Journal Logo


Comparison of the Effects of Cilostazol and Milrinone on Intracellular cAMP Levels and Cellular Function in Platelets and Cardiac Cells

Cone, James; Wang, Sheng; Tandon, Narendra; Fong, Miranda; Sun, Bing; Sakurai, Kazushi; Yoshitake, Masuhiro; Kambayashi, Jun-ichi; Liu, Yongge

Author Information
Journal of Cardiovascular Pharmacology: October 1999 - Volume 34 - Issue 4 - p 497-504
  • Free


Cilostazol (OPC-13013, 6-[4-(1-cyclohexyl-1H-tetrazol-5-yl)butoxy]-3,4-dihydrocarbostyril) has been used in Japan and other Asian countries for the treatment of lower-extremity peripheral arterial occlusive disease (PAOD) (1-3). Intermittent claudication is the most common debilitating symptom of PAOD (4). Patients with intermittent claudication are severely impaired in their ability to walk and perform daily activities (5). Recent clinical trials have shown that cilostazol significantly improves exercise performance in intermittent claudication patients (6-8), and it recently was approved by the Food and Drug Administration (FDA) for this indication (9). Although the mechanism of action by which cilostazol exerts its beneficial effects is not fully understood, it is generally believed that it involves the inhibition of cyclic nucleotide phosphodiesterase (PDE) type 3 (PDE3) and a corresponding increase in intracellular cyclic adenosine monophosphate (cAMP) levels in platelets (7,10,11). Elevation of intracellular cAMP in platelets inhibits platelet aggregation (12). PDE3 also is a major PDE isoform in cardiac ventricular myocytes (13-15). Elevation of intracellular cAMP in these cells increases contractility. Milrinone, as well as other PDE3 inhibitors, has been used as cardiotonic agents for heart failure patients (16,17). These inhibitors have shown beneficial effects on relieving the symptoms, but long-term use in severe congestive heart failure patients is associated with proarrhythmic activities (possibly caused by excessive increase of intracellular cAMP in the heart) (18) and increased mortality (19-21). Based on previous data of PDE3 inhibitors in heart failure patients, the FDA thus warned against the use of cilostazol by patients with heart failure. Although it is not known whether inhibition of PDE3 also would increase mortality in patients without severe congestive heart failure, the cardiac effects of cilostazol are still a concern, especially as the population of intermittent claudication patients may overlap with cardiac ischemic disease patients (22,23).

In eight placebo-controlled clinical trials involving 2,702 patients (patients with heart failure were excluded), cilostazol showed an excellent safety profile with a minimal increase in cardiac adverse events (e.g., palpitation, tachycardia) and no increase in mortality (24). Judging by the clinical experience, it may be possible that cilostazol is different from conventional PDE3 inhibitors, such as milrinone. To explore this possibility, we compared the ability of cilostazol and milrinone to increase intracellular cAMP in rabbit and human platelets, isolated rabbit ventricular myocytes, and rabbit coronary smooth muscle cells. We also studied the effects of these two agents on platelet aggregation and cardiac function in an isolated rabbit-heart model. Cilostazol has several metabolites, including OPC-13015, whose PDE3 inhibitory activity is sevenfold more potent than cilostazol with relatively high concentrations in plasma in vivo (24). Thus the cardiac effects of OPC-13015 also were studied. Our results indicate that cilostazol preferentially affects vascular circulation with minimal effects on cardiac ventricular function. This is consistent with its beneficial and safe effects on patients with PAOD.


This study was conducted in accordance with the "Guide for the Care and Use of Laboratory Animals," published by the National Research Council, 1996, Washington DC, and approved by the Institutional Animal Care and Use Committee. For human blood donors, they signed consent forms in conformity with the Declaration of Helsinki.


Cilostazol (lot 4576M) and OPC-13015 (lot 4F83M) were provided by Otsuka Pharmaceutical Co., Ltd., Tokushima, Japan. Collagenase (type II) was obtained from Worthington (Freehold, NJ, U.S.A.). Milrinone (lot 46H4097), ODQ (1H-[1,2,4]oxadiazolo[4,3-a]quinoxalin-1-one), and all the other agents were purchased from Sigma Chemical Co. (St. Louis, MO, U.S.A.). Cilostazol, OPC-13015, milrinone, and ODQ were dissolved in dimethyl sulfoxide (DMSO) at a stock concentration of 100 mM (10 mM for ODQ) and were diluted into buffer to achieve the final desired concentrations.

Cell preparation

Platelets. Peripheral blood samples were collected from six healthy volunteers (medication free) by a two-syringe technique, or five anesthetized rabbits (2-2.5 kg, New Zealand White) by using a 19G butterfly needle. Nine volumes of blood were drawn directly into a syringe containing one volume of 3.8% sodium citrate. Platelet-rich plasma (PRP) was obtained by centrifuging the whole blood at 150 g (800 rev/min in a benchtop centrifuge, model TJ-6, Beckman Instruments, Inc., Fullerton, CA, U.S.A.) for 20 min at room temperature. PRP was recovered without disturbing the buffy coat and packed red blood cells, and EDTA and EGTA were added (final concentration of 1 mM). The pH of the resultant PRP was reduced to 6.5 with citric acid (4 μl/ml of 1 M stock) before the sedimentation of platelets at 1,000 g (2,200 rpm) for 15 min. The platelet pellet was resuspended in 1/5 volumes of platelet-washing buffer (PWB, containing 128 mM NaCl, 5.5 mM dextrose, 4.26 mM NaH2PO4, 7.46 mM Na2HPO4, 4.77 mM trisodium citrate, and 2.35 mM citric acid, pH 6.5, containing 1 mM EDTA, 1 mM EGTA, and 0.5% BSA). The platelets were recovered by centrifugation at 800 g in a fixed-speed centrifuge (Clay Adams Sero-Fuge II, Becton Dickinson and Company, Sparks, MD, U.S.A.) for 3.5 min. The supernatant was discarded, and the platelets were washed once with PWB. Finally the platelets were resuspended in Tyrode's HEPES buffer (136.7 mM NaCl, 5.5 mM dextrose, 2.6 mM KCl, 13.8 mM NaHCO3, 1 mM MgCl2, 0.36 mM NaH2PO4, and 10 mM HEPES, pH 7.4), counted by using a Coulter ZM cell counter (Coulter Corporation, Hialeah, FL, U.S.A.), and platelet count was adjusted to 4 × 108 platelets/ml.

Rabbit ventricular myocytes. Isolated ventricular myocytes were obtained from rabbit hearts by conventional enzymatic dissociation methods (25). In brief, hearts were quickly excised from anesthetized (30 mg/kg, i.v., pentobarbital) rabbits (New Zealand White) weighing 2-3 kg, and the aortas were cannulated for coronary perfusion on a Langendorff apparatus (Radnoti Glass Inc., Monrovia, CA, U.S.A.). The heart was perfused at 75 mm Hg of pressure with modified Krebs-Henseleitbicarbonate buffer solution composed of 118 mM NaCl, 25 mM NaHCO3, 1.2 mM KH2PO4, 4.75 mM KCl, 1.2 mM MgSO4, 2 mM CaCl2, and 10 mM dextrose. The perfusate was bubbled with 95% O2 and 5% CO2 and maintained at 37°C. After 5-min perfusion, the heart was perfused without Ca2+ for another 5 min, after which the perfusion solution was switched to one containing collagenase (0.8 mg/ml) for 20-30 min. The heart was then removed from the Langendorff apparatus, and the atria were trimmed away. The ventricles were minced into small pieces, and the suspension was incubated in a shaking bath at 37°C for another 10 min in collagenase-containing solution. Cells were filtered through a polyethylene mesh filter with a pore size of 200 μm. After washing with Ca2+-free Krebs, Ca2+ was added back stepwise (0.25 mM every 5 min) to a final concentration of 1 mM. Cell viability (rod-shaped cells) was examined microscopically. Preparations with >60% viability were used for subsequent experiments.

Rabbit coronary smooth muscle cells. Cells were grown from explants of rabbit coronary arteries in Dulbecco's modified Eagle's medium (DMEM) with 10% bovine fetal serum (26). Smooth muscle cells were confirmed by "hill-and-valley" growth pattern and an anti-α-actin antibody (Sigma A-2547). Passages four and five of cells were plated in six-well Falcon plates and grown to confluence.

Intracellular cAMP measurement

Platelets. First, 1 μl of one of the test agents (cilostazol or milrinone at 0.2, 0.6, 2, 6, and 20 mM in DMSO) was aliquoted into separate polypropylene test tubes to final concentrations of 1, 3, 10, 30, and 100 μM. Tyrode's HEPES buffer and DMSO were used as controls. Each concentration point was assayed in duplicate. After prewarming at 37°C, 0.2 ml of platelet suspension was added to the tubes and mixed with a test agent by quickly vortexing. The tube was then incubated at 37°C for 3 min. The reaction was terminated by the rapid addition of 50 μl ice-cold 1.25N perchloric acid (PCA). After freeze-thawing once, the reaction mixtures were neutralized by mixing well with 50 μl of 1.25 M KHCO3 and then centrifuged at 3,500 rev/min (2000 g) for 15 min at 4°C. The resulting supernatants were collected and used for intracellular cAMP measurement. The cAMP concentration of the samples was measured by using a commercially available cAMP radioimmunoassay kit (NEK-033, DuPont NEN, Boston, MA, U.S.A.), according to the acetylated procedure described in the instructions from the manufacturer. Experiments were performed in duplicate.

Myocytes. Myocytes were rinsed twice with HEPES-buffered Tyrode's solution (140 mM NaCl, 5 mM KCl, 1 mM MgCl2, 2 mM CaCl2, 10 mM dextrose, and 10 mM HEPES, pH 7.4) and resuspended in the same buffer at a density of 2 × 105 cells/ml. Two microliters of one of the test agents (cilostazol and milrinone at 0.125, 0.375, 1.25, 3.75, and 12.5 mM in 12.5% DMSO, except 12.5 mM cilostazol in 62.5% DMSO for solubility reasons) were aliquoted separately into polypropylene test tubes to give a final concentration of 1, 3, 10, 30, and 100 μM. The corresponding concentrations of DMSO were used as controls. After prewarming at 37°C, 0.25 ml of the rabbit myocyte suspension was added to the tubes and mixed with the test agents by gentle vortexing. The tube was then incubated at 37°C for 10 min. The reaction was terminated by adding 50 μl of ice-cold 1.5N PCA. After freeze-thawing once, the reaction mixtures were neutralized by mixing well with 50 μl of 1.5 M KHCO3 and then centrifuged at 3,500 rev/min (2000 g) for 15 min at 4°C. The resulting supernatants were collected, and cAMP was measured by using the same method as mentioned for platelets. The myocytes from each rabbit were tested in duplicate.

Smooth muscle cells. Cells were washed once with warm assay buffer (D-PBS with Ca2+, Mg2+, and glucose; Gibco, Gaithersburgh, MD, U.S.A.). Cells were then treated with cilostazol, milrinone, or vehicle (DMSO) for 10 min at 37°C. The medium was then aspirated, and 1 ml of ice-cold 0.1N HCl was added. The plates were placed on a shaker for 1 h at 4°C to extract intracellular cAMP. To minimize any PDE activity during the extraction, 1 mM 3-isobutyl-1-methylxanthine (IBMX), a nonselective PDE inhibitor, was added into each well. The supernatants were obtained by centrifugation and used for cAMP measurement as mentioned for platelets. Experiments were performed in duplicate.

Platelet aggregation

Washed platelet suspensions in Tyrode's HEPES buffer (3-4 × 108 platelets/ml) were maintained at room temperature until use. All experiments were performed within 3-4 h after blood collection. The platelet suspension (300 μl) was transferred to the aggregation cuvettes and warmed at 37°C for 1 min before transferring the cuvette into the aggregation module of a four-channel aggregometer (AG10, Kowa Co., Nagoya, Japan) capable of measuring the kinetics of particle formation by laser scattering and aggregation by light transmission. Drug or vehicle was added to the sample such that the concentration of DMSO did not exceed 0.2%, and the sample was stirred with the drug for 1 min at 37°C before addition of the agonist. Aggregation was monitored for 6 min after the addition of 1 μg/ml collagen. Maximal light transmission at the end of 6 min was taken as a measure of aggregation. Aggregation values are presented as the percentage of control (vehicle) aggregation.

Isolated heart studies

Langendorff preparation. Male New Zealand White rabbits, weighing 1.4-2 kg, were used in this study. Rabbits were anesthetized with sodium pentobarbital (30 mg/kg) through a marginal ear vein, and the animals were then mechanically ventilated with a tracheal cannula. Hearts were exposed through a midline incision of the chest, quickly excised by an incision at the base of the heart, and put into ice-cold Krebs-Henseleit bicarbonate buffer. The heart was then attached to a Langendorff apparatus by the aortic root and perfused with nonrecirculating Krebs-Henseleit buffer at a constant pressure of 75 mm Hg. The perfusate was bubbled with 95% O2 and 5% CO2 gas mixture, and the bubbling rate was adjusted to maintain physiologic pH (7.35-7.45). Perfusate temperature was maintained at 38°C by a circulating water jacket surrounding the buffer reservoirs. The heart also was maintained at 38°C by a water-jacketed housing in which it was suspended. The open top of the jacket was covered with a piece of parafilm to maintain the humidity and temperature. The pulmonary artery was cannulated for coronary flow rate measurement. A saline-filled latex balloon, connected via a catheter to a pressure transducer, was inserted into the left ventricle and inflated to yield an end-diastolic pressure of 0-5 mm Hg. The pressure transducer was connected to a chart recorder (Model 7 polygraph, Grass Instrument Co., Quincy, MA, U.S.A.) to record left ventricular pressure and its first derivative (dp/dt) and heart rate. Coronary flow was measured by a timed collection of the effluent in a graduated cylinder. Hearts with left ventricular developed pressure <85 mm Hg at the end of the 20-min equilibrium period were not included in the study.

Protocol.Figure 1 shows the scheme of the experimental protocols. All the hearts had a 20-min stabilization period before the drug treatment. The drug concentrations tested were 1, 3, 10, and 30 μM. Cilostazol and OPC-13015 precipitated from the solution at 100 μM after several minutes of gassing; therefore 100 μM was not studied. For each concentration, hearts were exposed to 5 min of the drug followed by a 10-min washout period. The cardiac functional measurements were performed at the end of the 5-min drug perfusion and at the end of the 10-min washout period. At both these time points, cardiac function had stabilized. Each heart was exposed to only one drug but to all the concentrations tested in a stepwise manner. Because 10 min was sufficient to wash out the previous drug effect, the drug effect is expressed as the percentage change of values before and after each drug concentration. Equation (1)

FIG. 1
FIG. 1:
Experimental protocol for isolated rabbit heart studies. Arrow, where the measurements were taken. Milrinone, cilostazol, and OPC-13015 were tested by using a concentration escalating protocol from 1 to 30 μM.

OPC-13015 also was evaluated in this series of experiments, because it is the only metabolite with even higher potency of PDE3 inhibition than its parent compound.

To examine whether the nitric oxide (NO)-guanylyl cyclase pathway plays a role in the cardiac effects of cilostazol, we studied a group of rabbit hearts treated with a potent and selective NO-sensitive guanylyl cyclase inhibitor ODQ (27). Hearts were first perfused with 10 μM cilostazol for 5 min, followed by drug-free buffer for 10 min. Perfusion was then switched to ODQ (1 μM) for 5 min followed by 5 min of ODQ (1 μM) plus 10 μM cilostazol. The percentage changes induced by cilostazol were compared without or with ODQ.

Statistical analysis

Data are presented as the percentage change from baseline (mean ± SEM). A value of p < 0.05 was taken as the level of statistical significance (random-effect modeling approach, SAS). Because repeated measurements were taken from each volunteer, each rabbit or each heart, the measurements for each assay were correlated. Therefore in evaluating the treatment and concentration effects, the random-effects modeling approach was used according to Wald statistics (28,29). In these analyses, the intercept for each subject is considered random. PROC MIXED in SAS was used to fit the random-effects models.


cAMP studies

The basal levels of cAMP were comparable among experiments in each group, and the averaged values were higher in human platelets (1.9 ± 0.2 pmol/108 cells; five experiments) than those in rabbit platelets (0.3 ± 0.1 pmol/108 cells; five experiments). Cilostazol and milrinone both significantly increased intracellular cAMP in human (Fig. 2A) and rabbit platelets (Fig. 2B) in a concentration-dependent manner (p < 0.01 at each concentration vs. baseline value). Statistical analysis showed no significant difference between the two agents. The overall effect of cilostazol and milrinone on intracellular cAMP accumulation was greater in rabbit platelets than in human platelets.

FIG. 2
FIG. 2:
The percentage changes of intracellular cyclic adenosine monophosphate (cAMP) levels from baseline.A, B: Concentration-dependent elevation of cAMP in human platelets (average of six experiments) and rabbit platelets (average of five experiments), respectively. C: Concentration-dependent increases of cAMP in rabbit ventricular myocytes (average of seven experiments). #p < 0.05 versus the respective effect from milrinone. *p < 0.005 versus the respective effects from milrinone. D: Changes in rabbit coronary smooth muscle cells (average of seven experiments). *p < 0.01 versus the effects from milrinone. Open bar, milrinone. Solid bar, cilostazol.

In rabbit cardiac myocytes, preliminary time-course studies were conducted at 5, 10, or 30 min of drug incubation. All three treatments caused similar increases in cAMP (data not shown). Thus we chose 10 min of drug incubation for a subsequent comparison study (Fig. 2C). The basal levels of cAMP were comparable among experiments, and the averaged value was 78 ± 5.2 pmol/106 cells. Milrinone significantly increased cAMP from baseline at each concentration (p < 0.05), whereas cilostazol caused significant increases only at 10 μM and higher. Furthermore, compared with milrinone, the increases induced by cilostazol were significantly smaller at 10 μM (p < 0.05) and 30 and 100 μM (p < 0.005).

The basal levels of cAMP were comparable among experiments in rabbit coronary smooth muscle cells, and the averaged value was 70.1 ± 4.2 pmol/106 cells. Milrinone elevated cAMP levels at ≥10 μM (p < 0.01 vs. baseline), as shown in Fig. 2D. However, cilostazol did not increase intracellular cAMP in any of the concentrations tested.

Platelet aggregation

Consistent with the results of intracellular cAMP measurement, the potency of cilostazol and milrinone to inhibit human platelet aggregation was similar (Fig. 3). The median inhibitory concentration (IC50) values (obtained by sigmoidal fitting of the mean data) were 0.9 μM for cilostazol and 2 μM for milrinone.

FIG. 3
FIG. 3:
Concentration-dependent inhibition of collagen-induced human platelet aggregation by cilostazol and milrinone (average of four experiments). Cilostazol has a calculated IC50 of 0.9 μM, whereas the value for milrinone is 2 μM. The data were fitted with a sigmoidal curve and an R > 0.95. Open circles, milrinone. Solid circles, cilostazol.

Isolated heart studies

We performed vehicle control experiments in three hearts but did not observe any effect arising from any of the DMSO concentrations (maximal, 0.1%) used in our drug-treatment study. Cardiac function deterioration was <10% during the 90 min of perfusion. The basal cardiac functions were comparable among three groups, as shown in Table 1, and the percentage changes from baseline caused by milrinone, cilostazol, and OPC-13015 are summarized in Fig. 4.

Basal cardiac function
FIG. 4
FIG. 4:
Summarized data from cardiac function study. Open bar, milrinone (n = 8). Solid bar, cilostazol (n = 8). Hatched bar, OPC-13015 (n = 6). *p < 0.05 versus the respective effect from milrinone.#p < 0.05 versus the effect from milrinone or cilostazol.

Figure 4A shows the changes of LVDP. Milrinone increased LVDP at all the concentrations tested (p < 0.05 vs. baseline). Cilostazol did not have a significant effect on LVDP. Milrinone also increased dp/dt (maximal values of dp/dt as an index of contractility) from baseline values (p < 0.01; Fig. 4B). Cilostazol at 1 and 3 μM did not increase dp/dt (−2 ± −3% and 5 ± 2%, respectively). At concentrations of 10 and 30 μM, it did have a positive inotropic effect (13 ± 5% and 18 ± 6% increases, respectively; p < 0.05 vs. baseline). The increases of dp/dt seen with milrinone were significantly higher than those with cilostazol at all the concentrations tested (p < 0.05). There were no dose-dependent responses from OPC-13015 on LVDP and dp/dt. Although 1 μM OPC-13015 increased both LVDP and dp/dt (p < 0.05 vs. basal value), as did milrinone, there was no further increase at higher concentrations. Coronary-flow changes are shown in Fig. 4C. Milrinone and cilostazol concentration-dependently increased coronary flow at ≥3 μM (p < 0.01 vs. baseline), whereas OPC-13015 significantly increased coronary flow at all the concentrations tested (p < 0.01 vs. baseline). There was no significant difference among the three groups at any concentration. Heart rates did not change significantly after milrinone or cilostazol perfusion, although both agents showed a slightly positive chronotropic effect (Fig. 4D). The increase in heart rate seen with OPC-13015 was even less, compared with milrinone or cilostazol.

ODQ is a potent and selective NO-sensitive guanylyl cyclase inhibitor (27). Treating the heart with 1 μM ODQ did not affect the percentage changes of cardiac function caused by cilostazol. In the absence of ODQ, 10 μM cilostazol caused a 22 ± 5% increase in dp/dt and 55 ± 9% increase in coronary flow (n = 5). These values are similar to those obtained in the presence of 1 μM ODQ (25 ± 3% and 59 ± 13% increases, respectively; n = 5). A potent nitric oxide synthase inhibitor, NG-nitro-L-arginine (L-NNA), did not affect the cardiac effects of cilostazol either (data not shown).


Although progress has been made in surgical and non-pharmacologic interventional techniques for the management of PAOD in recent years, drug therapy is still an important noninvasive intervention to provide symptomatic relief by improving pain-free and overall walking distance (30). Cilostazol has been used successfully to treat patients with PAOD for >10 years in Japan and other Asian countries (1-3). Cilostazol also provided efficacy for the treatment of the symptoms of intermittent claudication in several recent clinical trials in the United States and Europe (6,7). The FDA recently approved cilostazol for such use (9). Although the underlying mechanisms of such treatment are not fully understood, the beneficial effects may be due to the inhibition of PDE enzymes, in particular PDE3 isoforms, resulting in elevation of intracellular cAMP levels and subsequent inhibition of platelet aggregation and vasodilation. PDE3 inhibitors also enhance cardiac contractility (15,31). The increased heart contractility and decreased peripheral resistance from PDE3 inhibitors were originally thought to be a perfect combination to treat heart failure. However, long-term use of PDE3 inhibitors in severe congestive heart failure patients has been associated with increased mortality (19-21). Although there are no data on the effects of PDE3 inhibitors on people without severe congestive heart failure, the possible adverse cardiac effects of cilostazol are still a safety concern.

In this study, we compared the potency of cilostazol with that of a conventional PDE3 inhibitor, milrinone, on the elevation of intracellular cAMP levels in platelets, cardiac ventricular myocytes, and coronary artery smooth muscle cells, and on platelet and cardiac function. Consistent with their equal inhibitory potencies on PDE3 (32), cilostazol and milrinone concentration-dependently increased intracellular cAMP in human and rabbit platelets and inhibited human platelet aggregation with similar potency. These data are consistent with those reported in the literature (10,11,33-35).

We next measured intracellular cAMP levels in rabbit ventricular myocytes and cardiac contractility, heart rate, and coronary flow in isolated rabbit hearts. Our results show that cilostazol did not increase cAMP as much as milrinone did. This is in agreement with the effects on cardiac contractility: whereas milrinone concentration-dependently increased cardiac contractility, the overall effect from cilostazol was significantly smaller. These results suggest that cilostazol is quite different from milrinone, and the contractility changes from cilostazol cannot be solely explained by PDE3 inhibition because the potencies of cilostazol and milrinone on PDE3 are very similar (32).

PDE3 inhibitors also have profound vasodilatory effects (36). To our surprise, although cilostazol increased coronary flow as potently as milrinone in our isolated rabbit heart model, it did not elevate intracellular cAMP levels in rabbit coronary smooth muscle cells. In the cAMP measurement study, we used 10 min for drug treatment (the same as for myocytes). It is unlikely that the effect of cilostazol on cAMP levels is delayed because we saw the increase of coronary flow within 1 min after the start of cilostazol perfusion. There were other studies on the effect of cilostazol on vascular smooth muscle cells. Ikeda et al. (37) showed that treatment with 10 μM cilostazol for 24 h increases cAMP in rat aorta smooth muscle cells. However, the marked difference of drug-exposure time makes it impossible to compare that study directly with ours. Consistent with our results, Tanaka et al. (38) did not see an increase of intracellular cAMP by 12-min incubation of 10 μM cilostazol with rabbit aortic artery smooth muscle cells. When Lindgren et al. (34) compared OPC-3911 (a PDE3 inhibitor structurally similar to cilostazol) with milrinone in platelet and rat aorta, they obtained results similar to what we found. In platelets, OPC-3911 and milrinone increased cAMP levels, but in the rat aorta, the increase was significant only for milrinone despite the similar potency in inhibition of PDE3 (39). Nevertheless, OPC-3911 and milrinone had similar relaxant profiles in human and rat coronary and renal arteries (40). We hypothesize that the poor correlation between cAMP level elevation and flow increase seen with cilostazol may reflect the compartmentation of cAMP in smooth muscle cells (36), and the whole-cell cAMP measurement may not readily detect the localized cAMP changes.

Considering that cilostazol also inhibits PDE5 (although to a lesser extent than PDE3), which exists in vascular smooth muscle cells, we examined whether NO and cGMP were involved in the flow increase caused by cilostazol. Our data show that neither NO synthase inhibitor L-NNA nor NO-sensitive guanylyl cyclase inhibitor ODQ affected the cilostazol-induced flow increase. L-NNA and ODQ did not alter the effects of cilostazol on cardiac contractility. These data suggest that the NO pathway does not account for the differences observed in this study between cilostazol and milrinone.

Of the several metabolites of cilostazol in vivo, the plasma concentration of OPC-13015 reaches as much as 20% of the parent compound, but it inhibits PDE3 with about sevenfold higher potency (24). Thus OPC-13015 may account for a substantial amount of biologic activity from cilostazol; therefore we studied its effect on cardiac function in our isolated rabbit heart model. Figure 4 shows that the cardiac effect of OPC-13015 is similar to that of cilostazol: both have significantly less impact on LVDP and contractility than does milrinone. In our preliminary study, the cAMP-elevating abilities of OPC-13015 and cilostazol were similar in isolated rabbit ventricular myocytes (data not shown), consistent with our cardiac contractility data. The effects on coronary flows among these three compounds are not different. OPC-13015 has almost no effect on heart rate. These data show that, although OPC-13015 has higher inhibitory effect on PDE3, its effect on cardiac function is not greater than that of cilostazol. This further supports the idea that the pharmacodynamic effects of PDE3 inhibitors may not all be alike. Although the cardiovascular effects from milrinone parallel PDE3 inhibition, there are large differences for cilostazol between PDE3 inhibition and its effect on cardiac cells. Future studies are needed to explain such differences and may shed light on the efficacy of cilostazol in the treatment of patients with PAOD.

In summary, we have shown that cilostazol increases intracellular cAMP in platelets and inhibits platelet aggregation as potently as does milrinone. However, the increase of intracellular cAMP and related cardiac functional changes from cilostazol are much less compared with those of milrinone. The data presented herein are consistent with the beneficial and safe outcomes in the treatment of patients with intermittent claudication and indicate that the biologic activities of cilostazol cannot be explained by PDE3 inhibition alone.


1. Yasuda K, Sakuma M, Tanabe T. Hemodynamic effect of cilostazol on increasing peripheral blood flow in arteriosclerosis obliterans. Arzneimittelforschung 1985;35:1198-200.
2. Kamiya T, Sakaguchi S. Hemodynamic effects of the antithrombotic drug cilostazol in chronic arterial occlusion in the extremities. Arzneimittelforschung 1985;35:1201-3.
3. Ohashi S, Iwatani M, Hyakuna Y, et al. Thermographic evaluation of the hemodynamic effect of the antithrombotic drug cilostazol in peripheral arterial occlusion. Arzneimittelforschung 1985;35:1203-8.
4. Hiatt WR. Current and future drug therapies for claudication. Vasc Med 1997;2:257-62.
5. Regensteiner J, Stiener J, Hiatt WR. Exercise training improves functional status in patients with peripheral arterial disease. J Vasc Surg 1996;23:104-15.
6. Money SR, Herd JA, Isaacsohn JL, et al. Effect of cilostazol on walking distances in patients with intermittent claudication caused by peripheral vascular disease. J Vasc Surg 1998;27:267-74.
7. Dawson DL, Cutler BS, Meissner MH, et al. Cilostazol has beneficial effects in treatment of intermittent claudication: results from a multicenter, randomized, prospective, double-blind trial. Circulation 1998;98:678-86.
8. Elam MB, Heckman J, Crouse JR, et al. Effect of the novel antiplatelet agent cilostazol on plasma lipoproteins in patients with intermittent claudication. Arterioscler Thromb Vasc Biol 1998;18:1942-7.
9. Food and Drug Administration. FDA approval of cilostazol. Washington, DC: FDA, 1999.
10. Uehara S, Hirayama A. Effects of cilostazol on platelet function. Arzneimittelforschung 1989;39:1531-4.
11. Tani T, Sakurai K, Kimura Y, et al. Pharmacological manipulation of tissue cyclic AMP by inhibitors: effects of phosphodiesterase inhibitors on the functions of platelets and vascular endothelial cells. Adv Second Messenger Phosphoprotein Res 1992;25:215-27.
12. Sheth SB, Colman RW. Regulatory and catalytic domains of platelet cAMP phosphodiesterases: targets for drug design. Semin Hematol 1995;32:110-9.
13. Nagaoka T, Shirakawa T, Kasuya J, et al. Cyclic nucleotide PDE-3: quantitation of PDE-3A and -3B mRNAs in rat tissues by RNase protection assay. Cell Biochem Biophys 1998;29:49-66.
14. Movsesian MA, Komas N, Krall J, et al. Expression and activity of low Km, cGMP-inhibited cAMP phosphodiesterase in cardiac and skeletal muscle. Biochem Biophys Res Commun 1996;225:1058-62.
15. Manganiello VC, Murata T, Taira M, et al. Diversity in cyclic nucleotide phosphodiesterase isoenzyme families. Arch Biochem Biophys 1995;322:1-13.
16. Boger JE, DeLuca SL, Watkins DF, et al. Infusion therapy with milrinone in the home care setting for patients who have advanced heart failure. J Intraven Nurs 1997;20:148-54.
17. Mehra MR, Ventura HO, Kapoor C, et al. Safety and clinical utility of long-term intravenous milrinone in advanced heart failure. Am J Cardiol 1997;80:61-4.
18. Lynch JJJ, Uprichard AC, Frye JW, et al. Effects of the positive inotropic agents milrinone and pimobendan on the development of lethal ischemic arrhythmias in conscious dogs with recent myocardial infarction. J Cardiovasc Pharmacol 1989;14:585-97.
19. Packer M, Carver JR, Rodeheffer RJ, et al. Effect of oral milrinone on mortality in severe chronic heart failure: the PROMISE Study Research Group. N Engl J Med 1991;325:1468-75.
20. Packer M. Treatment of chronic heart failure. Lancet 1992;340:92-5.
21. Thadani U, Roden DM. FDA panel report: January 1998. Circulation 1998;97:2295-6.
22. Criqui MH, Denenberg JO, Langer RD, et al. The epidemiology of peripheral arterial disease: importance of identifying the population at risk. Vasc Med 1997;2:221-6.
23. Brevetti G, Martone VD, Perna S, et al. Intermittent claudication and risk of cardiovascular events. Angiology 1998;49:843-8.
24. Otsuka America Pharmaceutical, Inc. NDA 20-863: cilostazol. Rockville: Otsuka America Pharmaceutical, 1997.
25. Liu Y, Gao WD, O'Rourke B, et al. Synergistic modulation of ATP-sensitive K+ currents by protein kinase C and adenosine: implications for ischemic preconditioning. Circ Res 1996;78:443-54.
26. Yasunari K, Kohno M, Kano H, et al. Mechanisms of action of troglitazone in the prevention of high glucose-induced migration and proliferation of cultured coronary smooth muscle cells. Circ Res 1997;81:953-62.
27. Garthwaite J, Southam E, Boulton CL, et al. Potent and selective inhibition of nitric oxide-sensitive guanylyl cyclase by 1H-[1,2,4]oxadiazolo[4,3-a]quinoxalin-1-one. Mol Pharmacol 1995;48:184-8.
28. Laird NM, Ware JH. Random effects models for longitudinal data. Biometrics 1982;38:963-74.
29. Laird NM, Lange N, Stram D. Maximum likelihood computations with repeated measures: application of the EM algorithm. J Am Stat Assoc 1987;82:87-105.
30. Darvin H, King T. Lower extremity arterial occlusive disease. Clin Podiatr Med Surg 1992;9:69-77.
31. Manganiello VC, Taira M, Degerman E, et al. Type III cGMP-inhibited cyclic nucleotide phosphodiesterases (PDE3 gene family). Cell Signal 1995;7:445-55.
32. Inoue Y, Tochizawa S, Toga K, Tachibana K, Nagamoto M, Sudo T. Effects of OPC-33509 and references on activity of cyclic nucleotide phosphodiesterase type 1, 2, 3A, 3B, 4, 5, and 7 in vitro. Tokushima, Japan: Otsuka Pharmaceutical Co., Ltd. In-House Report 01152, 1997.
33. Kimura Y, Tani T, Kanbe T, et al. Effect of cilostazol on platelet aggregation and experimental thrombosis. Arzneimittelforschung 1985;35:1144-9.
34. Lindgren SH, Andersson TL, Vinge E, et al. Effects of isozyme-selective phosphodiesterase inhibitors on rat aorta and human platelets: smooth muscle tone, platelet aggregation and cAMP levels. Acta Physiol Scand 1990;140:209-19.
35. Pattison A, Astley N, Eason CT, et al. A comparison of the effects of three positive inotropic agents (amrinone, milrinone and medorinone) on platelet aggregation in human whole blood. Thromb Res 1990;57:909-18.
36. Polson JB, Strada SJ. Cyclic nucleotide phosphodiesterases and vascular smooth muscle. Annu Rev Pharmacol Toxicol 1996;36:403-27.
37. Ikeda U, Ikeda M, Kano S, et al. Effect of cilostazol, a cAMP phosphodiesterase inhibitor, on nitric oxide production by vascular smooth muscle cells. Eur J Pharmacol 1996;314:197-202.
38. Tanaka T, Ishikawa T, Hagiwara M, et al. Effects of cilostazol, a selective cAMP phosphodiesterase inhibitor, on the contraction of vascular smooth muscle. Pharmacology 1988;36:313-20.
39. Komas N, Movsesian M, Kedev S, et al. cGMP-inhibited phosphodiesterase (PDE3). In: Schudt C, Dent G, Rabe KF, eds. Phosphodiesterase inhibitors. San Diego: Academic Press, 1996:89-102.
40. Lindgren S, Andersson KE. Effects of selective phosphodiesterase inhibitors on isolated coronary, lung and renal arteries from man and rat. Acta Physiol Scand 1991;142:77-82.

Cilostazol; Milrinone; OPC-13015; Phosphodiesterase; Platelet; Heart; Ventricular myocytes

© 1999 Lippincott Williams & Wilkins, Inc.