nature 1962;194:927-9.
18. Lugnier C, Schini VB. Characterization of cyclic nucleotide phosphodiesterases from cultured bovine aortic endothelial cells. Biochem Pharmacol 1990;39:75-84.
19. Akiyama H, Kudo S, Shimizu T. The absorption, distribution and excretion of a new antithrombotic and vasodilating agent, cilostazol, in rat, rabbit, dog and man. Arzneimittelforschung 1985;35:1124-32.
20. Bush LR, Shebuski RJ. In vivo models of arterial thrombosis and thrombolysis. FASEB J 1990;4:3087-98.
21. Romson JL, Bush LR, Haack DW, Lucchesi BR. The beneficial effects of oral ibuprofen on coronary artery thrombosis and myocardial ischemia in the conscious dog. J Pharmacol Exp Ther 1980;245:271-8.
21. Romson JL, Bush LR, Haack DW, Lucchesi BR. The beneficial effects of oral ibuprofen on coronary artery thrombosis and myocardial ischemia in the conscious dog. J Pharmacol Exp Ther 1980;245:271-8.
22. Weiss HJ, Turitto V. Prostacyclin (prostaglandin I2) inhibits platelet adhesion and thrombus formation on subendothelium. Blood 1979;53:244-50.
23. Lombroso M, Nicosia S, Paoletti R, Whittle BJR, Moncada S, Vane JR. The use of stable prostaglandins to investigate prostacyclin (PGI2)-binding sites and PGI2-sensitive adenylate cyclase in human platelet membranes. Prostaglandins 1984;27:321-33.
24. Schillinger E, Prior G. Prostaglandin I2 receptors in a particulate fraction of platelets of various species. Biochem Pharmacol 1980;29:2297-9.
25. Shepherd GL, Lewis PJ, Blair IA, de Mey C, McDermot J. Epoprostenol (prostacyclin, PGI2) binding and activation of adenylate cyclase in platelets of diabetic and control subjects. Br J Clin Pharmacol 1983;15:77-81.
26. Igawa T, Tani T, Chijiwa T, et al. Potentiation of anti-platelet aggregating activity of cilostazol with vascular endothelial cells. Thromb Res 1990;57:617-23.
27. Moncada S, Vane JR. Prostacyclin, platelet aggregation, and thrombosis. In: de Ganetano G, Garattini S, eds. Platelets: a multidisciplinary approach. New York: Raven Press, 1978;239-358.
28. Shafer AI, Handin RI. The role of platelets in thrombotic and vascular disease. Prog Cardiovasc Dis 1979;22:31-52.
29. Souness JE, Diocee BK, Martin W, Moodie SA. Pig aortic endothelial-cell cyclic nucleotide phosphodiesterase: use of phosphodiesterase inhibitors to evaluate their roles in regulating cyclic nucleotide levels in intact cells. Biochem J 1990;266:127-32.
30. Frishman WH, Burns B, Atac B, Alturk N, Altajar B, Lerrick K. Novel antiplatelet therapies for treatment of patients with ischemic heart disease: inhibitors of the platelet glycoprotein IIb/IIIa integrin receptor. Am Heart J 1995;130:877-92.
31. Tcheng JE. Enhancing safety and outcomes with the newer antithrombotic and antiplatelet agents. Am Heart J 1995;130:673-9.
32. Rodgers RP, Levin J. A critical reappraisal of the bleeding time. Semin Thromb Haemost 1990;16:1-20.
33. Kawamura K, Watanabe K, Kimura Y. Effect of cilostazol, a new antithrombotic drug, on cerebral circulation. Arzneimittelforschung 1985;35:1149-54.
34. Shintani S, Toba Y, Suzuki S, Ninomiya S, Umezato M, Hiyama T. General pharmacological properties of cilostazol: a new antithrombotic drug. Part 1: Effects on the central nervous system. Arzneimittelforshung 1985;35:1157-62.
35. Mochizuki N, Uchida S, Miyata H. Cardiovascular effects of NSP-804 and NSP-805, novel cardiotonic agents with vasodilator properties. J Cardiovasc Pharmacol 1993;21:983-95.
36. Silver PJ, Pagani ED. Biochemical and preclinical pharmacology of selective inhibitors of cardiovascular phosphodiesterase isozymes. In: Gwathmey JK, Briggs GM, Allen PD, eds. Heart failure: basic science and clinical aspects. New York: Marcel Dekker, 1993;387-98.
37. Dogterom P, Zbinden G. Cardiotoxicity of vasodilators and positive inotropic/vasodilating drugs in dogs: an overview. Crit Rev Toxicol 1992;22:203-41.
38. Spokas EG, Wang HH. Regional blood flow and cardiac responses to hydralazine. J Pharmacol Exp Ther 1980;212:294-303.
39. Lokhandwala MF. Pharmacological analysis of the tachycardia produced by histamine and specific H1- and H2-receptor agonists in anesthetized dogs. Eur J Pharmacol 1978;51:351-9.

NSP-513; Phosphodiesterase 3; Arterial thrombosis; Dog

© 2000 Lippincott Williams & Wilkins, Inc.