Reviews/Commentaries/Position Statements REVIEW ARTICLE

The Platelet in Focus on prevention of ischemic events

1 JOHN A. COLWELL, MD, PHD tive to other cardiovascular (CV) risk 2,3 RICHARD W. NESTO, MD factors, such as hypertension, albumin- uria, obesity, cigarette smoking, and dys- lipidemia, relative to nondiabetic patients with these comorbidities (4,5). Type 2 di- abetes is associated with resistance Accelerated atherosclerosis and the increased risk of thrombotic vascular events in diabetes may result and hyperinsulinemia and is often part of from dyslipidemia, endothelial dysfunction, platelet hyperreactivity, an impaired fibrinolytic balance, a metabolic syndrome called “Syndrome and abnormal blood flow. There is also a correlation between and cardiovascular (CV) events. The importance of platelets in the atherothrombotic process has led to investigation of using X,” which comprises hypertension, dys- antiplatelet agents to reduce CV risk. A meta-analysis conducted by the Antiplatelet Trialists’ Collab- lipidemia, decreased fibrinolysis, and in- oration demonstrated that aspirin reduced the risk of ischemic vascular events as a secondary pre- creased procoagulation factors (6). This vention strategy in numerous high-risk groups, including patients with diabetes. Based on results metabolic syndrome may also be seen in from placebo-controlled randomized trials, the American Diabetes Association recommends low- obese insulin-resistant individuals who dose enteric-coated aspirin as a primary prevention strategy for people with diabetes at high risk for do not yet have overt diabetes (7,8). CV events. Clopidogrel is recommended if aspirin allergy is present. There is occasionally a need for In addition to heightened risk for an alternative to aspirin or for additive antiplatelet therapy. Aspirin in low doses inhibits thromboxane CVD, patients with diabetes have a poorer production by platelets but has little to no effect on other sites of platelet reactivity. Agents such as prognosis than nondiabetic patients ticlopidine and clopidogrel inhibit ADP-induced platelet activation, whereas the platelet glycoprotein (Gp) IIb/IIIa complex receptor antagonists block activity at the fibrinogen binding site on the platelet. when they experience a major ischemic These agents appear to be useful in acute coronary syndromes (ACSs) in diabetic and nondiabetic vascular event. Diabetic patients who suf- patients. A combination of clopidogrel plus aspirin was more effective than placebo plus standard fer an acute myocardial infarction (MI) therapy (including aspirin) in reducing a composite CV outcome in patients with unstable angina and have a more complicated hospital course non–ST segment elevation myocardial infarction. In a meta-analysis of six trials in diabetic patients and a higher incidence of recurrent vas- with ACSs, intravenous GpIIb-IIIa inhibitors reduced 30-day mortality when compared with control cular events than nondiabetic patients subjects. Results from controlled prospective clinical trials justify the use of enteric-coated low-dose (9,10). For example, the 7-year incidence aspirin (81–325 mg) as a primary or secondary prevention strategy in adult diabetic individuals (aged of recurrent MI in a large population- Ͼ30 years) at high risk for CV events. Recent studies support the use of clopidogrel in addition to standard therapy, as well as the use of GpIIb-IIIa inhibitors in ACS patients. based study was 45% in diabetic patients versus 19% in nondiabetic patients (11). Diabetes Care 26:2181–2188, 2003 Moreover, the prognosis for diabetic pa- tients who undergo coronary revascular- ization procedures is worse than that for ardiovascular disease (CVD) is the coronary heart disease (CHD), stroke, nondiabetic subjects; patients with diabe- leading cause of disability and pre- and peripheral arterial disease (PAD) tes experience more postprocedural com- mature mortality in patients with di- from twofold to fourfold (2,3). The in- plications and have decreased infarct-free C survival (12,13). PAD is present in 8% of abetes (1). Diabetes increases the risk for creased risk is independent of and addi- patients at the time of diabetes diagnosis ●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●● but in 45% of patients who have had di- From the 1Diabetes Center, Medical University of South Carolina, Charleston, South Carolina; the 2Depart- abetes for 20 years (14). The presence of 3 ment of Cardiovascular Medicine, Lahey Clinic Medical Center, Burlington, Massachusetts; and Harvard symptomatic PAD is a marker for sys- Medical School, Boston, Massachusetts. Address correspondence and reprint requests to Richard W. Nesto, MD, Department of Cardiovascular temic atherosclerotic disease, including Medicine, Lahey Clinic Medical Center, 41 Mall Rd., Burlington, MA 01805. E-mail: richard.w.nesto@lahey. ischemic coronary and cerebrovascular org. events (8,15). Received for publication 9 December 2002 and accepted in revised form 2 April 2003. Both atherosclerosis and thrombosis J.A.C. has served on a clinical trial committee for Pfizer, has served on an advisory board for Takeda Pharmaceuticals America, and has received honoraria for speaking engagements from Bayer and Pfizer. appear to contribute significantly to the Abbreviations: ACC/AHA, American College of Cardiology/American Heart Association; ACS, acute increased CV risk of diabetic patients (7). coronary syndrome; ADA, American Diabetes Association; ATC, Antithrombotic Trialists’ Collaboration; The majority of ischemic coronary and ce- CAPRIE, Clopidogrel versus Aspirin in Patients at Risk of Ischemic Events; CHD, coronary heart disease; rebrovascular events are precipitated by CURE, Clopidogrel in Unstable Angina to Prevent Recurrent Events; CV, cardiovascular; CVD, cardiovas- cular disease; ETDRS, Early Treatment of Study; Gp, glycoprotein; HOPE, Heart vessel occlusion caused by atherosclerotic Outcomes Prevention Evaluation; HOT, Hypertension Optimal Treatment; MI, myocardial infarction; NO, plaque disruption, platelet aggregation, nitric oxide; PAD, peripheral arterial disease; RR, relative risk; TASS, Ticlopidine Aspirin Stroke Study; TTP, platelet adhesion, and resulting intravas- thrombotic thrombocytopenic purpura; TXA2, thromboxane A2; USPHS, U.S. Physicians’ Health Study; cular thrombosis. Several systems that vWF, von Willebrand factor. A table elsewhere in this issue shows conventional and Syste`me International (SI) units and conversion maintain the integrity and patency of the factors for many substances. vasculature are impaired in diabetes, in- © 2003 by the American Diabetes Association. cluding platelet and endothelial function,

DIABETES CARE, VOLUME 26, NUMBER 7, JULY 2003 2181 The platelet in diabetes coagulation, and fibrinolysis (16,17). shear stress imposed by blood flow and platelet aggregability and adhesiveness Thus, the balance in normal hemostasis is enable the platelet to remain attached to (16,22,27–33) are due to the following: shifted to favor thrombosis, increasing CV the vascular surface at the tissue injury risk. This review focuses on the role of site. Furthermore, these interactions ● Reduced membrane fluidity ϩ ϩ platelets in promoting thrombosis and ex- cause activated platelets to degranulate ● Altered Ca2 and Mg2 homeostasis 2ϩ plores how antiplatelet therapies can re- and release ADP and TXA2. Once re- (increased intracellular Ca mobiliza- 2ϩ duce the risk of ischemic events in leased, ADP and TXA2 bind to other spe- tion and decreased intracellular Mg ) diabetic patients. cific receptors, and those functions amplify ● Increased arachidonic acid metabolism ● the platelet aggregation process. In pa- Increased TXA2 synthesis PLATELET ACTIVATION tients with diabetes, platelet function is ● Decreased prostacyclin production AND THE altered in several ways, including an in- ● Decreased NO production ● ATHEROTHROMBOTIC creased release of TXA2 (22), accelerated Decreased antioxidant levels EVENT — Platelet activation is initi- platelet turnover (23,24), and an increase ● Increased expression of activation- ated by the binding of thrombogenic sub- in platelet aggregation (25). Platelet acti- dependent adhesion molecules (e.g., stances, such as collagen, thrombin, or vation and aggregation can be interrupted GpIIb-IIIa, P-selectin) components of atheromatous plaque, to at several of the steps detailed above (21). receptors located on the platelet surface. The formation of a platelet thrombus Platelets from patients with type 1 Receptor binding triggers a series of is accompanied by activation of the plasma and exhibit enhanced events that include hydrolysis of mem- coagulation cascade, leading to produc- platelet aggregation activity early in the brane phospholipids, mobilization of in- tion of thrombin. Thrombin stimulates disease course that may precede the de- tracellular calcium, and phosphorylation platelet aggregation, but more importantly, velopment of CVD (22,25,34). Numer- of important intracellular proteins (18). it mediates the conversion of fibrinogen ous biochemical abnormalities have been Notably, arachidonic acid is released from into fibrin, which polymerizes into an in- found that correlate with platelet hyper- membrane phospholipids and is con- soluble cross-linked gel that forms the reactivity. Platelets from diabetic patients verted into thromboxane A2 (TXA2), matrix for the thrombus, resulting in a exhibit reduced membrane fluidity, which produces vasoconstriction and fur- stable clot. (The processes of platelet ac- which may reflect changes in the lipid ther augments the platelet activation tivation, secretion, and aggregation lead- composition of the membrane or glyca- process. ing to intravascular thrombus formation tion of membrane proteins (16). Arachi- Calcium release and phosphorylation occur as a result of atherosclerotic plaque donic acid metabolism is increased in are critical for regulating the secretion of rupture.) platelets from diabetic patients; this leads ADP and several proteins—including von After plaque rupture, blood from the to enhanced TXA2 production and may Willebrand factor (vWF), fibronectin, lumen of the vessel comes into contact contribute to increased platelet sensitivity plasminogen activator inhibitor 1, throm- with the lipid-rich core of the plaque, (22,27). bospondin, platelet-derived growth fac- where it interacts with the highly pro- An increase in calcium mobilization tor, ␤-thromboglobulin, and platelet thrombotic substances collagen and tis- from intracellular storage pools, resulting factor 4—from platelet granules. ADP, in sue factor (26). Thrombus formation in increased intracellular calcium levels, turn, binds to platelet purinergic recep- leads to vessel obstruction. The thrombus has been correlated with the reduction in tors and triggers a conformational change may expand and result in total vessel oc- membrane fluidity (28). In addition to al- in glycoproteins IIb and IIIa on the plate- clusion, depending on the balance be- terations in platelet calcium homeostasis, let surface (19). This causes platelets to tween factors that promote and those that intracellular magnesium concentrations form a functional heterodimeric platelet oppose thrombosis. Factors that enhance are reduced, consistent with an increase glycoprotein (Gp) IIb-IIIa receptor com- thrombus progression include large in platelet hyperaggregability and adhe- plex that binds to fibrinogen and links plaque ruptures, lipid extrusion from the siveness (35). Magnesium supplementa- adjacent platelets as part of the platelet core, high-grade stenosis, low blood flow, tion can reduce these abnormal platelet aggregation process. high levels of fibrinogen, high levels of functions in people with diabetes (35). Prostacyclin and nitric oxide (NO), plasminogen activator inhibitor 1 (the Platelets from diabetic subjects pro- produced by normal endothelium, inhibit major inhibitor of factors that promote fi- duce less NO and prostacyclin, which platelet activation and relax vascular brinolysis, such as tissue plasminogen ac- normally inhibit platelet-endothelium in- smooth muscle to promote normal blood tivator), and reactive platelets. Factors teractions and promote endothelium- flow. People with diabetes have reduced that limit thrombus progression include mediated vasodilation. The concentration release of prostacyclin and NO (20). At high local blood flow, small and rapidly of NO synthase in platelets from patients sites of tissue injury, platelets adhere to healing plaque ruptures, and high levels with type 1 and type 2 diabetes is less than the subendothelium (21). In regions of of fibrinolysis. half that measured in platelets from non- low blood flow, subendothelial collagen diabetic individuals (29); however, insu- fibrils bind to the platelet collagen GpIa- ALTERED PLATELET lin will stimulate NO synthesis in platelets IIa receptor. This interaction is stabilized FUNCTIONS IN DIABETES — (30). Moreover, platelets from diabetic in regions of higher blood flow by vWF, Diabetic thrombocytopathy refers to dif- subjects contain reduced antioxidant lev- which links the collagen fibril to the plate- ferences in platelet function between dia- els, which tend to be associated with in- let GpIb-IX receptor. These endothelium- betic and nondiabetic individuals. creased aggregability and low platelet platelet interactions are resistant to the Among diabetic individuals, increased vitamin C levels (31). Platelet dysfunction

2182 DIABETES CARE, VOLUME 26, NUMBER 7, JULY 2003 Colwell and Nesto

Table 1—Aspirin as a primary prevention strategy in diabetes: evidence from collaborative trials

Patients with events Number of Years of Aspirin dose Trial diabetic subjects study CV end points (mg) Aspirin Placebo RR P USPHS (42) 533 5 MI 325 q.o.d. 4.0% 10.1% 0.39 NR ETDRS (43) 3,711 5 MI 650 q.o.d. 9.1% 12.3% 0.72 0.038 HOT (44) 1,501 3.8 MI 75 q.d. 2.3* 3.6* 0.64 0.002 Major CV events 8.9* 10.5* 0.85 0.03 *Events per 1,000 patient-years. NR, not reported. can be reversed by the addition of vitamin ● Microalbuminuria or macroalbumin- individuals, 1,501 of whom had diabetes E (36–38). uria (44). Subjects were randomized to either Patients with type 1 and type 2 diabe- ● Total cholesterol Ͼ200 mg/dl (LDL low-dose aspirin (75 mg/day) or placebo tes have increased populations of platelets cholesterol Ͼ100, HDL cholesterol therapy. Less than 10% of the participants that express activation-dependent adhe- Ͻ55 in women and Ͻ45 in men, and had clinical evidence for previous MI, sion molecules, such as activated GpIIb- triglycerides Ͼ200) stroke, or other CHD. The aspirin study, IIIa, lysosomal Gp53, thrombospondin, therefore, may be viewed as a primary and P-selectin (32). The increased expres- This body of evidence is further sup- prevention trial in high-risk individuals. sion of GpIIb-IIIa is consistent with the ported by the results of the recently com- Aspirin therapy resulted in an additional enhanced fibrinogen binding and aggre- pleted Primary Prevention Project in 15% reduction in the risk for CV events gability seen in platelets from diabetic which low-dose aspirin (100 mg/day) was over that seen with antihypertensive ther- subjects (33). Further, serum fibrinogen evaluated for the prevention of CV events apy (P ϭ 0.03). Fatal bleeding, including levels are also elevated in many patients in individuals with one or more of the cerebral, was equally common in the as- with type 1 or type 2 diabetes. In addition following: hypertension, hypercholester- pirin and placebo groups, whereas nonfa- to reflecting increased aggregability, the olemia, diabetes, obesity, family history tal bleeding was more common with enhanced surface expression of these ad- of premature MI, or being elderly (n ϭ aspirin therapy. Thus, the USPHS, the hesion molecules suggests that platelets 4,495). After a mean follow-up of 3.6 ETDRS, and the HOT trial support the can also communicate with leukocytes years, aspirin was found to significantly ADA’s position that low-dose aspirin ther- and possibly play a role in inflammation- lower the frequency of CV death (from apy is indicated in diabetic individuals mediated tissue damage in the vascula- 1.4 to 0.8%; relative risk [RR] 0.56 [CI who are at high risk for CV events. ture. Finally, platelets may interact with 0.31–0.99]) and total CV events (from plasma constituents, such as glycosylated 8.2 to 6.3%; RR 0.77 [0.62–0.95]) (41). Aspirin dosage in the primary LDLs, immune complexes, or vWF, to in- In addition to the Primary Prevention prevention strategy crease platelet aggregability or adhesion Project, there are three pertinent large- The ADA recommends that a dosage of (16,17). scale trials: the U.S. Physicians’ Health 81–325 mg of enteric-coated aspirin be Although antiplatelet agents, includ- Study (USPHS) (42), the Early Treatment used as a preventive strategy in high-risk ing aspirin and clopidogrel, irreversibly of Diabetic Retinopathy Study (ETDRS) diabetic individuals. The American Heart inactivate platelet function (39) for the 7- (43), and the Hypertension Optimal Association has recently issued similar to 10-day average life span of the platelet Treatment (HOT) study (44) (Table 1). guidelines (45). The American Heart As- in normal subjects, diabetic patients with The USPHS was a 5-year primary preven- sociation recommends 75–160 mg/day of vascular disease may have a greater rate of tion trial in 22,701 healthy men that in- aspirin as a primary prevention strategy in platelet turnover. cluded 533 men with diabetes. Among high-risk individuals, defined as those diabetic men, 4.0% of those treated with with a 10-year risk of CHD Ն10%. A sim- ASPIRIN AS A PRIMARY 325 mg aspirin every other day had an MI ilar view has been put forth by the U.S. PREVENTION STRATEGY IN versus 10.1% of those who received pla- Preventive Services Task Force (46). DIABETES — Based on collaborative cebo (RR 0.39). In the ETDRS, although trial data, the American Diabetes Associ- aspirin did not prevent progression of ret- ASPIRIN AS A SECONDARY ation (ADA) recommends that enteric- inopathy, it did produce a significant re- PREVENTION STRATEGY IN coated aspirin be used as a primary duction in risk for MI (28%) over 5 years DIABETES — The ADA recommends prevention strategy in patients with dia- (P ϭ 0.038) without an excess of retinal the use of aspirin (81–325 mg/day) as a betes who are classified as being at high or vitreous hemorrhage. This study may secondary prevention measure in diabetic risk for CV events on the basis of the fol- be viewed as a mixed primary and sec- patients with large vessel disease (history lowing risk factors (39–40): ondary prevention trial because 6% of of MI, vascular bypass procedure, stroke those enrolled had a history of MI and or transient ischemic attack, peripheral ● Family history of CHD Ͻ50% had elevated blood pressure or a vascular disease, claudication, and/or an- ● Cigarette smoking history of CVD. gina) (40). Two large meta-analyses of ● Hypertension The HOT trial studied antihyperten- major secondary prevention trials by the ● Weight Ͼ120% of ideal body weight sive treatment in 18,790 hypertensive Antithrombotic Trialists’ Collaboration

DIABETES CARE, VOLUME 26, NUMBER 7, JULY 2003 2183 The platelet in diabetes

(ATC) have concluded that aspirin (or an- other oral antiplatelet drug) is protective in most patients who are at high risk for CVD, including those with diabetes (47,48). The ATC meta-analysis of 287 second- ary prevention trials involved 212,000 high-risk patients who had acute or pre- vious vascular disease or another condi- tion that increased the risk of vascular occlusion (48). Aspirin was the most fre- quently used agent, with doses ranging from 75 to 325 mg/day. A low dose of aspirin (75–150 mg/day) was found to be at least as effective as higher daily doses, although it was noted that an initial Figure 1—Mechanism of action of antiplatelet agents. COX, cyclo-oxygenase. From Schafer (19). higher loading dose of at least 150 mg might be needed in acute settings. In the synthesis by endothelial cells and plate- ACE inhibitor therapy had a lower mor- main high-risk groups—acute MI, past lets, which might be expected to favor tality rate than those who were not treated history of MI, past history of stroke or thrombus formation. For this reason, with aspirin (51). To date, all these anal- transient ischemic attack, acute stroke, low-dose aspirin is favored for cardiopro- yses are observational. Prospective ran- and other relevant history of vascular dis- tection. This may be important in patients domized trials are needed for definitive ease—antiplatelet therapy significantly with diabetes, whose levels of prostacy- evidence. The current consensus is that reduced the incidence of vascular events clin are already reduced. Further, during low doses of aspirin can safely be used by 23%. In an earlier meta-analysis, the treatment with low-dose aspirin, platelets along with ACE inhibitors in patients with ATC reported that vascular event reduc- from diabetic patients released greater CHD, even in the presence of coronary tions remained significant in subgroup amounts of TXA2 than platelets from non- heart failure. analyses of middle- and older-aged, male diabetic individuals (24). The physiolog- and female, hypertensive and nonhyper- ical importance of this is unknown, and ADP receptor antagonists tensive, and diabetic and nondiabetic pa- many prospective clinical trials support Ticlopidine and clopidogrel are thienopy- tients (47). Low doses of aspirin were as the use of aspirin in people with diabetes ridine antiplatelet agents that inhibit the effective as high doses, but bleeding com- (14,24,39,41–48). binding of ADP to the platelet type 2 pu- plications were reduced at the lower dos- An important consideration affecting rinergic receptor (Fig. 1), an effect that age levels (47). the use of aspirin in patients with diabetes lasts for the lifetime of the platelet (7–10 In the Ͼ4,500 patients with diabetes is the interaction of aspirin with ACE in- days). By blocking the purinergic recep- studied in the ATC, the incidence of vas- hibitors. In the wake of the Heart Out- tor, these agents prevent the activation of cular events was reduced from 23.5% comes Prevention Evaluation (HOPE) the GpIIb-IIIa receptor and the subse- with control treatment to 19.3% with an- study, ACE inhibitors are emerging as a quent binding of fibrinogen. Clopidogrel tiplatelet therapy (P Ͻ 0.01) (47). By standard of care for many patients with and ticlopidine inhibit ADP-induced comparison, the lower incidence of vas- atherosclerosis and diabetes. In the 3,577 platelet aggregation and amplification of cular events among nearly 42,000 nondi- patients with diabetes in the HOPE study, aggregation induced by other agonists. abetic patients was reduced from 17.2 to treatment with ramipril (10 mg/day) low- The mechanism of action of clopidogrel 13.7% with antiplatelet therapy (P Ͻ ered the risk of MI, stroke, or CV death by and ticlopidine differs from that of aspirin. 0.00001). Although the overall incidence 25% (P ϭ 0.0004)—an effect that per- Ticlopidine was evaluated for its ef- of vascular events is much higher in pa- sisted after adjustments were made for fects on microvascular disease in diabetic tients with diabetes (consistent with their changes in systolic and diastolic blood patients in the Ticlopidine in Microangi- higher risk), the benefit of antiplatelet pressures (49). By blocking prostaglandin opathy of Diabetes study (52). A total of therapy in diabetic and nondiabetic pa- synthesis, aspirin has been hypothesized 435 patients with nonproliferative dia- tients was comparable: 42 vascular events to blunt the bradykinin-related vasodila- betic retinopathy were randomized to re- were prevented for every 1,000 diabetic tory effects of ACE inhibitors (50). Clini- ceive ticlopidine, 250 mg b.i.d., or patients and 35 events for every 1,000 cal data on the interaction of aspirin with placebo and were followed for up to 3 nondiabetic patients. ACE inhibitors have been mixed; retro- years. Ticlopidine significantly reduced Aspirin irreversibly inhibits the cyclo- spective analyses of two large heart failure annual microaneurysm progression by oxygenase pathway of arachidonic acid trials suggested that aspirin blunted the 67% based on fluorescein angiograms metabolism; it blocks platelet thrombox- benefit achieved with ACE inhibitor treat- (P ϭ 0.03), and among insulin-treated di- ane synthesis, resulting in inhibition of ment (50). On the other hand, in a retro- abetic patients, it reduced annual micro- platelet aggregation (19,22,39) (Fig. 1.). spective analysis of 11,575 patients with aneurysm progression by 85% (P ϭ However, aspirin has a paradoxical ef- coronary artery disease in the Bezafibrate 0.03). Moreover, the insulin-treated dia- fect—particularly at doses Ͼ325 mg/ Infarction Prevention Trial, patients who betic patients had a trend for developing day—because it also inhibits prostacyclin were treated with 250 mg/day aspirin and fewer new vessels. Overall progression of

2184 DIABETES CARE, VOLUME 26, NUMBER 7, JULY 2003 Colwell and Nesto retinopathy was significantly less severe with ticlopidine (P ϭ 0.04). This study supports the postulate that platelets are involved in the pathogenesis of micro- vascular disease in patients with diabetes. It was not designed to evaluate the effect of ticlopidine on CV events. A similar study with aspirin in diabetic individuals, the ETDRS, showed no effect on progres- sion of retinopathy (43). The Clopidogrel versus Aspirin in Pa- tients at Risk of Ischemic Events (CAP- RIE) trial examined the effects of 75 mg clopidogrel once daily versus 325 mg as- pirin once daily in a large secondary pre- vention population consisting of 19,185 patients with recent ischemic stroke, re- cent MI, or established PAD. Approxi- mately 20% of these patients were known to have diabetes. More than 6,300 pa- tients in each subgroup were randomized to receive 75 mg clopidogrel once daily or 325 mg aspirin once daily (53). The pri- mary end point was the combined inci- dence of ischemic stroke, MI, or vascular death. Patients were followed up for a mean of 1.9 years. The annual incidence of the primary end point was 5.32% with clopidogrel and 5.83% with aspirin, rep- resenting an 8.7% RR reduction with clo- pidogrel above aspirin (P ϭ 0.043) (Fig. 2A). Clopidogrel was more effective than aspirin in preventing fatal and nonfatal MI (incidence: 5.2 vs. 7.6%; risk reduction: 19.2%; P ϭ 0.008) (54,55). Clopidogrel provided an RR reduction of 5.2% for stroke and 7.6% for vascular death. In a recent report, Bhatt et al. (56) Figure 2—Kaplan-Meier analysis of the cumulative risk of primary end-point events in the retrospectively analyzed results in the di- CAPRIE (A) and CURE (B) studies. The primary end point in CAPRIE was the combined incidence abetic subgroup in the CAPRIE study. Of of ischemic stroke, MI, or vascular death; CAPRIE included patients with more severe macrovas- 1,914 diabetic patients randomized to cular disease (recent stroke or MI or symptomatic PAD). The CURE study included patients with clopidogrel, 15.6% had the composite unstable angina and nonϪQ-wave MI and who had a primary first outcome of CV death, nonfatal vascular primary end point versus 17.7% MI, or stroke. Modified with permission from CAPRIE (53) and CURE (58). of 1,952 diabetic patients randomized to aspirin therapy (P ϭ 0.042). Thus, clopi- dogrel appears to be an effective antiplate- mg thereafter) or placebo in addition to subgroups analyzed and was incremental let agent for secondary prevention in aspirin for up to 1 year. to the standard therapy, including GpIIb- diabetic subjects, whereas aspirin is more After the mean follow-up period of 9 IIIa inhibitors. The second primary out- cost-effective (57). months, patients on clopidogrel and aspi- come consisting of the first primary The Clopidogrel in Unstable Angina rin experienced a significant 20% reduc- outcome or refractory ischemia was also to Prevent Recurrent Events (CURE) tion in the first primary outcome, a significantly reduced (RR 0.84; P Ͻ study examined CV outcomes with clopi- composite of CV death, nonfatal MI, or 0.001) (58). Significantly more patients dogrel plus aspirin versus aspirin alone in stroke, compared with patients receiving in the clopidogrel and aspirin group had patients with acute ischemic heart disease aspirin and placebo (RR 0.80; 95% CI major bleeding (3.7 vs. 2.7%), but there (58). In the CURE study, 12,562 patients 0.72–0.90; P Ͻ 0.001) (58) (Fig. 2B). An was no increase in life-threatening bleeds. with unstable angina or nonϪQ-wave MI analysis of subgroups showed consis- Based in part on the CURE data, the who were being treated with aspirin, tency of effect on the primary outcome in American College of Cardiology and the ␤-blockers, and other conventional ther- patients with diabetes or those with no American Heart Association (ACC/AHA) apies were randomized to receive either diabetes. The trend favoring the addition task force on practice guidelines recently clopidogrel (300 mg loading dose and 75 of clopidogrel was consistent across all issued an updated version of their guide-

DIABETES CARE, VOLUME 26, NUMBER 7, JULY 2003 2185 The platelet in diabetes lines for the management of unstable an- in diabetic individuals was even greater in aspirin (0.17%) (53). Additionally, there gina and non–Q-wave MI and the the 1,279 patients who underwent percu- was no difference between aspirin and prevention of secondary CV events (59). taneous coronary intervention during the clopidogrel in the incidence of thrombo- Because clopidogrel is now recognized as index hospitalization; in these individu- cytopenia (53). Among the many millions efficacious in both the early phase and als, GpIIb-IIIa inhibitors reduced 30-day of individuals worldwide who have re- during long-term treatment of unstable mortality from 4 to 1.2% (P ϭ 0.002). ceived clopidogrel, there have been 11 angina and non–Q-wave MI, the guide- The large number of patients and events cases of possible TTP (62). Reportedly, lines recommend promptly adding clopi- (339 deaths) in this meta-analysis provide however, clopidogrel was implicated as a dogrel to aspirin in patients presenting strong evidence in support of the concept causative factor in only 5 of the 11 cases with unstable angina and nonϪQ-wave of platelet inhibition for diabetic patients (63). In these patients, TTP occurred MI. The guidelines recommend that clo- with ACS. within the first 2 weeks of therapy. Of the pidogrel be administered to patients who 11 patients, 10 recovered after receiving are hypersensitive or intolerant to aspirin. SAFETY OF ANTIPLATELET plasma exchange. This rate is essentially Further, clopidogrel should be used both DRUGS — Aspirin, like other nonste- the same as the background rate in the in patients being treated with medical roidal anti-inflammatory drugs, has the population at large. Physicians should be therapy or percutaneous coronary inter- potential to cause gastrointestinal adverse aware of the possibility of this syndrome vention for up to 9 months (59). effects, such as gastric mucosal injury or during the first 2 weeks after initiation of Although there were instances of ma- bleeding. These effects are dose related clopidogrel therapy. However, blood jor bleeds beyond 30 days, no cases of and may be minimized with enteric- monitoring is not required. life-threatening bleeds were reported. coated low-dose aspirin. However, gas- In the CURE study, clopidogrel plus Thus, the data suggest that the benefits of trointestinal toxicity may occur even at aspirin had a greater incidence of major taking clopidogrel plus aspirin for up to 9 low doses of aspirin (51). Although quite bleeding than aspirin alone (3.7 vs. 2.7%; months outweigh the risks (58). How- rare, the adverse effect of most concern is RR 1.38; P ϭ 0.001) (58). Bleeding was ever, in patients taking clopidogrel in intracerebral hemorrhage. It is important most common at puncture sites and in the whom coronary artery bypass graft is to recognize that there is no increase in gastrointestinal system (64). There was planned, the drug should be withheld for the risk for retinal or vitreous bleeding no increase in fatal hemorrhages or intra- at least 5 days and preferably for 7 days. with aspirin therapy (43). Additionally, if cranial hemorrhages. blood pressure is controlled in hyperten- In the meta-analysis of GpIIb-IIIa GP IIb-IIIa receptor antagonists sive patients, there is no increased risk of studies (48), there was an absolute excess Recent data from a meta-analysis of six intracerebral bleeding (44). of 23 major extracranial bleeding epi- trials of intravenous GpIIb-IIIa inhibitors A limitation to the use of ticlopidine is sodes per 1,000 patients treated. Fatal in acute coronary syndrome (ACS) pa- its potential to cause bone marrow sup- bleeding was rare. Intracranial hemor- tients with diabetes have reinforced the pression. Severe neutropenia and/or rhages were few in number (GpIIb-IIIa concept that treatment targeting the agranulocytosis have been observed in plus aspirin, 0.2% vs. aspirin alone, 0.1% platelet can play a major role in reducing ϳ1% of 2,048 stroke patients in clinical [NS]). CV events in diabetic individuals (60,61). trials (62). In the Ticlopidine Aspirin These inhibitors act at the fibrinogen Stroke Study (TASS), 2.4% of patients re- CONCLUSIONS AND binding site (Fig. 1). The meta-analysis ceiving ticlopidine had neutropenia (ab- CLINICAL included 6,458 diabetic patients in the solute neutrophil count Ͻ1,200 cells/ RECOMMENDATIONS — Macro- following trials: Platelet Receptor Inhibi- mm2) and 0.9% had severe neutropenia vascular complications are the leading tion for Ischemic Syndrome Manage- (absolute neutrophil count Ͻ450 cells/ causes of death in patients with diabetes ment (PRISM), Platelet Receptor mm2). Thrombotic thrombocytopenic and are responsible for a significant Inhibition for Ischemic Syndrome Man- purpura (TTP) has also been reported and amount of morbidity in this population. agement in Patients Limited by Unstable may be fatal if untreated. During clinical Moreover, patients with diabetes and es- Signs and Symptoms (PRISM-PLUS), trials of ticlopidine, one case was re- tablished atherosclerotic disease are at es- Platelet IIb/IIIa Antagonist for Reduction ported, but based on postmarketing data, pecially high risk for additional ischemic of Acute Coronary Events in a Global Or- U.S. physicians reported ϳ100 cases be- vascular events, including MI, stroke, and ganization Network (PARAGON) A and tween 1992 and 1997. The estimated in- vascular death. Diabetes and prediabetic B, Platelet IIb/IIIa in Unstable Angina: Re- cidence of ticlopidine-associated TTP conditions are associated with platelet ceptor Suppression Using Integrilin Ther- may be as high as one case in every and coagulation derangements. Because apy (PURSUIT), and Global Utilization of 2,000–4,000 patients exposed. This re- platelet aggregation plays an integral role Streptokinase and Tissue-Plasminogen sult necessitates hematological monitor- in thrombus formation, treatment strate- Activator for Occluded Arteries (GUSTO) ing for the first 3 months of treatment gies have focused on using antiplatelet IV. GpIIb-IIIa blockers significantly re- with ticlopidine (61). agents to prevent subsequent ischemic duced mortality at 30 days from 6.2 to Clopidogrel did not produce bone events. The efficacy of aspirin relative to 4.6% (P ϭ 0.007) in the diabetic patients. marrow toxicity in either tissue cultures placebo as a secondary prevention strat- Among the more than 22,000 patients in or animal models during preclinical stud- egy was established by a meta-analysis of these trials who did not have diabetes, ies. The incidence of neutropenia with clinical trials in the ATC that included a GpIIb-IIIa inhibitors did not improve sur- clopidogrel in the CAPRIE trial (0.10%) subgroup analysis of diabetic patients. vival. The effect of GpIIb-IIIa inhibitors was comparable to that associated with Three primary prevention trials have sup-

2186 DIABETES CARE, VOLUME 26, NUMBER 7, JULY 2003 Colwell and Nesto ported the use of aspirin as a primary pre- 10. MacDonald TM, Butler R, Newton RW, Platelet survival in streptozotocin- vention strategy in patients with diabetes Morris AD, for the DARTS/MEMO induced diabetic rats. Thromb Haemost who are at high risk for CV events. The Collaboration: Which drugs benefit dia- 51:307–312, 1984 ADP-receptor blocker clopidogrel has betic patients for secondary prevention of 24. DiMinno G, Silver MJ, Cerbone AM, been shown to produce additional risk re- myocardialinfarction?DiabetMed15:282– Murphy S: Trial of repeated low-dose as- 289, 1998 pirin in diabetic . Blood 68: duction relative to aspirin in two second- 11. Haffner SM, Lehto S, Ro¨nnemaa T, 886–891, 1986 ary prevention trials (CAPRIE and CURE). Pyo¨ra¨la¨ K, Laakso M: Mortality from cor- 25. Sagel J, Colwell JA, Crook L, Laimins M: Clinical practice recommendations onary heart disease in subjects with type 2 Increased platelet aggregation in early di- from the ADA indicate that a secondary diabetes and in nondiabetic subjects with abetes mellitus. Ann Intern Med 82:733– prevention strategy including antiplatelet and without prior myocardial infarction. 738, 1975 agents should be adopted for patients N Engl J Med 339:229–234, 1998 26. Davies MJ: Stability and instability: two with diabetes and evidence of macrovas- 12. Kip KE, Faxon DP, Detre KM, Yeh W, faces of coronary atherosclerosis: the Paul cular disease (40). This includes patients Kelsey SF, Currier JW: Coronary angio- Dudley White Lecture 1995. Circulation with a history of MI, vascular bypass pro- plasty in diabetic patients: the National 94:2013–2020, 1996 cedure, stroke or transient ischemic attack, Heart, Lung, and Blood Institute Percuta- 27. Mayfield RK, Halushka PV, Wohltmann neous Transluminal Coronary Angio- HJ, Lopes-Virella M, Chambers JK: Plate- peripheral vascular disease, claudication, plasty Registry. Circulation 94:1818–1825, let function during continuous insulin in- or angina. In addition, low-dose (81–325 1996 fusion treatment in insulin-dependent mg/day) enteric-coated aspirin therapy is 13. Stein B, Weintraub WS, Gebhart SP, diabetic patients. Diabetes 34:1127–1133, recommended by the ADA as a primary Cohen-Bernstein CL, Grosswald R, Liber- 1985 prevention strategy for people with diabe- man HA, Douglas JS Jr, Morris DC, King 28. Watala C, Boncer M, Golanski J, Koziol- tes who are at high risk for CV events (40). SB 3rd: Influence of diabetes mellitus on kiewcz W, Trojanowski Z, Walkowiak B: early and late outcome after percutaneous Platelet membrane lipid fluidity and in- transluminal coronary angioplasty. Circu- traplatelet calcium mobilization in type 2 References lation 91:979–989, 1995 diabetes. Eur J Haematol 61:319–326, 1. Nathan DM: Long-term complications of 14. Colwell JA: Clinical trials of antiplatelet 1998 diabetes mellitus. N Engl J Med 328: agents in diabetes mellitus: rationale and 29. Martina V, Bruno GA, Trucco F, 1676–1685, 1993 results. Semin Thromb Hemost 17:439– Zumpano E, Tagliabue M, Di Bisceglie C: 2. Stamler J, Vaccaro O, Neaton JD, Went- 444, 1991 Platelet cNOS activity is reduced in pa- worth D: Diabetes, other risk factors, and 15. Aronow WS, Ahn C: Prevalence of coex- tients with IDDM and NIDDM. Thromb 12-yr cardiovascular mortality for men istence of coronary artery disease, periph- Haemost 79:520–522, 1998 screened in the Multiple Risk Factor In- eral arterial disease, and atherothrom- 30. Trovati M, Anfossi G, Massucco P, Mat- tervention Trial. Diabetes Care 16:434– botic brain infarction in men and women tiello L, Costamagna C, Piretto V, Mul- 444, 1993 Ն62 years of age. Am J Cardiol 74:64–65, aroni E, Cavalot F, Bosia A, Ghigo D: 3. Brand FN, Abbott RD, Kannel WB: Dia- 1994 Insulin stimulates nitric oxide synthesis in betes, intermittent claudication, and risk 16. Jokl R, Colwell JA: Arterial thrombosis human platelets and, through nitric ox- of cardiovascular events. Diabetes 38: and atherosclerosis in diabetes. Diabetes ide, increases platelet concentrations of 504–509, 1989 Metab Rev 5:1–15, 1997 both guanosine-3Ј,5Ј-cyclic monophos- 4. Reaven GM: Banting Lecture 1988: Role 17. Jokl R, Colwell JA: Clotting disorders in phate and adenosine-3Ј,5Ј-cyclic mono- of in human disease. Di- diabetes. In International Textbook of Dia- phosphate. Diabetes 46:742–749, 1997 abetes 37:1595–1607, 1988 betes Mellitus. 2nd ed. Alberti KGMM, 31. Sarji KE, Kleinfelder J, Brewington P, 5. Pahor M, Psaty BM, Furberg CD: New ev- Zimmet P, DeFronzo RA, Keen H, Eds. Gonzalez J, Hempling H, Colwell JA: De- idence on the prevention of cardiovascu- Chichester, U.K., Wiley, 1997, p. 1543– creased platelet vitamin C in diabetes mel- lar events in hypertensive patients with 1557 litus: possible role in hyperaggregation. type 2 diabetes. J Cardiovasc Pharmacol 32 18. Handin RI: Bleeding and thrombosis. In Thromb Res 15:639–650, 1979 (Suppl. 2):S18–S23, 1998 Harrison’s Principles of Internal Medicine. 32. Tschoepe D, Rauch U, Schwippert B: 6. Serrano Rios M: Relationship between 14th ed. Fauci AS, Braunwald E, Issel- Platelet-leukocyte-cross-talk in diabetes obesity and the increased risk of major bacher KJ, Wilson JD, Martin JB, Kasper mellitus. Horm Metab Res 29:631–635, complications in non-insulin-dependent DL, Hauser SL, Longo DL, Eds. New York, 1997 diabetes mellitus. Eur J Clin Invest 28 McGraw-Hill, 1998, p. 339–345 33. Leet H, Paton RC, Passa P, Caen JP: (Suppl. 2):14–18, 1998 19. Schafer AI: Antiplatelet therapy. Am J Med Fibrinogen binding and ADP-induced 7. Colwell JA: Multifactorial aspects of the 101:199–209, 1996 aggregation in platelets from diabetic sub- treatment of the type II diabetic patient. 20. Klein RL, Colwell JA: Altered endothelial jects. Thromb Res 24:143–150, 1981 Metabolism 46 (Suppl. 1):1–4, 1997 function in diabetes mellitus: an over- 34. Mandal S, Sarode R, Dash S, Dash RJ: Hy- 8. Brand FN, Kannel WB, Evans J, Larson view. In Contemporary Endocrinology: En- peraggregation of platelets detected by MG, Wolf PA: Glucose intolerance, phys- docrinology of the Vasculature. Conn PM, whole blood platelet aggregometry in ical signs of peripheral artery disease, and Sowers JR, Eds. Totowa, NJ, Humana, newly diagnosed noninsulin-dependent risk of cardiovascular events: the Fra- 1996, p. 125–134 diabetes mellitus. Am J Clin Pathol 100: mingham Study. Am Heart J 136:919– 21. Boon GD: An overview of hemostasis. 103–107, 1993 927, 1998 Toxicol Pathol 21:170–179, 1993 35. Gawaz M, Ott I, Reininger AJ, Neumann 9. Cantrill JA, D’Emanuele A, Dornan TL, 22. Halushka PV, Rogers RC, Loadholt CB, FJ: Effects of magnesium on platelet ag- Garcia S: A survey of drug treatment and Colwell JA: Increased platelet thrombox- gregation and adhesion: magnesium outcomes in diabetic patients with acute ane synthesis in diabetes mellitus. J Lab modulates surface expression of glyco- myocardial infarcts. J Clin Pharm Ther 20: Clin Med 97:87–96, 1981 proteins on platelets in vitro and ex vivo. 207–213, 1995 23. Winocour PD, Laimins M, Colwell JA: Thromb Haemost 72:912–918, 1994

DIABETES CARE, VOLUME 26, NUMBER 7, JULY 2003 2187 The platelet in diabetes

36. Gerster H: Prevention of platelet dysfunc- or other atherosclerotic vascular diseases tients with evidence of cerebrovascular tion by vitamin E in diabetic atherosclero- (AHA Scientific Statement). Circulation 106: disease (Abstract). Neurology 51:332, 1998 sis. Z Ernahrungswiss 32:243–261, 1993 388–391, 2002 56. Bhatt D, Marso S, Hirsch A, Ringleb P, 37. Gisinger C, Watanabe J, Colwell JA: Vita- 46. U.S. Preventive Services Task Force: As- Hacke W, Topol E: Amplified benefitof min E and platelet eicosanoids in diabetes pirin for the primary prevention of cardio- clopidogrel versus aspirin in patients with mellitus. Prostaglandins Leukot Essent Fatty vascular events: recommendations and diabetes mellitus. Am J Cardiol 90:625– Acids 40:169–176, 1990 rationale. Ann Intern Med 136:157–160, 628, 2002 38. Jain SK, Krueger KS, McVie R, Jaramillo 2001 57. Gaspoz JM, Coxson PG, Goldman PA, JJ, Palmer M, Smith T: Relationship of 47. Antiplatelet Trialists’ Collaboration: Col- Williams LW, Duntz KM, Hunink MG, blood thromboxane-B2 (TxB2) with lipid laborative overview of randomised trials Goldman L: Cost effectiveness of aspirin, peroxides and effect of vitamin E and pla- of antiplatelet therapy. I. Prevention of clopidogrel, or both for secondary pre- cebo supplementation on TxB2 and lipid death, myocardial infarction, and stroke vention of coronary heart disease. N Engl peroxide levels in type 1 diabetic patients. by prolonged antiplatelet therapy in vari- J Med 346:1819–1821, 2002 Diabetes Care 21:1511–1516, 1998 ous categories of patients. BMJ 308:81– 58. The Clopidogrel in Unstable Angina to 39. Colwell JA: Aspirin therapy in diabetes. 106, 1994 Prevent Recurrent Events Trial Investiga- Diabetes Care 20:1767–1771, 1997 48. Antithrombotic Trialists’ Collaboration: tors: Effects of clopidogrel in addition to 40. American Diabetes Association: Aspirin Collaboration meta-analysis of random- aspirin in patients with acute coronary therapy in diabetes (Position Statement). ised trials of antiplatelet therapy for pre- syndromes without ST-segment eleva- Diabetes Care 26 (Suppl. 1):S87–S88, vention of death, myocardial infarction, tion. N Engl J Med 345:494–502, 2001 2003 and stroke in high risk patients. BMJ 324: 59. American College of Cardiology website: 41. Collaborative Group of the Primary Pre- 71–86, 2002 ACC/AHA Guideline update for the man- vention Project: Low-dose aspirin and vi- 49. Heart Outcomes Prevention Evaluation agement of patients with unstable angina tamin E in people at cardiovascular risk: a (HOPE) Study Investigators: Effects of and non-ST-segment elevation myocar- randomized trial in general practice. Lan- ramipril on cardiovascular and microvas- dial infarction: unstable angina/non-ST cet 357:89–95, 2001 cular outcomes in people with diabetes elevation myocardial infarction. Available 42. Steering Committee of the Physicians’ mellitus: results of the HOPE study and at www.acc.org. Accessed 29 September Health Study Research Group: Final re- MICRO-HOPE substudy. Lancet 355: 2002 port on the aspirin component of the on- 253–259, 2000 60. Roffi M, Chew DP, Mukherjee D, Bhatt going Physicians’ Health Study. N Engl 50. Stys T, Lawson WE, Smaldone G, Stys A: DL, White JA, Heeschen C, Hamm CW, J Med 321:129–135, 1989 Does aspirin attenuate the beneficial ef- Moliterno DJ, Califf RM, White HD, 43. ETDRS Investigators: Aspirin effects on fects of angiotensin converting enzyme Kleiman NS, Theroux P, Topol EJ: Platelet mortality and morbidity in patients with inhibition in heart failure? Arch Intern Med glycoprotein IIb/IIIa inhibitors reduce diabetes mellitus. JAMA 268:1292–1300, 160:1409–1413, 2000 mortality in diabetic patients with non- 1992 51. Leor J, Reicher-Reiss H, Goldbourt U, ST-segment-elevation acute coronary 44. Hansson L, Zanchetti A, Carruthers SG, Boyko V, Gottlieb S, Battler A, Behar S: syndromes. Circulation 104:2767–2771, Dahlof B, Elmfeldt D, Julius S, Menard J, Aspirin and mortality in patients treated 2001 Rahn KH, Wedel H, Westerling S, for the with angiotensin-converting enzyme in- 61. Ticlid (package insert). Nutley, NJ, Roche HOT Study Group: Effects of intensive hibitors: a cohort study of 11,575 patients Laboratories, 1995 blood-pressure lowering and low-dose as- with coronary artery disease. J Am Coll 62. Bennett CL, Connors JM, Carwile JM, pirin in patients with hypertension: prin- Cardiol 33:1920–1925, 1999 Moake JL, Bell WR, Tarantolo SR, cipal results of the Hypertension Optimal 52. TIMAD Study Group: Ticlopidine treat- McCarthy LJ, Sarode R, Hatfield AJ, Feld- Treatment (HOT) randomised trial. Lan- ment reduces the progression of non- man MD, Davidson CJ, Tsai HM: Throm- cet 351:1755–1762, 1998 proliferative diabetic retinopathy. Arch botic thrombocytopenic purpura associated 45. Pearson TA, Blair SN, Daniels SR, Eckel Ophthalmol 108:1577–1583, 1990 with clopidogrel. N Engl J Med 342:1773– RH, Fair JM, Fortmann SP, Franklin BA, 53. CAPRIE Steering Committee: A random- 1777, 2000 Goldstein LB, Greenland P, Grundy SM, ised, blinded, trial of clopidogrel versus 63. Paradiso-Hardy FL, Papastergiou J, Lanc- Hong Y, Miller NH, Lauer RM, Ockene IS, aspirin in patients at risk of ischaemic tot KL, Cohen EA: Thrombotic thrombo- Sacco RL, Sallis JF Jr, Smith SC Jr, Stone events (CAPRIE). Lancet 348:1329–1339, cytopenic purpura associated with NJ, Taubert KA: AHA guidelines for pri- 1996 clopidogrel: further evaluation. Can J Car- mary prevention of cardiovascular disease 54. Gent M: Benefit of clopidogrel in patients diol 18:771–773, 2002 and stroke: 2002 update. Consensus with coronary disease (Abstract). Circula- 64. Plavix (package insert). New York, Bris- panel guide to comprehensive risk reduc- tion 96 (Suppl. 1):467, 1997 tol-Myers Squibb/Sanofi Pharmaceuticals tion for adult patients without coronary 55. Easton JD: Benefit of clopidogrel in pa- Partnership, 2002

2188 DIABETES CARE, VOLUME 26, NUMBER 7, JULY 2003