effect of platelet glycoprotein iib/iiia receptor blockade with tirofiban on adverse cardiac events...

Effect of platelet glycoprotein IIb/IIIa receptorblockade with tirofiban on adverse cardiac eventsin women with unstable angina/non-ST–elevationmyocardial infarction (PRISM-PLUS Study)Thao Huynh, MD,a Pierre Theroux, MD,b Steven Snapinn, PhD,c and Ying Wan, MSc,c for the PRISM-PLUSInvestigators Montreal, Quebec, Canada, and Blue Bell, Pa

Background Previous trials demonstrated the efficacy of platelet glycoprotein IIb/IIIa receptors blockade with tirofi-ban in reducing acute ischemic events in patients with unstable angina/non–ST-elevation myocardial infarction. Little isknown about the effect of tirofiban among women with acute coronary syndromes.

Objective We aimed to determine the benefit and safety of tirofiban plus heparin versus heparin alone on cardiacischemic events among women with unstable angina/non–ST-elevation myocardial infarction.

Method and Results We performed a post hoc analysis of all women enrolled in the PRISM-PLUS trial. At earlytime points, there appeared to be a reduction of the primary composite end point of death, myocardial infarction, or re-fractory ischemia among women treated with tirofiban plus heparin (RR, 0.78 and 0.67) compared with women treatedwith heparin alone. However, at 30 and 180 days, there was no significant reduction of events with the combinationtherapy of tirofiban plus heparin (treatment-by-sex interaction, P � .05). Death or myocardial infarction was not signifi-cantly reduced by the combination therapy among women at all time points.

Conclusions Although the effects of tirofiban in reducing the primary composite outcome were similar among menand women early in the study, there appeared to be a difference at the later time points. In particular, tirofiban was effec-tive among men, but there was no clear effect among women at 30 and 180 days. (Am Heart J 2003;146:668–73.)

Platelet activation and aggregation are pivotal in thepathophysiology of acute coronary syndromes. Numer-ous clinical trials have demonstrated the effectivenessof platelet glycoprotein IIb/IIIa inhibitors in reducinglife-threatening complications in patients with acutecoronary syndromes.1–4 Tirofiban had been shown tobe effective in reducing death, myocardial infarction,or refractory ischemia in patients with unstable angi-na/non–ST-elevation myocardial infarction (UA/NSTEMI).1,2 However, little is known about the benefitand safety of tirofiban among women with coronarysyndromes. The objective of this study was to exam-ine, in a post hoc analysis of the PRISM-PLUS data,

whether the addition of tirofiban to standard anti-thrombotic therapy would be effective in women.

MethodsThe study population, design, and the main findings of the

PRISM-PLUS study have been described in detail elsewhere.2

The initial study design involved double-blind random assign-ment to 1 of 3 treatment groups: tirofiban alone, tirofibanplus heparin, or heparin alone. At an interim analysis, thetirofiban-alone arm of the study was prematurely discontin-ued on the recommendation of the Data Safety MonitoringBoard, secondary to an excessive 7-day mortality rate in thesepatients.

Study end pointsThe primary end point of PRISM-PLUS was a composite of

death from any cause, new myocardial infarction (MI), or re-fractory ischemia within 7 days after random assignment. Re-hospitalization for UA/NSTEMI was also counted in the com-posite primary end point when assessed at 7 days, 30 days,and 6 months. Predefined secondary end points included thesame composite end point 48 hours and 30 days after ran-dom assignment, the 3 components of this end point as sepa-rate measures, and a composite of death or MI.

From aMontreal General Hospital, Montreal, Quebec, bMontreal Heart Institute, Mon-treal, Quebec, Canada, and cMerck Research Laboratories, Blue Bell, Pa.The PRISM-PLUS trial was sponsored by Merck Research Laboratories. Dr Snapinn andMs Wan are employees of Merck Research Laboratories.Submitted August 19, 2002; accepted January 15, 2003.Reprint requests: Thao Huynh, MD, Montreal General Hospital, 1650 Cedar Ave,Room E-5200, Montreal, Quebec H3G-1A4, Canada.E-mail: [email protected]© 2003, Mosby, Inc. All rights reserved.0002-8703/2003/$30.00 � 0doi:10.1016/S0002-8703(03)00255-2

MI was defined as a new episode of chest pain of at least20 minutes in duration with new Q waves or a rise in theserum creatine kinase level to 2 times the upper limit of nor-mal. Refractory angina included the following symptoms:chest pain �20 minutes in duration, repetitive chest painswith transient ST-T changes, or recurrent ischemia with pul-monary edema or hypotension or requiring intra-aortic coun-terpulsation or urgent coronary intervention. All events wereevaluated by an end point committee blinded to treatmentassignment.

Assessment of safetyBleeding events were assessed throughout the duration of

study drug infusion and for 24 hours after infusion by usingthe criteria developed by the Thrombolysis in Myocardial In-farction trial group. Major bleeding was defined as a decreasein the hemoglobin levels of �40 g/L, the need for transfusionof �2 units of blood, the need for corrective surgery, theoccurrence of an intracranial or retroperitoneal hemorrhageor cardiac tamponade, or any combination of these events.Minor bleeding was defined as a decrease of �30g/L.5

Statistical analysisThe present report focused only on the comparison be-

tween the patients treated with tirofiban plus heparin andheparin alone. We excluded the patients who were randomlyassigned to the tirofiban-alone therapy from analysis becausethis arm was discontinued prematurely. The Pearson �2 testwas used to compare the baseline characteristics and bleed-ing complications except for the mean age, which was com-pared by Student t test. Clinical events were analyzed bymeans of a Cox proportional hazards model, and the risk ra-tios presented are based on these models. Tests for the effectof tirofiban were based on a model with an indicator fortreatment; tests for treatment by sex additionally included anindicator for sex and the product of the treatment and sexindicators. All Cox regression models included adjustmentsfor prior use of heparin and antecedent use of aspirin. Allprobability values were 2-sided, with values �.05 consideredsignificant.

ResultsA total of 1915 patients were enrolled in the main

study. We excluded 345 patients enrolled in the tirofi-ban-alone arm. Of the remaining 1570 patients, therewere 252 women (31.6%) in the group who receivedtirofiban plus heparin and 254 women (32.9%) in theheparin arm.

Table I describes the demographic characteristics ofwomen and men. The mean age was comparable be-tween the two sexes. Although women had more hy-pertension (P � .0001) and diabetes mellitus (P �.04), they had less prior MI and coronary bypass sur-gery than men (P � .0001). High-risk clinical featuressuch as transient ST elevation and NSTEMI were foundmore often in men (P � .03), whereas women hadmore T-wave inversion (P � .03). However, the meanTIMI risk score6 was similar in both sexes (3.87 for

men and 3.88 for women). The baseline characteristicswere similar among the women randomly assigned tothe two treatment arms.

The incidences of the primary end point and thesecondary combined end point of MI and death inwomen versus men, irrespective of treatment, are pre-sented in Table II. Despite considerable differences inbaseline characteristics, women and men had similarrates of adverse outcomes, which is consistent withtheir similar TIMI risk scores.

Figure 1 presents the risk ratios for the effect of tiro-fiban on the primary end point and the combined endpoint of death and MI among women and men. Tirofi-ban was clearly beneficial in men, with a reduction ofboth the composite primary end point and the com-bined end point of death and MI, at all time points.Men treated with tirofiban plus heparin had a risk ratioof 0.67, 0.68, 0.73, and 0.71 for the composite endpoint at 48 hours and 7, 30, and 180 days.

Table III compares the outcomes among womenwho received tirofiban plus heparin compared withthe women who were receiving heparin alone. Atearly time points, women appeared to have similarreduction of the composite end point with tirofiban asmen. The risk ratios for the composite end point were0.78 and 0.67 at 2 and 7 days, respectively, for womenwho received heparin plus tirofiban. However, therewas no clear effect of tirofiban on death and MIamong women at all time points. A statistical test forquantitative interaction between tirofiban and sex wassignificant, with P � .05 for the 180-day composite

Table I. Baseline characteristics of patients by sex

CharacteristicsMen

(n � 1064)Women

(n � 506) P

Mean age (y � SD) 61.9 � 11.62 65.8 � 11.30 .46Prior myocardial infarction

(%)498 (46.8) 155 (30.6) �.001

Prior congestive heart failure(%)

100 (9.4) 50 (9.9) .76

Prior percutaneous coronaryintervention (%)

107 (10.6) 39 (7.7) .13

Prior coronary artery bypasssurgery (%)

187 (17.6) 44 (8.7) �.001

Smoking (%) 862 (81.3) 245 (48.5) �.001Hypertension (%) 525 (49.3) 348 (68.8) �.001Hypercholesterolemia (%) 524 (49.3) 257 (50.8) .57Diabetes mellitus (%) 229 (21.5) 133 (26.3) .036Unstable angina (%) 558 (52.4) 298 (58.9) .02Non ST-elevation

myocardial infarction (%)506 (47.6) 208 (41.1)

Transient ST elevation (%) 162 (15.2) 56 (11.1) .026ST depression (%) 614 (57.7) 305 (60.3) .33T waves inversions (%) 534 (50.2) 283 (55.9) .033Prior therapy: aspirin (%) 547 (51.4) 267 (52.8) .62Prior heparin (%) 700 (65.8) 307 (60.7) .048

American Heart JournalVolume 146, Number 4

Huynh et al 669

primary end point, indicating that the effect of tirofi-ban among men was significantly greater than its effectamong women.

Bleeding complicationsTable IV presents the overall bleeding complications

among men and women, irrespective of treatment.Major bleeds were rare in both sexes. Women hadmore minor bleeds than men. The incidences of bleed-ing complications among men and women, by treat-ment allocation, are shown in Table V. Although thereappeared to be an increase in bleeding complicationsin women treated with tirofiban and heparin, this didnot reach statistical significance.

DiscussionWomen with UA/NSTEMI who were enrolled in

PRISM-PLUS had similar rates of reduction in the com-

posite end point of death, MI, and refractory ischemiawith tirofiban as men at 2 and 7 days. However, con-trary to men, there was no clear benefit of tirofibanseen at 30 and 180 days among women treated withtirofiban plus heparin. There was also no significantreduction of death and myocardial infarction inwomen by tirofiban at all time points. Was this sexdifferential response to tirofiban the result of true bio-logical difference or of chance? Can this be explainedby different pathophysiological mechanisms of acutecoronary syndromes or biological response to the gly-coprotein IIb/IIIa receptor inhibitors?

Sex differences in platelet reactivityPrevious studies suggested sex differences in platelet

reactivity, with a greater sensitivity of the platelets ofwomen to aggregating stimuli.7,8 Faraday et al8 demon-strated a 50% to 80% increase of activated glycoproteinIIb/IIIa receptors in women compared with men giventhe same agonist. Given their increased platelet reac-tivity, it is possible that women may have differentresponses to glycoprotein receptors IIb/IIIa blockade.

Effects of glycoprotein IIb/IIIa antagonism in womenundergoing percutaneous coronary intervention

Cho et al9 and Fernandes et al10 showed the benefitsof abciximab and eptifibatide, respectively, in reduc-tion of adverse outcomes among women who under-went percutaneous coronary intervention. They didnot find any sex-related response to treatment. How-ever, these studies focused exclusively on patientswho had percutaneous coronary intervention. The ef-fect of glycoprotein IIb/IIIa receptor inhibitors inwomen with acute coronary syndromes is less wellstudied.

Although the benefit of eptifibatide was not seen inthe whole group of women enrolled in PURSUIT,4 ep-tifibatide was beneficial in reducing acute ischemicevents in American women.11 This may reflect thehigh revascularization rate in the United States4,11 and

Table II. Outcomes at 2, 7, 30, and 180 day by sex

Time Events Women, n � 506 (%)Men,

n � 1064 (%)Risk ratios(95% CI) P

48 Hours Composite 34 (6.7) 72 (6.8) 1.02 (0.68–1.53) .94MI/Death 7 (1.4) 21 (2.0) 0.71 (0.30–1.66) .42

7 Days Composite 82 (16.2) 161 (15.1) 1.09 (0.83–1.42) .55MI/Death 30 (5.9) 74 (7.0) 0.84 (0.55–1.29) .43



Composite, Death or myocardial infarction or refractory ischemia; MI, myocardial infarction.

Figure 1

Effects of treatment (T�H vs H) on end points at various timepoints in women and men. T�H, Tirofiban plus heparin; H, hepa-rin; MI, myocardial infarction.

American Heart JournalOctober 2003

670 Huynh et al

the well-proven benefit of glycoprotein IIb/IIa receptorblockade in patients who underwent coronary revascu-larization.11–17

Glycoprotein IIb/IIIa blockade in women with acutecoronary syndromes

Are women with acute coronary syndromes a lower-risk group, and thus the benefit of potent antiplateletdrugs is less evident? Hochman et al18 demonstratedthat among patients with unstable angina, female sexwas associated with an independent protective effectfor death or MI (OR, 0.65; CI, 0.49 to 0.87). Amongthe patients enrolled in PURSUIT, Boersma et al19

showed that women were at lower risk than men for30-day death/MI (OR, 0.56; CI, 0.45 to 0.89).

Boersma et al20 performed a meta-analysis of plateletglycoprotein IIb/IIIa receptor inhibitors in acute coro-nary syndromes that included the PRISM-PLUS datapresented in this article. They observed a differentialtreatment effect between men and women. In women,the odds ratio and confidence interval were consistentwith a possible risk increase with glycoprotein IIb/IIIareceptor inhibitors (11.5% versus 10.4%, 30-day events;OR, 1.15; CI, 1.02 to 1.30). However, the difference intherapeutic response between men and women wasno longer present when patients were stratified ac-cording to their baseline troponin level. A reduction of30-day death/MI was seen in both men and womenwith elevated troponin level, whereas no benefit wasobserved in patients with normal troponin level, irre-

Table III. Treatment effects in women

Events Tirofiban � heparin (%) Heparin (%)Risk ratios(95% CI) P

Sextreatment

interactionP

48 Hours Composite 15 (5.9) 19 (7.5) 0.78 (0.40–1.53) .47 .72MI/Death 3 (1.2) 4 (1.6) 0.73 (0.16–3.3) .69 .23

7 Days Composite 34 (13.4) 48 (19.0) 0.67 (0.43–1.04) .08 .98MI/Death 14 (5.5) 16 (6.3) 0.86 (0.42–1.78) .69 .18



Table IV. Bleeding complications among men and women

Women,n � 506 (%)

Men,n � 1064 (%)

Risk ratios(95% CI) P

TIMI major bleeding 8 (1.6) 9 (0.8) 1.87 (0.73–4.82) .19TIMI minor bleeding 64 (12.7) 81 (7.6) 1.66 (1.22–2.27) .001All TIMI bleeding 77 (15.2) 97 (9.1) 1.67 (1.26–2.21) .0003

TIMI, Thrombolysis in Myocardial Infarction trial.

Table V. Incidence of bleeding complications among men and women by treatment

Heparin � tirofiban (%) Heparin (%)Risk ratios(95% CI) P

Treatment-sexinteraction

P

Women 254 252TIMI major bleeding 6 (2.4) 2 (0.8) 2.98 (0.61–14.61) .16 .43TIMI minor bleeding 37 (14.6) 27 (10.7) 1.36 (0.85–2.16) .19 .76All TIMI bleeding 44 (17.3) 33 (13.1) 1.32 (0.87–2.01) .19 .83

Men 519 545TIMI major bleeding 5 (1.0) 4 (0.7) 1.31 (0.35–4.86) .68 .68TIMI minor bleeding 44 (8.5) 37 (6.8) 1.25 (0.82–1.90) .30 .30All TIMI bleeding 53 (10.2) 44 (8.1) 1.26 (0.86–1.85) .23 .23


Huynh et al 671

spective of sex. Serum troponin elevation appeared tobe a reliable tool to identify women at risk for acuteischemic events, who will benefit from glycoproteinIIb/IIIa receptors inhibitors.

Compared with the above meta-analysis,20 we hadmore long-term data with 6-month outcomes, in-hospi-tal refractory ischemia, and rehospitalization for acutecoronary syndromes. Our patients were a more ho-mogenous group, with only one glycoprotein IIb/IIIareceptor inhibitor being studied.

The lack of an obvious benefit of tirofiban of death,MI, and refractory ischemia among women at 30 and180 days in PRISM-PLUS could be explained by thesmall sample size. There was also no effect of tirofibanon death and MI among women at all time points. Thesignificant sex and treatment interaction at 180 days (P� .05) raised the possibility of a true different sex re-sponse with tirofiban and that tirofiban was less pro-tective in reducing late ischemic events in women.The explanation for this finding remained unclear.

Could it be that the women enrolled in PRISM-PLUSwere lower-risk patients, so that the long-term benefitof tirofiban was not shown? This did not appear to bethe case. These women appeared to be as high-risk asthe male patients, with similar TIMI scores (3.87 formen and 3.88 for women) and rates of cardiac adverseoutcomes. The rates of revascularization were also sim-ilar between the two sexes (55.1% for men and 52.4%for women).

Angiographic data among women with acutecoronary syndromes

Angiographic data from many clinical trials and regis-tries21–25 showed less clinically significant coronaryartery disease and multivessel coronary artery diseasein women than in men. This did not necessarily implythat the diagnosis of an acute ischemic event was in-correct in these patients. Twenty-five percent ofwomen with insignificant coronary stenosis in PUR-SUIT had confirmed MI. Myocardial infarction was con-firmed by enzymatic criteria in 41.1% of the womenenrolled in PRISM-PLUS.2 Thrombus formation mayplay a less important role in the pathophysiology ofacute coronary syndromes in women, and the underly-ing coronary disease in these patients may be less se-vere.

Zhao et al25 reported the angiographic results of pa-tients enrolled in PRISM-PLUS. Female sex was predic-tive, by univariate analysis, of less coronary arterythrombus. However, it was no longer a significant pre-dictor by multivariate analysis. This finding impliedthat although female sex by itself was not an indepen-dent predictor, there was less active coronary throm-bus in women. A potential explanation for our resultsmay be that our female patients had less intracoronary

thrombus and/or less severe underlying coronary dis-ease and thus derived less benefit from tirofiban.

Bleeding events among womenWomen enrolled in this study had significantly more

minor bleeds than men, irrespective of treatmentgroup. We did not observe a significant increase ofbleeding complications in women treated with tirofi-ban and heparin; however, this can be due to lack ofpower secondary to the small sample size.

LimitationsFirst, our study had the inherent limitations of a post

hoc subgroup analysis.26 The sample size was likelytoo small to show a significant benefit of tirofiban inwomen; however, our results were in concordancewith Boersma meta-analysis of glycoprotein IIb/IIIareceptor inhibitors in acute coronary syndromes.20

Sex-treatment interaction confirming the differentialtherapeutic response between men and women wassignificant at 180 days. These findings raised the possi-bility of a true differential sex response to tirofiban.

Second, PRISM-PLUS enrollment took place during1994 to 1996, before widespread implementation ofserum troponin measurement. Since serum troponinwas not available, we could not validate the previousreport that elevation of serum troponin was a moreimportant predictor of response to glycoprotein recep-tor IIb/IIIa blockade than sex.20

Third, although the majority of patients did undergodiagnostic coronary angiography (78.3%), coronaryanatomy was not available in 112 women (22.2% ofthe women enrolled).25 We could not validate the hy-pothesis that tirofiban may be less beneficial in womenbecause of less severe coronary artery disease in thesepatients. Finally, although we did not observe a signifi-cant increase in bleeding complications among womenassigned to heparin plus tirofiban, this can be due tothe lack of power of the small sample size.

ConclusionsAt early time points, women with UA/NSTEMI who

were enrolled in PRISM-PLUS had similar rates of re-duction of the primary composite outcome as menwith tirofiban. However, at 30 and 180 days, there wasno significant reduction of events with the combina-tion therapy. Death or MI was not significantly re-duced by the combination therapy of tirofiban plusheparin at all time points.

Although the explanation was unclear, it may be thatwomen had less severe underlying coronary artery dis-ease with less active thrombus. Serum troponin eleva-tion may be helpful in identifying high-risk womenwho will benefit more from glycoprotein IIb/IIIa re-ceptor inhibitors. Future studies focusing on platelet

American Heart JournalOctober 2003

672 Huynh et al

glycoprotein IIb/IIIa receptor inhibitors and serum tro-ponin elevation in women with acute coronary syn-dromes are needed.

We thank Dr Peter Dibattiste for his invaluableinput, guidance, and help in the completion of themanuscript.

References1. PRISM Study Investigators. A comparison of aspirin plus tirofiban

with aspirin plus heparin for unstable angina. N Engl J Med1998;338:1498–505.

2. PRISM-PLUS Study Investigators. Inhibition of the platelet glycopro-tein IIb/IIIa receptor with tirofiban in unstable angina and non-Q-wave myocardial infarction: Platelet Receptor Inhibition in IschemicSyndrome Management in Patients Limited by Unstable Signs andSymptoms. N Engl J Med 1998;338:1488–97.

3. PARAGON Investigators. International, randomized, controlledtrial of lamifiban (a platelet glycoprotein IIb/IIIa inhibitor) or hepa-rin, or both in unstable angina: platelet IIb/IIIa Antagonism for theReduction of Acute coronary events in a Global Organization Net-work. Circulation 1998;97:2386–95.

4. PURSUIT Trial Investigators. Inhibition of platelet glycoprotein IIb/IIIa with eptifibatide in patients with acute coronary syndromes.N Engl J Med 1998;339:436–43.

5. Cheesebro JH, Knatterud G, Roberts R, et al. Thrombolysis in Myo-cardial Infarction (TIMI) Trial, phase 1: a comparison between in-travenous tissue plasminogen activator and intravenous streptoki-nase: clinical findings through hospital discharge. Circulation1987;76:142–54.

6. Antman E, Cohen M, Bernink P, et al. The TIMI risk score for un-stable angina/non-ST-elevation myocardial infarction. JAMA2000;284:835–42.

7. Johnson M, Ramey E, Ramwell PW. Sex and age differences inhuman platelet aggregation. Nature 1975;253:353–7.

8. Faraday N, Goldschmidt-Clermont PJ, Bray P. Gender differencesin platelet GPIIb/IIIa activation. Thromb Haemost 1997;77:748–54.

9. Cho L, Topol EJ, Balog C, et al. Clinical benefit of glycoproteinIIb/IIIa blockade with abciximab is independent of gender: pooledanalysis from EPIC, EPILOG and EPISTENT trials. J Am Coll Car-diol 2000;35:381–6.

10. Fernandes LS, Tcheng JE, O’Shea JC, et al. Is glycoprotein IIb/IIIaantagonism as effective in women as in men following percutane-ous coronary intervention? Lessons from the ESPRIT study. J AmColl Cardiol 2002:401085–91.

11. Lincoff AM, Harrington RA, Califf RM, et al. Management of pa-tients with acute coronary syndromes in the United States by plate-let glycoprotein IIb/IIIa inhibition: insights from the platelet glyco-protein IIb/IIIa in unstable angina: receptor suppression usingintegrelin therapy (PURSUIT) Trial. Circulation 2000;102:1093–100.

12. CAPTURE Investigators. Randomised placebo-controlled trial ofabciximab before and during coronary intervention in refractoryunstable angina: the CAPTURE trial. Lancet 1997;349:1429–35.

13. Lincoff AM, Califf RM, Anderson KM, et al. Evidence for preven-tion of death and myocardial infarction with platelet membraneglycoprotein IIb/IIIa receptor blockade by abciximab (c7E3 Fab)among patients with unstable angina undergoing percutaneouscoronary intervention. J Am Coll Cardiol 1997;30:149–56.

14. Kleiman NS, Lincoff ML, Flaker GC, et al. Early percutaneous cor-onary intervention, platelet inhibition with eptifibatide, and clinicaloutcomes in patients with acute coronary syndromes. Circulation2000;101:751–7.

15. IMPACT-II Investigators. Randomized placebo-controlled trial ofeffect of eptifibatide on complications of percutaneous coronaryintervention: IMPACT-II. Lancet 1997;3349:1422–8.

16. ESPRIT Investigators. Novel dosing regimen of eptifibatide inplanned coronary stent implantation (ESPRIT): a randomized, pla-cebo-controlled trial. Lancet 2000;356:2037–44.

17. Boersma E, Akkerhuis M, Theroux P, et al. Platelet glycoproteinIIb/IIIa receptor inhibition in non-ST-elevation acute coronary syn-dromes: early benefit during medical treatment only, with addi-tional protection during percutaneous coronary intervention. Circu-lation 1999;100:2045–8.

18. Hochman JS, Tams JE, Thompson TD. Sex, clinical presentation,and outcome in patients with acute coronary syndromes. N EnglJ Med 1999;341:226–32.

19. Boersma E, Pieper KS, Steyerberg EW, et al. Predictors of outcomein patients with acute coronary syndromes without persistent ST-Segment elevation: results from an international trial of 9461 pa-tients. Circulation 2000;101:2557–67.

20. Boersma E, Harrington RA, Moliterno DJ, et al. Platelet glycopro-tein IIb/IIIa inhibitors in acute coronary syndromes: a meta-analy-sis of all major randomized clinical trials. Lancet 2002;359:189–98.

21. GUSTO-IIb Investigators. A comparison of recombinant hirudinwith heparin for the treatment of acute coronary syndromes.N Engl J Med 1996;335:775–82.

22. Roe MT, Harrington RA, Prosper DM, et al. Clinical and therapeu-tic profile of patients presenting with acute coronary syndromeswho do not have significant coronary artery disease. Circulation2000;102:1101–6.

23. Hochman JS, McCabe CH, Stone PH, et al. Outcome and profileof women and men presenting with acute coronary syndromes: areport from TIMI IIIB. J Am Coll Cardiol 1997;30:1141–8.

24. Scirica BM, Moliterno DJ, Every NR, et al. Differences betweenmen and women in the management of unstable angina pectoris(the GUARANTEE Registry). Am J Cardiol 1999;84:1145–50.

25. Zhao XQ, Theroux P, Snapinn SM, et al. Intracoronary thrombusand platelet glycoprotein IIB/IIIa receptor blockade with tirofibanin unstable angina or non-Q-Wave myocardial infarction. Circula-tion 1999;100:1609–15.

26. Yusuf S, Wittes J, Probstfield J, et al. Analysis and interpretation oftreatment effects in subgroups of patients in randomized clinicaltrials. JAMA 1991;266:93–8.


Huynh et al 673

effect of platelet glycoprotein iib/iiia receptor blockade with tirofiban on adverse cardiac events...

Documents