Effect of caffeine on platelet inhibition by clopidogrel in healthy subjects and patients with coronary artery disease

Effect of caffeine on platelet inhibition by clopidogrel in healthy subjects and patients with coronary artery disease Eli I. Lev, MD, a Mehmet E. Arikan, MD, a Muthiah Vaduganathan, BS, a Carlos L. Alviar, MD, a Armando Tellez, MD, a Nilesh Mathuria, MD, b Angela Builes, MD, a Juan F. Granada, MD, a Ian del Conde, MD, c and Neal S. Kleiman, MD a Houston, TX; and Boston, MA Background Clopidogrel inhibits the platelet P2Y12 receptor, leading to increased intracellular cyclic AMP (camp) levels. Caffeine also causes a rise in platelet camp. We aimed to test the effect of acute caffeine administration on platelet inhibition by clopidogrel, in healthy volunteers and patients with coronary artery disease. Methods Cohort 1: 12 healthy subjects were enrolled in a 2-week crossover study. Blood samples were drawn at baseline, 2, 4, and 24 hours after 300 mg clopidogrel intake. At the first week, 6 subjects received caffeine (300 mg pill, equivalent to a medium sized coffee drink) 30 minutes after clopidogrel. At week 2, the other 6 subjects received caffeine. One month later the effect of caffeine alone was tested. Platelet function was evaluated by aggregation in response to 5, 10, and 20 μmol/l adenosine diphosphate, 1 μg/ml collagen, and flow cytometric determination of P-selectin expression, PAC-1 binding, and vasodilator-stimulated phosphoprotein phosphorylation. Cohort 2: 40 patients with coronary artery disease receiving aspirin and clopidogrel (75 mg daily) for 1 week were tested at baseline and 2.5 hours after caffeine (300 mg). Results In cohort 1 (crossover study), caffeine was associated with lower adenosine diphosphate induced aggregation at 4 hours, lower activation markers at 2 hours, and lower vasodilator-stimulated phosphoprotein phosphorylation at 4 hours after clopidogrel. Caffeine alone had no effect on the assessed platelet surface biomarkers. In cohort 2, caffeine administration was associated with lower platelet activation markers (P-selectin, PAC-1 binding), without significant effect on aggregation. Conclusions Acute caffeine administration after clopidogrel loading appears to be associated with enhanced platelet inhibition 2 to 4 hours after clopidogrel intake. The mechanism probably involves synergistic increase in camp levels. (Am Heart J 2007;154:694.e1-694.e7.) Background Clopidogrel is currently administered to several million patients in the United States for periods of months to years, especially after coronary stenting. Clopidogrel has been shown to reduce cardiovascular complications in patients with acute coronary syndromes 1 and patients who have undergone coronary stenting. 2 The mechanism of action of clopidogrel's active metabolite involves inhibition of the purinergic adenosine diphosphate (ADP) receptor P2Y12 on the platelet membrane. Blockade of this receptor prevents uncoupling of the associated Gi2 From the a The Methodist Hospital Research Institute and the Methodist DeBakey Heart Center, Houston, TX, affiliated with Weill Medical College, Cornell University, NY, b Department of Cardiology, Baylor College of Medicine, Houston, TX, c Department of Medicine, Brigham and Women's Hospital, Boston, MA. Submitted May 25, 2007; accepted July 24, 2007. Reprint requests: Neal S. Kleiman, MD, Section of Cardiology, The Methodist DeBakey Heart Center, 6565 Fannin St/Mail Station F-1090, Houston, TX 77030. E-mail: nkleiman@tmh.tmc.edu 0002-8703/$ - see front matter 2007, Mosby, Inc. All rights reserved. doi:10.1016/j.ahj.2007.07.014 protein which ultimately leads to increased platelet cyclic AMP (camp) formation. 3 Cyclic AMP is a key signaling molecule in inhibiting platelet aggregation, but its intracellular levels are affected by several other commonly used compounds. For instance, methylxanthines, such as caffeine, theophylline, and theobromine (an ingredient of chocolate), all cause elevation of intracellular camp levels by inhibiting adenosine receptors (types A 1 and A 2 ) on the platelet membrane. 4,5 Caffeine affects mainly the A 2A receptor and, to a lesser extent, the other receptor subtypes. 5 As both blockade of P2Y12 receptors by clopidogrel and inhibition of adenosine receptors by methylxanthines lead to increase in intracellular platelet camp levels, we hypothesized that coadministration of clopidogrel and caffeine may have synergistic effects on platelet aggregation. The effect of caffeine consumption on platelet reactivity depends on the caffeine dose and duration of administration. Chronic caffeine consumption ( 7 days) appears to be associated with inhibition of platelet aggregation, probably through upregulation of adenosine receptors. 4-6 The effect of acute caffeine administration

694.e2 Lev et al American Heart Journal October 2007 on platelet function is less clear. Different studies have shown either an increase, decrease, or no change in platelet reactivity after acute caffeine administration. 7-9 To our knowledge, there is no information on the effect of acute caffeine administration in conjunction with clopidogrel on platelet reactivity. Accordingly, our aim was to examine the effect of acute caffeine consumption, at a dose equivalent to commercial coffee drinks, on the antiplatelet effect of clopidogrel, in healthy volunteers and in patients with coronary artery disease (CAD). Methods Subjects Cohort 1. Twelve healthy volunteers (mean age 33.5 ± 6 years, 3 women and 9 men) who were not taking any medications in the month before the study were included. Cohort 2. Forty patients with CAD 1 day after uncomplicated elective percutaneous coronary intervention (PCI) were enrolled. Inclusion criteria were aspirin treatment for at least a week (81-325 mg a day), clopidogrel treatment for at least a week (75 mg daily after the loading dose), and use of bivalirudin during the PCI. Exclusion criteria were use of glycoprotein (GP) IIb/IIIa inhibitors during the PCI, acute coronary syndrome within 1 week, hemodynamic instability, any contraindications to aspirin or clopidogrel, thrombocytopenia (b100 10 3 cells per cubic millimeter), anemia (hemoglobin b10 g/dl), or renal failure (creatinine N2.5 mg/dl). The first cohort of the study was approved by the Investigational Review Board (ethics committee) of the Baylor College of Medicine, Houston, TX, and the second cohort of the study was approved by the Investigational Review Board (ethics committee) of the Methodist Hospital Research Institute, Houston, TX. All subjects in each cohort provided written informed consent. Study design Cohort 1. Subjects in the first cohort underwent a 2-week crossover study. In the first week, after 2 days without consumption of any caffeine-containing beverages or foods (eg, coffee, tea, cola, chocolate, etc) all subjects were given clopidogrel 300 mg orally at 9:00 AM. Six randomly selected subjects were given 300 mg of caffeine in the form of pills (1.5 pills, each pill containing 200 mg caffeine, Prolab Nutrition, Inc, Chatsworth, CA) 30 minutes after the clopidogrel. The other 6 subjects were not given caffeine. The 300-mg dose of caffeine was chosen because it is equivalent to the caffeine content of a 16-oz commercial coffee drink. Four blood samples were taken from all subjects during a period of 24 hours: (a) baseline just before clopidogrel administration; (b) 2 hours after clopidogrel administration; (c) 4 hours after clopidogrel; (d) 24 hours after clopidogrel treatment. After a washout period of 1 week (from clopidogrel) all subjects were again given 300 mg of clopidogrel at 09:00 AM (the same day of the week). In the last 2 days of the washout period, subjects once more refrained from ingesting any caffeine-containing beverages or foods. In the second week of the study, the 6 subjects who did not receive caffeine the first week received 300 mg of caffeine (30 minutes after clopidogrel intake), whereas the other 6 subjects did not take caffeine. Four blood samples were drawn at the same time points as the first week. The investigators who performed the platelet function assays were blinded to the randomization of the subjects. One month after the completion of the crossover study all subjects were tested for the effect of caffeine alone on platelet function. After 2 days without caffeine-containing beverages or foods, 2 blood samples were drawn: at baseline (09:00 AM) and 2 hours after taking 300 mg of caffeine (without clopidogrel). The 2-hour point was chosen because peak blood levels of caffeine are obtained within 1 to 2 hours after an oral dose 10,11 and to enable comparison to previous studies which tested the acute effects of caffeine 1 to 3 hours after administration. 8,9 Cohort 2. Patients with CAD 1 day after uncomplicated elective PCI had a baseline blood sample taken at 06:00 AM. They were asked to refrain from any caffeine-containing beverages or foods during the day before blood drawing. Immediately after the baseline blood sample was drawn the patients were given 300 mg of caffeine (with pills similar to cohort 1). Two and a half hours (±15 minutes) after caffeine administration a second blood sample was drawn, before the daily clopidogrel dose (administered at 09:00 AM). This time point was chosen for several reasons: (a) After an interim analysis of cohort 1, a caffeine effect was observed 2 to 4 hours after clopidogrel loading, which was 1.5 to 3.5 hours after the caffeine. Therefore, a midpoint was chosen. (b) We wished to test the effect of caffeine in proximity to the time that peak blood levels of caffeine are achieved (1-2 hours) after an oral dose. 10,11 (c) This design enabled both blood samples to be drawn before the administration of the daily clopidogrel dose. Blood sampling. All blood samples were obtained from an antecubital vein, using a 21-gauge needle, and collected in tubes containing 3.2% citrate. The tubes were filled to capacity and then gently mixed. The blood samples were processed within 2 hours of blood collection. Platelet function testing Platelet aggregation. Turbidimetric platelet aggregation was performed in platelet-rich plasma with a platelet count adjusted to 250 10 3 /mm 3. Platelets were stimulated with 5, 10, and 20 μmol/l ADP and 1 μg/ml collagen. Aggregation was performed with a BioData PAP-8 platelet aggregometer (BioData Corp, Horsham, PA). The extent of aggregation was defined as the maximal amount of light transmission reached 6 minutes after addition of the agonist, with platelet-poor plasma used as a reference. Platelet activation. Platelet activation was determined by assessing platelet surface expression of activated GP IIb/IIIa receptors and P-selectin in response to ADP stimulation, using flow cytometry as previously described. 12 Briefly, GP IIb/IIIa activation was assessed using a fluorescein isothiocyanate conjugated PAC-1 antibody (Becton Dickinson, San Jose, CA), whereas P-selectin expression was determined using an R-Phycoerythrin conjugated anti-cd62p antibody (BD Pharmingen, San Jose, CA). Citrated whole blood was diluted with Tyrodes buffer and stimulated for 5 minutes with 10 μmol/l ADP (final concentration). After adding the corresponding antibody and incubating for 20 minutes at room temperature and dark conditions, the mixture was fixated with phosphate-buffered saline containing 1% paraformaldehyde. The samples were analyzed with a Coulter Epics XL MCL flow cytometer (Beckman-Coulter, Miami, FL).

American Heart Journal Volume 154, Number 4 Lev et al 694.e3 Nonstimulated samples served as negative controls. Both PAC-1 binding and P-selectin were expressed as log mean fluorescence intensity (MFI). Vasodilator-stimulated phosphoprotein phosphorylation. Vasodilator-stimulated phosphoprotein (VASP) phosphorylation was measured only in cohort 1, time points 0, 2, and 4 hours. Vasodilator-stimulated phosphoprotein phosphorylation levels reflect the degree of clopidogrel-induced P2Y12 receptor inhibition. 13,14 To determine the VASP phosphorylation state of whole blood, we used a standardized flow cytometric assay (PLT VASP/P2Y12 kit; Biocytex, Marseille, France). 13 Briefly, a citrated blood sample was incubated with PGE 1 or with PGE 1 and ADP for 10 minutes and fixed with paraformaldehyde, after which the platelets were permeabilized with nonionic detergent. The cells were then labeled with a primary monoclonal antibody against serine 239-phosphorylated VASP (16C2), followed by a secondary fluorescein isothiocyanate conjugated polyclonal goat antimouse antibody. Analyses were performed with a Beckman-Coulter flow cytometer, and a P2Y12 reactivity index (PRI) was calculated from the MFI of samples incubated with PGE 1 or PGE 1 and ADP, according to the manufacturer's instructions. P2Y12 reactivity index, expressed as a percentage, is the relative difference in VASP MFI between resting (+PGE1) and activated (PGE1 + ADP) platelets ([MFI PGE 1 ] [MFI PGE 1 + ADP] [MFI PGE 1 ]) 100%). A reduced PRI is indicative of more enhanced clopidogrel-induced inhibition. 15 Cyclic AMP levels. Plasma levels of camp were measured in 20 randomly selected patients in cohort 2 (baseline + postcaffeine samples). Levels of camp were measured from frozen plasma samples using a camp enzyme immunoassay kit (camp Biotrak EIA, GE Healthcare, Bio-Sciences Corp, Piscataway, NJ), according to the manufacturer's instructions. Statistical analysis. Continuous variables are presented as mean ± SD. In cohort 1, analysis of platelet function with versus without caffeine at the various time points was performed by repeated measures analysis of variance with 2 repeated measures factors: caffeine and time. When the interaction between the repeated measures factors was significant, post hoc analysis for specific pairwise comparisons was performed by paired Student t tests. Intragroup comparisons for the analysis of the effect of caffeine alone (cohort 1) and the analysis of cohort 2 were performed by paired Student t tests. Analyses were performed using Sigmastat version 3.1 statistical software (SPSS, Inc, Chicago, IL), and statistical significance was set at P b.05. Results Cohort 1 Crossover study Platelet aggregation in response to 5 and 10 μmol/l ADP after clopidogrel intake, with and without caffeine, is presented in Figure 1. With both ADP concentrations caffeine administration was associated with lower platelet aggregation (increased platelet inhibition) 4 hours after clopidogrel intake (at 4 hours: 5 ADP: 49.3% ± 13% without caffeine vs 35% ± 10% with caffeine, P =.02; 10 ADP: 60.5% ± 13% without caffeine vs 43.7% ± 12% with caffeine, P =.03). At 24 hours after clopidogrel intake the differences diminished. Similar results were obtained using 20 μmol/l ADP for stimulation of platelet Figure 1 Cohort 1 Platelet activation markers. Platelet activation markers P-selectin expression and PAC-1 binding, reflecting GP IIb/IIIa activation after clopidogrel intake (300 mg), with and without caffeine (300 mg taken 30 min after clopidogrel) (n = 12). Caffeine administration was associated with lower levels of both activation markers 2 hours after clopidogrel intake (P =.01). aggregation (significant difference observed only at 4 hours after clopidogrel intake: 64.2% ± 12% without caffeine vs 50.1% ± 13% with caffeine, P =.04). No significant differences were observed in collagen-induced aggregation between the effect of clopidogrel with and without caffeine. Platelet activation markers P-selectin expression and PAC-1 binding after clopidogrel intake, in the presence and absence of caffeine, are presented in Figure 2. As observed for ADP-induced platelet aggregation, lower levels of activation markers were detected when the subjects received caffeine in addition to clopidogrel. However, the differences between the absence and presence of caffeine were noted at an earlier time point 2 hours after clopidogrel intake (at 2 hours: P-selectin: 7.7 ± 3.6 MFI without caffeine vs 5.6 ± 3.3 MFI with caffeine, P =.01; PAC-1 binding: 4.6 ± 1.6 MFI without caffeine vs 3.3 ± 1.6 MFI with caffeine, P =.01). At 4 hours after clopidogrel intake, the differences were no longer significant and at 24 hours they diminished (Figure 2).

694.e4 Lev et al American Heart Journal October 2007 Figure 2 Figure 3 Cohort 1 PRI. P2Y12 reactivity index, reflecting VASP phosphorylation, at baseline and 2 and 4 hours after clopidogrel intake, with and without caffeine (n = 12, healthy subjects). At both 2 and 4 hours, PRI was lower with caffeine intake. However, the differences (with vs without caffeine) reached statistical significance only at the 4 hours time point (P =.025). Cohort 1 Platelet activation markers. Platelet activation markers P-selectin expression and PAC-1 binding, reflecting GP IIb/IIIa activation after clopidogrel intake (300 mg), with and without caffeine (300 mg taken 30 min after clopidogrel) (n = 12). Caffeine administration was associated with lower levels of both activation markers 2 hours after clopidogrel intake (P =.01). P2Y12 reactivity index at time points 0, 2, and 4 hours with and without caffeine is presented in Figure 3. At both 2 and 4 hours, PRI was lower with caffeine intake, reflecting a higher degree of clopidogrel-induced inhibition with caffeine. However, the differences (with vs without caffeine) reached statistical significance only at the 4 hour time point (without caffeine: 65.4 ± 13% vs with caffeine: 54.2 ± 15%, P =.025) (Figure 3). Cohort 1 Effect of caffeine alone The effect of 300 mg caffeine on platelet function was tested 2 hours after caffeine administration. Caffeine had no effect on platelet aggregation or activation markers (there were no differences between the baseline and postcaffeine samples) (Table I). Cohort 2 Patients with CAD Forty patients with CAD who were taking clopidogrel for at least a week were included in the study. Their clinical characteristics and concurrent medications are presented in Table II. Platelet aggregation and activation markers at baseline and 2.5 hours after caffeine administration are presented in Figures 4 and 5. Caffeine induced a significant reduction in platelet activation markers (PAC-1 binding: 3.1 ± 1.4 MFI decreased to 2.6 ± 1.6 MFI after caffeine, P =.01; P-selectin expression: 6.4 ± 3.6 MFI decreased to 5.6 ± 3.5 MFI after caffeine, P =.02) (Figure 5). However, platelet aggregation did not differ significantly between the preand postcaffeine samples, although there was a trend for decrease in 20 μmol/l ADP-induced aggregation after caffeine intake (P =.1) (Figure 4). Cyclic AMP levels Plasma levels of camp, measured in 20 randomly selected patients, increased by 32% after caffeine (baseline: 1170 ± 471 vs 2.5 hours post-caffeine: 1550 ± 353 fmol per well, P =.0002). Discussion This is the first study to examine the effect of acute caffeine administration on the degree of clopidogrelinduced platelet inhibition. We have found that in healthy volunteers caffeine administration was associated with enhanced platelet inhibition 2 to 4 hours after clopidogrel treatment (loading dose), as reflected by platelet aggregation and activation markers, as well as PRI. In patients with CAD who were receiving long-term clopidogrel treatment this effect was less pronounced, but nonetheless, caffeine administration was associated with increased inhibition of platelet activation markers. Both these findings suggest that caffeine augments the antiplatelet effect of clopidogrel.

American Heart Journal Volume 154, Number 4 Lev et al 694.e5 Table I. Effect of caffeine alone in normal subjects (n = 12) Baseline After caffeine (2 h) Aggregation 5 μmol/l ADP (%) 80.5 ± 20 76.4 ± 15 10 μmol/l ADP (%) 91.5 ± 14 94.4 ± 7 20 μmol/l ADP (%) 92.6 ± 10 95.3 ± 6 1 μg/ml collagen (%) 75.8 ± 21 75.3 ± 16 PAC-1 binding (MFI) 3.9 ± 2 3.7 ± 2 P-Selectin expression (MFI) 9.6 ± 4 9.1 ± 4 The mechanism of the augmentation of clopidogrelinduced platelet inhibition by caffeine most likely involves an increase in the platelet levels of camp, which is a key signaling molecule involved in inhibition of platelet activation and aggregation. Although we did not measure directly the intracellular platelet levels of camp, we did observe a rise in plasma levels of camp (measured in cohort 2), which probably reflects intracellular camp changes as well. It should be emphasized that, in our study (cohort 1), caffeine alone, at a dose of 300 mg equivalent to the content of a 16-oz commercial coffee drink had no significant effect on platelet reactivity after 2 hours. Therefore, the increased platelet inhibition observed when caffeine was administered after clopidogrel in cohort 1 does not appear to be just an additive phenomenon. It is possible that the rise in intracellular camp levels attained by caffeine alone (at the dose given in our study) was below a certain threshold and not enough to cause significant platelet inhibition. However, in conjunction with clopidogrel, possibly the rise in camp levels, achieved by the 2 separate pathways, may have been high enough to surpass the effect of clopidogrel alone. The rise in camp caused by caffeine and clopidogrel most likely induced the phosphorylation of VASP by camp-dependent protein kinases, 13,16 as reflected by the marked decrease in PRI when clopidogrel and caffeine were coadministered. The augmenting effect of caffeine on clopidogrelinduced platelet inhibition was more robust in the healthy volunteers (cohort 1). This tendency may be explained by a number of factors. First, the patients with CAD in cohort 2 also received aspirin, in addition to clopidogrel. Most were also treated with statins, which have been shown to have antiplatelet properties. 17 The additional medications with antiplatelet effects may have blunted the ability to detect subtle effects of caffeine administration. Second, the effect of caffeine addition was tested at only 1 time point in cohort 2 (2.5 hours after caffeine). In cohort 1, significant differences in platelet aggregation were observed only 4 hours after the clopidogrel loading dose, which was 3.5 hours after the caffeine administration. Therefore, a later time point may have also yielded wider differences in platelet aggregation in cohort 2. Third, platelet activation markers, assessed by flow cytometry, may be more sensitive to Table II. Clinical characteristics and concurrent medications in cohort 2 patients (n = 40) Mean or frequency Age (y) 64.6 ± 9 Women (%) 11 (27.5%) BMI (kg/m 2 ) 28.9 ± 5 Diabetes (%) 10 (25%) Hypertension (%) 30 (75%) Hyperlipidemia (%) 26 (65%) Active smoking (%) 4 (10%) Prior MI (%) 11 (27.5%) Prior CABG (%) 8 (20%) Laboratory data Hemoglobin 13.9 ± 1.4 WBC 8.1 ± 3.5 Platelets 201 ± 63 Creatinine 1.1 ± 0.2 Concurrent medications Aspirin 325 mg (%) 33 (82.5%) 81 mg (%) 7 (17.5%) Clopidogrel 75 mg (%) 40 (100%) Statins (%) 28 (70%) β-blockers (%) 22 (55%) ACEI/ARB (%) 16 (40%) BMI, Body mass index; MI, myocardial infarction; CABG, coronary artery bypass graft; WBC, white blood cells; ACEI, angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blocker. subtle changes in platelet reactivity than platelet aggregation. Unlike turbidimetric platelet aggregation assays, flow cytometric studies assess platelets in their physiological state (ie, in whole blood) after minimal manipulation and separation, thus preventing artificial in vitro activation and possible loss of platelet subpopulations. 18 Furthermore, specific activation-dependent changes in platelets can be assessed in the present study surface P- selectin expression, PAC-1 binding to αiibβ3 integrins on activated platelets, and VASP phosphorylation state. These potential advantages of flow cytometry may explain why platelet activation markers revealed an augmenting effect of caffeine at an earlier time point than platelet aggregation in cohort 1 and, contrary to aggregation, also exhibited significant differences in cohort 2 (baseline vs caffeine). The current study has several clinical implications. First, the observed effect of caffeine on the antiplatelet effect of clopidogrel may contribute to the wide interindividual variability found in the response to clopidogrel. 14,19,20 The augmenting effect of caffeine is likely to also contribute to intra-individual variability (at different time points) in the response to clopidogrel in subjects consuming caffeine. Second, given the findings of our study, conditions for testing the response and antiplatelet effects of clopidogrel in human subjects should include refraining from caffeine-containing beverages and foods for at least 24 hours before testing. Third, other methylxanthines, such as theophylline-containing

694.e6 Lev et al American Heart Journal October 2007 Figure 4 Figure 5 Cohort 2 Platelet aggregation. Platelet aggregation in response to 5, 10, and 20 μmol/l ADP and 1 μg/ml collagen at baseline and 2.5 hours after 300 mg caffeine intake (n = 40, patients taking clopidogrel and aspirin for at least a week). Differences between the baseline and postcaffeine samples were not significant. Cohort 2 Platelet activation markers. Platelet activation markers P-selectin expression and PAC-1 binding, reflecting GP IIb/IIIa activation at baseline and 2.5 hours after 300 mg caffeine intake (n = 40, patients taking clopidogrel and aspirin for at least a week). Both activation markers were lower after caffeine intake (P =.01 for PAC-1 binding and P =.02 for P-selectin). medications, may also have a similar augmenting effect on platelet inhibition by clopidogrel. In addition, phosphodiesterase inhibitors, such as cilostazol, which inhibit camp degradation and therefore cause an increase in platelet camp levels, may also enhance platelet inhibition by clopidogrel. Although, theoretically, increased platelet inhibition is desirable after coronary stenting and/or acute coronary syndromes, it may lead to increased bleeding complications. Therefore, possible effects of other methylxanthines or phosphodiesterase inhibitors on platelet inhibition by clopidogrel require further investigation. There are several limitations to our study. Only 1 dose of caffeine was examined. This dose (300 mg) is equivalent to the caffeine content of medium-sized commercial coffee drink (16 oz) and also falls well within the range of doses tested in previous studies of acute caffeine administration. 7-9 Nonetheless, our findings are specific for the caffeine dose we used. In addition, only 1 time point 2.5 hours after caffeine administration was examined in cohort 2. This specific time point was chosen because peak blood levels of caffeine are obtained within 1 to 2 hours after an oral dose 10,11 and to enable comparison to previous studies which tested the acute effects of caffeine (alone) 1 to 3 hours after administration. 8,9 However, given the results of the first cohort, a later time point may have yielded more robust results. Finally, both cohorts were of relatively limited size. Despite these limitations, the findings of our study support a unique interaction, in which caffeine appears to potentiate the antiplatelet effects of clopidogrel. As clopidogrel is taken by millions of patients for prolonged periods, and many of them probably consume caffeine-containing beverages and foods on a regular basis, further examination is required to determine the extent, characteristics, and significance of this augmenting effect. We wish to thank Mr Tim DeLao for his excellent technical assistance and Dr. Htut K. Win for his valuable statistical analysis. References 1. Yusuf S, Zhao F, Mehta SR, et al. Effects of clopidogrel in addition to aspirin in patients with acute coronary syndromes without ST-segment elevation. N Engl J Med 2001;345:494-502. 2. Steinhubl SR, Berger PB, Mann III JT, et al. Clopidogrel for the reduction of events during observation. Early and sustained dual oral antiplatelet therapy following percutaneous coronary intervention: a randomized controlled trial. JAMA 2002;288: 2411-20. 3. Savi P, Herbert JM. Clopidogrel and ticlopidine: P2Y12 adenosine diphosphate-receptor antagonists for the prevention of atherothrombosis. Semin Thromb Hemost 2005;31:174-83. 4. Varani K, Portaluppi F, Merighi S, et al. Caffeine alters A 2a adenosine receptors and their function in human platelets. Circulation 1999;19: 2499-502. 5. Varani K, Portaluppi F, Gessi S, et al. Dose and time effects of caffeine intake on human platelet adenosine A(2A) receptors: functional and biochemical aspects. Circulation 2000;102:285-9. 6. Paul S, Kurunwune B, Biaggioni I. Caffeine withdrawal: apparent heterologous sensitization to adenosine and prostacyclin actions in human platelets. J Pharmacol Exp Ther 1993;267:838-43. 7. Cavalcante JW, Santos Jr PR, Menezes MG, et al. Influence of caffeine on blood pressure and platelet aggregation. Arq Bras Cardiol 2000;75:97-105.

American Heart Journal Volume 154, Number 4 Lev et al 694.e7 8. Ammaturo V, Perricone C, Canazio A, et al. Caffeine stimulates in vivo platelet reactivity. Acta Med Scand 1988;224: 245-7. 9. Galli C, Colli S, Gianfranceschi G, et al. Acute effects of ethanol, caffeine, or both on platelet aggregation, thromboxane formation, and plasma-free fatty acids in normal subjects. Drug Nutr Interact 1984;3:61-7. 10. Culm-Merdek KE, von Moltke LL, Harmatz JS, et al. Fluvoxamine impairs single-dose caffeine clearance without altering caffeine pharmacodynamics. Br J Clin Pharmacol 2005;60:486-93. 11. Kamimori GH, Karyekar CS, Otterstetter R, et al. The rate of absorption and relative bioavailability of caffeine administered in chewing gum versus capsules to normal healthy volunteers. Int J Pharm 2002;234:159-67. 12. Michelson AD, Barnard MR, Krueger LA, et al. Evaluation of platelet function by flow cytometry. Methods 2000;21:259-70. 13. Aleil B, Ravanat C, Cazenave JP, et al. Flow cytometric analysis of intraplatelet VASP phosphorylation for the detection of clopidogrel resistance in patients with ischemic cardiovascular diseases. Flow cytometric analysis of intraplatelet VASP phosphorylation for the detection of clopidogrel resistance in patients with ischemic cardiovascular diseases. J Thromb Haemost 2005; 3:85-92. 14. Grossmann R, Sokolova O, Schnurr A, et al. Variable extent of clopidogrel responsiveness in patients after coronary stenting. Thromb Haemost 2004;92:1201-6. 15. Gurbel PA, Bliden KP, Samara W, et al. Clopidogrel effect on platelet reactivity in patients with stent thrombosis: results of the CREST Study. J Am Coll Cardiol 2005;46:1827-32. 16. Waldmann R, Nieberding M, Walter U. Vasodilator-stimulated protein phosphorylation in platelets is mediated by camp- and cgmp-dependent protein kinases. Eur J Biochem 1987;167:441-8. 17. Bonetti PO, Lerman LO, Napoli C, et al. Statin effects beyond lipid lowering are they clinically relevant? Eur Heart J 2003;24:225-48. 18. Michelson AD. Evaluation of platelet function by flow cytometry. Pathophysiol Haemost Thromb 2006;35:67-82. 19. Gurbel PA, Bliden KP, Hiatt BL, et al. Clopidogrel for coronary stenting: response variability, drug resistance, and the effect of pretreatment platelet reactivity. Circulation 2003;23:2908-13. 20. Serebruany VL, Steinhubl SR, Berger PB, et al. Variability in platelet responsiveness to clopidogrel among 544 individuals. J Am Coll Cardiol 2005;2:246-51.