C

A

1

gtmgtbStsoreEni1Eati

taawprfwi4r9wwavtio

bo

0d

Leukemia Research 35 (2011) 286–289

Contents lists available at ScienceDirect

Leukemia Research

journa l homepage: www.e lsev ier .com/ locate / leukres

ommissioned Editorial Debates

spirin in low-risk essential thrombocythemia, not so simple after all?

. Harrison

Despite landmark advances in our understanding of the patho-enesis of what are now termed myeloproliferative ‘neoplasms’here have been relatively few significant advances in their treat-

ent. Indeed some conflicting opinions reflected in nationaluidance or consensus documents also exist an example beinghe use of aspirin for patients with “low risk” essential throm-ocythemia (ET). The recently published British Committee fortandards in Haematology (BCSH) “Guidelines for the investiga-ion and management of thrombocytosis in adults and children”uggest all patients should be treated with low dose aspirin unlesstherwise contraindicated but recognise that the strength of theecommendation is only evidence level IIb grade B [1]. In fact thevidence for the benefit of aspirin in ET is based mostly upon theCLAP study in patients with polycythemia vera [2]. The BCSH defi-ition of low risk ET also differs from that previously published and

s patients aged less than 40 years with a platelet count of less than500 × 109/L, and no prior thrombosis or haemorrhage related toT. Other authors have suggested a higher upper age limit and it islso unclear whether the presence of indisputable vascular risk fac-ors (such as hypertension or diabetes requiring pharmacotherapy)s consistent with a low risk category [3].

A recent publication from Alvarez-Lavaran et al. [4] ques-ions the benefit of low dose aspirin in low risk ET. Theseuthors retrospectively assessed, the incidence rates of arterialnd venous thrombosis in 300 low-risk ET patients either treatedith antiplatelet drugs as monotherapy (n = 198, follow up 802erson-years) or observation only (n = 102, 848 person years). Theyeport that while overall rates of thrombotic events did not dif-er between these patient groups two subgroups of patients didorse with observation only: JAK2 V617F-positive patients had an

ncreased risk of venous thrombosis (incidence rate ratio [IRR]:.0; 95% CI: 1.2–12.9; p = 0.02); and patients with cardiovascularisk factors had increased rates of arterial thrombosis (IRR: 2.5;5% CI: 1.02–6.1; p = 0.047). An increased risk of major bleedingas observed in patients with platelet count >1000 × 109/L treatedith antiplatelet therapy (IRR: 5.4; 95% CI: 1.7–17.2; p = 0.004). The

uthors conclude that antiplatelet therapy reduces the incidence ofenous thrombosis in JAK2-positive patients and the rate of arterialhrombosis in patients with associated cardiovascular risk factors;n the remaining low-risk patients observation may be an adequate

ption.

For patients without haematological disease low dose aspirin isoth of definite and substantial benefit for secondary preventionf occlusive vascular disease but evidence of its role in primary

145-2126/$ – see front matter. Crown Copyright © 2010 Published by Elsevier Ltd. All rioi:10.1016/j.leukres.2010.10.009

prevention of vascular occlusive disease is controversial with therisks of haemorrhage outweighing any benefit in several stud-ies. The most recent of these studies was a meta-analysis fromthe Antithrombotic Trialists Collaboration of low dose aspirin forprimary prevention in healthy volunteers in which the rates ofthrombosis were 5.4 per 1000 person-years and yet in this settingaspirin yielded a 12% proportional reduction in serious vascularattacks (0.51% aspirin vs 0.57% control, per year, p = 0.0001) duemainly to a reduction in non-fatal myocardial infarction but thiswas offset by increased major gastrointestinal and extracranialbleeds (0.1% vs 0.07% per year, p < 0.0001) [5].

For patients with ET where there is a risk of the paradoxicalcombination of a bleeding tendency in addition to an enhancedrisk of occlusive vascular disease, the role of aspirin has tradition-ally been controversial with the documented ability of aspirin (atdoses of 300 mg thrice daily) to uncover a latent bleeding tendency[6]. However subsequent studies of low dose aspirin (100 mg/day)in ET documented its safety (if platelets <1000 × 109/L and no priorhaemorrhage) as well as potential ability to reduce arterial throm-botic complications [7]. A recent Cochrane review for both ET andPV concluded that the use of aspirin was associated with a sta-tistically non-significant reduction in the risk of fatal thromboticevents, without an increased risk of major bleeding, when com-pared with no treatment in patients with PV who have no clearindication or contraindication to aspirin therapy; however therewas no adequate evidence for ET [8]. The other potential and yetalso controversial benefits of low dose aspirin such as prevention ofcolorectal polyp recurrence, breast cancer recurrence, preventionof dementia among others (vida infra) are less likely to be pertinentto young low risk ET patients.

While the results of the study from the Spanish group areprovocative the data are limited by its retrospective nature, thepotential bias in that it is unclear why certain patients wereallocated “antiplatelet” agents or not, and that the “antiplatelet”agents used were not uniformly low dose aspirin. Furthermorethe rates of thrombotic events were 21.2 and 17.7 per 1000person-years for antiplatelet therapy and observation, respectively(p = 0.6). These are lower than those previously reported from twohaematology institutions in France (20 patients, total thromboticcomplications 5.1%/patient year) and Italy (40 cases, 4.1%/patientyear) but there is variability in the literature with another studysuggesting the rates of thrombosis in low risk patients was

not statistically different from controls (1.91%/patient year vs1.5%) (reviewed in [3,9,10]). Meanwhile an on-going prospec-tive observational study in low risk ET (http://clinicaltrials.gov/ct2/show/NCT00175838) should further inform the question of

ghts reserved.

Leuke

sE

caa

2

modparPm

1

Commissioned Editorial Debates /

afety and efficacy of low dose aspirin in the setting of low riskT.

However several interesting questions and challenges for clini-al practice remain such as the management of the phenomena ofspirin resistance and intolerance and for which patients low dosespirin should be contraindicated.

. Barbui

The first step in deciding whether to recommend aspirin for pri-ary prophylaxis in Essential Thrombocythemia is an assessment

f whether the annual risk for that individual of developing a car-iovascular event exceeds the rate of harm. The balance betweenreventing vascular occlusion and causing excess bleeding withspirin establishes the net benefit. Translating evidence from theesults of ECLAP (European Collaboration on Low-dose Aspirin inolycythemia) study [3] in Polycythemia Vera (PV) patients to ETay be questionable for at least two main reasons

. The thrombotic risk in PV is higher than in ET. The populationinvolved in the ECLAP trial was selected on the basis of uncer-tain indication of aspirin in PV at low-cardiovascular risk. Thus,ECLAP should be considered a primary prophylaxis trial. In fact,those patients with a clear indication for aspirin prophylaxiswere excluded. Only 532 out of 1638 subjects were randomisedto receive 100 mg ASA or placebo. After a follow-up of about3 years, there was a statistically significant 59% reduction ofmajor thromboses (both arterial and venous) without a signif-icant increase of hemorrhagic complications in the ASA group.These results suggest that low dose ASA should be used in all PVsubjects having no contraindication to this treatment. Of note isthat all the 1638 ECLAP patients were followed in an observa-tional study [11] that prospectively evaluated the vascular riskof several subgroups and the impact of some factors such as ageand thrombotic history on this risk. The vascular risk in asymp-tomatic patients younger than 65 years was around 3% patientsper year. In the subjects with a previous thrombosis and in thoseaged >65 was approximately 6% patients per year while thosewith a combination of these two factors had a vascular risk ashigh as 10% patients per year, in spite of the standard cytore-ductive treatment as well as of the administration of aspirin.These risk levels are comparable to those reported in subjectswith a recent acute coronary or cerebrovascular event [12]. InET no large prospective studies have determined the rates ofincident major vascular events. Major thrombosis and risk fac-tors were evaluated in a large retrospective cohort of 1063 ETpatients in whom the role of patient related (age and previ-ous vascular events) factors upon vascular risk was ascertained[13,14]. Patients were classified as being at low- or high-risk forthrombosis according to standard risk factors (age ≥60 yearsand/or a previous major thrombotic event). Low risk patients(N = 517, 49%) were followed with no cytoreductive therapywhereas high-risk patients (N = 546, 51%) were given hydrox-yurea (HU) in the great majority (90%) The target of therapy wasto keep platelet number below 600 × 109/L. Low dose aspirin(100 mg daily) was prescribed in 703 patients (66%) accordingto the inclination of the treating physician. During up to 38years of follow up (median 4.8 years), 118 major thrombosis(2.3%/patients/year) were objectively diagnosed and included48 ischemic cerebral strokes or transient ischemic attacks. In

low risk the rate of vascular complications was around 1.5–2.0%patients per year. Multivariable analysis confirmed that ageand previous thrombosis were independent factors for occlu-sive events (HR = 1.7, 95% CI = 1.1–2.6, p = 0.01). Thus, the rate ofmajor thrombosis to be prevented in ET is significantly lower

mia Research 35 (2011) 286–289 287

than in PV. This notion should be considered when evaluatingthe balance between efficacy and safety.

2. The rate of bleeding in ET is reported to be higher than in PV[15]. In the whole ECLAP population prospectively followed inthe observational study (Marchioli et al. [11]) the rates of totalhemorrhages and major bleeding were 2.9 and 0.8 events per 100persons per year, respectively. Risk factors for bleeds were age,disease duration, use of antiplatelet agents and previous historyof bleeding. Thus, there is no point in debating the utility of low-dose aspirin in PV since the net risk/benefit ratio is favorable inall risk categories.

In ET, serious hemorrhagic problems can occur at any ageincluding children [16] and aspirin may unmask a bleedingtendency in patients with severe functional platelet defects asdemonstrated in ET [17]. The overall rate of severe bleeding inuntreated patients is 0.6% person-years while it becomes 1.26%year if patients are treated long-term with aspirin. These ratesapply to asymptomatic and younger than 60 year patients, acategory otherwise at low risk for thrombosis. In these patients, arecent retrospective analysis [4] reported that only patients withcardiovascular risk factors or with JAK2 V617F mutation, had afavorable risk/benefit ratio by low-dose aspirin. These authorsconcluded that aspirin should not be given as primary prophylaxisin all the category of low risk (age <60 years and no prior throm-bosis). Furthermore, thrombocytosis (>1.0 million × 109/L) wasassociated with bleeding so that these patients require a carefulattention for risks of possible life threatening hemorrhage. Thus,evidence is insufficient to assess the net balance of benefits andharms of low-dose aspirin in primary prevention of major vascularevents in younger and asymptomatic ET patients.

3. Harrison

We agree that there is a lack of evidence for the use of aspirin inpatients with ET and also that we should be cautious in extrap-olating results of studies performed in PV to patients with ET.However it is interesting that the limited data comparing patientswith ET and the JAK2 V617F mutation with those diagnosed withPV suggests that these disease entities are in fact very similar withindistinguishable event rates for thrombosis [18,19]. The num-ber of low risk patients in these studies is unclear and we shouldexplore this idea in more detail. This similarity between JAK2 V617Fpositive-ET and PV in combination with the finding from the Span-ish group that low-dose aspirin may be of particular benefit in JAK2V617F positive patients is also interesting. However it is curiousthat a meta-analysis of the influence of the JAK2 V617F muta-tion upon thrombotic risk in ET was required as the magnitudeof increased risk was small [20].

The issue of what constitutes low risk ET remains a difficultyand while different definitions are used this will cause on-goingconfusion for treating clinicians. In the UK those patients with vas-cular risk factors would certainly have qualified as high risk patientsin the PT-1 trial [21] and internationally may in fact be regardedas intermediate risk patients [3]. Thus the definition of low riskET merits further research (and ideally consensus) as indeed doesidentification of factors that could substratify low risk patients intohigher and lower risk. Candidate factors include leucocyte count,JAK2 V617F allele burden as well as others but the role of the formertwo remains unproven.

You mention some other areas for discussion which I think

are also important, they include: When is low dose aspirin con-traindicated? What are the risk factors for bleeding in ET? Canwe define patients who are likely to be aspirin resistant? In thoselatter patients is the best treatment dual antiplatelet therapy orcytoreductive drugs with low dose aspirin?

2 Leuke

ut1bnubf(Ateaodcdatbaa7

dsrwpegmieofifif[woiwnbtcrwsfaswtsde

4

am

88 Commissioned Editorial Debates /

The BCSH guideline group suggests that aspirin should besed with caution in patients with a history of bleeding and inhose with extreme thrombocytosis (platelet count in excess of000–1500 × 109/L) [1] especially in the context of a bone marrowiopsy [22]. The presence of acquired von Willebrand’s disease doesot predict bleeding as such but should suggest caution with these of even low dose aspirin. Furthermore whether aspirin shoulde used in combination with warfarin where the indication for war-arin is not secondary prevention of arterial or venous thrombosise.g. for atrial fibrillation or valvular heart disease) is unknown.dditionally aspirin should be used with care in children due to

he risks of Reye’s syndrome. This risk of Reye’s syndrome is great-st in younger children (<15 years), and those on high doses ofspirin (>20–45 mg/kg) [23]. Some believe that there is no safe dosef aspirin in children, although few cases have been reported withoses <10 mg/kg per day [24]. Aspirin should also be used withaution in those with history of prior peptic ulcer disease. Currentata suggest that such patients however benefit from aspirin inddition to a proton pump inhibitor (such as esomeprazole) ratherhan use of clopidogrel where the cumulative incidence of recurrentleeding during a 12 month period was 8.6% (95% CI: 4.1–13.1%)mong patients who received clopidogrel and 0.7% (95% CI: 0–2.0%)mong those who received aspirin plus esomeprazole (difference:.9% points; 95% CI: 3.4–12.4; p = 0.001) in the pivotal study [25].

Regarding so-called aspirin resistance this is a subject whoseetail is beyond the scope of our correspondence but has been theubject of much debate and heralded by some as an “elephant in theoom”. It is important however to acknowledge that some patientsill sustain thrombotic events despite aspirin therapy and theseatients may be “aspirin resistant” due to potential platelet or dis-ase related factors such as impaired sensitivity of COX-1; or COX-1ene polymorphisms; or platelet stimulation by aspirin insensitiveechanisms, e.g. shear stress; COX-2 expression, etc. (reviewed

n [26]). Two additional potential explanations for lower thanxpected response to low-dose aspirin have been identified bothf which are important factors for clinicians to acknowledge. Therst being compliance with therapy; indeed data outside the MPNeld suggests that up to 40% of patients with cardiovascular risk

actors are unable/unwilling to tolerate regular low-dose aspirin27]. This finding is consistent with other estimates suggesting thatithout special efforts, patients follow prescribed dosage regimens

nly about 50% of the time. Secondly the recently characterizednteraction with non-steroidal anti-inflammatory drugs (NSAIDs)

here concomitant administration of aspirin with ibuprofen oraproxen is now known to result in competition for a commoninding site on COX-1 that prevents aspirin from gaining accesso, and acetylating, the target serine [28]. This can readily be over-ome by separating the administration of the two drugs, but thisequires knowledge and patient education. Hence the situationith regard to treating “aspirin resistance” as manifest by thrombo-

is occurring in the face of apparent aspirin therapy is not straightorward and will involve addressing issues such as compliancend concomitant medications. However most patients who demon-trate clinical aspirin resistance and as such develop a thrombosisill then merit cytoreductive therapy. If they then have a further

hrombotic event then consideration should probably be given toearching for uncontrolled additional vascular risk factors, use ofual antiplatelet therapy (unproven safety and little data regardingfficacy in ET) or more aggressive cytoreduction.

. Barbui

My conclusions are that the risk-benefit ratio of low-dosespirin for primary prophylaxis of major vascular events (stroke,yocardial infarction, peripheral arterial thrombosis and venous

mia Research 35 (2011) 286–289

thromboemolism) in asymptomatic and younger than 60 year ETpatients remains unclear. This issue is becoming more and morerelevant as asymptomatic and young patients are now being diag-nosed with higher frequency. As suggested by Girodon et al. [29] theincrease of ET diagnosis was noted after 2005, mainly due to the useof WHO criteria that include JAK2 V617F determination and lowerthreshold of platelet count from 600 × 109/L to 450 × 109/L. Thereare now data showing that in these low risk patients, subgroups atdifferent thrombotic risk can be identified.

Whether in the subgroup with hereditary thrombophilic states,such as the congenital deficiencies of natural anticoagulants(antithrombin, protein C and protein S) and genetic mutations (fac-tor V Leiden and prothrombin G20210A) the thrombotic risk isincreased remains uncertain. Italian guidelines [3] suggest to con-sider for prophylaxis only patients with a clear family history ofthrombosis. In the recent Spanish paper [4], it was recognized thatthe ET low-risk subgroup presenting with concomitant cardiovas-cular factors such as hypertension, diabetes and dyslipidemia hada net advantage by the use of antiaggregating agents. Moreover,recent studies demonstrated that novel disease related biologicalfactors, including JAK2 V617F mutational status and/or mutationalburden and leukocytosis, might be independent predictors of futuremajor thrombosis. This was also shown in low-risk ET patients[30–33] in whom leukocytosis, over the median value (8.7 × 109)was associated with occurrence of major thrombosis at significantlyhigher frequency than in cases with lower leukocyte values (rela-tive risk 2.7, 95% confidence interval 1.2–6.4, p = 0.01).

Interestingly, the prognostic role of leukocytosis was indepen-dent from platelet counts suggesting that, in so-called low risk ET,subgroups at higher thrombotic risk exist, in which one should testthe hypothesis that low dose aspirin might lead to a reduction ofmajor vascular events without increasing the hemorrhagic com-plications. In the mean time, a safer alternative is an aggressivemanagement of modifiable thrombotic risk factors such as diabetes,hypertension, obesity, metabolic syndrome and patients should berequested to stop smoking.

If, in spite of these uncertainties, the physician decision is toprescribe aspirin, patients should be informed about the possibledrug-related bleeding, stressing that this is significantly associatedwith age, previous history of gastrointestinal bleeds and extremethrombocytosis. It should be emphasized that, for the time being,the net benefit of low-dose aspirin is favorable only in high riskpatients who already had major vascular events. The combinationof low-dose aspirin and anagrelide [21] was demonstrated to leadto an excess of significant bleeds indicating caution in prescribingconcomitant drugs in the particular setting of myeloproliferativedisorders characterized by platelet functional and biochemicalabnormalities and prolonged bleeding time exacerbated by the useof aspirin [17]. However, in the light of aspirin efficacy in high riskpatients, previous history of upper gastrointestinal bleeding shouldnot be viewed as an absolute contraindication. Aspirin should bewithdrawn during the acute episode and resumed as soon as gas-tric ulcer or other intestinal problems have been controlled. Assuggested in the previous section, the combination of low-doseaspirin and proton-pump inhibitors should be considered in thesehigh risk patients. The persistence of thrombotic recurrences inspite of cytoreductive therapy and aspirin would suggest to changeor intensify antiplatelet therapy. However, combining aspirin withclopidogrel may significantly enhance the bleeding tendency espe-cially in patients over 70–75 year and the evidence of benefit in ETpatients has not yet been proven.

In conclusion, until additional data become available, manyuncertainties limit the use of aspirin in low-risk patients with ETand long-term aspirin in asymptomatic cases should be indicatedwith great caution in the routine use. For each patient an individ-ual assessment of thrombotic and hemorrhagic risk should be made

Leuke

agtbtch

C

A

a

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

Commissioned Editorial Debates /

nd, in particular, previous conditions with a high bleeding risk (e.g.astric ulcers or esophageal varices secondary to splanchnic veinhrombosis and portal hypertension), and perhaps extreme throm-ocytosis, should limit the use of antiaggregating agents. Many ofhese uncertainties deserve to be solved by designing appropriatelinical trials that should stratify patients by their predicted risk ofemostatic complications.

onflict of interest statement

All authors have no conflict of interest to report.

cknowledgement

None. No funding to declare.Contributions. Both authors wrote, revised and gave final

pproval of the version submitted.

eferences

[1] Harrison CN, Bareford D, Butt N, et al. Guideline for investigation and manage-ment of adults and children presenting with a thrombocytosis. Br J Haematol2010;149:352–75.

[2] Landolfi R, Marchioli R, Kutti J, et al. Efficacy and safety of low-dose aspirin inpolycythemia vera. N Engl J Med 2004;350:114–24.

[3] Barbui T, Barosi G, Grossi A, et al. Practice guidelines for the therapy of essentialthrombocythemia. A statement from the Italian Society of Hematology, the Ital-ian Society of Experimental Hematology and the Italian Group for Bone MarrowTransplantation. Haematologica 2004;89:215–32.

[4] Alvarez-Larran A, Cervantes F, Pereira A, et al. Observation versus antiplatelettherapy as primary prophylaxis for thrombosis in low-risk essential thrombo-cythemia. Blood 2010;116(8):1205–10.

[5] Antithrombotic Trialists’ (ATT) Collaboration, Baigent C, Blackwell L, CollinsR, Emberson J, Godwin J, Peto R, Buring J, Hennekens C, Kearney P, MeadeT, Patrono C, Roncaglioni MC, Zanchetti A. Aspirin in the primary andsecondary prevention of vascular disease: collaborative meta-analysis of indi-vidual participant data from randomised trials. Lancet 2009;373(May 30(9678)):1849–60.

[6] Tartaglia AP, Goldberg JD, Berk PD, Wasserman LR. Adverse effects of antiaggre-gating platelet therapy in the treatment of polycythemia vera. Semin Hematol1986;23:172–6.

[7] van Genderen PJ, Michiels JJ, van SR, Lindemans J, van Vliet HH. Platelet con-sumption in thrombocythemia complicated by erythromelalgia: reversal byaspirin. Thromb Haemost 1995;73:210–4.

[8] Squizzato A, Romualdi E, Middeldorp S. Antiplatelet drugs for polycythaemiavera and essential thrombocythaemia. Cochrane Database Syst Rev 2008;(April16 (2)).

[9] Ruggeri M, Finazzi G, Tosetto A, et al. No treatment for low-risk thrombo-cythaemia: results from a prospective study. Br J Haematol 1998;103:772–7.

10] Pearson TC, Bareford D, Craig J, et al. The management of ‘low-risk’ and‘intermediate-risk’ patients with primary thrombocythaemia. MPD (UK) StudyGroup. Br J Haematol 1999;106:833–4.

11] Marchioli R, Finazzi G, Landolfi R, et al. Vascular and neoplastic risk in alarge cohort of patients with polycythemia vera. J Clin Oncol 2005;23:2224–

32.

12] Patrono C, Garcia Rodriguez LA, Landolfi R, Baigent C. Low-dose aspirin for theprevention of atherothrombosis. N Engl J Med 2005;353:2373–83.

13] Carobbio A, Finazzi G, Antonioli E, et al. Thrombocytosis and leukocytosisinteraction in vascular complications of essential thrombocythemia. Blood2008;112:3135–7.

mia Research 35 (2011) 286–289 289

14] Carobbio A, Finazzi G, Antonioli E, et al. Hydroxyurea in essential thrombo-cythemia: rate and clinical relevance of responses by European LeukemiaNetcriteria. Blood 2010;19(116 (7)):1051–5.

15] van Genderen PJ, Michiels JJ. Erythromelalgic, thrombotic and haemorrhagicmanifestations of thrombocythaemia. Presse Med 1994;23:73–7.

16] Mitus AJ, Barbui T, Shulman LN, et al. Hemostatic complications in youngpatients with essential thrombocythemia. Am J Med 1990;88:371–5.

17] Barbui T, Buelli M, Cortelazzo S, Viero P, De Gaetano G. Aspirin and risk ofbleeding in patients with thrombocythemia. Am J Med 1987;83:265–8.

18] Finazzi G, Rambaldi A, Guerini V, Carobbo A, Barbui T. Risk of thrombosis inpatients with essential thrombocythemia and polycythemia vera according toJAK2 V617F mutation status. Haematologica 2007;92:135–6.

19] Carobbio A, Finazzi G, Antonioli E, et al. JAK2V617F allele burden and thrombo-sis: a direct comparison in essential thrombocythemia and polycythemia vera.Exp Hematol 2009;37:1016–21.

20] Dahabreh IJ, Zoi K, Giannouli S, et al. Is JAK2 V617F mutation more than adiagnostic index? A meta-analysis of clinical outcomes in essential thrombo-cythemia. Leuk Res 2009;33:67–73.

21] Harrison CN, Campbell PJ, Buck G, et al. Hydroxyurea compared with anagrelidein high-risk essential thrombocythemia. N Engl J Med 2005;353:33–45.

22] Bain BJ. Bone marrow biopsy morbidity: review of 2003. J Clin Pathol2005;58:406–8.

23] Belay ED, Bresee JS, Holman RC, et al. Reye’s syndrome in the United Statesfrom 1981 through 1997. N Engl J Med 1999;340:1377–82.

24] Pinsky PF, Hurwitz ES, Schonberger LB, Gunn WJ. Reye’s syndrome and aspirin.Evidence for a dose–response effect. JAMA 1988;260:657–61.

25] Chan FK, Ching JY, Hung LC, et al. Clopidogrel versus aspirin and esomeprazoleto prevent recurrent ulcer bleeding. N Engl J Med 2005;352:238–44.

26] Ben-Dor I, Kleiman NS, Lev E. Assessment, mechanisms, and clinical implica-tion of variability in platelet response to aspirin and clopidogrel therapy. Am JCardiol 2009 Jul;104(2):227–33.

27] Dawson J, Quinn T, Rafferty M, Higgins P, Ray G, Lees KR, et al. Aspirin resistanceand compliance with therapy. Cardiovasc Ther 2010;(June 11).

28] MacDonald TM, Wei L. Effect of ibuprofen on cardioprotective effect of aspirin.Lancet 2003;361:573–4.

29] Girodon F, Bonicelli G, Schaeffer C, et al. Significant increase in the appar-ent incidence of essential thrombocythemia related to new WHO diagnosticcriteria: a population-based study. Haematologica 2009;94:865–9.

30] Carobbio A, Finazzi G, Guerini V, et al. Leukocytosis is a risk factor for throm-bosis in essential thrombocythemia: interaction with treatment, standard riskfactors, and Jak2 mutation status. Blood 2007;109:2310–3.

31] Carobbio A, Antonioli E, Guglielmelli P, et al. Leukocytosis and risk stratificationassessment in essential thrombocythemia. J Clin Oncol 2008;26:2732–6.

32] Barbui T, Carobbio A, Rambaldi A, Finazzi G. Perspectives on thrombosis inessential thrombocythemia and polycythemia vera: is leukocytosis a causativefactor? Blood 2009;114:759–63.

33] Barbui T, Carobbio A, Cervantes F, et al. Thrombosis in primary myelofibrosis:incidence and risk factors. Blood 2010;115:778–82.

Claire Harrison a,∗

Tiziano Barbui b

a Guy’s and St Thomas’ NHS Foundation Trust, GreatMaze Pond, London SE1 9RT, UK

b Division of Hematology, Ospedali Riuniti diBergamo, Bergamo, Italy

∗ Corresponding author. Tel.: +44 2071882742; fax:+44 2071882728.

E-mail address: Claire.Harrison@gstt.nhs.uk (C.Harrison)

27 September 2010Available online 5 November 2010