Recent advance in antiplatelet therapy: The mechanisms, evidence and approach to the problems

Abstract

Antiplatelet therapy has been established as a preventive medicine for ischemic cardiovascular diseases both at acute and chronic phases. This therapy is also crucial for the prevention of thrombotic events after coronary stent implantation. So far, many lines of clinical evidence have demonstrated the beneficial effects of aspirin (an irreversible cyclooxygenase inhibitor) and thienopyridine derivatives (adenosine diphosphate (ADP)‐receptor P2Y₁₂ inhibitors). Recently, it has been reported that the cardiovascular risk is elevated in patients with platelets resistant to these drugs, compared to the good responders. One of the current problems to be solved in antiplatelet therapy is to find out patients resistant to the antiplatelet therapy and improve its preventive effects. In addition to aspirin and thienopyridines, several types of drugs with antiplatelet effects are currently available in clinical practice. Clinical evidence has recently been accumulating for these drugs that can be potential alternatives in patients with aspirin or thienopyridine resistance. In this review, the mechanisms, evidence and approach to the present problems of drugs with antiplatelet effects are discussed.

• Antiplatelet drugs
• aspirin
• aspirin resistance
• clopidogrel
• clopidogrel resistance
• platelet activation

Abbreviations
5‐HT	=	5‐hydroxytriptamine
ADP	=	adenosine diphosphate
ATP	=	adenosine triphosphate
cAMP	=	cyclic adenosine monophosphate
cGMP	=	cyclic guanosine monophosphate
COX	=	cyclooxygenase
DHA	=	docosahexaenoic acid
EPA	=	eicosapentaenoic acid
GP	=	glycoprotein
HMG‐CoA	=	3‐hydroxy‐3‐methylglutaryl coenzyme A
NSAID	=	non‐steroidal anti‐inflammatory drug
PAD	=	peripheral artery disease
PAR	=	protease‐activated receptor
PCI	=	percutaneous coronary intervention
PDE	=	phosphodiesterase
PDGF	=	platelet‐derived growth factor
PKA	=	cAMP‐dependent protein kinase
PKG	=	cGMP‐dependent protein kinase
PRP	=	platelet‐rich plasma
RR	=	relative risk
SNP	=	single nucleotide polymorphism
TF	=	tissue factor
VASP	=	vasodilator‐stimulated phosphoprotein
vWF	=	von Willebrand factor

Introduction: Molecular mechanism of platelet activation

Acute coronary syndrome and cerebral infarction are caused by thrombotic occlusion of arteries. When plaque rupture or erosion occurs in coronary arteries, subendothelial tissue is exposed to the lumen 1–4. Collagen in this tissue recruits von Willebrand factors (vWFs) from the blood stream, and collagen‐bound vWFs change its conformation in a high‐shear condition to become competent to interact with glycoprotein (GP) Ib/V/IX complex on the platelet surface (Figure 1) 3,4. This interaction causes ‘rolling’ of platelets on the vessel wall. Subsequently, platelets adhere tightly through collagen receptors such as integrin α₂β₁ (GPIa/IIa) and GPVI 3,4. There, platelets are activated and a series of reactions including granule secretion, aggregation and shape change occur, resulting in the formation of ‘white’ thrombus (Figure 1). In activated platelets, phosphatidyl serine emerges on the surface by a flip‐flop mechanism and provides a place for coagulation cascades (Figure 2). Cells, such as macrophages, in subendothelial tissues of atheromatous plaque express tissue factors on their surface. Further, circulating tissue factors 5 possibly present on monocyte‐derived microparticles gather at the growing thrombus through the interaction with P‐selectin on the plasma membrane of activated platelets, which is translocated by α‐granule secretion 3,4. These tissue factors trigger the coagulation cascade to generate thrombin at the last step. The functions of thrombin, a serine protease, are not only to generate fibrin by cleaving fibrinogen, but also to activate platelets through thrombin receptors. Among four protease‐activated receptor (PAR) family members, PAR‐1 is the major thrombin receptor in human platelets 6. Thus, platelets initially play a central role in arterial thrombus formation and, at later stages, collaborate with the coagulation system to form ‘red’ thrombus, resulting in occlusion of arteries (Figure 2).

Figure 1. Molecular mechanism of platelet activation. Activated platelets exert functions such as adhesion, granule secretion and aggregation as described in the text.

Figure 2. Synergistic interaction between platelet activation and coagulation system. The interaction of platelets and coagulation system following plaque rupture is schematically shown.

Platelets contain three types of vesicles including α‐granules, dense granules and lysosomes 3,4. They are secreted in response to increased intracellular calcium ion concentration in activated platelets (Figure 1) 7. α‐Granules contain protein factors such as platelet‐derived growth factor (PDGF), vWF and fibrinogen. P‐selectin and GPIIb/IIIa (integrin α_IIbβ₃) are membrane proteins stored on α‐granule membranes. These are translocated to the plasma membrane upon α‐granule exocytosis. Secreted hydrolytic enzymes stored in lysosomes would contribute to vascular remodeling. Dense granules contain adenosine diphosphate (ADP) and 5‐hydroxytryptamine (5‐HT or serotonin). Through receptors on platelets, ADP and 5‐HT enhance platelet activation in positive feedback manners. Activated platelets produce thromboxane A₂ through the arachidonic acid cascade. Thromboxane A₂ is also a self‐agonist. These agonists together with thrombin play roles in synergistic activation of platelets at the site of thrombus formation. Therefore, drugs that block one receptor pathway are used as antiplatelet drugs (Figure 3).

Figure 3. Molecular targets of antiplatelet drugs. Target molecules of the drugs with antiplatelet activity are shown as described in the text.

Platelet aggregation is mediated by GPIIb/IIIa that is abundantly expressed in platelets (Figure 1) 3,4. More than half GPIIb/IIIa molecules are present on the plasma membrane and some are in α‐granules in unstimulated platelets. Although GPIIb/IIIa has no ligand‐binding activity in unstimulated platelets, it undergoes a conformational change upon platelet activation to become able to bind its ligands, vWF and fibrinogen 3,4. Both ligands have multiple binding sites for activated GPIIb/IIIa and thereby induce platelet aggregation by bridging adjacent platelets. Activation of this integrin is considered the ‘common final pathway’ since all the signaling pathways utilize this molecule at the last step toward aggregation.

Antiplatelet therapy has been established as a preventive medicine for cardiovascular events. So far, many lines of clinical studies have demonstrated beneficial effects in the therapy with aspirin and thienopyridine derivatives. Several types of other drugs with antiplatelet effects are also available in clinical practice and evidence has been accumulating for such drugs. Recently, it has been noticed that the cardiovascular risk is elevated in patients with platelets poorly responsive to aspirin or thienopyridine, compared to the good responders, a phenomenon called antiplatelet drug resistance. This problem should be solved by establishing a strategy to find out which patients are resistant to the therapy and to improve its preventive effects. In this review, the mechanisms, evidence and present topics of antiplatelet drugs are discussed.

Key messages

• Antiplatelet therapy has been established as a preventive medicine for ischemic cardiovascular disease.

• In this review, concerning antiplatelet drugs, not only aspirin and thienopyridine derivatives but also other types of drugs, the molecular mechanisms, recent evidence, and current problems such as aspirin resistance are discussed for the better antiplatelet therapy.

Drugs with antiplatelet effects

Aspirin

Mechanism and evidence

In stimulated platelets, phospholipase A₂ is activated to generate arachidonic acid by cleaving phospholipids. Thromboxane A₂ is produced from arachidonic acid through a series of reactions including an initial step mediated by cyclooxygenase (COX) 3. Thromboxane A₂ activates platelets through its receptor (TP), a seven‐transmembrane receptor coupled to a heterotrimeric G‐protein, Gq 8.

Three isoforms of COXs are so far identified. COX‐1, expressing ubiquitously and constitutively, is the major COX in platelets 9–12 (Figure 3). COX‐2 is an inducible COX expressing in cells at the site of inflammation. COX‐2 is also expressed constitutively in several tissues such as endothelial cells 9–12. COX‐3 is a splicing variant of COX‐1, and sensitive to acetoaminophen with little effect on COX‐1 or COX‐2 11. Aspirin, acetylsalicylic acid, is an irreversible inhibitor of COX‐1 and COX‐3, but not COX‐2 at low concentrations. Therefore, the effect of aspirin on platelet function is due to inhibition of the initial step for thromboxane A₂ production.

Prostacyclin is produced mainly by the COX‐2 pathway in endothelial cells and inhibits platelet functions (see below) 12. The functions of thromboxane A₂ and prostacyclin are yin and yang in the regulation of platelet functions. Moreover, cells in vascular walls such as smooth muscle cells express both receptors. Recently, it has been demonstrated that the balance between thromboxane A₂ and prostacyclin also regulates atherogenesis since suppression of the thromboxane A₂ pathway inhibited atherosclerosis, whereas suppression of the prostacyclin pathway enhanced it in animal models 12. Therefore, aspirin would potentially suppress atherogenesis by decreasing the thromboxane A₂/prostacyclin ratio.

Aspirin is an antiplatelet drug most widely used in the world. A meta‐analysis by the Antithrombotic Trialist's Collaboration revealed that 22% of cardiovascular events including myocardial infarction, stroke, and vascular death were prevented by aspirin therapy in high‐risk patients 13. Aspirin therapy has, thus, become the standard for the secondary prevention of cardiovascular events in high‐risk patients, while it is controversial for the primary prevention in low risk patients.

Aspirin resistance

The efficacy of aspirin in the inhibition of platelet functions differs from patient to patient. Cardiovascular events preferentially occur in patients with low responses to aspirin therapy 14–16. This low response is referred to as ‘aspirin resistance’ 17–20. Aspirin therapy would bring more effects on the risk reduction if we could find out such patients and modify the therapy by dose‐up or changing aspirin into another drug. However, even the definition of aspirin resistance has not been determined, and the prevalence varies between 5%–60% from report to report depending on laboratory tests used and definitions 17.

With the word ‘aspirin resistance’, we should consider biochemical and clinical aspirin resistance 10. Biochemical aspirin resistance implies insufficient inhibition of platelet function by aspirin in laboratory tests, while clinical aspirin resistance indicates an increased risk of cardiovascular events in patients under aspirin treatment. Biochemical values in laboratory tests would acquire a strong clinical impact when their correlation with clinical outcome is established. Developing valid and comprehensive values indicating aspirin's efficacy, which predict a cardiovascular risk with a clear cut‐off point, is currently expected.

The light transmission (optical) aggregometer has been conventionally used for assessing aggregability of platelets in platelet‐rich plasma (PRP) 21. This method is the standard most widely used in this field. Gum et al. examined platelet aggregability by this method in 326 patients with stable cardiovascular diseases under 325 mg/day aspirin treatment and prospectively followed them for approximately 2 years 22. They observed aspirin resistance in 17 cases (5.2%) defined as a mean aggregation of ⩾70% with 10 µM ADP and ⩾20% with 0.5 mg/mL arachidonic acid and a 4.14‐times increase of cardiovascular risk in the aspirin resistant group 22.

On the other hand, the optical aggregometer requires preparation of PRP. This procedure may affect the aggregation. Another limitation of this ex vivo examination is evaluating aggregation of platelets alone. In other words, this method eliminates the effects of other blood cells that may affect thrombus formation in vivo. It has been shown that the presence of erythrocytes promotes platelet aggregation 23, and leukocytes produce platelet activating factors 24.

For a more physiological and simpler procedure, several systems to examine whole‐blood aggregation have been developed. Ultra Rapid Platelet Function assay (URPF)‐ASA (Accumetrics Inc., San Diego, California, USA) is a whole‐blood aggregometer, where whole blood is added to a cartridge containing fibrinogen‐coated beads, and the optical transmission is measured after removing the beads. Using this method, Chen et al. reported that percutaneous coronary intervention (PCI)‐related increase of cardiac troponin I and MB isoenzyme of creatine kinase were more frequently elevated in aspirin resistant patients 25.

Thromboxane A₂ is an unstable molecule and its stable metabolite, 11‐dehydro‐thromboxane B₂, is measured as a value reflecting thromboxane A₂ production 26. Therefore, it is expected that 11‐dehydro‐thromboxane B₂ level is reduced by aspirin intake, and that this value indicates the efficacy of aspirin although thromboxane A₂ is produced by other types of cells through the COX‐2 pathway 12,17. The Heart Outcomes Prevention Evaluation (HOPE) Study is a randomized, placebo‐controlled, 2×2 factorial trial of ramipril and vitamin E for the secondary prevention of cardiovascular diseases. Reevaluation of the HOPE Study patients treated with aspirin revealed that the highest quartile group of urine 11‐dehydro‐thromboxane B₂ level (>33.8 ng/urine mmol creatinine) had a 1.8‐times higher risk for myocardial infarction, stroke and cardiovascular death, compared with the lowest quartile (<15.1 ng/urine mmol creatinine) 26. On the other hand, urine 11‐dehydro‐thromboxane B₂ levels were widely distributed from individual to individual although they were reduced by aspirin intake in each individual. Urine 11‐dehydro‐thromboxane B₂ levels in healthy volunteers were reported to be 72.4–625.9 (179.5±142.0) ng/urine mmol creatinine before aspirin intake, and 12.9–118.0 (39.8±22.9) ng/urine mmol creatinine after aspirin intake 27. Further prospective studies would be required to establish the urine 11‐dehydro‐thromboxane B₂ level as a marker of aspirin efficacy linking to clinical outcome.

Mechanisms of aspirin resistance

Several mechanisms are considered as causes of aspirin resistance. The dose of aspirin might be insufficient in some patients with aspirin resistance. Platelet Function Analyzer (PFA)‐100 is a whole‐blood aggregometer (Dade Behring, Deerfield, Illinois, USA) that measures the ‘closure time’, a period before closing the special cartridges coated with platelet agonists (collagen plus either ADP or norepinephrine) by clot under a high‐shear conditions 28. Using this system, Gonzalez‐Conejero et al. reported that the ‘closure times’ in 29% of 24 subjects were not efficiently prolonged by a 100 mg/day aspirin intake, but completely inhibited by a 500 mg/day aspirin intake 27.

Drug interaction could be a cause of aspirin resistance. Non‐steroidal anti‐inflammatory drugs (NSAIDs) are commonly used for their analgesic, anti‐inflammatory, and antipyretic effects. They inhibit not only COX‐2 but also COX‐1. In a cross‐over study in healthy subjects, concomitant administration of ibuprofen was reported to antagonize platelet inhibition induced by aspirin, while that of rofecoxib (a COX‐2 inhibitor), acetoaminophen, or dicrofenac showed no effect 29. Since aspirin and NSAIDs bind the same site of COX‐1, the interaction of aspirin with COX‐1 would be competitively inhibited in the presence of other NSAIDs. Aspirin binds COX‐1 irreversibly while ibuprofen does reversibly. Since the lifespan of aspirin is very short, the COX‐1 activity would be recovered after the effect of ibuprofen is lost. This hypothesis is supported by the observation that ibuprofen antagonized aspirin's effect when ibuprofen was administered 2 hours before aspirin, but not 2 hours after aspirin 29. The reason why dicrofenac, another NSAID, did not affect aspirin's function is unclear 29.

The effect of NSAIDs on the aspirin therapy was examined in 7,107 patients with cardiovascular diseases 30. This study revealed that the hazard ratio of the risk of cardiovascular mortality was 1.73 (95% CI: 1.05–2.84) in the aspirin plus ibuprofen group compared to that in the aspirin alone group 30. Such increase of risk was not observed in the aspirin plus dicrofenac or aspirin plus other NSAID group 30.

Physicians' Health Study, a completed randomized trial of aspirin for the primary prevention of cardiovascular diseases, was reevaluated from the aspect of NSAID intake 31. In the aspirin‐treated group, the use of NSAIDs on more than 60 days/year was associated with the risk of myocardial infarction (relative risk (RR) 2.86; 95% CI: 1.25–6.56) compared with no use of NSAIDs. These data suggest that NSAIDs, especially ibuprofen, may reduce the benefit of aspirin for the prevention of cardiovascular diseases.

It has been reported that a 75‐mg daily aspirin intake causes platelet inhibition without attenuation during 2 years, measured by the optical aggregometer using 1 µg/mL collagen as a stimulus 32. However, a recent study showed an opposite finding that aspirin's effects progressively decreased during aspirin therapy for a long period 33. In this study, the maximal aggregation rates induced by 2 µg/mL collagen at 0, 2 and 24 months after aspirin intake were 88.2±21.8%, 37.9±24.4% and 61.9±23.9%, respectively, while the inhibition of 2 µM ADP‐induced platelet aggregation by ticlopidine did not decrease at 24 months 33. Time‐dependent decrease of aspirin's effects might be a cause of aspirin resistance. This mechanism is unknown, but it might be due to compensation by enhanced other signaling pathways in platelets. If this is the case, this phenomenon is clinically quite critical and we would have to modify aspirin therapy after 2‐year treatment. Further study is urgently required.

It has been reported that aspirin did not inhibit platelet aggregation or thromboxane A₂ production in vitro at Day 5 or Day 10 after coronary bypass surgery 24. COX‐2, a poor substrate of aspirin, was transiently induced in platelets at Day 5. However, expression of COX‐2 was attenuated to the baseline level at Day 10 when the aspirin resistance was still detected. Furthermore, in vitro addition of a selective COX‐2 inhibitor celecoxib did not affect arachidonic acid‐induced thromboxane A₂ formation or aggregation of PRP in aspirin‐resistant patients after bypass surgery 24. These results suggest that induced COX‐2 may contribute to the aspirin resistance, but it would not be the only factor responsible for the resistance immediately after bypass surgery.

Thromboxane A₂ might be produced by a ‘transcellular’ mechanism 26. Macrophages and vascular endothelial cells produce prostaglandin H₂ from arachidonic acid by COX‐2. This prostaglandin H₂ is released to outside of the cells and absorbed into platelets. Then, platelets could utilize prostaglandin H₂ as a source of thromboxane A₂ production. This type of production of thromboxane A₂ would bypass COX‐1.

Furthermore, other habitual and circumstantial conditions also affect the aspirin efficacy. Collagen‐induced platelet aggregability measured under aspirin treatment is positively correlated with serum total cholesterol levels 34. Norepinephrine‐induced platelet activation is only partly inhibited by aspirin 35. Cigarette smoking increased platelet aggregability in habitual smokers with coronary heart disease and this enhanced aggregability is not inhibited by aspirin 36. These reports suggest that high serum cholesterol, physical and mental stress and/or cigarette smoking may cause aspirin resistance in vivo, although their molecular mechanisms remain unclear.

There are more than 100 single nucleotide polymorphisms (SNP) in the thromboxane A₂ production pathway, which might be involved in aspirin resistance 37,38. SNPs in COX‐1, GPIa, GPIIIa, and FXIII have been shown to affect aspirin's effects on platelet functions 27,37,38, while their impact on the clinical outcome remains to be determined.

ADP receptor antagonists

ADP is stored in dense granules and secreted from activated platelets by regulated exocytosis (Figure 1). Platelets express three types of P2 purinergic receptors: metabotropic ADP receptors P2Y₁ and P2Y₁₂, and an ionotropic P2X₁ adenosine triphosphate (ATP) receptor involved in calcium ion influx 18,39,40. The P2Y₁ and P2Y₁₂ receptors are required for ADP‐induced full aggregation, while it remains unclear whether the P2X₁ receptor plays a role in platelet aggregation 18,39,40. The P2Y₁ receptor is associated with a heterotrimeric G‐protein, Gq, which induces phospholipase C activation to cause protein kinase C activation, an increase of intracellular calcium ion concentrations and transient aggregation 41. On the other hand, the P2Y₁₂ receptor is coupled to Gi which causes a decrease of the cyclic adenosine monophosphate (cAMP) level and activation of phosphoinositide 3‐kinase 40. The P2Y₁₂ signaling is crucial for stabilizing thrombus under high‐shear condition demonstrated by in vitro flow models 42,43. Furthermore, in vivo observation of experimental thrombus formation by a real‐time intravital microscopy revealed that blockade of the P2Y₁₂ receptor reduced the initial size of thrombus and frequency of embolization 44.

Irreversible P2Y₁₂ antagonists, ticlopidine and clopidogrel, are prodrugs that are activated through oxidation by liver enzyme cytochrome P450 10,18. In several trials, ticlopidine has been demonstrated to reduce cardiovascular events 10, 45. In the CAPRIE (Clopidogrel versus Aspirin in Patients at Risk of Ischemic Events) Study, patients with vascular diseases were followed under treatment with either clopidogrel (75 mg/day) or aspirin (325 mg/day) for 1–3 years (mean 1.9 years) 46. Cardiovascular events occurred in 5.3% in the clopidogrel group and in 5.8% in the aspirin group (P = 0.043) 46. This study revealed superior effects of clopidogrel to those of aspirin although the difference was not large. The CURE (Clopidogrel in Unstable angina to prevent Recurrent Event trial) Study demonstrated that treatment with aspirin plus clopidogrel reduced cardiovascular events by 20% compared with aspirin alone 47. Thus, P2Y₁₂ blockers have similar, or even stronger, preventive effects for ischemic cardiovascular events compared to aspirin. During treatment with these drugs, especially with ticlopidine, attention should be paid to adverse effects such as liver dysfunction, neutropenia and thrombocytopenic purpura caused by autoimmunity against a vWF‐cleaving enzyme, ADAMTS13 48.

Clopidogrel resistance has also been reported. Sixty patients with acute myocardial infarction were treated with clopidogrel and prospectively followed for 6 months 49. Patients were stratified into four groups according to relative reduction of the ADP‐induced maximal aggregation rates at Day 6 compared with those at Day 0 measured by the optical aggregometer. Vascular events occurred in 6 patients out of 15 (40%) in the group with the lowest response to clopidogrel, while in only 1 out of 45 (2.2%) in the other group 49. Thus, vascular events preferentially occurred in the clopidogrel‐resistant group.

Furthermore, clopidogrel is a prodrug activated by cytochrome P450 including CYP3A4 in the liver. The inhibitory effects of clopidogrel on platelet aggregation are correlated with activity of CYP3A4 50. Therefore, effects of clopidogrel are inhibited by erythromycin and atorvastatin that are also metabolized by CYP3A4, and enhanced by CYP3A4 inducers such as rifampicin 50. Polymorphisms of CYP3A4 and the P2Y₁₂ ADP‐receptor could affect the effect of clopidogrel to inhibit platelet aggregation and might be involved in the clopidogrel resistance.

Drugs that increase intracellular cAMP and cyclic guanosine monophosphate (cGMP)

Prostacyclin, mainly produced by endothelial cells, inhibits platelet aggregation and secretion by elevating intracellular cAMP levels through its specific Gs‐coupling receptor, IP 51. Prostacyclin also has a vasodilatory effect. It reduces an ischemic pain in peripheral artery disease (PAD) patients and is used for the treatment of pulmonary hypertension 52.

Since the half‐life of prostacyclin is only 3 minutes 53, an orally‐mediated prostacyclin‐analogue, beraprost, has been developed as a stable compound whose half‐life is approximately 1 hour 54. It inhibits platelet aggregation for several hours after oral intake 54. A placebo‐controlled randomized trial enrolling patients with PAD revealed that beraprost improved PAD‐related symptoms such as pain‐free walking distance 55, although this improvement was not observed in another study 56. However, a meta‐analysis of these trials has shown that beraprost significantly reduced combined endpoints including cardiovascular events and critical events in lower‐limb, such as deterioration of the ischemia and amputation, by 39% 57. Other prostacyclin analogues, iloprost and remodulin, have been developed as injection drugs.

Nitric oxide stimulates guanylate cyclase and increases cyclic guanosine monophosphate (cGMP) levels in platelets, causing inhibition of platelet functions 58. Effects of cAMP and cGMP are mainly mediated by cAMP‐dependent protein kinase (PKA) and cGMP‐dependent protein kinase (PKG), respectively. Both kinases phosphorylate many proteins in platelets and one of the common PKA and PKG substrates involved in the regulation of platelet aggregation is vasodilator‐stimulated phosphoprotein (VASP), a regulator of actin‐cytoskeletal reorganization 59. Recently, the measurement of phosphorylation levels of VASP is under development to monitor the efficacy of antiplatelet drugs affecting PKA and/or PKG 59.

Levels of cAMP and cGMP are also regulated by their degrading enzymes cyclic‐nucleotide phosphodiesterases (PDE) 58. Major PDEs in platelets are PDE2, PDE3A and PDE5. While PDE5 is specific for cGMP, PDE2 and PDE3A degrade both cAMP and cGMP. There is a feedback regulation among some PDEs. For example, cGMP stimulates the PDE2 activity and inhibits the PDE3A activity 58. Importantly, these PDEs are also expressed in cardiovascular systems such as vascular smooth muscle cells and cardiomyocytes. For this reason, PDE‐inhibitors have effects on these cells in addition to platelets.

Cilostazol is an inhibitor of PDE3A, a major cAMP‐degrading enzyme expressed in platelets 60,61. Since cilostazol improves symptoms and walking distance in PAD patients 62,63, it is recommended for the treatment of PAD 64. Recently, cilostazol has been demonstrated to inhibit recurrent ischemic cerebrovascular events by 42% in a prospective randomized trial performed in Japan 65. It should be noted, however, that cilostazol increases cAMP levels in cardiomyocytes, causing side effects such as tachycardia in some patients.

Dipyridamole is an inhibitor of PDE5 to increase cGMP levels in platelets and inhibits their aggregation 66. European Stroke Prevention Study 2 (ESPS‐2) is a randomized placebo‐controlled double‐blind trial enrolling 6,602 patients with prior cerebral infarction, analyzing the effect of dipyridamole on the secondary prevention of ischemic stroke 67. It has demonstrated that dipyridamole alone reduced the stroke risk compared to placebo by 16%, aspirin alone by 18%, and dipyridamole plus aspirin by 36% 67, indicating that the efficient risk reduction was obtained by dipyridamole. However, a Cochrane review analyzing 19,842 patients in 26 studies including ESPS‐2 revealed that dipyridamole did not prevent vascular death (RR: 1.02; 95% CI: 0.90–1.17) and that it prevented vascular events (RR: 0.90; 95% CI: 0.83–0.98) although it only reached statistical significance due to the ESPS‐2 results 68. Further trials would be required for establishing a preventive effect of dipyridamole.

Omega‐3 fatty acids

It is well known that fish oil contains omega‐3 fatty acids such as eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA). The GISSI‐Prevention Study has demonstrated that treatment with omega‐3 fatty acids reduced cardiovascular events (death, non‐fatal myocardial infarction and non‐fatal stroke) by 15% in patients with prior coronary heart disease 69. The Japan EPA Lipid Intervention Study (JELIS) was performed as a prospective open‐labeled randomized trial where approximately 18,000 hypercholesterolemic patients treated with HMG‐CoA reductase inhibitors were followed with or without EPA (1,800 mg/day) for 5 years 70. This study revealed that EPA reduced major coronary events by 19% 70, indicating that EPA exerted additional preventive effects for patients treated with cholesterol‐lowering therapy.

Omega‐3 fatty acids not only decrease triglyceride levels but also inhibit platelet functions 71. EPA is incorporated into phospholipids, components of cellular membrane, by replacing arachidonic acid. Non‐functional thromboxane A₃ is generated from EPA in platelets and functional prostacyclin (PGI₃) is produced in endothelial cells 72. Therefore, EPA decreases the functional thromboxane A₂/prostacyclin ratio, resulting in an inhibition of platelet functions. This effect might contribute to the risk reduction by omega‐3 fatty acids.

Serotonin receptor antagonists

Serotonin (5‐HT) is synthesized in neuronal cells in brain and enterochromaffin cells in the gastrointestinal tract 73. Platelets take in enteric cell‐released serotonin through serotonin transporters and store it in dense granules. Serotonin is released from activated platelets at the site of bleeding and vascular damage such as plaque rupture, while neuronal serotonin is stored in specific types of neurons and used as a neurotransmitter.

Serotonin receptors are divided into 7 distinct classes (5‐HT₁ to 5‐HT₇) and 14 subfamilies. Serotonin secreted from activated platelets plays a role as a self‐agonist to amplify platelet activation and as a vasoconstrictor for vascular smooth muscle cells via the Gq‐coupling 5‐HT_2A receptor 74. As selective 5‐HT_2A receptor antagonists, naftidrofuryl 75,76 and sarpogrelate 77 have been developed. It has been demonstrated in randomized placebo‐controlled studies 75,76 that naftidrofuryl improved clinical symptoms such as walking distance in patients with PAD. Therefore, it is recommended for the treatment of PAD 64.

GPIIb/IIIa antagonists

Abciximab, a Fab fragment of a monoclonal antibody recognizing activated integrin GPIIb/IIIa, is used as an adjunctive therapy for percutaneous coronary intervention (PCI) 78. Non‐antibody type inhibitors against activated GPIIb/IIIa, eptifibatide and tirofiban, have also been developed as injection drugs and their usefulness has been demonstrated in patients undergoing PCI 79,80. Oral GPIIb/IIIa inhibitors were examined by several large‐scale clinical trials. However, the meta‐analysis of these trials revealed that these drugs brought no benefit for reducing coronary events, but increased adverse bleeding events 81,82. Their narrow therapeutic range of concentration and partial agonist effects of these drugs might contribute to the results 81,83.

Conclusions

Antiplatelet therapy has been established to prevent cardiovascular events in high‐risk patients. It is well known, however, that efficacies of antiplatelet drugs vary from patient to patient. Furthermore, the cardiovascular risk is elevated in patients with platelets resistant to these drugs compared to the good responders, even though the definition of the drug resistance is not yet fixed. Platelet function is not usually monitored in ordinary clinical practice. We need to develop a simple and reliable laboratory test for monitoring the effect of antiplatelet therapy, like blood pressure in antihypertensive therapy and serum cholesterol concentration in cholesterol lowering therapy. A diagnostic value of laboratory examination acquires valid clinical significance when it is associated with clinical outcome. Further prospective clinical trials are required to establish such a value in this field. Several drugs with antiplatelet effects are currently available in addition to aspirin and thienopyridine derivatives, and evidence has been accumulating for these drugs. For patients with insufficient response to an antiplatelet drug, we could improve the therapy by increasing the dose or changing the antiplatelet drug into others.

Acknowledgements

I am grateful to Dr Susana Prat, Barcelona, a visiting scientist in our Department; Dr Yasuyuki Fujita, London; Dr Arata Tabuchi, Berlin; Dr Hidenori Arai, Kyoto; and members of my laboratory for critical reading and discussion of this manuscript. I also thank Professor Dr Toru Kita for continuous encouragement of my research projects. This work was supported by Ministry of Education, Culture, Sports, Science, and Technology Research Grants No. 16209031, by Health and Labour Sciences Research Grant for Cardiovascular Research H17‐006 from the Ministry of Health Labour and Welfare, Japan and in part by grants from the Takeda Science Foundation, Japan Cardiovascular Research Foundation, and a grant from The Kato Memorial Trust for Nambyo Research.