eCr

VP

0d

Impact of Anemia on Platelet Response to Clopidogrel in PatientsUndergoing Percutaneous Coronary Stenting

Catalin Toma, MDa,*, Firas Zahr, MDa, Diego Moguilanski, MDa, Sheree Grate, MSa,Roy W. Semaan, MDa, Nicole Lemieux, MDa, Joon S. Lee, MDa, Andrea Cortese-Hassett, PhDb,

Suresh Mulukutla, MDa, Sunil V. Rao, MDc, and Oscar C. Marroquin, MDa

High residual platelet reactivity (HRPR) on clopidogrel is a predictor of recurrent ischemicevents in patients undergoing percutaneous coronary interventions (PCI). Significant in-traindividual variability in platelet aggregation on repeat testing has been reported. Tounderstand factors contributing to the variability in platelet aggregation testing, we exam-ined clinical and laboratory elements linked to HRPR in 255 consecutive patients tested>12 hours after PCI using light transmission aggregometry (LTA) in response to adenosinediphosphate 5 �mol/L and VerifyNow P2Y12 assay (VNP2Y12; Accumetrics). HRPR wasdefined as >46% residual aggregation for LTA and >236 P2Y12 response units (PRUs) forVNP2Y12. On multivariate analysis the only variable independently associated with HRPRwith both LTA and VNP2Y12 was laboratory-defined anemia. Prevalences of HRPR byLTA were 34.3% in anemic patients, 15.6% in patients with normal hemoglobin levels, and59.8% versus 25.9% by VNP2Y12 (p <0.005 for the 2 comparisons). In a subgroup of 50patients, testing was done before and after the clopidogrel loading dose. At baseline there wereno differences in platelet aggregation with either assay; however, absolute decrease in reactivityafter the clopidogrel load was significantly less in anemic patients compared to patients withnormal hemoglobin (change in residual aggregation by LTA 15.8 � 5.8% vs 28.8 � 3.2%,p <0.05; change in PRU by VNP2Y12 56.5 � 35.5 vs 145.0 � 14.2 PRUs, p <0.05, respec-tively). In conclusion, anemia is an important contributor to apparent HRPR on clopidogreland may explain some of the intraindividual variability of platelet aggregation testing. © 2012

Elsevier Inc. All rights reserved. (Am J Cardiol 2012;109:1148–1153)

hsaso

The variability in the inhibitory effect of clopidogrel onplatelet aggregation has been well documented. In additionto genetic variation of cytochrome P450 enzymes involvedin the conversion of the drug to its active metabolite,1

platelet response to clopidogrel is likely affected by addi-tional factors, in particular the overall level of platelet ac-tivation, platelet turnover, and density of the P2Y12 recep-tors, which cannot be predicted by genetic testing.2 Thus, itis likely that the clinically relevant assessment would be thatof inhibition of platelet function rather than genetic testing.Large prospective clinical studies in patients with coronaryartery disease using aggregation-based measurements ofplatelet reactivity indeed have demonstrated increased re-current ischemic events in patients who are considered tohave high residual platelet reactivity on clopidogrel(HRPR), i.e., “low responders” to clopidogrel.3–5 In addi-

aHeart and Vascular Institute, University of Pittsburgh Medical Center,Pittsburgh, Pennsylvania; bInstitute for Transfusion Medicine (ItXM),Pittsburgh, Pennsylvania; cDuke University Medical Center, Cardiac Cath-terization Laboratories, Durham VA Medical Center, Durham, Northarolina. Manuscript received September 26, 2011; revised manuscript

eceived and accepted November 30, 2011.This work was supported by internal funding from the Heart and

ascular Institute, University of Pittsburgh Medical Center, Pittsburgh,ennsylvania.

*Corresponding author: Tel: 412-647-0211; fax: 412-647-8117.

lE-mail address: tomac@upmc.edu (C. Toma).

002-9149/12/$ – see front matter © 2012 Elsevier Inc. All rights reserved.oi:10.1016/j.amjcard.2011.11.049

tion to interindividual variability, recent data have indicateda significant intraindividual variability of response to clopi-dogrel.6,7 This further underscores the dynamic nature ofplatelet reactivity, in particular in the periprocedural setting.In the present study we sought to identify clinical andlaboratory variables that would predict HRPR after percu-taneous coronary interventions (PCIs) in a large cohort ofconsecutive patients with coronary artery disease undergo-ing PCI, with a longitudinal subgroup of patients testedbefore and after clopidogrel loading at the time of theprocedure.

Methods

The patients included in this study were enrolled in a pro-spective ongoing PCI registry at the University of PittsburghMedical Center from November 2008 through March 2010 inaccordance with the University of Pittsburgh institutional re-view board. Platelet function data were collected routinely inall patients undergoing PCI as part of their clinical risk assess-ment. For the analyzed cohort blood was collected the morningafter the intervention (12 to 24 hours after the clopidogrel loadat time of PCI; if eptifibatide was used, testing was done �24ours after drug discontinuation). The nursing staff was in-tructed on proper blood collection techniques for plateletggregation testing. Blood samples were couriered to the In-titute for Transfusion Medicine and processed within 4 hoursf collection. In a subgroup of patients, blood was also col-

ected at the time of PCI immediately after obtaining arterial

www.ajconline.org

mmkcpc

tlC5

arwaaU

AMCB

MWHAM

c

1149Coronary Artery Disease/Anemia and Platelet Reactivity

access and before administration of antiplatelet agents. Thelimiting factor in obtaining consecutive datasets in all thepatients who underwent PCI in this interval was the limitedwindow of availability of the reference laboratory (weekdaysonly from 8 A.M. to 2 P.M.). We included patients in whomeptifibatide was used if testing was done �24 hours after itsdiscontinuation (and that the base P2Y12 response unit [PRU]�190 for the VerifyNow P2Y12 [VNP2Y12] assay [Accumet-rics, San Diego, California] was done according to the manu-facturer’s instructions). Patients in whom abciximab was usedand patients with missing data from 1 of the assays wereexcluded.

Clinical and laboratory data were collected at the sametime. Chronic kidney disease was defined as an estimatedglomerular filtration rate �60 ml/min/1.73 m2 (calculated us-ing the Modification of Diet in Renal Disease equation). Ane-mia was defined according to our laboratory hemoglobin con-centration cutoffs �11.7 g/dl for women and �12.9 g/dl for

en. Patients’ clinical histories including a history of diabetesellitus, hypertension, stroke (cerebrovascular accident),

nown coronary artery disease, peripheral vascular disease,ancer, smoking status, current medications (including protonump inhibitor use and clopidogrel and aspirin use) wereollected prospectively at the time of presentation.

Platelet-rich plasma adjusted using platelet-poor plasmao 250,000 platelets/ml in all samples was used for standardight transmission aggregometry (LTA; PAP4, Bio/Dataorp., Horsham, Pennsylvania) with adenosine diphosphate�mol/L as the agonist. Peak residual platelet aggregation

was recorded. Parallel samples were analyzed using theVNP2Y12 assay with clopidogrel response tested using aden-osine diphosphate 20 �mol/L with prostaglandin E for selec-tive P2Y12 activation according to the manufacturer’s instruc-tions. The VNP2Y12 assay also generates a base value, whichrepresents a relatively adenosine diphosphate–independent ag-gregation response to thrombin receptor activating peptide, andit is used as an individual reference point of maximal plateletaggregation. By normalizing the aggregation amount (mea-sured in PRUs), the VNP2Y12 assay generates a “percentinhibition.” All samples were collected and processedwithin 4 hours from blood collection. Standard accepteddefinitions for HRPR using the 2 assays were used (�235PRUs for VNP2Y12 and �46% for LTA using adenosinediphosphate 5 �mol/L).8

Continuous variables are expressed as median (interquar-tile range) and categorical variables are expressed as percent-ge. To avoid the assumption of normality in the distribution ofesults of platelet testing, nonparametric statistical methodsere used when applicable. We first performed univariate

nalysis using logistic regression for categorical variablesnd Spearman rank correlation for continuous variables.nivariate predictors with a p value �0.1 were included in

a multivariate analysis using logistic regression in separatemodels with HRPR by LTA and by VNP2Y12 as outcomes.In the subgroup of patients who underwent testing beforeand after PCI, a nonparametric Mann–Whitney test wasused to compare variables between groups. Statistical sig-nificance was defined at a p value �0.05 (2-sided). Statis-tical analyses were performed with STATA 10.0 (STATACorp., College Station, Texas) and Prism 5 (GraphPad,

LaJolla, California).

Results

In total 255 consecutive patients who had platelet testingresults by LTA and VNP2Y12 after PCI were analyzed; ofthese 50 clopidogrel-naive patients had testing done beforeand after PCI. Baseline characteristics of patients enrolledare presented in Table 1. In total 30.6% of patients werealready on long-term clopidogrel therapy at time of presenta-tion, and 70.5% of these patients received an additional loadingdose of clopidogrel at time of PCI (median 300 mg, interquar-tile range 300 to 600). All clopidogrel-naive patients weretreated with a loading dose of clopidogrel in the peri-PCIsetting (median 600 mg, interquartile range 600 to 600).

Platelet aggregation by LTA in response to adenosinediphosphate correlated well with the VNP2Y12 assay(Spearman r � 0.64, p �0.0001). The overall proportions ofpatients with HRPR in our entire studied population were23.4% by LTA and 40.4% by VNP2Y12. Univariate anal-ysis was then performed for all clinical and laboratoryvariables collected as listed in Table 1 and using plateletreactivity testing results as a categorical variable. Onlyvariables associated with HRPR on univariate analysis witha p value �0.1 were then included in a multivariate logisticregression in 2 separate models for LTA and VNP2Y12(Figure 1). Goodness of fit in these 2 models was thenconfirmed using a Hosmer–Lemeshow test (p � 0.32 and0.83 for VNP2Y12 and LTA, respectively, with p �0.05

Table 1Baseline clinical characteristics (n � 255)

Characteristic Number ofPatients (%)

Median (IQR)

ge (years) 65 (57–73)en 171 (67.1%)aucasian 242 (94.9%)ody mass index (kg/m2) 28.7 (25.5–33.1)

Smokers 45 (17.6%)Diabetes mellitus 87 (34.1%)Known coronary artery disease 122 (47.8%)Chronic kidney disease* 69 (27.1%)Hypertension 212 (83.5%)Hypercholesterolemia 186 (72.9%)Cancer 30 (11.8%)Congestive heart failure 54 (21.2%)Stroke 13 (5.1%)Peripheral vascular disease 41 (16.1%)Presentation as myocardial infarction 93 (36.5%)Eptifibatide use 31 (12.2%)Bivaluridin use 123 (48.8%)Platelets (106/ml) 189 (158–234)

ean platelet volume (fL) 8.1 (7.6–8.7)hite blood cell count (106/ml) 8.0 (6.4–10.0)emoglobin (g/dl) 12.7 (11.5–14.1)nemia 95 (37.3%)edications on presentation

Long-term clopidogrel 78 (30.7%)Long-term aspirin 176 (69.0%)Statin 152 (59.6%)Proton pump inhibitor 82 (32.3%)

IQR � interquartile range.* Estimated glomerular filtration rate �60 ml/min/1.73 m2 by Modifi-

ation of Diet in Renal Disease equation.

indicating lack of fit). For the LTA model only anemia,

ahaan2

Eci

1150 The American Journal of Cardiology (www.ajconline.org)

gender, and administration of a clopidogrel load at time ofPCI emerged as independent predictors of platelet reactivity(Figure 1). For the VNP2Y12 logistic regression model theonly independent variables for HRPR were anemia anddiabetes mellitus (Figure 1). Ultimately, the only commondenominator between these assays that was independentlypredictive of HRPR was anemia. Using LTA proportions ofpatients with HRPR were 34.3% in anemic patients and15.6% in patients with normal hemoglobin levels. ForVNP2Y12 prevalences of HRPR were 59.8% in anemicpatients and 25.9% in patients with normal hemoglobin(p �0.005 for the 2 comparisons).

We also investigated whether anemia influences plateletresponse to clopidogrel at an individual patient level usingthe subgroup of clopidogrel-naive patients who had testingdone before and after clopidogrel therapy initiation. Of

Figure 1. Results of multivariate analysis for light transmission aggregomresidual platelet reactivity (defined as �46% residual aggregation by light trwith a p value �0.01 on univariate analysis are included. Hazard ratios and

lements significantly associated with high residual platelet reactivity on mulhronic kidney disease; CVA � cerebrovascular accident (stroke); DM � dinfarction; PPI � proton pump inhibitor.

these 14 were anemic (average hemoglobin 11.3 � 0.3 g/dl) b

and 36 patients had normal hemoglobin (14.2 � 0.2 g/dl)based on postprocedure hemoglobin values. In the 2 sub-groups before clopidogrel administration there was no dif-ference in percent aggregation by LTA or PRU valuesbetween anemic and nonanemic patients. However, thechange in platelet response to adenosine diphosphate wassignificantly decreased in anemic patients compared to pa-tients with normal hemoglobin with either assay (Figure 2).Percent maximal aggregation by LTA in response to aden-osine diphosphate 5 �mol/L decreased by 15.8 � 5.8% innemic patients and by 28.8 � 3.2% in patients with normalemoglobin (p �0.05; Figure 2). Likewise, change in PRUsfter the thienopyridine load was lower in patients withnemia (56.5 � 35.5 PRUs) compared to patients withormal hemoglobin (145.0 � 14.2 PRUs, p �0.05; Figure). The same was observed when comparing percent inhi-

) and VerifyNow P2Y12 assay (B). Only elements associated with highsion aggregometry and �236 P2Y12 response units by VerifyNow P2Y12)onfidence intervals for multivariate analysis are listed for each parameter.analysis for each assay (black circles) are shown. ASA � aspirin; CKD �ellitus; HTN � history of hypertension; MI � presentation as myocardial

etry (Aansmis95% c

tivariateabetes m

ition by the VNP2Y12 assay, which was significantly

pPl denosin

1151Coronary Artery Disease/Anemia and Platelet Reactivity

higher in patients with normal hemoglobin versus anemicpatients (43.0 � 4.1% vs 16.9 � 8.8%, p �0.05). Beforeclopidogrel there was a very weak linear correlation be-tween hemoglobin levels before the procedure and percentmaximum aggregation by LTA (Spearman r � �0.23, p �0.1; Figure 3) or PRUs by VNP2Y12 (Spearman r � �0.30,p � 0.04; Figure 3) before the procedure. After the clopi-dogrel load, however, a significant correlation of plateletaggregation response to adenosine diphosphate relative tohemoglobin levels after the procedure was present with the2 assays (LTA, Spearman r � �0.37, p � 0.006;VNP2Y12, Spearman r � �0.46, p � 0.0007; Figure 3).

Figure 2. Comparison between results of light transmission aggregometclopidogrel-naive patients who underwent testing before and after a clpercutaneous coronary intervention were comparable between anemic (ercutaneous coronary intervention there was significantly lower platelet i2Y12 assay (C), and VerifyNow P2Y12 percent inhibition (E) (p �0.005

oad was also significantly lower in anemic patients (p �0.05). ADP � a

The same robust correlation between hemoglobin and LTA

or VNP2Y12 results was obtained after PCI when using theentire dataset (Spearman r � �0.25 for LTA, Spearman r ��0.45 for VNP2Y12, p �0.0001; Figure 3).

Discussion

To understand patient-related elements that lead to indi-vidual HRPR, we performed a multivariate analysis of clin-ical and laboratory parameters in an unselected cohort ofconsecutive patients undergoing PCI including a significantproportion of patients with acute coronary syndromes. Uni-variate analyses for the 2 assays used agree with previous

VerifyNow P2Y12 assay as a continuous variable in a subgroup of 50rel load at time of percutaneous coronary intervention. Results beforeymbols) and patients with normal hemoglobin (closed symbols). Aftern in anemic patients by light transmission aggregometry (A), VerifyNow). (B, D, F) Absolute decrease in platelet aggregation after the clopidogrele diphosphate; Hb � hemoglobin.

ry andopidogopen snhibitioto 0.01

data. Female gender, hypertension, diabetes mellitus,

oicVhapapma

aaa

car

1152 The American Journal of Cardiology (www.ajconline.org)

chronic kidney disease, and use of proton pump inhibitorswere also associated with HRPR using these 2 aggregation-based assays in previous larger cohorts of patients.4–6,9 Inour multivariate analysis using either assay, anemiaemerged as the strongest independent predictor of apparentHRPR after PCI, an effect that has been noted by oth-ers.4,10,11 The impact of anemia on HRPR is further under-scored in our study by the comparison between before andafter thienopyridine loading at an individual patient level,indicating that the individual response to clopidogrel loadwas decreased in anemic patients.

The key question is whether the observed impact ofanemia on results of the aggregation-based assays is a truebiological phenomenon occurring in vivo or an artifactualeffect on the platelet function tests employed. Whole bloodaggregometry can be influenced by a low hematocrit,12

which would render the blood more optically transparentand thus artifactually lead to an apparent increased aggre-gation. This is less likely for LTA, which employs platelet-

Figure 3. Correlation between hemoglobin level and platelet aggregation beunderwent testing before and after a clopidogrel load at time of percutaneoubefore thienopyridine administration is nonsignificant for light transmissionassay (B) (p � 0.04), whereas after the load there is a significant associatioorrelation �0.001). A similar correlation between hemoglobin and resultst the entire cohort of patients after percutaneous coronary interventionegression for each dataset is shown (continuous line).

rich plasma devoid of erythrocytes. Furthermore, changes in i

hematocrit can affect calcium concentration in citratedblood used for the 2 assays, resulting in decreased adenosinediphosphate–induced aggregation at higher hematocrit val-ues13,14; however, these differences are likely meaningfulnly over wide hematocrit values. Despite these concerns,n our subgroup of patients tested before administration oflopidogrel, there was no difference in the results of LTA orNP2Y12 in anemic patients versus those with normalemoglobin; however, after the clopidogrel load plateletggregation was markedly higher in anemic patients com-ared to patients with normal hemoglobin. This key findingnd the common effect of anemia on the 2 assays raise theossibility that the effect of anemia is not artifactual butay be reflective of a true biological interaction between

nemia and response to clopidogrel.Our study is limited in that it does not provide a mech-

nism for the association between anemia and platelet re-ctivity. Increased erythropoietin levels have been shown toctivate platelets15,16 and reticulated immature platelets are

pidogrel loading and �12 hours after loading in subgroup of patients whoary intervention. Correlation between hemoglobin and platelet aggregationometry (A) (p � 0.1) and of borderline significance for VerifyNow P2Y12moglobin and residual platelet aggregation with the 2 assays (C, D) (p fortelet testing (expressed as a continuous variable) was found when lookinght transmission aggregometry; F, VerifyNow P2Y12 assay). The linear

fore clos coronaggregn of heof pla(E, lig

ncreased in anemic patients,17 although the lack of corre-

arde

bdaabt

AacL

1

1

1

1

1

1

1

1

1

2

2

1153Coronary Artery Disease/Anemia and Platelet Reactivity

lation with mean platelet volume, a less robust marker ofthrombocytosis, in our study does not lend support to thisassociation. Because the 1/2 life of the active metabolite ofclopidogrel in plasma is relatively short, the proportion ofyounger platelets generated after the clopidogrel loadingdose (�12 hours in our analysis) may not be inhibited bythe drug. Alternatively, the metabolism of clopidogrel to itsactive metabolite may be different in anemic patients. Re-gardless of the basis for this interaction, however, the effectof anemia in our study was major: the proportion of patientswith apparent HRPR was 2 times in anemic patients com-pared to patients with normal hemoglobin. It is also impor-tant to note that degree of anemia in our study was relativelymild, with median hemoglobin in this group of 11.3 g/dl.

There are several aspects in which the influence of anemiaon the present results of aggregation-based platelet testing maybear clinical relevance. First, 2 recent large studies have re-cently demonstrated that up to 50% of patients with initialHRPR after PCI had improved platelet response at 30-dayfollow-up despite no change in clopidogrel dose.6,7 Many ofthese patients may have had mild anemia on presentation thatresolved at follow-up. Second, multiple studies have indicatedthat anemia is strongly associated with increased recurrentischemic events and cardiac death.17–20 Given the interactionbetween red blood cells and platelets,21 it is plausible thatnemia influences platelet reactivity in vivo. The decreasedesponse to clopidogrel in anemic patients demonstrated by ourata may represent the basis for the increased recurrent isch-mic events in anemic patients.

Further studies are required to explore the mechanistic linketween anemia and platelet aggregation. Nevertheless, ourata provide evidence that even mild anemia is independentlyssociated with a significant increase in apparent HRPR todenosine diphosphate on clopidogrel when using aggregation-ased assays. This represents an important consideration toake into account when performing platelet-testing after PCI.

cknowledgment: We thank Rosalyn Rapsinski, MA, forssistance with data collection and our clinical research co-oordinators Mellisa Enlow, MSN, Laurie Dennis, BS, andisa Baxendel, BS.

1. Erlinge D, Varenhorst C, Braun OO, James S, Winters KJ, JakubowskiJA, Brandt JT, Sugidachi A, Siegbahn A, Wallentin L. Patients withpoor responsiveness to thienopyridine treatment or with diabetes havelower levels of circulating active metabolite, but their platelets respondnormally to active metabolite added ex vivo. J Am Coll Cardiol2008;52:1968–1977.

2. Sweeny JM, Gorog DA, Fuster V. Antiplatelet drug “resistance”. Part1. Mechanisms and clinical measurements. Nat Rev Cardiol 2009;6:273–282.

3. Marcucci R, Gori AM, Paniccia R, Giusti B, Valente S, Giglioli C,Buonamici P, Antoniucci D, Abbate R, Gensini GF. Cardiovasculardeath and nonfatal myocardial infarction in acute coronary syndromepatients receiving coronary stenting are predicted by residual plateletreactivity to ADP detected by a point-of-care assay: a 12-monthfollow-up. Circulation 2009;119:237–242.

4. Breet NJ, van Werkum JW, Bouman HJ, Kelder JC, Ruven HJ, BalET, Deneer VH, Harmsze AM, van der Heyden JA, Rensing BJ,Suttorp MJ, Hackeng CM, ten Berg JM. Comparison of platelet func-

tion tests in predicting clinical outcome in patients undergoing coro-nary stent implantation. JAMA 2010;303:754–762.

5. Price MJ, Endemann S, Gollapudi RR, Valencia R, Stinis CT, LevisayJP, Ernst A, Sawhney NS, Schatz RA, Teirstein PS. Prognostic sig-nificance of post-clopidogrel platelet reactivity assessed by a point-of-care assay on thrombotic events after drug-eluting stent implantation.Eur Heart J 2008;29:992–1000.

6. Price MJ, Berger PB, Teirstein PS, Tanguay JF, Angiolillo DJ, SpriggsD, Puri S, Robbins M, Garratt KN, Bertrand OF, Stillabower ME,Aragon JR, Kandzari DE, Stinis CT, Lee MS, Manoukian SV, CannonCP, Schork NJ, Topol EJ; GRAVITAS Investigators. Standard- vshigh-dose clopidogrel based on platelet function testing after percuta-neous coronary intervention: the gravitas randomized trial. JAMA2011;305:1097–1105.

7. Campo G, Parrinello G, Ferraresi P, Lunghi B, Tebaldi M, Miccoli M,Marchesini J, Bernardi F, Ferrari R, Valgimigli M. Prospective evaluationof on-clopidogrel platelet reactivity over time in patients treated withpercutaneous coronary intervention relationship with gene polymor-phisms and clinical outcome. J Am Coll Cardiol 2011;57:2474–2483.

8. Bonello L, Tantry US, Marcucci R, Blindt R, Angiolillo DJ, Becker R,Bhatt DL, Cattaneo M, Collet JP, Cuisset T, Gachet C, Montalescot G,Jennings LK, Kereiakes D, Sibbing D, Trenk D, Van Werkum JW,Paganelli F, Price MJ, Waksman R, Gurbel PA; Working Group onHigh On-Treatment Platelet Reactivity. Consensus and future direc-tions on the definition of high on-treatment platelet reactivity to aden-osine diphosphate. J Am Coll Cardiol 2010;56:919–933.

9. Price MJ, Nayak KR, Barker CM, Kandzari DE, Teirstein PS. Predic-tors of heightened platelet reactivity despite dual-antiplatelet therapyin patients undergoing percutaneous coronary intervention. Am J Car-diol 2009;103:1339–1343.

10. Elsenberg EH, van Werkum JW, van de Wal RM, Zomer AC, BoumanHJ, Verheugt FW, Berg JM, Hackeng CM. The influence of clinicalcharacteristics, laboratory and inflammatory markers on “high on-treatment platelet reactivity” as measured with different platelet func-tion tests. Thromb Haemost 2009;102:719–727.

1. Voisin S, Bongard V, Tidjane MA, Lhermusier T, Carrié D, Sié P. AreP2Y12 reaction unit (PRU) and % inhibition index equivalent for theexpression of P2Y12 inhibition by the VerifyNow assay? Role of haema-tocrit and haemoglobin levels. Thromb Haemost 2011;106:227–229.

2. Cecchi E, Marcucci R, Paniccia R, Bandinelli B, Valente S, Giglioli C,Lazzeri C, Gensini GF, Abbate R, Mannini L. Effect of blood hemat-ocrit and erythrocyte deformability on adenosine 5=-diphosphate plate-let reactivity in patients with acute coronary syndromes on dual anti-platelet therapy. Am J Cardiol 2009;104:764–768.

3. Bell DN, Spain S, Goldsmith HL. Extracellular-free Ca�� accountsfor the sex difference in the aggregation of human platelets in citratedplatelet-rich plasma. Thromb Res 1990;58:47–60.

4. Hardisty RM, Hutton RA, Montgomery D, Rickard S, Trebilcock H.Secondary platelet aggregation: a quantitative study. Br J Haematol1970;19:307–319.

5. Dale GL, Alberio L. Is there a correlation between raised erythropoietinand thrombotic events in sickle-cell anaemia? Lancet 1998;352:566–567.

6. Tàssies D, Reverter JC, Cases A, Calls J, Escolar G, Ordinas A. Effectof recombinant human erythropoietin treatment on circulating reticu-lated platelets in uremic patients: association with early improvementin platelet function. Am J Hematol 1998;59:105–109.

7. Manoukian SV, Feit F, Mehran R, Voeltz MD, Ebrahimi R, Hamon M,Dangas GD, Lincoff AM, White HD, Moses JW, King SB III, OhmanEM, Stone GW. Impact of major bleeding on 30-day mortality andclinical outcomes in patients with acute coronary syndromes: an anal-ysis from the acuity trial. J Am Coll Cardiol 2007;49:1362–1368.

8. Tsujita K, Nikolsky E, Lansky AJ, Dangas G, Fahy M, Brodie BR, DudekD, Möckel M, Ochala A, Mehran R, Stone GW. Impact of anemia onclinical outcomes of patients with st-segment elevation myocardial infarc-tion in relation to gender and adjunctive antithrombotic therapy (from theHORIZONS-AMI trial). Am J Cardiol 2010;105:1385–1394.

9. Sattur S, Harjai KJ, Narula A, Devarakonda S, Orshaw P, Yaeger K.The influence of anemia after percutaneous coronary intervention onclinical outcomes. Clin Cardiol 2009;32:373–379.

0. Greenberg G, Assali A, Vaknin-Assa H, Brosh D, Teplitsky I, FuchsS, Battler A, Kornowski R, Lev EI. Hematocrit level as a marker ofoutcome in st-segment elevation myocardial infarction. Am J Cardiol2010;105:435–440.

1. Aarts PA, Heethaar RM, Sixma JJ. Red blood cell deformability

influences platelets—vessel wall interaction in flowing blood. Blood1984;64:1228–1233.