Intravenous Thrombolysis inAcute Myocardial Infarction

E. Magnus Ohman, MD, FCCP; Robert A. Harrington, MD;Christopher P. Cannon, MD; Giancarlo Agnelli, MD;John A. Cairns, MD; and J.Ward Kennedy, MD

Abbreviations: APTT 5 activated partial thromboplastin time;ASSENT 5 Asessment of the Safety and Efficacy of a NewThrombolytic Agent; CI 5 confidence interval; FTT 5 Fibrino-lytic Therapy Trialists’ Collaborative Group; GISSI 5 GruppoItaliano per lo Studio Streptokinasi nell’Infarto Miocardico;GP 5 glycoprotein; GUSTO 5 Global Utilization of Streptoki-nase and TPA (alteplase) for Occluded Coronary Arteries; HIT 5Hirudin for the Improvement of Thrombolysis; INJECT 5International Joint Efficacy Comparison of Thrombolytics;ISIS 5 International Study of Infarct Survival; MI 5 myocardialinfarction; n-PA 5 lanoteplase; NS 5 not significant; RAPID 5Reteplase Angiographic Phase II International Dose-finding;r-PA 5 reteplase; rt-PA 5 recombinant tissue-type plasminogenactivator; scu-PA 5 single-chain urokinase-type plasminogen activa-tor; TIMI 5 Thrombosis in Myocardial Infarction; TNK-tPA 5tenecteplase; t-PA 5 tissue plasminogen activator

(CHEST 2001; 119:253S–277S)

F ibrinolytic therapy for acute myocardial infarction(MI) has been one of the most potent treatments ever

developed for a condition that kills more patients world-wide than any other. This review will focus on approvedagents and the randomized trials that have led to theirwidespread use. The use of adjunctive antithrombotictherapies, such as aspirin and heparin, also will be dis-cussed, but a more complete discussion of these therapiesis located in other sections of this supplement. A briefoverview of new adjunctive therapies, such as plateletglycoprotein (GP) IIb/IIIa inhibitors and direct thrombininhibitors, will be provided. This section also will explorehow therapeutic success can be evaluated, cost-effective-ness analysis, and complications with this therapy. Finally,we will provide a set of recommendations for the use offibrinolytic therapy in acute MI based on the publishedliterature.

Brief History of Reperfusion Therapy forAcute MI

The first description of the use of a prolonged infusionof streptokinase for patients with acute MI appeared in1958.1 Several smaller studies followed, but they generallyshowed no clear therapeutic advantage over standardtherapy. When these early trials were combined in ameta-analysis2 in the mid-1980s, a significant reduction inmortality was revealed. This observation, coupled withearly angiographic observations of reperfusion amongpatients receiving intracoronary streptokinase,3 usheredin the modern era of reperfusion therapy for acute MI.It was soon recognized that intracoronary fibrinolytictherapy could salvage myocardium and that earlierrestoration of patency of the infarct-related arteryresulted in better preservation of left ventricular function.4–6

This discovery led to the transition from intracoronary ther-apy (with its inherent delay in establishing coronary reperfu-sion) back to IV therapy.

The use of large clinical trials, which focused onmortality and safety profiles through adequate samplesizes, led to worldwide adoption of IV thrombolysis foracute MI.7–10 Observations from these trials made itpossible to define the therapeutic windows and risk/benefit ratios for an array of patient subgroups. Theimportance of therapy that could be administered rapidlyand easily to allow earlier treatment also became evident.11

Continued evaluation of mechanistic components of ther-apy served as an underpinning for the large mortality trialsand led to development of adjunctive therapies to enhancereperfusion rates and safety profiles.12 Although adjunc-tive therapies such as aspirin and IV heparin have beenfound to be important to reduce mortality and to maximizetherapeutic efficacy, other approaches to these ends,including direct thrombin inhibition and complete plateletpacification by GP IIb/IIIa inhibitors, remain to be real-ized. The future of reperfusion therapies, for example,should include strategies compatible with primary angio-plasty, which has shown improved mortality over IVthrombolysis in a meta-analysis of 10 small trials13 but islimited by a lack of widespread availability and expertise.

Pharmacology of Fibrinolytic Agents

The term thrombolytic agents really is a misnomerwhen applied to plasminogen activators. All of theseagents convert plasminogen to plasmin that then degradesfibrin, a major structure component of the thrombus; thus,the more correct term is fibrinolytic therapy. The agentscan be grouped into direct and indirect plasminogenactivators (Table 1). The relatively nonspecific agentsinclude streptokinase, anistreplase (anisoylated plasmino-gen streptokinase activator complex), and urokinase. Thenewer (or second-generation) plasminogen activators in-clude recombinant tissue-type plasminogen activator (rt-PA) (alteplase or duteplase) and several variants of tissue-type plasminogen activator: reteplase (r-PA), tenecteplase(TNK-tPA), and lanoteplase (n-PA).

StreptokinaseStreptokinase is an indirect fibrinolytic agent that binds

to plasminogen, thereby converting plasminogen into aplasmin-like molecule capable of converting plasminogento plasmin because streptokinase activates both circulatingand fibrin-bound plasminogen to plasmin; it producessystemic plasminemia with resultant depletion of fibrino-gen, plasminogen, and factors V and VIII. This “systemiclytic state” creates a sustained hypocoagulable state thatmay reduce the risk of rethrombosis. Patients who receivestreptokinase can develop antistreptococcal antibodies.The rate has varied in the literature, but it has been welldocumented that patients treated with streptokinase haveallergic reactions. In the milder form, these include chills,fever, and rigors. In more severe forms, anaphylaxis hasbeen described. Generation of bradykinin also may con-tribute to the hypotension that occurs in patients receivingstreptokinase. This hypotensive reaction usually responds

Correspondence to: E. Magnus Ohman, MD, FCCP, Duke Clin-ical Research Institute, PO Box 17969, Durham, NC 27715;e-mail: ohman001@mc.duke.edu

CHEST / 119 / 1 / JANUARY, 2001 SUPPLEMENT 253S

well to vasopressors and IV fluids. Because of thesereactions and serologic confirmation of antibodies, therehas been some controversy about the repeat administra-tion of streptokinase. It has been suggested that patientsshould not receive a second dose of streptokinase within 1year of initial therapy. However, many patients havereceived a second dose of streptokinase without sufferingsevere allergic or anaphylactic reactions.

AnistreplaseAnistreplase is a modified streptokinase molecule,

bound to lys-plasminogen to form an activator complex.Because lys-plasminogen has affinity for fibrin, it washoped that the complex would be more fibrin specific thanstreptokinase. The modified streptokinase undergoes ac-tivation after deacylation, providing a much longer half-life(approximately 100 min) and allowing single-bolus dosing.The antigenicity and side effects profile of anistreplase aresimilar to those seen with streptokinase.

UrokinaseThis naturally occurring plasminogen activator has been

used to treat acute MI for . 30 years.2 Urokinase issignificantly less antigenic than streptokinase, and, unlikestreptokinase, urokinase directly activates plasminogen.Despite its years of use, however, it has not been devel-oped as a standard treatment for acute MI. Althoughurokinase was used extensively for treating peripheralvascular occlusions, production problems have curtailed itsavailability.

Saruplase (Prourokinase)This agent, also known as single-chain urokinase-type

plasminogen activator (scu-PA), is a single-chain precursorto urokinase that has little intrinsic enzymatic activity.scu-PA has relative fibrin specificity similar to tissueplasminogen activator (t-PA). The amino-acid sequence ofscu-PA resembles that of t-PA, and thus this agent has avery short half-life. Scu-PA circulates bound to a specificinhibitor; under this condition, the catalytic activity ofscu-PA is inactivated. In the presence of fibrin, thecomplex between scu-PA and its specific inhibitor isdissociated and thus scu-PA is able to express its fibrino-lytic activity.

rt-PAsrt-PAs are naturally occurring, serine proteases that are

physiologically identical to the naturally occurring endog-enous plasminogen activator in humans. In its naturalstate, t-PA is produced by vascular endothelium. Plasmin-ogen activator inhibitors counteract its effect in humans.The rt-PA molecule was cloned by Pennica and col-leagues14 and is produced by recombinant DNA technol-ogy. Two forms have been manufactured commercially.Alteplase is predominantly a single-chain rt-PA molecule.In the mid-1980s, manufacturing began for a predomi-nantly two-chain rt-PA molecule, duteplase. Althoughthese drugs never were compared directly in trials, thelatter compound was dropped after large trials failed toshow its superiority over streptokinase.

As opposed to streptokinase, alteplase is not antigenicand appears not to be associated with allergic reactions.The fibrin specificity of alteplase is considerably greaterthan that of streptokinase. Nevertheless, alteplase pro-duces mild fibrinogen depletion. Another theoretical ad-vantage of alteplase over streptokinase is its ability to lysemore highly cross-linked fibrin. This theoretical advantageis more of a consideration among patients who have hadsymptoms for a longer duration.

r-PATruncated forms of rt-PA have been developed; the first

was r-PA. This is a single-chain deletion mutant that lacksthe finger, epidermal growth factor, and Kringle-1 do-mains (Fig 1).15 This mutation results in a half-life abouttwice that of native t-PA, permitting double-bolus therapyof 10 U, 30 min apart. The fibrinogen depletion with r-PAis greater than that with alteplase but less than that withstreptokinase. Because the finger domain mediates thehigh-affinity interaction of rt-PA with fibrin, r-PA haslower affinity for fibrin than rt-PA. No antigenicity hasbeen reported with this compound. Its mode of actionotherwise is similar to that of naturally occurring t-PA.

TNK-tPATNK-tPA is similar to wild-type t-PA, but has amino-

acid substitutions at three sites (Fig 1).15 A threonine isreplaced by asparagine, which adds a glycosylation site toposition 103; an asparagine is replaced by a glutamine,thereby removing a glycosylation site at position 117; andfour amino acids in the protease domain (lysine, histidine,arginine, and arginine) are replaced by four alanine resi-dues. These mutations, in animal models, produce aprolonged half-life,16,17 increased fibrin specificity,16 andincreased resistance to inhibition by plasminogen activatorinhibitors-1 compared with naturally occurring t-PA.18–20

In humans, TNK-tPA has shown slower plasma clearancerelative to values for alteplase; its plasma eliminationhalf-life ranges from 11 to 20 min compared with 3.5 minreported for alteplase.21 TNK-tPA is more fibrin-specificthan alteplase, which itself is more fibrin specific thanstreptokinase or r-PA. Systemic fibrinogen and plasmino-gen levels fell by only 5 to 15% over the first 6 h ofTNK-tPA dosing of 30 to 50 mg, compared with 40 to 50%reductions after alteplase dosing. Similarly, the consump-tion of a2-antiplasmin, the fluid-phase inhibitor of plas-min, and the resultant increase in plasmin-a2-antiplasmincomplexes were four to five times greater with alteplase vsTNK-tPA. The greater fibrin specificity of TNK-tPAcompared with alteplase helps explain its efficacy whengiven as a 5-s to 10-s single bolus,22 and the fact that it doesnot induce the “plasminogen steal” phenomenon.23

n-PAn-PA is another deletion mutant of naturally occurring

t-PA. It combines deletions of the finger and epidermalgrowth factor domains with substitution of the asparagineresidue at position 117 with a glycine residue to removethe glycosylation site in the first Kringle domain. It has oneof the longest half-lives of the mutant t-PA molecules

254S Sixth ACCP Consensus Conference on Antithrombotic Therapy

permitting a single bolus administration. Because it lacksthe finger domain, it exhibits lower affinity for fibrin thant-PA.

Evaluation of Therapeutic SuccessThe completeness and success of reperfusion therapy

have been extensively studied, as measured by angiogra-phy and ST-segment resolution on the 12-lead ECG. Inaddition, the therapeutic risk/benefit ratio has been wellestablished in large mortality trials. The field of thrombol-ysis developed traditionally; that is, the first studies eval-uated a new agent vs a placebo or an established treat-ment, in trials that tested for superiority. After the firstlarge, placebo-controlled trials showed an unequivocalbenefit with fibrinolytic therapy, it became unethical totest newer therapies against placebo therapy. Further,agents that may have shown similar efficacy in superioritytrials appeared to have other characteristics that couldoffer a clinical advantage (such as bolus therapy vs infu-sion). Thus, a different trial method was developed. Thisstatistical approach, called equivalence testing, is based onthe assumption that differences between outcomes withtwo treatments should be less than a certain predeter-mined value to be considered equivalent. Recent studiessuggest that this threshold should be defined as a , 1%absolute difference in mortality or a relative difference of, 14% between the agents. This threshold is based on thehalf of the benefit on mortality observed with fibrinolytictherapy over placebo in placebo-controlled trials (a 2%absolute difference in mortality)24 and the full benefitobserved with alteplase over streptokinase (1% absolutemortality difference).9 To establish equivalence, the 95%confidence intervals (CIs) are calculated based on theabsolute difference in mortality rates. A new therapy thathas an absolute rate and 95% CI within the 1% is deemedto be of equivalent efficacy to an established therapy. Thisconcept will hold true only if a similar safety profile isobserved.

To establish equivalence, a number of different statis-tical approaches have been taken in randomized clinicaltrials. There appears not to be any consensus on the exactboundaries to accept. However, it should be recognizedthat the rationale might differ among different clinicaltrials depending on what the goal is with the intendedtherapy. In the International Joint Efficacy Comparison ofThrombolytics (INJECT) trial, in which noninferioritywith r-PA was tested against streptokinase, a 1% absolutedifference (half of the benefit) in mortality was used as thelower boundary (see section on r-PA), the assumptionbeing that streptokinase is associated with a 2% lowermortality against placebo. However, in a study comparingtwo different ways of administration of t-PA, a muchtighter 95% boundary of 0.4% was used, the assumptionbeing that with similar therapies, half of the benefit thatwas observed with a bolus plus infusion of t-PA overstreptokinase (1%) should be maintained. In this particu-lar case, the observed 95% CI extended to 0.49%, sug-gesting that the modes of administrations of t-PA were notequivalent (see section on alteplase). Another approach

Tab

le1—

Cha

ract

eris

tics

ofF

ibri

noly

tic

Age

nts*

Age

nts

Sour

ceM

olec

ular

Wei

ght,

kDa

Mut

atio

nSt

ruct

ure

Fib

rin

Spec

ific

Met

abol

ism

Hal

f-lif

e,m

inM

ode

ofA

ctio

nA

ntig

enic

Dos

ing

Oth

erPr

oper

ties

Hos

pita

lC

ost/D

ose

($U

S)†

Stre

ptok

inas

eG

roup

Ast

rept

ococ

ci47

Hep

atic

18–2

3A

ctiv

ator

com

plex

Yes

1-h

infu

sion

$300

Uro

kina

seR

ecom

bina

nt,h

uman

feta

l,ki

dney

35–5

5H

epat

ic14

–20

Dir

ect

No

1-h

infu

sion

$2,0

00

Ani

stre

plas

eG

roup

Ast

rept

ococ

cipl

asm

inog

en,

anis

oyla

ted

131

Hep

atic

70–1

20D

irec

tYe

s10

-min

sing

lebo

lus

$1,8

00

t-PA

Rec

ombi

nant

,hum

an63

–70

F,E

,K,P

dom

ains

11

Hep

atic

3–8

Dir

ect

No

Bol

us,9

0-m

inin

fusi

on1

fibri

nbi

ndin

g$2

,200

r-PA

Rec

ombi

nant

,hum

anm

utan

tt-

PA39

Sing

le-c

hain

dele

tion

Lac

ksF

,E,K

1do

mai

ns;

has

K2,

Pdo

mai

ns1

Ren

al15

–18

Dir

ect

No

Dou

ble

bolu

s2

fibri

nbi

ndin

g$2

,200

TN

K-t

PAR

ecom

bina

ntpl

usm

utat

ion

Poin

tsu

bstit

utio

nsT

103N

,N11

7Q,K

HH

RR

296–

299

AA

AA

11

1H

epat

ic18

–20

Dir

ect

No

Sing

lebo

lus

Res

ista

nce

toPA

I-1

$2,2

00

n-PA

Rec

ombi

nant

plus

mut

atio

nD

omai

nde

letio

ns,

poin

tsu

bstit

utio

n

Lac

ksF

,Edo

mai

ns,G

in11

7su

bstit

utio

n1

30–4

5D

irec

tN

oSi

ngle

bolu

s2

fibri

nbi

ndin

gN

A

*PA

I5

plas

min

ogen

activ

ator

inhi

bito

r;N

A5

not

avai

labl

e;F

5fin

ger

dom

ain;

E5

epid

erm

algr

owth

fact

ordo

mai

n;K

5K

ring

ledo

mai

n;P

5se

rine

prot

ease

dom

ain.

†Cur

rent

pric

esof

usua

ldos

e.

CHEST / 119 / 1 / JANUARY, 2001 SUPPLEMENT 255S

has been to use an odds ratio of , 1.5 (less than a relative50% increase in mortality) compared with streptokinase(see section on scu-PA).

Angiographic EvaluationSince the mid-1980s, the angiographic assessment of

epicardial coronary flow has been critical in the develop-ment of new reperfusion strategies, as seen in Tables2–4.25–75 The angiographic substudy of the Global Utili-zation of Streptokinase and TPA (alteplase) provided thelink between Thrombosis in Myocardial Infarction (TIMI)grade 3 flow (normal antegrade flow) and 30-day survival;Simes and colleagues76 showed that the superior survivalwith alteplase vs streptokinase resulted from the superiorepicardial blood flow provided by alteplase. In the 1990s,Gibson and colleagues77 moved us beyond a fixation onepicardial coronary flow with the development of conceptssuch as the corrected TIMI frame count and the myocar-dial perfusion blush score, which take into considerationboth epicardial and microvascular flow. As the field movestoward combining fibrinolytic therapy with antiplatelet therapy,the notion of improved myocardial perfusion through improve-ment in microcirculatory flow has become an important tenetof reperfusion investigation.

ECG EvaluationInvestigation over the last 25 years into the concepts of

reperfusion and myocardial salvage have now come fullcircle. Work by Braunwald and colleagues77a in the mid-1970s concentrated on the importance of restoring myocar-dial perfusion with the goal of preserving left ventricularfunction. There was a shift in the 1980s, as investigation ofacute MI came to focus on the restoration of epicardialcoronary flow. Quantification of this flow during initial workby the TIMI investigators formed the basis of much investi-gation over the next 10 to 12 years. With the introduction ofthe platelet GP IIb/IIIa inhibitors into clinical practice, wehave witnessed a shift back to actual myocardial perfusion,and investigation now centers on not only restoration ofepicardial coronary flow but also preservation of microcircu-latory flow. As fibrinolytic therapy moved from clinical

investigation (with its reliance on angiography to determinetherapeutic success) into clinical practice, clinical measuresof reperfusion were needed to gauge the success or failure oftherapy. Unfortunately, simple measures of reperfusion, suchas resolution of chest pain and of ECG findings, were notedto be unreliable in their predictive accuracy.78 Investigatorsreported that although complete resolution of chest pain withcomplete resolution of ST-segment elevation was highlypredictive of a patent coronary artery, this combination offindings was extremely unusual and thus of relative littleclinical utility among patients being treated with fibrinolytictherapy.

Braunwald and others78a championed the use of the12-lead ECG as a way to quantify left ventricular dysfunc-tion after MI. It was not until the mid-1990s, however,that Schroder and colleagues79 more completely charac-terized ST-segment resolution and its relationship toclinical outcomes. Data from five recent randomized trialsof almost 6,000 patients, outlining the relationship be-tween ST-segment resolution after fibrinolytic therapy and30-day mortality, are shown in Table 5.79–82 Clearly, thereis a linear relationship between the degree of ST-segmentresolution and later mortality that is highly statisticallysignificant and very relevant clinically, as this is an easilyemployed bedside technique. Observations from primaryangioplasty experiences have confirmed the value of thestatic ECG as a tool to predict clinical outcomes afterreperfusion. While rapid early resolution of ST-segmentelevation is better than later resolution, the optimal timingof static ECG tracings is unknown. Furthermore, the ECGundoubtedly provides important information about tissueperfusion and not just epicardial coronary flow; however,how best to incorporate serial tracings into the acuteevaluation of these patients requires further investigation.

Continuous ST-segment monitoring has evolved overthe last 10 to 15 years as an alternative to static ECGmeasurement in the evaluation of reperfusion. The obvi-ous advantage is that it provides insight into the contin-uum of reperfusion, providing information on cyclic flow,reperfusion, and reocclusion. The main limitation is that itrequires special equipment. Until recent technical im-provements, its use was limited by the fact that patientswere required to be at bed rest. Continuous ST-segmentmonitoring can provide predictive information about pa-tency of the epicardial infarct artery, but there are com-plexities because of the issues of tissue-level perfusion andcollateral blood flow. The ST-segment evaluation appearsto be best at providing data reflecting myocardial perfu-sion rather than epicardial coronary flow. Work from theGlobal Utilization of Streptokinase and TPA (alteplase) forOccluded Coronary Arteries (GUSTO)-1 trial83 has shownthat this technique can be used to predict clinical out-comes. Additionally, it is an important tool in the nonin-vasive evaluation of new drugs and reperfusion strategies.Use of this technology may allow a smaller sample sizethan angiographic trials typically require.

Mortality and Intracranial Hemorrhage End PointsThe large amount of information available from fibrino-

lytic mortality trials has provided ample opportunity toFigure 1. Structure of t-PA and mutant forms. Reprinted fromBrener and Topol15 with permission.

256S Sixth ACCP Consensus Conference on Antithrombotic Therapy

create outcome models based on patient characteristics.Lee and colleagues84 have used the GUSTO-1 trial toprovide insight into the predictors of 30-day mortalityamong patients being treated with fibrinolytic therapy foracute MI (Table 6). Certain patient characteristics (age,Killip class, and infarct location) are associated with muchhigher 30-day mortality.

The major risk of fibrinolytic therapy is intracranialhemorrhage, which usually occurs in the first 24 h afterstarting therapy. Several patient characteristics are associ-ated with a higher risk of intracranial bleeding (Table7).85–89 The level of concomitant anticoagulation with IVheparin, as measured by activated partial thromboplastintime (APTT), also can influence the rate of intracranialhemorrhage. An APTT of between 50 s and 75 s, mea-sured between 6 h and 24 h after starting fibrinolytictherapy and IV heparin therapy, appears to offer the bestbalance between optimal outcomes and the risk of intra-cranial hemorrhage.90

1. Fibrinolytic Therapy

Trials of StreptokinaseThe field of IV fibrinolytic therapy was transformed

when it became possible to examine the efficacy of therapywith angiography during the acute phase of MI. The initialtrials of urokinase and streptokinase were carried out withdosages established on a theoretical basis. In fact, IVstreptokinase was not subjected to a true form of dose-ranging angiographic trial until well into the 1980s.50

Nevertheless, the placebo-controlled trials of streptoki-nase were very powerful in showing a significant mortalityreduction with IV streptokinase for acute MI.

Several trials25–42 have reported patency of the infarct-related artery at different time points among patients notreceiving fibrinolytic therapy (Table 2). Most patients didreceive aspirin and heparin, although aspirin was notstandard therapy for acute MI until the InternationalStudy of Infarct Survival (ISIS)-2 trial10 results werepublished in 1988.

Several angiographic trials also were undertaken toexplore the patency and recanalization rates with IVstreptokinase. The first trials explored rates of recanaliza-tion. In those studies, patients underwent urgent angiog-raphy. If the vessel was occluded (TIMI grade 0 or 1 flow),IV fibrinolytic therapy was given and the rate of openingwas established. The delays inherent in this approach,however, suggested adverse effects on left ventricularfunction, as described. Thus patency trials were devel-oped, wherein patients with acute MI were given fibrino-lytic therapy as soon as possible and then underwentangiography, where patency (defined as TIMI grade 2 or3 flow) was examined. The results for both patency andrecanalization trials of streptokinase are shown in Table3.26,30,34,37,39–59 Overall, the angiographic data suggest pa-tency rates with streptokinase of approximately 44% at 60min and 48% at 90 min. Approximately 2 to 3 h afterbeginning therapy, patency rates were 72%, achieving arate of between 75% and 85% at 24 h to 21 days aftertherapy from a pooled meta-analysis.42 These rates are

substantially higher than those of control patients (Fig 2)42

but less than those with accelerated alteplase and similarto those with anistreplase or alteplase given . 3 h.

The efficacy of streptokinase with regard to mortalitywas evaluated in four large, placebo-controlled trials (Ta-ble 8).8,10,91,92 The first true mortality trial for streptoki-nase was the Gruppo Italiano per lo Studio Streptokinasinell’Infarto Miocardico (GISSI-1) trial.8 This was an open-label, randomized trial of 11,806 patients. Of note, only14% of patients received aspirin and only 62% receivedany heparin in this study; all adjunctive therapies were atthe investigator’s discretion. Nevertheless, in-hospitalmortality (14 to 21 days) was reduced by 18% comparedwith standard therapy (10.7% vs 13.0%; p 5 0.002). Thereduction in mortality was time dependent, decreasingfrom a 47% reduction in patients treated within 1 h, to23% for those treated within 3 h, and to 17% for thosetreated within 6 h of symptom onset. The reduction inmortality was maintained . 12 months (17.2% with strep-tokinase vs 19.0% for control subjects; p 5 0.008). TheIntravenous Streptokinase in Acute Myocardial infarctionstudy91 was a double-blind, randomized trial of streptoki-nase vs placebo in 1,741 patients with ST-segment-eleva-tion MI. Consistent with the GISSI-1 findings, there wasan 11% reduction in 21-day mortality, although this didnot achieve statistical significance.

The second ISIS trial was a large, double-blind, place-bo-controlled study of IV streptokinase in patients withsuspected MI.10 There were no specific entry criteriaother than the physician’s clinical suspicion of an acuteMI. Most patients, however, had ST-segment elevation orleft bundle-branch block on the presenting ECG. In all,17,187 patients were randomized in 417 hospitals world-wide. Patients were enrolled up to 24 h after symptomonset, but most were randomized in the first 12 h. Thestudy used a 2 3 2 factorial design testing aspirin alone(162.5 mg/d for 1 month), streptokinase alone (1.5 MU. 1 h), both, or neither. Patients randomized to strep-tokinase had a 25% reduction in 35-day vascular mor-tality compared with those who received placebo (9.2%vs 12.0%; p , 0.001). The study also showed thataspirin alone could reduce mortality by a relative 23%(p , 0.001). The most important part of this trial was thesynergistic effects of aspirin with streptokinase, whichproduced a 42% reduction in vascular mortality (8.0% vs13.2%; p , 0.001). Additional benefits with aspirin in thistrial included reduced rates of reinfarction, cardiac arrest,cardiac rupture, and stroke. Similar to the GISSI-1 study,there was clear evidence of time dependency for treat-ment benefit. Patients treated within 6 h of symptoms hadsignificantly improved survival, and this benefit persistedfor treatment that began within up to 12 h after symptomonset.

A smaller, South American trial (Estudio MulticentricoEstreptoquinasa Republicas de America del Sur)92 alsowas a double-blind, placebo-controlled trial of streptoki-nase. This study was altered to include only patients whopresented at least 6 h after but within 24 h of symptomonset, once the ISIS-2 results were reported. Mortality at

CHEST / 119 / 1 / JANUARY, 2001 SUPPLEMENT 257S

35 days did not differ significantly in the 3,568 patientsenrolled between 6 h and 24 h (11.2% for streptokinase vs11.8% for placebo).

Consistent across these trials was that the treatmentbenefit observed in the first 21 to 42 days was maintainedup to 1 year. The Fibrinolytic Therapy Trialists’ Collabo-rative Group (FTT) combined these and other trials in ameta-analysis.24 The overall benefit was observed amongpatients with ST-segment elevation or bundle-branchblock irrespective of age, sex, BP, heart rate, prior MI, ordiabetic status. Furthermore, the treatment benefit wasgreater the earlier that treatment began. For patientstreated within 6 h, the absolute reduction in mortality was30 lives saved for 1,000 patients treated; for patientstreated within the first 7 to 12 h after symptom onset, itwas 20 lives saved per 1,000 treated. For patients treatedbetween 13 h and 18 h after symptom onset, there was anuncertain trend toward mortality reduction of approxi-mately 10 lives saved per 1,000 treated. Fibrinolytictherapy was associated with approximately four extrastrokes per 1,000 patients treated; most of that occurredwithin 2 days. Approximately 50% were associated with anearly death and so were already accounted for in theoverall mortality reduction. Of the remaining patients withstroke, 25% were moderately or severely disabled and theother 25% were not. The overview thus suggested a

treatment benefit for most patients who present withacute MI within 12 h of symptom duration.

Trials of Alteplase

After two small series of patients had been treated withalteplase,93,94 the first comparative trial was conductedbetween alteplase and streptokinase. In the TIMI-1 trial,95

290 patients with acute MI underwent diagnostic coronaryangiography and then were treated with either streptoki-nase or alteplase, in addition to IV heparin. The primaryend point, reperfusion of an initially occluded coronaryartery after 90 min, was achieved in 62% of alteplase-treated patients compared with 31% of streptokinase-treated patients (p , 0.001; Fig 3).22,26,95 The patency rateat 90 min, independent of findings on the baseline angio-gram, was 70% for alteplase vs 43% for streptokinase(p , 0.001). The European Study Group reported nearlyidentical results.44 Alteplase was then studied in numerousangiographic trials,26,31,32,35,36,38,40,42,44,53–55,60–75 as re-viewed by Granger and colleagues (Table 4).96 These earlytrials observed that the 3-h dosing regimen of alteplaseresulted in superior patency and TIMI grade 3 flow resultsat both 60 min and 90 min compared with streptokinase oranistreplase.96 Neuhaus and colleagues97 developed an“accelerated” 90-min dosing regimen for alteplase, which

Table 2—Infarct-Artery Patency Without Fibrinolytic Therapy*

StudyPatients,

No.Time to

Treatment,† minMean Catheterization

TimePatency, %

(No./Total Patients) 95% CI, %

BaselineAnderson et al25 23 168 0 min 9 (2/23)Chesebro et al26 289 285 0 min 20 (57/289)Timmis et al27 40 186 0 min 28 (11/40)Pooled 352 20 (70/352) 16–24

1 hCollen et al28 14 284 45 min 7 (1/14)Topol et al29 25 228 68 min 13 (3/23)Cribier et al30 23 208 74 min 22 (5/23)Pooled 62 15 (9/60) 6–24

90 minVerstraete et al31 65 198 75–90 min 21 (13/62) 11–31

2 to 3 hGuerci et al32 66 192 2.5 h 24 (15/62) 14–35

1 dArmstrong et al33 56 180 17 h 29 (7/24)Durand et al34 29 149 35 h 13 (3/24)Pooled 85 21 (10/48) 9–32

3 to 21 dBassand et al35 119 188 4.1 d 35 (38/105)NHFA36 71 195 5–7 d 41 (26/64)Armstrong et al33 56 180 10 d 45 (47/105)Kennedy et al37 177 210 10 d 78 (259/334)de Bono38 366 168 10–21 d 78 (259/334)White et al39 112 180 21 d 54 (50/92)O’Rourke et al40 71 114 21 d 64 (40/63)Bassand et al41 55 190 21 d 67 (31/46)Pooled 1,027 61 (508/838) 57–64

*NHFA 5 National Heart Foundation of Australia Coronary Thrombosis Group. Reprinted from Granger et al42 with permission.†Time from symptom onset to start of thrombolytic or control treatment.

258S Sixth ACCP Consensus Conference on Antithrombotic Therapy

was found to achieve even higher rates of early reperfusionthan did the 3-h regimen of alteplase,61 anistreplasetreatment,98,99 or streptokinase treatment.12

Given the importance of rapid reperfusion, a fibrino-lytic regimen that achieves a higher rate of early infarct-artery patency would be expected to be associated withlower mortality. This notion was called into question withthe disappointing results of the GISSI-2 and InternationalStudy Group100 and the ISIS-3 trial (Table 9).101 The lackof benefit seen in these trials, however, may have beenbecause of the use of subcutaneous heparin (rather thanIV heparin) and the use of duteplase as opposed toalteplase, as will be discussed.

The relationship of early reperfusion and improved sur-vival was strongly supported by the results of GUSTO-1 trial,which set out to evaluate several promising fibrinolyticregimens.9,12 The reference arms of the trial both usedstreptokinase, one with subcutaneous heparin (12,500 Uq12h beginning at 4 h) and one with IV heparin, as notedearlier in this article. The third arm was front-loaded (accel-erated) alteplase and IV heparin. The fourth arm was com-bination fibrinolytic therapy, which involved about two thirdsof the typical doses of alteplase and streptokinase with IVheparin. All patients received aspirin, 325 mg/d.

A total of 41,021 patients were enrolled in GUSTO-1, ofwhich the primary end point was 30-day mortality (Table

Table 3—Patency With IV Streptokinase*

Study†Patients,

No. Dose, MUTime to

Treatment

MeanCatheterization

TimePatency, % (No./

Total Patients)95%

CI, %

1 hPRIMI43 203 1.5 140 min 61 min 48 (82/171)Cribier et al30 21 1.5 115 min 74 min 52 (11/21)Pooled 224 48 (93/192) 41–56

90 minECSG-2 Verstraete et al44 65 1.0 156 min 75–90 min 55 (34/62)Stack et al45 216 1.5 180 min 90 min 44 (95/216)Chesebro et al26 159 1.5 286 min 90 min 43 (63/146)Lopez-Sendon et al46 25 1.5 , 6 h 90 min 60 (15/25)Hogg et al47 63 1.5 209 min 90 min 53 (31/58)PRIMI43 203 1.5 140 min 91 min 64 (124/194)Charbonnier et al48 58 1.5 168 min 93 min 51 (27/53)Pooled 789 51 (389/754) 48–55

2 to 3 hTEAM-249 182 1.5 158 min 126 min 73 (129/176)Monnier et al49 11 1.5 135 min 150 min 64 (7/11)Six et al50 56 1.5 150 min 168 min 60 (32/53)Golf et al51 31 1.5 138 min 176 min 70 (21/30)Pooled 280 70 (189/270) 65–70

1 dPRIMI43 203 1.5 140 min 14–36 h 88 (160/181)Hogg et al47 63 1.5 209 min 24 h 88 (49/56)Lopez-Sendon et al46 25 1.5 , 6 h 24 h 75 (18/24)Durand et al34 35 1.5 149 min 39 h 82 (27/33)Ribeiro et al52 50 1.2 180 min 48 h 80 (40/50)Pooled 376 86 (294/344) 82–89

3 to 21 dPAIMS53 85 1.5 127 min 4 d 77 (57/74)Kennedy et al37 191 1.5 210 min 10 d 69 (89/130)Cherng et al54 63 1.5 294 min 10–14 d 57 (16/28)Lopez-Sendon et al46 25 1.5 , 6 h 15–21 d 90 (17/19)White et al39 107 1.5 180 min 21 d 75 (74/99)White et al55 135 1.5 150 min 21 d 75 (87/116)Bassand et al41 52 1.5 MU/0.5 h 210 min 21 d 68 (32/47)Pooled 658 73 (372/513) 70–78

Hillis et al56 34 1.5 60 min 90 min 32‡Spann et al57 43 1.5 60 min 60 min 49‡Rogers et al58 16 1.0 45 min 45 min 44‡TIMI-1 Chesebro et al26 119 1.5 60 min 90 min 31‡de Marneffe et al59 10 1.5 30 min 60 min 80‡

*PRIMI 5 Pro-Urokinase in Myocardial Infarction trial; PAIMS 5 Plasminogen Activator Italian Multicentre Study. Adapted from Granger etal,42 with permission.

†Studies with streptokinase dose $ 1.0 MU.‡Percent with recanalization.

CHEST / 119 / 1 / JANUARY, 2001 SUPPLEMENT 259S

10).9,12 Mortality at 30 days was significantly lower in theaccelerated alteplase arm compared with each of the threeother arms.9 The improvement in mortality was present asearly as 24 h after treatment began, with alteplase-treatedpatients having a significantly lower mortality rate. Othermajor complications also were reduced in patients treatedwith alteplase, including less cardiogenic shock, congestiveheart failure, and ventricular arrhythmias.

Some physicians have stated that the relative 14%

reduction in mortality with alteplase vs streptokinase (anabsolute 1% difference) is a small, clinically irrelevantimprovement. Thrombolysis itself, however, widely con-sidered to be a revolution in the treatment of acute MI,led to a 25% relative (or approximately 2% absolute)improvement in mortality compared with placebo. Incounting the absolute benefit in terms of the number oflives saved, standard fibrinolysis is estimated to save 26lives per 1,000 patients treated compared with placebo in

Table 4—Patency With IV Alteplase*

Study†Patients,

No. DoseTime to

TreatmentMean Catheterization

TimePatency, % (No./

Total Patients) 95% CI, %

1 hde Bono38 183 100 mg/3 h 156 min 42 min 60 (108/180)Smalling et al60 91 1.25 mg/kg/3 h 228 min 56 min 45 (40/89)Carney et al61 138 100 mg/3 h 168 min 62 min 63 (54/86)Topol et al62 75 1.25 mg/kg/3 h 216 min 68 min 57 (40/70)Pooled 487 57 (242/425) 52–61

90 minVerstraete et al44 64 0.75 mg/kg/1.5 h 180 min 75–90 min 70 (43/61)TIMI-IIA63 133 100 mg/6 h 168 min 84 min 75 (98/131)Chesebro et al26 157 80 mg/3 h 287 min 90 min 70 (100/143)Carney et al61 138 100 mg/3 h 168 min 90 min 77 (94/122)Verstraete et al31 64 0.75 mg/kg/1.5 h 204 min 90 min 61 (38/62)Topol et al64 50 100 mg/3 h 243 min 90 min 52 (26/50)Califf et al65 95 100 mg/3 h 200 min 90 min 71 (67/95)Topol et al62 75 1.25 mg/kg/3 h 216 min 90 min 69 (49/71)Topol et al62 142 1 mg/kg/h 190 min 90 min 72 (102/142)Johns et al66 68 1 mg/kg/1.5 h 180 min 90 min 76 (52/68)Whitlow and Bashore67 206 100 mg/3 h , 6 h 90 min 63 (126/199)Smalling et al60 91 1.25 mg/kg/3 h 228 min 90 min 70 (61/87)Topol et al68 134 1.5 mg/kg/4 h 168 min 92 min 79 (104/131)Grines et al69 107 100 mg/3 h 180 min 95 min 64 (65/102)Neuhaus et al70 124 70 mg/1.5 h , 4 h 97 min 69 (84/121)Pooled 1,648 70 (1,109/1,585) 68–72

2 to 3 hTopol et al62 75 1.25 mg/kg/3 h 216 min 120 min 79 (59/75)Guerci et al32 72 80–100 mg/3 h 192 min 150 min 66 (44/67)Pooled 147 73 (103/142) 65–80

1 dNeuhaus et al70 124 70 mg/1.5 h , 4 h 24 h 78 (82/105)Anderson et al71 164 100 mg 168 min 33 h 86 (128/149)TIMI-IIA63 128 100–150 mg/6 h 174 min 33 h 82 (93/114)TIMI-II72 1,366 100 mg 156 min 33 h 85 (1,040/1,229)Pooled 1,782 84 (1,343/1,597) 82–86

3 to 21 dde Bono et al73 652 100 mg/3 h 170 min 3.4 d 79 (410/518)Magnani53 86 100 mg/3 h 124 min 4.1 d 81 (63/78)Bassand et al35 93 100 mg/3 h 172 min 5.4 d 76 (64/84)NHFA36 73 100 mg/3 h 195 min 5–7 d 70 (43/61)TIMI-IIA63 389 100–150 mg/6 h 174 min 7–10 d 79 (240/303)Thompson et al74 241 100 mg/3 h 155 min 7–10 d 80 (157/196)Neuhaus et al70 124 70 mg/3 h , 4 h 10 d 73 (63/86)Cherng et al54 59 100 mg/3 h 312 min 10–14 d 77 (23/30)Rapold et al75 34 100 mg/3 h 186 min 12.5 d 81 (25/31)de Bono38 367 100 mg/3 h 156 min 15 d 87 (283/327)White et al55 135 100 mg/3 h 150 min 21 d 76 (94/124)O’Rourke et al40 74 100 mg/3 h 120 min 21 d 81 (55/68)Pooled 2,327 80 (1,520/1,906) 78–81

*See Table 2 for abbreviation. Reprinted from Granger et al,42 with permission.†Studies with 0.75- to 1.5-mg/kg or 70- to 100-mg alteplase.

260S Sixth ACCP Consensus Conference on Antithrombotic Therapy

patients with ST-segment elevation. Treatment with acceler-ated alteplase and IV heparin saves an additional 10 lives per1,000 treated, representing a 40% improvement over stan-dard fibrinolytic regimens. Otherwise stated, the use ofalteplase instead of streptokinase prevents one of the sevendeaths that would have occurred among 100 patients treated.As such, accelerated alteplase does afford an importantbenefit over previous standard fibrinolytic regimens.

Despite the aggressive regimens of thrombolysis, aspi-rin, and heparin, intracranial hemorrhage occurred onlyrarely in GUSTO-1. For each of the streptokinase arms,0.5% of patients suffered an intracranial hemorrhagecompared with 0.7% of patients treated with acceleratedalteplase and 0.9% of patients treated with combinationfibrinolytic therapy.9 To put the results in full perspective,the GUSTO-1 investigators developed the concept of “netclinical benefit,” that is, the avoidance of either death ornonfatal, disabling stroke. When comparing the net clini-cal benefit among the four regimens, accelerated alteplasestill provided a clear benefit compared with the otherthree regimens.

The benefit of accelerated alteplase was seen in nearly

every subgroup analyzed, including patients with anterioror inferior MI and in the young and the elderly. Theabsolute benefit was greater in higher-risk patients, forexample, those with anterior MI.

To fully understand the benefits of the various fibrino-lytic regimens, an angiographic substudy was carried out.12

Over 2,400 patients were randomized to undergo angiog-raphy at 90 min, 180 min, 24 h, or 5 days. At theimportant, 90-min time point, the alteplase-treated pa-tients had a significantly higher patency rate and a muchhigher rate of TIMI grade 3 flow, which, as noted above,is associated with the best outcomes (Table 10).9,12 At theother three time points, there were no significant differ-ences among the four fibrinolytic regimens. Thus thebenefit of accelerated alteplase was associated with earlyopening of the infarct-related artery. The improved pa-tency at 90 min was associated with improved survival atboth 24 h and at 30 days, thus highlighting the benefits ofrapid reperfusion.76,102

TIMI-4: The TIMI-4 trial99 was a double-blind trialcomparing accelerated alteplase, anistreplase, and theircombination. All patients received aspirin and IV heparin.Accelerated alteplase was found to have a 78% patencyrate after only 60 min compared with only 60% foranistreplase or combination fibrinolytic therapy.99 At 90min, patency and TIMI grade 3 flow rates both weresignificantly better in the accelerated alteplase arm. Over-all clinical outcomes, using a composite end point and1-year survival, also were better with alteplase. Thus, thisdouble-blind trial confirmed the results found in theGUSTO-1 trial.

Table 5—Mortality Prediction With ST-Segment Resolution After Fibrinolysis*

TrialPatients,

No.Time of

Analysis, min Follow-up, d

Resolution of ST-Segment Elevation,† %

p ValueComplete Partial None

ISAM79 1,516 180 35 2.8 4.3 9.2 , 0.001INJECT80 1,398 180 35 2.5 4.3 17.5 , 0.001HIT-481 998 180 35 2.8 6.0 14.3 , 0.001GUSTO-III81a 1,577 90 30 2.7 4.8 13.0 , 0.001TIMI-1481b 444 90 30 1.0 4.2 5.9 , 0.001

*Reprinted from Roe et al82 with permission.†Complete resolution was $ 70% of baseline; partial resolution was 31 to 69% of baseline; no resolution was # 30% of baseline.

Table 6—Predictors of 30-d Mortality AfterFibrinolysis for Acute MI: the GUSTO-1 Model*

Variables Adjusted x2 p Value

Age 717† , 0.00001Systolic BP 550† , 0.00001Killip class 350 (3 df)† , 0.00001Heart rate, beats/min 272 (2 df)† , 0.00001Location of infarction 143 (2 df)† , 0.00001Previous infarction 64† , 0.00001Age-by-Killip class interaction 29† , 0.00001Height, cm 31 (4 df)† , 0.00001Time to treatment 23† , 0.00001Diabetes 21† , 0.00001Weight, kg 16‡ , 0.0001Smoking 22 (2 df)‡ , 0.0001Choice of fibrinolytic therapy 15 (3 df)‡ , 0.0001Previous bypass surgery 16‡ , 0.0001Hypertension 14§ , 0.01Prior cerebrovascular disease 10§ , 0.01

*Adapted from Lee et al84 with permission; df 5 degrees of freedom.†p , 0.00001.‡p , 0.0001.§p , 0.001.

Table 7—Predictors of Intracranial Hemorrhage AfterFibrinolysis for Acute MI: the GUSTO-1 Model*

Variables x2 p Value

Age 64.94 0.0001Weight 34.63 0.0001Prior cerebrovascular disease 20.50 0.0001Diastolic BP 14.07 0.0002Combination alteplase-streptokinase therapy 14.01 0.0002Hypertension 7.44 0.0065Hypertension times age interaction term 5.90 0.0152Accelerated alteplase treatment 3.98 0.0460Systolic BP 3.88 0.0488

*C-index value, 0.53; adapted from Gore et al89 with permission.

CHEST / 119 / 1 / JANUARY, 2001 SUPPLEMENT 261S

The benefits of accelerated alteplase seen in theGUSTO-1 and TIMI-4 trials vs the lack of benefit seen inGISSI-2 and ISIS-3 reflects two factors: the alteplaseregimen and the heparin dosing. The former trials usedthe accelerated alteplase regimen, which results in ahigher rate of early patency compared with the older, 3-hregimen42 and early, IV heparin, which improves lateinfarct-artery patency. In contrast, the GISSI-2 and ISIS-3trials used the slower infusion of alteplase or duteplase anddelayed, subcutaneous heparin, which does not elevate theAPTT level until approximately 24 h after the start oftreatment. Reocclusion of an open infarct-related artery,which is associated with a threefold increase in mortality,occurs most often during this period. Thus, a subcutane-ous heparin regimen cannot prevent this important pre-dictor of poor outcomes.103,104

The benefits of accelerated alteplase and IV heparin,then, are based on the ability to achieve rapid, sustainedinfarct-artery patency after acute MI. This link betweenearly reperfusion, especially achievement of TIMI grade 3flow, and improved survival was established in theGUSTO-1 angiographic substudy.12,76 Thus, if newer reg-imens, such as fibrinolytic therapy plus platelet GP IIb/IIIa inhibition,105,106 can further improve early reperfu-sion, further reductions in mortality can be expected.

Double-Bolus Alteplase: Initial interest in a double-bolus regimen of alteplase came from a series of patientsto whom two 50-mg boluses of alteplase were given 30 minapart. TIMI grade 3 flow was achieved in 88% of patients,a considerably higher rate than in previous studies.107 In alater randomized trial,108 however, double-bolus alteplaseresulted in TIMI grade 3 flow in only 58% of patients

Figure 2. Patency rates (TIMI grade 2 or 3 flow) in patients treated with streptokinase, acceleratedalteplase, standard-dose alteplase, or anistreplase for acute MI, vs control subjects. Adapted fromGranger et al,42 with permission.

Table 8—Placebo-Controlled Mortality Trials of Streptokinase*

Variables GISSI-18 ISAM91 ISIS-210 EMERAS92

Patients, No. 11,806 1,741 17,187 3,568Sites, No. 176 38 417 236Dose/duration 1.5 MU/1 h 1.5 MU/1 h 1.5 MU/1 h 1.5 MU/1 hEnrollment period, mo/yr 02/1984–06/1985 03/1982–03/1985 03/1985–12/1987 01/1988–01/1991Placebo blinding No Yes Yes YesAge criteria All , 75 yr All AllSymptom duration, h , 12 , 6 , 24 6–24ECG criteria ST-segment 1 or 2 ST-segment 1 None NoneAspirin 6 Yes Randomized YesHeparin 6 IV 6 6Mortality follow-up, d In-hospital 21 35 35

*EMERAS 5 Estudio Multricentrico Estreptoquinasa Republicas de America del Sur.

262S Sixth ACCP Consensus Conference on Antithrombotic Therapy

compared with a 66% rate in patients treated with theaccelerated, 90-min infusion of alteplase. Further, theContinuous Infusion vs Double-Bolus Administration ofAlteplase trial,109 which compared double-bolus vs accel-erated infusion dosing of alteplase, was terminated earlybecause of concern about the safety of the double-bolusregimen. Thirty-day mortality tended to be higher in thedouble-bolus group than in the accelerated-infusion group(7.98% vs 7.53%). Statistically, double-bolus alteplase wasnot equivalent to the infusion regimen. Rates of hemor-rhagic stroke were 1.12% after double-bolus alteplasecompared with 0.81% after accelerated infusion of alte-plase (p 5 0.23).109 Based on these data, double-bolusalteplase is not recommended for general clinical use, andthe accelerated, 90-min infusion remains the currentstandard dosing for alteplase treatment of acute MI.

Cost-Effectiveness of Alteplase: A formal cost-effective-ness analysis was incorporated into the GUSTO-1 protocolas a substudy to be carried out in the United States andCanada.110 At 1 year, alteplase-treated patients had bothhigher costs ($2,845) and higher survival (an absolute 1.1%higher rate, or 11 more patients surviving per 1,000patients treated) compared with streptokinase-treated pa-tients. The incremental cost-effectiveness ratio was$32,678 per year of life saved.110 The use of alteplase inpatients with anterior MI yielded even more favorablecost-effectiveness values but less in inferior infarction andyoung patients. Thus, the cost-effectiveness of alteplasecompared with streptokinase compares favorably with thatof other therapies, such as hemodialysis for end-stagerenal disease ($35,000 to $50,000 per year of life saved).

Potential Advantages of Bolus Fibrinolytic AgentsAdministration of fibrinolytic agents as a bolus has

several potential advantages. First, its ease of administra-tion could aid in more rapid treatment of acute MI, whichhas been shown to improve survival.8,111 Reducing thetime to treatment, particularly the “door-to-drug” time,has been identified as a critical target by the NationalHeart Attack Alert Program.112 An increased door-to-drugtime has recently been shown to relate directly to in-creased mortality.113 The time from “the decision” to “thestart of drug” can be reduced if a simple, bolus fibrinolyticagent is available. The advantage of single-bolus therapy inrelationship to compliance was established in ISIS-3, inthat 95% of patients assigned to anistreplase treatmentactually received the drug compared with only 89% and90% of patients in the alteplase and streptokinase groups,respectively.101 Further, in a study by Hilleman andSeyedrondbari,114 patients given double-bolus r-PA ther-apy received the drug 15 min sooner than did thosetreated with alteplase infusion.

Second, bolus fibrinolytic may make more feasible thepromising strategy of prehospital therapy.115–117 In anoverview of all trials, a 19% reduction in mortality withprehospital treatment compared with standard, hospital-based treatment was observed.115 In a single study,117

using bolus anistreplase treatment, mortality was reducedby . 50% when given before patients arrived at thehospital.

Another potential advantage of bolus fibrinolytic isfewer medication errors, which are associated with ad-verse outcomes and longer hospital stays in this popula-tion.118–120 With the bolus-plus-infusion regimen for alte-plase, for example, there is a surprisingly high percentageof patients with medication errors (an incorrect dose orinfusion duration). In GUSTO-1, 12% of the 41,021patients treated with alteplase or streptokinase infusionhad a medication error, although the 30-day mortality wassignificantly higher in patients with a medication errorthan in those given the correct dose (for alteplase, 7.7% vs5.5%; for streptokinase, 11.3% vs 6.4%; both p , 0.001).This observation is limited by the difficulty in adjusting forrisk factors for adverse events in this population.121 In theNational Registry of Myocardial Infarction involving. 71,000 patients, patients who received a dose of alte-plase . 1.5 mg/kg had a 2.3-fold increase in intracranialhemorrhage, with a multivariate risk ratio of 1.49, suggest-ing that medication errors with bolus and infusion fibrino-lytic therapy may be important.88

In the Intravenous n-PA for Treatment of InfarctingMyocardium Early (InTIME)-II trial,122 there were moredosing errors in the alteplase group than in the single-bolus n-PA group (7.3% vs 5.7%; p , 0.001). As was seenin GUSTO-1, mortality was higher among alteplase-treated patients with medication errors vs those receivingthe correct alteplase dose (12.5% vs 5.9%; p , 0.001).Interestingly, the same relationship was not seen forweight-adjusted n-PA. Intracranial hemorrhage also wassignificantly increased among alteplase-treated patientswith medication errors (1.4% vs 0.6% with the correctalteplase dose).122 For the double-bolus agent r-PA, therate of medication errors also has been low; only 1% ofpatients did not receive the full r-PA dose in one studycompared with 4% for alteplase (p 5 0.03).114

In summary, these data from several large trials showthat (1) bolus fibrinolytic therapy can reduce the rate ofmedication errors, and (2) medication errors may beassociated with increased rates of both mortality andintracranial hemorrhage.

Trials of r-PAr-PA was one of the first mutant t-PA molecules to

undergo extensive clinical testing. Early observations sug-gested that optimal therapeutic efficacy resulted whenr-PA was divided into two boluses (10 U 1 10 U) given 30min apart.123 This was followed by two angiographic trialscomparing alteplase with r-PA. The first, the ReteplaseAngiographic Phase II International Dose-finding (RAP-ID)-1 trial, examined three dosing strategies for r-PA(Table 11).124 These were compared with an infusion ofalteplase (100 mg delivered . 3 h). The TIMI grade 3flow rate at 90 min was 63% with r-PA compared with 49%with alteplase (p , 0.05). A second, larger trial (RAPID-2)compared the best regimen from RAPID-1 with acceler-ated alteplase.125 Once again, r-PA was found to besuperior to accelerated alteplase. When these two trialswere combined, the rate of TIMI grade 3 flow at 90 minwas 61% for r-PA (10 U 1 10 U) compared with 45% forthe accelerated alteplase regimen (p , 0.01). The 16%

CHEST / 119 / 1 / JANUARY, 2001 SUPPLEMENT 263S

absolute increase in TIMI grade 3 rate with r-PA overaccelerated alteplase was less than the 24% increase seenwith alteplase over streptokinase in the GUSTO-1 angio-graphic substudy, but this smaller difference translatedinto a much larger difference in mortality in the RAPIDtrials (3.1% for r-PA vs 8.4% for alteplase) in the twoRAPID trials.

The INJECT study125a examined whether double-bolusr-PA was at least equivalent to streptokinase in reducingmortality (Table 12). The 35-day mortality rate with r-PAwas 9% compared with 9.5% with streptokinase. The 95%CIs for the absolute mortality difference (0.5%; 95% CI,2 1.98 to 0.96) did not extend to a possible 1% higher

mortality rate with r-PA compared with streptokinase.This would suggest that r-PA is at least equivalent tostreptokinase and therefore superior to placebo in thisstudy. These findings served as the basis for US Food andDrug Administration approval for r-PA.

The GUSTO-3 study126 compared double-bolus r-PAwith accelerated alteplase. This was a superiority trial totest whether the reported 16% increase in TIMI grade 3flow with r-PA compared with t-PA would translate intoimproved 30-day mortality. A total of 15,059 patientspresenting within 6 h of MI symptom onset were enrolled.The primary end point of 30-day mortality was reached in7.47% of r-PA-treated patients and in 7.24% of alteplase-

Figure 3. Incidence of TIMI grade 2 or 3 flow 90 min after TNK-tPA or accelerated alteplasetreatment in the TIMI-10B trial. Reprinted from Cannon et al,22 with permission.

Table 9—Results From the GISSI-2/International Study and the ISIS-3 Trial*

TrialsStreptokinase,

sq HeparinStreptokinase,No Heparin

Alteplase,‡sq Heparin

Alteplase,‡No Heparin

Anistreplase,sq Heparin

Anistreplase,No Heparin

GISSI-2/International Study100

Patients, No. 10,396 10,372 10,361 10,407 — —35-d mortality, % 9.2 9.6 9.3 9.4 — —Total stroke, % 1.0 1.3 — —Intracranial hemorrhage, % 0.3 0.4 — —

ISIS-3101†Patients, No. 13,780 13,746 13,77335-d mortality, % 10.6 10.3 10.5Total stroke, % 1.1 1.5 1.4Intracranial hemorrhage, % 0.25 0.22 0.72 0.59 0.71 0.40

*sq 5 subcutaneous.†In the ISIS-3 study, the mortality in patients randomized to subcutaneous heparin therapy across the three arms was similar to those randomizedto no heparin therapy (10.3% vs 10.6%). When ISIS-3 and GISSI-2 were combined, the 35-day mortality was identical with t-PA andstreptokinase (10.0% vs 10.0%).

‡3-h infusion; duteplase was used in ISIS-3.

264S Sixth ACCP Consensus Conference on Antithrombotic Therapy

treated patients (p 5 0.6; Table 12).126 Therefore r-PAwas not superior to alteplase in this study. The 95% CI forthe absolute mortality difference of 0.23% ranged from1.11% in favor of alteplase to 0.66% in favor of r-PA. Usingthe 1% cutoff, these observations did not prove theequivalence of the two agents. The mortality rates weresimilar when patients were categorized into subgroups,including by age, infarct location, and enrolling region.There was an interaction between symptom duration andoutcomes with r-PA vs alteplase, which was of borderlinesignificance (p 5 0.05). The rates of stroke, bleeding, andintracranial hemorrhage did not differ significantly. Thelack of superiority of r-PA over alteplase in clinical out-come is consistent with the concept that at least a 20%absolute increase in TIMI grade 3 flow incidence isrequired to substantially improve mortality in acute MI.

Trials of TNK-tPAClinical testing of TNK-tPA began in the TIMI-10A

trial,16 with doses ranging from 5 to 50 mg. The trialshowed a greater incidence of TIMI grade 3 flow at 90 min(57 to 64%) in patients given 30 to 50 mg of TNK-tPA thanthose treated with lower doses (p 5 0.032). In TIMI-10B,22 a total of 886 patients were randomized to receiveeither accelerated alteplase or a 5-s to 10-s bolus of 30 mgor 50 mg of TNK-tPA. The 50-mg dose was discontinuedbecause of increased bleeding and replaced with a 40-mgdose. The 40-mg dose of TNK-tPA produced an incidenceof TIMI grade 3 flow at 90 min similar to that withalteplase (Fig 4)127; the 30-mg dose produced a signifi-cantly lower rate (54.6%, p 5 0.04 vs alteplase), and the50-mg dose produced a rate of 65.8% (p 5 not significant[NS]).22 The rate of TIMI grade 2 or 3 flow and TIMIframe counts at 90 min were similar between TNK-tPAand alteplase. At 60 min, there was no difference in therates of TIMI grade 3 flow or overall patency.

Weight-Adjusted Dosing: Both TIMI-10B and the As-sessment of the Safety and Efficacy of a New Thrombo-

lytic Agent (ASSENT)-1 study called for prospectivelydefined, weight-based analyses of efficacy.22,128,129 Therate of TIMI grade 3 flow was 62 to 63% for doses ofTNK-tPA of approximately 0.5 mg/kg or higher, but wasonly 51 to 54% at lower doses (p 5 0.028 across quintiles).When dose/weight was stratified into tertiles, the mediancorrected TIMI frame count was significantly lower (thatis, coronary flow was faster) in patients who received thehigher “weight-corrected” dose.129

Safety Results in TIMI-10B: During the first phase ofthe trial (that is, before the reduction in heparin dosagedescribed below), there were three intracranial hemor-rhages among 78 patients (3.8%; 95% CI, 0.8 to 10.8%)treated with the 50-mg dose of TNK-tPA. In the parallelASSENT-I trial, however, no intracranial hemorrhagesoccurred at this dose. This dose was dropped from furthertesting in TIMI-10B, and, at the same time, the doses ofheparin were reduced. Further analysis showed that theconcomitant heparin dose could have played a larger rolethan that of the TNK-tPA dose in defining the rate ofintracranial hemorrhage.

Initially in TIMI-10B and ASSENT-I, heparin dosingwas at the discretion of the treating physicians. However,a protocol amendment mandated that patients receive theafter dose of heparin: for patients . 67 kg, a 5,000-U bolusand 1,000 U/h infusion; for patients weighing # 67 kg, a4,000-U bolus and 800 U/h infusion. The amendment alsomandated that the heparin dose be adjusted according tothe nomogram beginning with the 6-h APTT.

The rates of both intracranial hemorrhage and seriousbleeding were lower after the protocol amendment, from2.2 to 0% for the 30-mg dose of TNK-tPA (p 5 0.047), andfrom 2.8 to 1.2% for alteplase (p 5 0.29; overall combinedp 5 0.04).22 The rates of intracranial hemorrhage weresimilarly and significantly reduced in the overall TNK-tPAexperience, combining the TIMI-10B and ASSENT-1data.130 Severe bleeding also decreased with the reducedheparin dosing, from 3 to 0% for the 30-mg dose of

Table 10—Results From the GUSTO-1 Trial*

GUSTO-19,12

Streptokinase PlusSubcutaneous

HeparinStreptokinase Plus

IV HeparinAccelerated Alteplase

Plus IV HeparinCombination Therapy

Plus IV Heparin† p Value‡

Patients, No. 9,796 10,377 10,344 10,328 —30-d mortality 7.2 7.4 6.3† 7.0 0.005†Net clinical benefit§ 7.7 7.9 6.9† 7.6 0.006†24-h mortality 2.8 2.9 2.3† 2.8 0.005†Intracranial hemorrhage 0.5 0.5 0.7† 0.9 0.03Congestive heart failure 17.5 16.8 15.2† 16.8 , 0.001Cardiogenic shock 6.9 6.3 5.1 6.1 , 0.001Angiographic substudy

Patients, No. 293 283 292 299 —Infarct-artery patency, 90 min\ 54 60 81 73 , 0.001TIMI grade 3 flow, 90 min 29 32 54 38 , 0.001

*Data are presented as % unless otherwise indicated.†Reduced-dose alteplase and reduced-dose streptokinase.‡For alteplase vs both streptokinase arms combined.§Death or nonfatal disabling stroke.\TIMI grade 2 or 3 flow.

CHEST / 119 / 1 / JANUARY, 2001 SUPPLEMENT 265S

TNK-tPA (p 5 0.02), and from 8 to 2% for alteplase(p 5 0.01; combined, p 5 0.001).22 Thus, the subsequentphase-III trial, ASSENT-2, used the lower-dose heparinregimen.

The rate of serious bleeding (noncerebral bleedingrequiring transfusion) was lower with TNK-tPA comparedwith alteplase in TIMI-10B. For alteplase, 7.0% of pa-tients required transfusion compared with 1.0% of pa-tients treated with 30 mg of TNK-tPA (p , 0.001) and1.3% of those treated with 40 mg of TNK-tPA(p , 0.01).131 Similar low rates were observed in theASSENT-1 trial.131 Thus, there was early evidence that thevery fibrin-specific agent TNK-tPA might be associatedwith lower rates of bleeding than alteplase.

ASSENT-I: ASSENT-I was a randomized trial of threedoses of TNK-tPA, with its primary goal to determine therate of intracranial hemorrhage with each dose, to assist indetermining the appropriate dose for a large, phase IIItrial. A total of 3,235 patients were randomized to receive30 mg of TNK-tPA (n 5 1,705), 40 mg of TNK-tPA(n 5 1,457), or 50 mg of TNK-tPA (n 5 73).128 As notedabove, the 50-mg dose was discontinued and replaced by40 mg because of increased bleeding observed in theTIMI-10B study. Intracranial hemorrhage occurred in0.77% of patients overall: 0.94% in the 30-mg arm and0.62% in the 40-mg arm. No strokes were found in the 73patients treated with 50 mg of TNK-tPA. Among patientstreated within 6 h of symptom onset, the rates of intracra-nial hemorrhage were 0.56% with 30 mg of TNK-tPA and0.58% with 40 mg of TNK-tPA. Death, nonfatal stroke, orsevere bleeding complications occurred in low proportionsof patients: 6.4%, 7.4%, and 2.8%, in the 30-mg, 40-mg,and 50-mg arms, respectively.

ASSENT-2: TNK-tPA was compared with acceleratedalteplase in this large equivalence mortality trial of pa-tients with acute ST-segment elevation MI presentingwithin 6 h of the onset of chest pain. The study enrolled16,950 patients worldwide. TNK-tPA was given as aweight-adjusted dose of 0.53 mg/kg given in 5-mg incre-ments, ranging from 30 to 50 mg.127

Overall mortality essentially was identical between thetwo agents: 6.17% for TNK-tPA vs 6.15% for alteplase.This trial was an “equivalence” trial,132 and under itsprospectively defined criteria, TNK-tPA was shown to beequivalent to alteplase. The relative risk of 30-day mortal-ity was 1.00 for TNK-tPA vs alteplase (90% CI, 0.91 to1.10; p value for equivalence 5 0.028). The equivalence ofthe agents in reducing mortality was shown in nearly everysubgroup tested.

Intriguingly, patients treated . 4 h after symptom onsethad improved outcomes compared with such patientstreated with alteplase. This benefit may relate to thegreater fibrin specificity of TNK-tPA. The first observationof a benefit of greater fibrin specificity in patients treated. 4 h came from the TIMI-1 trial, in that 90-min patencywas preserved in patients treated with alteplase, whethertreated before or after 4 h of symptoms, but patency wassignificantly worse in patients who received streptokinaseafter 4 h rather than before.26,95 Similar findings were seenin an analysis of German angiographic fibrinolytic tri-als.133,134 The same pattern favoring the more fibrin-

specific agent was seen in the GUSTO-III trial, in thatpatients treated . 4 h after symptom onset had signifi-cantly lower mortality with alteplase compared with r-PA,a less fibrin-specific agent.126 The occlusive clot may bemore resistant the longer it has been able to mature, andthe greater fibrin specificity of a fibrinolytic agent mayenhance its ability to lyse the clot.

Safety Observations: In ASSENT-2, the rates of intra-cranial hemorrhage were nearly identical for TNK-tPAand alteplase (0.93% and 0.94%, respectively), as were theoverall rates of stroke (1.78% and 1.66%).127 Of note,patients aged . 75 years showed a trend toward reducedintracranial hemorrhage with TNK-tPA vs alteplase treat-ment (1.7% vs 2.6%). The group at the highest risk forintracranial hemorrhage was elderly female patientsweighing # 67 kg,135 which has been noted in two previ-ous multivariable analyses.88,136 It is encouraging that therate of intracranial hemorrhage in this high-risk group wasonly 1.1% after treatment with TNK-tPA compared with3.0% for those treated with alteplase (multivariate ad-justed odds ratio, 0.30; 95% CI, 0.09 to 0.98; p , 0.05).135

In all other patients, the intracranial hemorrhage rateswere similar between the two groups.

These benefits with regard to intracranial hemorrhagewere paralleled by significantly lower rates of majorbleeding. In the trial as a whole, the rates of majorbleeding were 4.7% for TNK-tPA and 5.9% for alteplase(p 5 0.0002).127 Overall bleeding likewise occurred infewer patients treated with TNK-tPA (p 5 0.0003).127,135

Similarly, the rate of bleeding requiring transfusion wassignificantly lower with TNK-tPA.

In summary, the single-bolus agent TNK-tPA showspromise based on the data from the ASSENT-2 trial.Mortality with this agent is similar to that with alteplase,and major bleeding is lower. The US Food and DrugAdministration recently approved TNK-tPA for use.

Other Fibrinolytic AgentsA number of other fibrinolytic agents have been tested

in humans. The agents described in this section aregenerally not approved for use or are in limited use aroundthe world. However, some of these agents have beenextensively tested and some are currently being evaluated.

Urokinase: This agent was one of the first testedfibrinolytic agents in humans. Despite this, it has under-gone relatively sparse clinical trials evaluation. A numberof smaller angiographic trials were carried out in the1980s.65,70,137,138 These collectively showed angiographicpatency and TIMI grade 3 flow rates that in general weresuperior to streptokinase and similar to those observedwith t-PA, particularly when the higher dose of 3 millionunits was used. Urokinase was tested in a trial of 2,201patients with a dose of 2 million units of urokinase plusheparin against heparin alone.139 At 16 days, the mortalitywas 8% in the urokinase plus heparin group and 8.3% inthe heparin-alone group. This agent therefore offers littleclinical advantage over existing agents and is predomi-nantly used for peripheral vascular thrombotic occlusions.

scu-PA: scu-PA is a naturally occurring protein that isproduced through recombinant DNA technology. It has a

266S Sixth ACCP Consensus Conference on Antithrombotic Therapy

Kringle domain that is homologous to plasminogen andnaturally occurring t-PA that makes it relatively fibrinspecific. Limited hydrolysis by plasmin converts the mol-ecule to a two-chain urokinase-type plasminogen activator.This agent has undergone limited clinical testing. In aninitial angiographic trial, scu-PA was associated with ahigher patency rate compared with streptokinase alone(72% vs 48%; p , 0.001) at 60 min after initiation oftherapy.43 The effect was only modest at the 90-minangiography (71% vs 64%). An angiographic trial compar-ing it against 3 h of t-PA was also carried out in 473patients with acute MI.140 At 60 min, the patency rateswere similar with either regimen (scu-PA, 80%; t-PA,75%). This was followed by a randomized equivalencetrial.141 A total of 3,089 patients were randomized totreatment with either 80 mg of scu-PA or 1.5 million unitsof streptokinase with IV heparin in both groups.142 The30-day mortality rates were 5.7% for scu-PA and 6.7% forstreptokinase (absolute difference, 1.01%; 95% CI,2 2.78% to 0.75%). The CIs do not extend to 1%;therefore, a placebo response can be excluded. However,there was a significant higher rate of intracranial hemorrhagewith scu-PA (0.9%) compared with streptokinase (0.3%;p 5 0.038), calling into question the validity in calling scu-PAequivalent to streptokinase.142

n-PA: This is a single-bolus agent that has undergoneextensive testing. It is a modified t-PA molecule that hashad the fibronectin finger-like and the epidermal growthfactor domains deleted. In addition, a point mutation atsite 117 has been used. These changes have led to asubstantial longer half-life compared with other agents.143

It was initially tested in an angiographic trial144 with dosesranging from 15 to 120 KU/kg in 602 patients with MI. Atthe highest dose, the patency rate was higher with n-PAcompared with accelerated t-PA (83% vs 71%; p , 0.05).There was a trend toward higher TIMI grade 3 flow rates(57% vs 46%, respectively). n-PA was also tested in a largephase-III randomized equivalence trial.144 A total of15,078 patients were treated with n-PA using the 120-KU/kg dose vs accelerated t-PA. Overall 30-day mortalitywas similar between the two agents (n-PA, 6.7% vs t-PA,6.6%; p , 0.05 for equivalence). However, intracranialhemorrhage was significantly higher with n-PA comparedwith t-PA (1.13% vs 0.62%; p , 0.003). This agent is notbeing developed for commercial use.

Staphylokinase: Recombinant staphylokinase is a 136amino acid single chain that activates plasminogen toplasmin with high fibrin specificity, which has undergone

limited testing in humans.145 In a randomized angio-graphic trial,146 48 patients were allocated to staphyloki-nase (double bolus of either 10 mg or 20 mg) and had aTIMI grade 3 flow rate of 62% compared with 58% for 52patients given t-PA. The highest dose (20 mg) had thehighest TIMI grade 3 rate (74%). In a further study of 82patients, a bolus and infusion were tested (15 mg, 30 mg,or 45 mg). The 90-min TIMI grade 3 flow rates showed noevidence of a dose response, with rates from 62% with thelowest dose to 63% with the highest dose.147

Complications of IV Fibrinolytic TherapyThe main complication of fibrinolytic therapy is bleed-

ing, with the most severe bleeding complication beingintracranial hemorrhage. The FTT reported, in theiroverview of nine trials that randomized 58,600 patients, anexcess of 3.9 strokes per 1,000 patients treated withfibrinolysis vs placebo (Table 13).24 The excess stroke riskassociated with fibrinolytic therapy largely is attributableto the excess risk of intracranial hemorrhage that occurs inthe first day after such treatment. In the GUSTO-1 trial of41,021 patients, 268 patients suffered an intracranialhemorrhage, of whom 160 patients (59.7%) died by 30days.89,148 Clinical predictors of intracranial hemorrhageare shown in Table 7.89 Multivariable predictors of mor-tality after an intracranial hemorrhage included Glasgowcoma scale score, shorter time from fibrinolytic therapy tostroke onset, and total hemorrhage volume,148 baselineclinical predictors of overall mortality in this population,84

of that age of the patient was the most important.Regarding noncerebral bleeding, the FTT24 defined

major bleeding events as those that were considered lifethreatening or required blood transfusion. They reporteda 1.1% incidence among patients receiving fibrinolytictherapy compared with 0.4% among those receiving pla-cebo, an increase of seven major bleeds per 1,000 patientsso treated.24 In the most comprehensive report on nonce-rebral bleeding after fibrinolytic therapy to date, Berkow-itz and other GUSTO-1 colleagues149 defined severebleeding as that causing substantial hemodynamic com-promise requiring intervention, and moderate bleeding asthat requiring transfusion but without associated hemody-namic compromise. Table 14 displays bleeding accordingto treatment assignment in GUSTO-1. The most commoncause of bleeding in GUSTO-1 was the use of coronaryrevascularization procedures. The most powerful multiva-riable predictors of moderate or severe bleeding in GUS-

Table 11—Dosing Regimens and Results of the RAPID-1 and RAPID-2 Trials

Study Fibrinolytic AgentPatients,

No.Patency at90 min, %

TIMI Grade 3 Flowat 90 min, %

Smalling et al124 r-PA 15 U 146 63 41r-PA 10 U plus 5 U 152 67 46r-PA 10 U plus 10 U 154 85 63Alteplase 100 mg/3 h 154 77 49

Bode et al125 r-PA 10 U plus 10 U 169 83 60Alteplase (accelerated) 155 73 45

CHEST / 119 / 1 / JANUARY, 2001 SUPPLEMENT 267S

TO-1 were advanced age, lighter body weight, and femalesex (Table 15).149 These variables remained the mostpotent predictors of bleeding risk even among patientswho did not undergo an in-hospital cardiac procedure.Regarding the issue of fibrin specificity, even after adjust-ment for the APTT level, the bleeding was greater withstreptokinase use than with alteplase use. More recently,similar observations have been made with TNK-tPA overalteplase.143

2. Adjunctive Treatment With DirectThrombin Inhibitors

Antithrombin Agents

Thrombin plays a central role in arterial thrombosis. Itconverts fibrinogen to fibrin in the final common pathwayfor clot formation, it is a powerful stimulus for plateletaggregation, and it activates factor XIII, which leads tocross-linking and stabilization of the fibrin clot. Thrombinmolecules are incorporated into coronary thrombi and canform the nidus of rethrombosis (that is, reocclusion orreinfarction) as the thrombus undergoes fibrinolysis.

Heparin: Heparin is unable to inhibit clot-boundthrombin, largely because the heparin-binding site ofthrombin is not accessible when it is bound to the fibrinclot.150 Nevertheless, heparin is an important adjunctiveagent after treatment with t-PA-derived agents (TNK-tPA,alteplase, and r-PA) to maintain infarct-related arterypatency.73,151,152 An important lesson learned in the TIMI-10B and ASSENT-2 trials,130 also evident in TIMI-9 andGUSTO-2 studies,153–156 is that lower doses of heparinwith thrombolysis are associated with reduced rates ofintracranial and major hemorrhage. A recent overview157

of all major fibrinolytic trials, and of detailed informationfrom the TIMI and InTIME-2 trials, confirmed that lowerdoses of heparin are associated with reduced intracranialhemorrhage. Based on the emerging data, the 1999 updateto the American College of Cardiology/American HeartAssociation guidelines for the management of acute MIrecommends a new, lower dose of heparin: a bolus of 60U/kg (up to 4,000 U) and an initial infusion of 12 U/kg/h(up to 1,000 U/h).158

Clinical trials159–163 involving unfractionated heparinhave used universal therapeutic APTT ranges—typically

Figure 4. Dosing schedule of weight-adjusted TNK-tPA used in the ASSENT-2 trial. Reprinted withpermission from Lancet.127

Table 12—Large, Comparative Trials of r-PA

Study Fibrinolytic AgentPatients,

No.Mortality at30–35 d, %

IntracranialHemorrhage, %

INJECT125a r-PA 10 U plus 10 U 3,004 9.02 0.77Streptokinase 1.5 MU 3,006 9.53 0.37

GUSTO-3126 r-PA 10 U plus 10 U 10,138 7.47 0.91Accelerated alteplase 4,921 7.24 0.87

268S Sixth ACCP Consensus Conference on Antithrombotic Therapy

50 to 70 s—regardless of the responsiveness of thethromboplastin reagent in use at the participating institu-tions. This responsiveness has been shown to have signif-icant variation, similar to that of prothrombin time re-agents.

Consensus conferences of the College of AmericanPathologists,164 and the American College of Chest Phy-sicians,165 and other sources159,161,162,166 have recom-mended against generalization of therapeutic APTTranges. There is wide agreement that therapeutic APTTranges should be customized for the specific thromboplas-tin reagent in use. Because clinical trials have failed to doso, evidence-based recommendations for use of unfrac-tionated heparin for cardiac indications are difficult tomake. We have made recommendations based on theAPTT ranges as they are described in published studies.However, institutions that have established therapeuticAPTT ranges in the recommended fashion are encouragedto continue using them. The implementation of a discrep-ant universal therapeutic range at such an institution willlikely lead to systematic errors in heparin dosing.

2.2. Direct Thrombin Inhibitors: Direct thrombin in-hibitors also have undergone extensive evaluation in con-junction with fibrinolytic therapy. The prototypic agent ishirudin, a 65-amino-acid polypeptide derived from theleech Hirudo medicinalis, which acts as a potent, selectivethrombin inhibitor compared with heparin.166 Hirudinselectively binds in a 1:1 fashion to thrombin at two sites:the carboxy-terminus of hirudin binds to the substraterecognition site, the domain of thrombin that recognizesfibrinogen167 or the platelet,168 and the amino- terminus ofhirudin binds to the catalytic site of thrombin.167 Hirudindoes not inhibit other enzymes in the coagulation orfibrinolytic pathways, such as factor Xa, factor IX, kal-likrein, activated protein C, plasmin, or t-PA.166 Hirudindoes not bind covalently to thrombin; however, the disso-ciation rate is extremely slow, making hirudin an essen-tially irreversible inhibitor of thrombin.166,169 Hirudin isproduced in yeast by recombinant DNA technology. Sev-eral different hirudin preparations are available, includingdesirudin and lepirudin.170 Other direct thrombin inhibi-tors also are available, including bivalirudin,171–173 argatro-

ban,174 efegatran,175 and inogatran.176 Bivalirudin, which isbeing tested in a large mortality trial with streptokinase,contains three domains: the 12-amino-acid carboxy-termi-nus derived from hirudin; a four-amino-acid sequence(D-Phe-Pro-Arg-Pro), which binds to the catalytic site ofthrombin; and a linker region with the optimal length toallow binding of both inhibitory sites.171

Direct thrombin inhibitors inhibit all the major actionsof thrombin, including thrombin-induced generation offibrin, thrombin-induced platelet activation, and the auto-catalytic reaction of thrombin.166,177 Potential advantagesof hirudin over heparin are that hirudin can inhibitclot-bound thrombin,150 it is not inhibited by activatedplatelets,178 and it does not require a cofactor. Thus, it mayprovide a more stable anticoagulant response.177

The effects of desirudin with thrombolysis were testedin the TIMI-5, TIMI-6, TIMI-9, and GUSTO-2 tri-als.99,155,156,179 Hirudin provided a more stable APTT,which was within the target range almost twice as often. Noepisodes of thrombocytopenia were reported for hirudin.

In TIMI-5, a lower rate of reinfarction was observedwith hirudin than heparin (4.3% vs 11.9% for heparin;p 5 0.03) and a trend toward less reocclusion (1.6% vs6.7%; p 5 0.07).99 In the phase III, TIMI-9B trial, asimilar trend toward less reinfarction was noted duringhospitalization (2.3% vs 3.4%; p 5 0.07), but there was nodifference in the primary end point of death, MI, severecongestive heart failure, or shock at 30 days (12.9% forhirudin vs 11.9% for heparin; p 5 NS).155 Similarly, theincidence of death or MI did not differ between the twoanticoagulants (9.7% vs 9.5% for heparin; p 5 NS). Hiru-din was tested in . 12,000 patients across the spectrum ofacute coronary syndromes in the GUSTO-2b trial. Therewas significantly less reinfarction with hirudin (5.4% vs6.3% for heparin; p 5 0.04), but only a trend towardreduction in death or MI at 30 days (8.9% vs 9.8%;p 5 0.06).156 In patients with ST-segment elevation MI,the incidence of death or MI was slightly lower withhirudin (9.9% vs 11.3%; p 5 0.13). There was an intrigu-ing trend toward a greater benefit of hirudin in patientstreated with streptokinase vs alteplase in GUSTO-2b,180

but this was not observed in TIMI-9B.155

Table 13—Incidence and Etiology of Stroke With Fibrinolytic Therapy*

VariablesFibrinolytic

(n 5 29,315), No. (%)Control (n 5 29,285),

No. (%)Excess per 1,000 treated,

No. (SD)

Day 0 or 1Probable hemorrhage 78 (0.3) 4 (0.0) 2.5 (0.3)Other 88 (0.3) 35 (0.1) 1.8 (0.4)Total 166 (0.6) 39 (0.1) 4.3 (0.5)

Days 2 to 35Probable hemorrhage 31 (0.1) 11 (0.0) 0.7 (0.2)Other 139 (0.5) 172 (0.6) 2 1.1 (0.6)Total 170 (0.6) 183 (0.6) 2 0.4 (0.7)

Days 0 to 35Probable hemorrhage 111 (0.4) 16 (0.1) 3.2 (0.4)Other 229 (0.8) 208 (0.7) 0.7 (0.7)Total 340 (1.2) 224 (0.8) 3.9 (0.8)

*Adapted with permission by Lancet.24

CHEST / 119 / 1 / JANUARY, 2001 SUPPLEMENT 269S

In the Hirudin for the Improvement of Thrombolysis(HIT)-3 trial, excess intracranial hemorrhage was ob-served with lepirudin (3.4% vs 0%).181 In the HIT-4 trial,which enrolled 1,208 patients and used a lower dose oflepirudin, TIMI flow grade 3 was observed in 40.7% in thelepirudin and in 33.5% in the heparin group (p 5 0.16).182

No differences were seen between lepirudin and heparinin the rate of hemorrhagic stroke (0.2% vs 0.3%), reinfarc-tion (4.6% vs 5.1%), or mortality (6.8% vs 6.4%) at 30 days.Thus, lepirudin in conjunction with streptokinase did notsignificantly improve reperfusion or clinical outcomes inthis study.

Angiographic trials with other direct thrombin inhibi-tors also have been conducted. In a pilot study and theHirulog Early Reperfusion/Occlusion trial,172,173 a trendtoward greater early (90 to 120 min) TIMI grade 3 flowwas observed with the higher dose of bivalirudin com-pared with heparin in patients receiving streptokinase.Testing with other agents found modest or no improve-ments compared with heparin.174,175,183

Thus, there has not been a dramatic improvement inclinical outcomes with direct thrombin inhibitors as ad-juncts to fibrinolytic therapy in acute MI. Nevertheless, a

large trial now is evaluating the direct thrombin inhibitorbivalirudin, which is being compared with heparin as anadjunctive agent to streptokinase in the ongoing HirulogEarly Reperfusion/Occlusion-2 study.

Adjunctive Therapy With GP IIb/IIIa ReceptorBlockers

The platelet GP IIb/IIIa receptor antagonists have beenshown to be effective and safe in reducing the ischemiccomplications of percutaneous coronary intervention andreducing the composite of death or MI among patientspresenting with acute coronary syndromes without ST-segment elevation.184 The success of these agents in thesegroups of patients has led to a number of investigationscombining GP IIb/IIIa-receptor blockers with fibrinolytictherapy. Initial trials were performed with full doses ofboth agents (Table 16).185–188 These trials uniformlyshowed improvement in the angiographic or ECG mea-sures of reperfusion, but concerns were raised aboutbleeding risks with this combination therapy. More re-cently, data have been reported on two trials that havecombined the monoclonal antibody GP IIb/IIIa inhibitor,abciximab, with half-doses of fibrinolytic therapy (Table17).105,106 These investigations105,106 have been very prom-ising in showing that reperfusion can be achieved rapidlybeyond what is seen with fibrinolysis alone, and thatmyocardial perfusion can be improved as measured byECG findings. Several similar, dose-finding trials areevaluating the small-molecule inhibitors of GP IIb/IIIa inconjunction with reduced-dose fibrinolysis. Larger mortal-ity trials will need to be performed to provide definitiveevidence as to whether such combination therapies im-prove mortality without increasing the risk of intracranialhemorrhage to an unacceptable level.

Recommendations

1. Fibrinolytic Therapy1.1. We recommend that all patients with acute MI

who receive fibrinolytic therapy receive aspirin (165to 325 mg) on arrival to the hospital and dailythereafter (grade 1A).

1.2. We recommend that patients with ischemicsymptoms characteristic of acute MI for # 12 h whohave ST-segment elevation or left bundle-branchblock on the ECG receive IV fibrinolytic therapyunless they have contraindications (grade 1A).

Table 14—Moderate and Severe Bleeding by Fibrinolytic Assignment in GUSTO-1*

Bleeding RateStreptokinase Plus Subcutaneous

Heparin (n 5 9,608)

Streptokinase PlusIV Heparin(n 5 10,387)

Alteplase Plus IV Heparin(n 5 10,366)

Combination† Plus IV Heparin(n 5 10,341)

Moderate or severe 1,160 (11.8) 1,451 (14.0) 1,155 (11.1) 1,388 (13.4)Severe 117 (1.2) 151 (1.5) 92 (0.9) 135 (1.3)

*Data are presented as No. (%). Adapted from Berkowitz et al149 with permission.†Reduced-dose alteplase and streptokinase infusions.

Table 15—Multivariable Baseline Predictors ofModerate or Severe Bleeding from GUSTO-1

Variables x2 Odds Ratio (95% CI)

Enrollment in the United States 520.78 1.76 (1.08–2.85)Age (60 yr vs 50 yr) 222.05 1.30 (1.26–1.35)Weight, kg 163.88 —

100 vs 90 — 0.83 (0.73–0.95)85 vs 75 — 0.81 (0.78–0.85)

Female sex 72.84 1.42 (1.31–1.53)Streptokinase/IV heparin therapy† 32.82 —

US patients — 1.13 (1.01–1.26)Non-US patients — 1.63 (1.36–1.95)

Combination fibrinolytic therapy‡ 31.15 —US patients — 1.02 (0.92–1.14)Non-US patients — 1.66 (1.39–1.95)

Diastolic BP (90 mm Hg vs80 mm Hg)

28.51 0.94 (0.92–0.96)

Black race 10.17 1.33 (1.12–1.57)Alteplase treatment† 9.20 —

US patients — 0.85 (0.76–0.96)Non-US patients — 1.14 (0.94–1.38)

*Adapted from Berkowitz et al149 with permission.†vs streptokinase with subcutaneous heparin.‡Reduced-dose alteplase and streptokinase infusions.

270S Sixth ACCP Consensus Conference on Antithrombotic Therapy

1.3. For patients with symptoms characteristic ofacute MI and duration of 12 to 24 h who haveST-segment elevation or left bundle-branch block onthe ECG, we recommend that IV fibrinolytic therapyshould be considered (grade 2B).

1.4. We recommend that in patients with priorintracranial hemorrhage, any stroke within the pastyear, or active bleeding, clinicians do not administerIV fibrinolytic therapy (grade 1B).

1.5. We recommend that all patients with acute MIwho are candidates for fibrinolytic therapy receive itwithin 30 min after arrival to the hospital (grade 1A).

For patients with symptom duration # 12 h, werecommend administration of one of the fibrinolyticagents: streptokinase, anistreplase, or alteplase (allgrade 1A in comparison to placebo).

Remark: r-PA is equivalent to streptokinase.1.6. For patients with symptom duration 6 h, we

recommend the administration of alteplase overstreptokinase (grade 1A).

Remark: TNK-tPA is equivalent to alteplase.1.7. We recommend that patients with known

allergy or sensitivity to streptokinase receive alte-plase, TNK-tPA, or r-PA (grade 1C1).

2. Adjunctive Treatment With ThrombinInhibitors

2.1. Heparin2.1.1. For patients receiving streptokinase, we

recommend administration of subcutaneous unfrac-tionated heparin (12,500 U q12h for 48 h) (grade 2A).

2.1.2. For patients given streptokinase or anis-treplase, we recommend administration of IV unfrac-

tionated heparin only if they are at high risk ofsystemic or venous thromboembolism (anterior MI,existing heart failure, previous embolus, atrial fibril-lation, or left ventricular thrombus) (grade 1C).

Remark: Heparin should be given not earlier than4 h after therapy and when the APTT is , 70 s. Thetarget APTT should be 50 to 70 s, and the infusionshould continue for $ 48 h.

2.1.3. For patients receiving alteplase, r-PA, orTNK-tPA, we recommend administration of IV un-fractionated heparin for 48 h (grade 1B).

2.1.4. For patients receiving IV heparin with alte-plase, r-PA, or TNK-tPA, we recommend adminis-tration of either standard-dose unfractionated hepa-rin (5,000-U bolus followed by 1,000 U/h) (grade 2A)or weight-adjusted dosing (60-U/kg bolus [4,000 Umaximum] followed by 12 U/kg/h [1,000 U/h maxi-mum]) (grade 2C), both adjusted to maintain anAPTT of 50 to 70 s.

2.2. Direct Thrombin Inhibitors2.2.1. For patients with known or suspected hep-

arin-induced thrombocytopenia or thrombosis whoare receiving fibrinolytic therapy (either alteplase orstreptokinase), we recommend administration of IVhirudin (lepuridin 0.1-mg/kg bolus followed 0.15-mg/h infusion) (grade 2A).

References1 Fletcher AP, Alkjaersig N, Smyrniotis FE, et al. Treatment

of patients suffering from early, myocardial infarction withmassive and prolonged streptokinase therapy. Trans AssocAm Physicians 1958; 71:287–296

2 Yusuf S, Collins R, Peto R, et al. Intravenous and intracoro-

Table 16—Clinical Trials of Full-Dose Fibrinolytic Therapy With Platelet GP IIb/IIIa Inhibition*

Trial Year Patients, No. GP IIb/IIIa Agent Fibrinolytic Agent Primary Findings

Kleiman et al185 1993 68 m7E3 Alteplase First test of concept; 1 angiographic patencyOhman et al186 1997 180 Eptifibatide Alteplase 1 acute TIMI grade 3 flow and faster ECG

resolution vs alteplase alonePARADIGM187 1998 345 Lamifiban Alteplase or streptokinase 1 bleeding at highest doses but also faster

ST-segment resolution vs fibrinolytic alone

*Adapted from Topol188 with permission; PARADIGM 5 Platelet Aggregation Receptor Antagonist Dose Investigation and Reperfusion Gain inMyocardial Infarction.

Table 17—Completed Trials of Reduced-Dose Fibrinolysis and Platelet GP IIb/IIIa Inhibition*

Trials Year Patients, No. Primary Treatment Adjunctive Heparin Bolus Plus InfusionTIMI Grade 3 Flow

at 90 min, %

Antman et al106 1999 888 Alteplase plus abciximab 60-U/kg bolus; 7-U/kg/h infusion 78Alteplase plus abciximab 30-U/kg bolus; 4-U/kg/h infusion 69Alteplase alone 70-U/kg bolus; 15-U/kg/h infusion 62

SPEED105 2000 528 Abciximab plus r-PA 60-U/kg boluses 61†Abciximab plus r-PA 40-U/kg boluses 51†r-PA alone 70-U/kg boluses 47†

*SPEED 5 Strategies for Patency Enhancement in the Emergency Department.†60-min angiography.

CHEST / 119 / 1 / JANUARY, 2001 SUPPLEMENT 271S

nary fibrinolytic therapy in acute myocardial infarction:overview of results on mortality, reinfarction and side-effectsfrom 33 randomized controlled trials. Eur Heart J 1985;6:556–585

3 Rentrop KP, Blanke H, Karsch KR, et al. Acute myocardialinfarction: intracoronary application of nitroglycerin andstreptokinase. Clin Cardiol 1979; 2:354–363

4 Kennedy JW, Ritchie JL, Davis KB, et al. Western Wash-ington randomized trial of intracoronary streptokinase inacute myocardial infarction. N Engl J Med 1983; 309:1477–1482

5 Ritchie JL, Davis KB, Williams DL, et al. Global andregional left ventricular function and tomographic radionu-clide perfusion: the Western Washington IntracoronaryStreptokinase in Myocardial Infarction Trial. Circulation1984; 70:867–875

6 Mathey DG, Sheehan FH, Schofer J, et al. Time from onsetof symptoms to thrombolytic therapy: a major determinantof myocardial salvage in patients with acute transmuralinfarction. J Am Coll Cardiol 1985; 6:518–525

7 Yusuf S, Wittes J, Friedman L. Overview of results ofrandomized clinical trials in heart disease: 1. Treatmentsafter myocardial infarction. JAMA 1988; 260:2088–2093

8 Gruppo Italiano per lo Studio della Streptochinasinell’Infarto Miocardico (GISSI). Effectiveness of intrave-nous thrombolytic treatment in acute myocardial infarction.Lancet 1986; 1:397–402

9 The GUSTO Investigators. An international randomizedtrial comparing four thrombolytic strategies for acute myo-cardial infarction. N Engl J Med 1993; 329:673–682

10 ISIS-2 (Second International Study of Infarct Survival)Collaborative Group. Randomised trial of intravenous strep-tokinase, oral aspirin, both, or neither among 17,187 cases ofsuspected acute myocardial infarction: ISIS-2. Lancet 1988;2:349–360

11 AIMS Trial Study Group. Effect of intravenous APSAC onmortality after acute myocardial infarction: preliminary re-port of a placebo-controlled clinical trial. Lancet 1988;1:545–549

12 The GUSTO Angiographic Investigators. The effects oftissue plasminogen activator, streptokinase, or both on cor-onary-artery patency, ventricular function, and survival afteracute myocardial infarction. N Engl J Med 1993; 329:1615–1622

13 Weaver WD, Simes RJ, Betriu A, et al. Comparison ofprimary coronary angioplasty and intravenous thrombolytictherapy for acute myocardial infarction: a quantitative re-view. JAMA 1997; 278:2093–2098

14 Pennica D, Holmes WE, Kohr WJ, et al. Cloning andexpression of human tissue-type plasminogen activatorcDNA in E coli. Nature 1983; 301:214–221

15 Brener S, Topol E. Third-generation thrombolytic agents foracute myocardial infarction. In: Topol E, ed. Acute coronarysyndromes. New York, NY: Marcel Dekker, 1998; 167–192

16 Cannon CP, McCabe CH, Gibson M, et al. TNK-tissueplasminogen activator in acute myocardial infarction: resultsof the thrombolysis in myocardial infarction (TIMI) 10Adose-ranging trial. Circulation 1997; 95:351–356

17 Modi NB, Eppler S, Breed J, et al. Pharmacokinetics of aslower clearing tissue plasminogen activator variant, TNK-tPA, in patients with acute myocardial infarction. ThrombHaemost 1998; 79:134–139

18 Collen D, Stassen JM, Yasuda T, et al. Comparative throm-bolytic properties of tissue-type plasminogen activator andof a plasminogen activator inhibitor-1-resistant glycosylationvariant, in a combined arterial and venous thrombosis modelin the dog. Thromb Haemost 1994; 72:98–104

19 Keyt BA, Paoni NF, Refino CJ, et al. A faster-acting andmore potent form of tissue plasminogen activator. Proc NatlAcad Sci USA 1994; 91:3670–3674

20 Benedict CR, Refino CJ, Keyt BA, et al. New variant ofhuman tissue plasminogen activator (TPA) with enhancedefficacy and lower incidence of bleeding compared withrecombinant human TPA. Circulation 1995; 92:3032–3040

21 Tanswell P, Tebbe U, Neuhaus KL, et al. Pharmacokineticsand fibrin specificity of alteplase during accelerated infu-sions in acute myocardial infarction. J Am Coll Cardiol 1992;19:1071–1075

22 Cannon CP, Gibson CM, McCabe CH, et al. TNK-tissueplasminogen activator compared with front-loaded alteplasein acute myocardial infarction: results of the TIMI 10B trial.Circulation 1998; 98:2805–2814

23 Torr SR, Nachowiak DA, Fujii S, et al. “Plasminogen steal”and clot lysis. J Am Coll Cardiol 1992; 19:1085–1090

23a The Assessment of the Safety, and Efficacy of a NewThrombolytic (ASSENT-2) Investigators. Single-bolus te-necteplase compared with front-loaded alteplase in acutemyocardial infarction: the ASSENT-2 double-blind ran-domised trial. Lancet 1999; 354:716–722

24 Fibrinolytic Therapy Trialists’ (FTT) Collaborative Group.Indications for fibrinolytic therapy in suspected acute myo-cardial infarction: collaborative overview of early mortalityand major morbidity results from all randomised trials of. 1000 patients. Lancet 1994; 343:311–322

25 Anderson JL, Marshall HW, Askins JC, et al. A randomizedtrial of intravenous and intracoronary streptokinase in pa-tients with acute myocardial infarction. Circulation 1984;70:606–618

26 Chesebro JH, Knatterud G, Roberts R, et al. Thrombolysisin Myocardial Infarction (TIMI) Trial: Phase I. A compari-son between intravenous tissue plasminogen activator andintravenous streptokinase. Circulation 1987; 76:142–154

27 Timmis AD, Griffin B, Crick JC, et al. Anisoylated plasmin-ogen streptokinase activator complex in acute myocardialinfarction: a placebo-controlled arteriographic coronary re-canalization study. J Am Coll Cardiol 1987; 10:205–210

28 Collen D, Topol EJ, Tiefenbrunn AJ, et al. Coronarythrombolysis with recombinant human tissue-type plasmin-ogen activator: a prospective, randomized, placebo-con-trolled trial. Circulation 1984; 70:1012–1017

29 Topol EJ, O’Neill WW, Langburd AB, et al. A randomized,placebo-controlled trial of intravenous recombinant tissue-type plasminogen activator and emergency coronary angio-plasty in patients with acute myocardial infarction. Circula-tion 1987; 75:420–428

30 Cribier A, Berland J, Saoudi N, et al. Intracoronary strep-tokinase, OK! . . . intravenous streptokinase first? Heparin orintravenous streptokinase in acute infarction: preliminaryresults of a prospective randomized trial with angiographicevaluation in 44 patients. Haemostasis 1986; 16:122–129

31 Verstraete M, Bleifeld W, Brower RW, et al. Double-blindrandomised trial of intravenous tissue-type plasminogenactivator versus placebo in acute myocardial infarction(ECSG-1). Lancet 1985; 2:965–969

32 Guerci AD, Gerstenblith G, Brinker JA, et al. A randomizedtrial of intravenous tissue plasminogen activator for acutemyocardial infarction with subsequent randomization toelective coronary angioplasty. N Engl J Med 1987; 317:1613–1618

33 Armstrong PW, Baigrie RS, Daly PA, et al. Tissue plasmin-ogen activator: Toronto (TPAT) placebo-controlled random-ized trial in acute myocardial infarction. J Am Coll Cardiol1989; 13:1469–1476

34 Durand P, Asseman P, Pruvost P, et al. Effectiveness of

272S Sixth ACCP Consensus Conference on Antithrombotic Therapy

intravenous streptokinase on infarct size and left ventricularfunction in acute myocardial infarction. Clin Cardiol 1987;10:383–392

35 Bassand JP, Machecourt J, Cassagnes J, et al. Multicentertrial of intravenous anisoylated plasminogen streptokinaseactivator complex (APSAC) in acute myocardial infarction:effects on infarct size and left ventricular function. J Am CollCardiol 1989; 13:988–997

36 National Heart Foundation of Australia Coronary Throm-bolysis Group. Coronary thrombolysis and myocardial sal-vage by tissue plasminogen activator given up to 4 hoursafter onset of myocardial infarction. Lancet 1988; 1:203–208

37 Kennedy JW, Martin GV, Davis KB, et al. The WesternWashington Intravenous Streptokinase in Acute MyocardialInfarction Randomized Trial. Circulation 1988; 77:345–352

38 de Bono DP. The European Cooperative Study Group trialof intravenous recombinant tissue-type plasminogen activa-tor (rt-PA) and conservative therapy versus rt-PA and im-mediate coronary angioplasty. J Am Coll Cardiol 1988;12:20A–23A

39 White HD, Norris RM, Brown MA, et al. Effect of intrave-nous streptokinase on left ventricular function and earlysurvival after acute myocardial infarction. N Engl J Med1987; 317:850–855

40 O’Rourke M, Baron D, Keogh A, et al. Limitation ofmyocardial infarction by early infusion of recombinant tis-sue-type plasminogen activator. Circulation 1988; 77:1311–1315

41 Bassand JP, Faivre R, Becque O, et al. Effects of earlyhigh-dose streptokinase intravenously on left ventricularfunction in acute myocardial infarction. Am J Cardiol 1987;60:435–439

42 Granger CB, White HD, Bates ER, et al. A pooled analysisof coronary arterial patency and left ventricular functionafter intravenous thrombolysis for acute myocardial infarc-tion. Am J Cardiol 1994; 74:1220–1228

43 PRIMI Trial Study Group. Randomised double-blind trial ofrecombinant pro-urokinase against streptokinase in acutemyocardial infarction. Lancet 1989; 1:863–868

44 Verstraete M, Bernard R, Bory M, et al. Randomised trial ofintravenous recombinant tissue-type plasminogen activatorversus intravenous streptokinase in acute myocardial infarc-tion (ECSG-2). Lancet 1985; 1:842

45 Stack RS, O’Connor CM, Mark DB, et al. Coronary reper-fusion during acute myocardial infarction with a combinedtherapy of coronary angioplasty and high-dose intravenousstreptokinase. Circulation 1988; 77:151–161

46 Lopez-Sendon J, Seabra-Gomes R, Macaya C, et al. Intra-venous anisoylated plasminogen streptokinase activatorcomplex versus intravenous streptokinase in myocardialinfarction: a randomized multicenter study. Circulation1988; 78:II-277

47 Hogg KJ, Gemmill JD, Burns JM, et al. Angiographicpatency study of anistreplase versus streptokinase in acutemyocardial infarction. Lancet 1990; 335:254–258

48 Charbonnier B, Cribier A, Monassier JP, et al. Etudeeuropeenne multicentrique et randomisee de l’APSAC ver-sus streptokinase dans l’infarctus du myocarde. Arch MalCoeur Vaiss 1989; 82:1565–1571

49 Monnier P, Sigwart U, Vincent A, et al. Anisoylated plas-minogen streptokinase activator complex versus streptoki-nase in acute myocardial infarction. Drugs 1987; 33(suppl3):175–178

50 Six AJ, Louwerenburg HW, Braams R, et al. A double-blindrandomized multicenter dose-ranging trial of intravenousstreptokinase in acute myocardial infarction. Am J Cardiol1990; 65:119–123

51 Golf S, Vogt P, Kaufmann U, et al. Intravenous thrombolytictreatment for acute myocardial infarction: effects of earlyintervention and early examination. Acta Med Scand 1988;224:523–529

52 Ribeiro EE, Silva LA, Carneiro R, et al. A randomized trialof direct PTCA vs intravenous streptokinase in acute myo-cardial infarction. J Am Coll Cardiol 1991; 17:152A

53 Magnani B. Plasminogen Activator Italian Multicenter Study(PAIMS): comparison of intravenous recombinant single-chain human tissue- type plasminogen activator (rt-PA) withintravenous streptokinase in acute myocardial infarction.J Am Coll Cardiol 1989; 13:19–26

54 Cherng WJ, Chiang CW, Kuo CT, et al. A comparisonbetween intravenous streptokinase and tissue plasminogenactivator with early intravenous heparin in acute myocardialinfarction. Am Heart J 1992; 123:841–846

55 White HD, Rivers JT, Maslowski AH, et al. Effect ofintravenous streptokinase as compared with that of tissueplasminogen activator on left ventricular function after firstmyocardial infarction. N Engl J Med 1989; 320:817–821

56 Hillis LD, Borer J, Braunwald E, et al. High dose intrave-nous streptokinase for acute myocardial infarction: prelimi-nary results of a multicenter trial. J Am Coll Cardiol 1985;6:957–962

57 Spann JF, Sherry S, Carabello BA, et al. High-dose, briefintravenous streptokinase early in acute myocardial infarc-tion. Am Heart J 1982; 104:939–945

58 Rogers WJ, Mantle JA, Hood WP Jr, et al. Prospectiverandomized trial of intravenous and intracoronary streptoki-nase in acute myocardial infarction. Circulation 1983; 68:1051–1061

59 de Marneffe M, Van Thiel E, Ewalenko M, et al. High-doseintravenous thrombolytic therapy in acute myocardial infarc-tion: efficiency, tolerance, complications and influence onleft ventricular performance. Acta Cardiol 1985; 40:183–198

60 Smalling RW, Schumacher R, Morris D, et al. Improvedinfarct-related arterial patency after high dose, weight-adjusted, rapid infusion of tissue-type plasminogen activatorin myocardial infarction: results of a multicenter randomizedtrial of two dosage regimens. J Am Coll Cardiol 1990;15:915–921

61 Carney RJ, Murphy GA, Brandt TR, et al. Randomizedangiographic trial of recombinant tissue-type plasminogenactivator (alteplase) in myocardial infarction. J Am CollCardiol 1992; 20:17–23

62 Topol EJ, Morris DC, Smalling RW, et al. A multicenter,randomized, placebo-controlled trial of a new form ofintravenous recombinant tissue-type plasminogen activator(activase) in acute myocardial infarction. J Am Coll Cardiol1987; 9:1205–1213

63 TIMI Study Group. Immediate versus delayed catheteriza-tion and angioplasty after thrombolytic therapy for acutemyocardial infarction. N Engl J Med 1988; 260:2849–2858

64 Topol EJ, Ellis SG, Califf RM, et al. Combined tissue-typeplasminogen activator and prostacyclin therapy for acutemyocardial infarction. J Am Coll Cardiol 1989; 14:877–884

65 Califf RM, Topol EJ, Stack RS, et al. Evaluation of combi-nation thrombolytic therapy and timing of cardiac catheter-ization in acute myocardial infarction: results of thromboly-sis and angioplasty in myocardial infarction phase 5randomized trial. Circulation 1991; 83:1543–1556

66 Johns JA, Gold HK, Leinbach RC, et al. Prevention ofcoronary artery reocclusion and reduction in late coronaryartery stenosis after thrombolytic therapy in patients withacute myocardial infarction: a randomized study of mainte-nance infusion of recombinant human tissue-type plasmin-ogen activator. Circulation 1988; 78:546–556

CHEST / 119 / 1 / JANUARY, 2001 SUPPLEMENT 273S

67 Whitlow PL, Bashore TM. Catheterization/Rescue Angio-plasty After Thrombolysis (CRAFT) study: acute myocardialinfarction treated with recombinant tissue plasminogen ac-tivator versus urokinase. J Am Coll Cardiol 1991; 17:276A

68 Topol EJ, George BS, Kereiakes DJ, et al. A randomizedcontrolled trial of intravenous tissue plasminogen activatorand early intravenous heparin in acute myocardial infarction.Circulation 1989; 79:281–286

69 Grines CL, Nissen SE, Booth DC, et al. A prospective,randomized trial comparing combination half-dose tissue-type plasminogen activator and streptokinase with full-dosetissue-type plasminogen activator. Circulation 1991; 84:540–549

70 Neuhaus KL, Tebbe U, Gottwik M. Intravenous recombi-nant tissue plasminogen activator (rt-PA) and urokinase inacute myocardial infarction: results of the German activatorurokinase study (GAUS). J Am Coll Cardiol 1988; 12:581–587

71 Anderson JL, Becker LC, Sorensen SG, et al. Anistreplaseversus alteplase in acute myocardial infarction: comparativeeffects on left ventricular function, morbidity and 1-daycoronary artery patency. J Am Coll Cardiol 1992; 20:753–766

72 TIMI Study Group. Comparison of invasive and conserva-tive strategies after treatment with intravenous tissue plas-minogen activator in acute myocardial infarction: results ofthe Thrombolysis In Myocardial Infarction (TIMI) phase IItrial. N Engl J Med 1989; 320:618–627

73 de Bono DP, Simoons ML, Tijssen J, et al. Effect of earlyintravenous heparin on coronary patency, infarct size, andbleeding complications after alteplase thrombolysis: resultsof a randomised double blind European Cooperative StudyGroup trial. Br Heart J 1992; 67:122–128

74 Thompson PL, Aylward PE, Federman J, et al. A random-ized comparison of intravenous heparin with oral aspirin anddipyridamole 24 hours after recombinant tissue-type plas-minogen activator for acute myocardial infarction. Circula-tion 1991; 83:1534–1542

75 Rapold HJ, Kuemmerli H, Weiss M, et al. Monitoring offibrin generation during thrombolytic therapy of acute myo-cardial infarction with recombinant tissue-type plasminogenactivator. Circulation 1989; 79:980–989

76 Simes RJ, Topol EJ, Holmes DR Jr, et al. Link between theangiographic substudy and mortality outcomes in a largerandomized trial of myocardial reperfusion: importance ofearly and complete infarct artery reperfusion. Circulation1995; 91:1923–1928

77 Gibson CM, Cannon CP, Daley WL, et al. TIMI framecount: a quantitative method of assessing coronary arteryflow. Circulation 1996; 93:879–888

77a Braunwald E, Maroko PR, Libby P. Reduction of infarctsize following coronary occlusion. Circ Res 1974; 35(suppl3):192–201

78 Califf RM, O’Neill W, Stack RS, et al. Failure of simple clinicalmeasurements to predict perfusion status after intravenousthrombolysis. Ann Intern Med 1988; 108:658–662

78a Braunwald E, Maroko PR. ST-segment mapping realisticand unrealistic expectations. Circ 1976; 54:529–532

79 Schroder R, Dissmann R, Bruggemann T, et al. Extent ofearly ST segment elevation resolution: a simple but strongpredictor of outcome in patients with acute myocardialinfarction. J Am Coll Cardiol 1994; 24:384–391

80 Schroder R, Wegscheider K, Schroder K, et al, for theINJECT Trial Group. Extent of early ST segment elevationresolution: a strong predictor of outcome in patients withacute myocardial infarction and a sensitive measure to

compare thrombolytic regimens. J Am Coll Cardiol 1995;26:1657–1664

81 Neuhaus KL, Zeymer U, Tebbe U, et al. Resolution of STsegment elevation is an early predictor of mortality inpatients with acute myocardial infarction: meta analysis ofthree thrombolysis trials. Circulation 1998; 98:I-632

81a Anderson R, White H, Ohman EM, et al. Resolution ofST-segment elevation 90 minutes after thrombolysis foracute myocardial infarction predicts outcome: a GUSTO-IIIsubstudy. J Am Coll Cardiol 1998; 31(suppl A):371A

81b de Lemos JA, Antman EM, Giugliano RP. Comparison of a60-versus 90-minute determination of ST-segment resolu-tion after thrombolytic therapy for acute myocardial infarc-tion. Am J Cardiol 2000; 86:1235–1237

82 Roe MT, Ohman EM, Maas ACP, et al. Shifting theopen-artery hypothesis downstream: the quest for optimalreperfusion. J Am Coll Cardiol 2001; 37:9–18

83 Klootwijk P, Langer A, Meij S, et al. Non-invasive predictionof reperfusion and coronary artery patency by continuous STsegment monitoring in the GUSTO-I trial. Eur Heart J1996; 17:689–698

84 Lee KL, Woodlief LH, Topol EJ, et al. Predictors of 30-daymortality in the era of reperfusion for acute myocardialinfarction: results from an international trial of 41,021patients. Circulation 1995; 91:1659–1668

85 Gore JM, Sloan M, Price TR, et al. Intracerebral hemor-rhage, cerebral infarction, and subdural hematoma afteracute myocardial infarction and thrombolytic therapy in theThrombolysis In Myocardial Infarction study: ThrombolysisIn Myocardial Infarction, phase II; pilot and clinical trial.Circulation 1991; 83:448–459

86 Sloan MA, Gore JM. Ischemic stroke and intracranialhemorrhage after thrombolytic therapy for acute myocardialinfarction: a risk-benefit analysis. Am J Cardiol 1992; 69:21A–38A

87 Goldberg RJ, Gore JM, Gurwitz JH, et al. The impact of ageon the incidence and prognosis of initial acute myocardialinfarction: the Worcester Heart Attack Study. Am Heart J1989; 117:543–549

88 Gurwitz JH, Gore JM, Goldberg RJ, et al. Risk for intracra-nial hemorrhage after tissue plasminogen activator treat-ment for acute myocardial infarction: participants in theNational Registry of Myocardial Infarction 2. Ann InternMed 1998; 129:597–604

89 Gore JM, Granger CB, Sloan MA, et al. Stroke afterthrombolysis: mortality and functional outcomes in theGUSTO-I trial. Circulation 1995; 92:2811–2818

90 Granger CB, O’Connor CM, Bleich SD, et al. An overviewof the effect of intravenous heparin on clinical endpointsafter thrombolytic therapy for acute myocardial infarction.Circulation 1992; 86:I-259

91 The ISAM Study Group. A prospective trial of intravenousstreptokinase in acute myocardial infarction (ISAM): mor-tality, morbidity, and infarct size at 21 days. N Engl J Med1986; 314:1465–1471

92 EMERAS (Estudio Multicentrico Estreptoquinasa Repub-licas de America del Sur) Collaborative Group. Randomisedtrial of late thrombolysis in patients with suspected acutemyocardial infarction. Lancet 1993; 342:767–772

93 Van de Werf F, Ludbrook PA, Bergmann SR, et al. Coro-nary thrombolysis with tissue-type plasminogen activator inpatients with evolving myocardial infarction. N Engl J Med1984; 310:609–613

94 Van de Werf F, Bergmann SR, Fox KA, et al. Coronarythrombolysis with intravenously administered human tissue-type plasminogen activator produced by recombinant DNAtechnology. Circulation 1984; 69:605–610

274S Sixth ACCP Consensus Conference on Antithrombotic Therapy

95 The TIMI Study Group Thrombolysis in Myocardial Infarc-tion (TIMI) trial phase I findings. N Engl J Med 1985;312:932–936

96 Granger CB, Califf RM, Topol EJ. Thrombolytic therapy foracute myocardial infarction: a review. Drugs 1992; 44:293–325

97 Neuhaus KL, Feuerer W, Jeep-Tebbe S, et al. Improvedthrombolysis with a modified dose regimen of recombinanttissue-type plasminogen activator. J Am Coll Cardiol 1989;14:1566–1569

98 Neuhaus KL, von Essen R, Tebbe U, et al. Improvedthrombolysis in acute myocardial infarction with front-loaded administration of alteplase: results of the rt-PA-APSAC patency study (TAPS). J Am Coll Cardiol 1992;19:885–891

99 Cannon CP, McCabe CH, Diver DJ, et al. Comparison offront-loaded recombinant tissue-type plasminogen activator,anistreplase and combination thrombolytic therapy for acutemyocardial infarction: results of the Thrombolysis in Myo-cardial Infarction (TIMI) 4 trial. J Am Coll Cardiol 1994;24:1602–1610

100 GISSI-2, and International Study Group. Six-month survivalin 20,891 patients with acute myocardial infarction random-ized between alteplase and streptokinase with or withoutheparin. Eur Heart J 1992; 13:1692–1697

101 ISIS-3 (Third International Study of Infarct Survival Collab-orative Group). ISIS-3: a randomised comparison of strep-tokinase vs tissue plasminogen activator vs anistreplase andof aspirin plus heparin vs aspirin alone among 41,299 casesof suspected acute myocardial infarction. Lancet 1992;339:753–770

102 Cannon CP, Braunwald E. GUSTO, TIMI, and the case forrapid reperfusion. Acta Cardiol 1994; 49:1–8

103 Ohman EM, Califf RM, Topol EJ, et al. Consequences ofreocclusion after successful reperfusion therapy in acutemyocardial infarction. Circulation 1990; 82:781–791

104 Ohman EM, Topol EJ, Califf RM, et al. An analysis of thecause of early mortality after administration of thrombolytictherapy. Coron Artery Dis 1993; 4:957–964

105 Trial of abciximab with and without low-dose reteplase foracute myocardial infarction: strategies for Patency Enhance-ment in the Emergency Department (SPEED) Group.Circulation 2000; 101:2788–2794

106 Antman EM, Giugliano RP, Gibson CM, et al. Abciximabfacilitates the rate and extent of thrombolysis: results of thethrombolysis in myocardial infarction (TIMI) 14 trial. Cir-culation 1999; 99:2720–2732

107 Purvis JA, McNeill AJ, Siddiqui RA, et al. Efficacy of 100 mgof double-bolus alteplase in achieving complete perfusion inthe treatment of acute myocardial infarction. Am J Cardiol1994; 2:6–10

108 Bleich SD, Adgey AA, McMechan SR, et al. An angiographicassessment of alteplase: double-bolus and front-loaded infu-sion regimens in myocardial infarction. Am Heart J 1998;136:741–748

109 The COBALT Investigators. A comparison of continuousinfusion of alteplase with double-bolus administration foracute myocardial infarction. N Engl J Med 1997; 337:1124–1130

110 Mark DB, Hlatky MA, Califf RM, et al. Cost effectiveness ofthrombolytic therapy with tissue plasminogen activator ascompared with streptokinase for acute myocardial infarc-tion. N Engl J Med 1995; 332:1418–1424

111 TIMI IIIB Investigators. Effects of tissue plasminogenactivator and a comparison of early invasive and conservativestrategies in unstable angina and non-Q wave myocardialinfarction: results of the TIMI IIIB trial. Circulation 1994;89:1545–1556

112 National Heart Attack Alert Program Coordinating Commit-tee 6MtTWG. Emergency department: rapid identificationand treatment of patients with acute myocardial infarction.Ann Emerg Med 1994; 23:311–329

113 Cannon CP, Gibson CM. Longer thrombolysis door-to-needle times are associated with increased mortality in acutemyocardial infarction: an analysis of 85,589 patients in theNational Registry of Myocardial Infarction 213. J Am CollCardiol 2000 (in press)

114 Hilleman DE, Seyedroudbari A. Cost-minimization analysisof direct injection and infusion fibrinolytic therapy in acutemyocardial infarction: a multicenter study. Circulation 1999;98:I-381

115 European Myocardial Infarction Project Group. Prehospitalthrombolytic therapy in patients with suspected acute myo-cardial infarction. N Engl J Med 1993; 329:383–389

116 Weaver WD, Cerqueira M, Hallstrom AP, et al. Prehospital-initiated vs hospital-initiated thrombolytic therapy: the Myo-cardial Infarction Triage and Intervention Trial. JAMA 1993;270:1211–1216

117 Rawles J. Halving of mortality at 1 year by domiciliarythrombolysis in the Grampian Region Early AnistreplaseTrial (GREAT). J Am Coll Cardiol 1994; 23:1–5

118 Bates DW, Boyle DL, Vander Vliet MB, et al. Relationshipbetween medication errors and adverse drug events. J GenIntern Med 1995; 10:199–205

119 Bates DW, Cullen DJ, Laird N, et al. Incidence of adversedrug events and potential adverse drug events: implicationsfor prevention. JAMA 1995; 274:29–34

120 Bates DW, Spell N, Cullen DJ, et al. The costs of adversedrug events in hospitalized patients. JAMA 1997; 277:307–311

121 Vorchheimer DA, Baruch L, Thompson TD, et al. NorthAmerican vs Non-North American streptokinase use inGUSTO-I: impact of protocol deviation on mortality benefitof TPA. Circulation 1997; 96:I-535

122 Alexander JH, Al-Khatib S, Cantor W, et al. Highlights fromthe American College of Cardiology 48th Annual ScientificSessions, March 7 to 10, 1999. Am Heart J 1999; 138:175–190

123 Bode C, Nordt TK, Peter K, et al. Patency trials withreteplase (r-PA): what do they tell us? Am J Cardiol 1996;78:16–19

124 Smalling RW, Bode C, Kalbfleisch J, et al. More rapid,complete, and stable coronary thrombolysis with bolusadministration of reteplase compared with alteplase infusionin acute myocardial infarction. Circulation 1995; 91:2725–2732

125a International Joint Efficacy Comparison of Thrombolytics.Randomised, double-blind comparison of reteplase double-bolus administration with streptokinase in acute myocardialinfarction (INJECT): trial to investigate equivalence. [seecomments] [published erratum appears in Lancet 1995 Oct7;346(8980):980]. Lancet 1995; 346:329–336

125 Bode C, Smalling RW, Berg G, et al. Randomized compar-ison of coronary thrombolysis achieved with double-bolusreteplase (recombinant plasminogen activator) and front-loaded, accelerated alteplase (recombinant tissue plasmino-gen activator) in patients with acute myocardial infarction.Circulation 1996; 94:891–898

126 The Global Use of Strategies to Open Occluded CoronaryArteries (GUSTO III) Investigators. A comparison of rete-plase with alteplase for acute myocardial infarction. N EnglJ Med 1997; 337:1118–1123

127 Deleted in proof128 Van de Werf F, Cannon CP, Luyten A, et al. Safety

assessment of single-bolus administration of TNK tissue-

CHEST / 119 / 1 / JANUARY, 2001 SUPPLEMENT 275S

plasminogen activator in acute myocardial infarction: theASSENT- 1 trial. Am Heart J 1999; 137:786–791

129 Gibson CM, Cannon CP, Murphy SA, et al. Weight-adjusteddosing of TNK-tissue plasminogen activator and its relationto angiographic outcomes in the Thrombolysis In Myocar-dial Infarction 10B trial. Am J Cardiol 1999; 84:976–980

130 Giugliano RP, Cannon CP, McCabe CH, et al. Lower doseheparin with thrombolysis is associated with lower rates ofintracranial hemorrhage: results from TIMI 10B and AS-SENT I. Circulation 1997; 96:I-535

131 Fox NL, Cannon CP, Berioli S, et al. Rates of seriousbleeding events requiring transfusion in AMI patientstreated with TNK-tPA. J Am Coll Cardiol 1999; 33:279

132 Ware JH, Antman EM. Equivalence trials. N Engl J Med1997; 337:1159–1161

133 Cannon CP. Timely thrombolysis: synergism to open arter-ies and reduce mortality. Am Heart J 1999; 137:1–3

134 Zeymer U, Tebbe U, von Essen R, et al. Influence of time totreatment on early infarct-related artery patency after dif-ferent thrombolytic regimens. Am Heart J 1999; 137:34–38

135 Barron HV, Fox NL, Berioli S, et al. Comparison ofintracranial hemorrhage rates in patients treated with rt-PAand TNK-tPA: impact of gender, age and low body weight.Circulation 1999; 100:I-1

136 Simoons ML, Maggioni AP, Knatterud G, et al. Individualrisk assessment for intracranial hemorrhage during throm-bolytic therapy. Lancet 1993; 342:1523–1528

137 Mathey DG, Schofer J, Sheehan FH, et al. Intravenousurokinase in acute myocardial infarction. Am J Cardiol 1985;55:878–882

138 Wall TC, Phillips HRI, Stack RS, et al. Results of high doseintravenous urokinase for acute myocardial infarction. Am JCardiol 1990; 65:124–131

139 Rossi P, Bolognese L. Comparison of intravenous urokinaseplus heparin versus heparin alone in acute myocardialinfarction. Am J Cardiol 1991; 68:585–592

140 Bar FW, Meyer J, Vermeer F, et al. Comparison of saruplaseand alteplase in acute myocardial infarction: SESAM StudyGroup; the Study in Europe with Saruplase and Alteplase inMyocardial Infarction. Am J Cardiol 1997; 79:727–732

141 Tebbe U, Michels R, Adgey J, et al. Randomized, double-blind study comparing saruplase with streptokinase therapyin acute myocardial infarction: the COMPASS equivalencetrial. J Am Coll Cardiol 1998; 31:487–493

142 White HD. Thrombolytic therapy and equivalence trials[editorial]. J Am Coll Cardiol 1998; 31:494–496

143 den Heijer P, Vermeer F, Ambrosioni E, et al. Evaluation ofa weight-adjusted single-bolus plasminogen activator in pa-tients with myocardial infarction: a double-blind, random-ized angiographic trial of lanoteplase versus alteplase. Cir-culation 1998; 98:2117–2125

144 The InTIME-II Investigators. Intravenous NPA for thetreatment of single-bolus lanoteplase vs accelerated alte-plase for the treatment of patients with acute myocardialinfarction. Eur Heart J 2000; 2005–2013

145 Collen D, Van de Werf F. Coronary thrombolysis withrecombinant staphylokinase in patients with evolving myo-cardial infarction. Circulation 1993; 87:1850–1853

146 Vanderschueren S, Barrios L, Kerdsinchai P, et al. Arandomized trial of recombinant staphylokinase versus alte-plase for coronary artery patency in acute myocardial infarc-tion. Circulation 1995; 92:2044–2049

147 Armstrong PW, Burton JR, Palisaitis D, et al. Collaborativeangiographic patency trial of recombinant staphylokinase(CAPTORS). Am Heart J 2000; 139:820–823

148 Sloan MA, Sila CA, Mahaffey KW, et al. Prediction of 30-daymortality among patients with thrombolysis-related intracra-

nial hemorrhage. Circulation 1998; 98:1376–1382149 Berkowitz SD, Granger CB, Pieper KS, et al. Incidence and

predictors of bleeding after contemporary thrombolytictherapy for myocardial infarction. Circulation 1997; 95:2508–2516

150 Weitz JI, Hudoba M, Massel D, et al. Clot-bound thrombinis protected from inhibition by heparin-antithrombin III butis susceptible to inactivation by antithrombin III-indepen-dent inhibitors. J Clin Invest 1990; 86:385–391

151 Bleich SD, Nichols TC, Schumacher RR, et al. Effect ofheparin on coronary arterial patency after thrombolysis withtissue plasminogen activator in acute myocardial infarction.Am J Cardiol 1990; 66:1412–1417

152 Hsia J, Hamilton WP, Kleiman N, et al. A comparisonbetween heparin and low-dose aspirin as adjunctive therapywith tissue plasminogen activator for acute myocardial in-farction. N Engl J Med 1990; 323:1433–1437

153 Antman EM. Hirudin in acute myocardial infarction: safetyreport from the Thrombolysis and Thrombin Inhibition InMyocardial Infarction (TIMI) 9A trial. Circulation 1994;90:1624–1630

154 The Global Use of Strategies to Open Occluded CoronaryArteries (GUSTO) IIa Investigators. Randomized trial ofintravenous heparin versus recombinant hirudin for acutecoronary syndromes. Circulation 1994; 90:1631–1637

155 Antman EM. Hirudin in acute myocardial infarction:Thrombolysis and Thrombin in Myocardial Infarction(TIMI) 9B trial. Circulation 1996; 94:911–921

156 The Global Use of Strategies to Open Occluded CoronaryArteries (GUSTO) IIb Investigators. A comparison of re-combinant hirudin with heparin for the treatment of acutecoronary syndromes. N Engl J Med 1996; 335:775–782

157 Giugliano RP, Cutler SS, Llevadot J. Risk of intracranialhemorrhage with accelerated tPA: importance of the hepa-rin dose. Circulation 1999; 100:I-650

158 Ryan TJ, Antman EM, Brooks NH, et al. Update: ACC/AHAguidelines for the management of patients with acute myo-cardial infarction: a report of the American College ofCardiology/American Heart Association Task Force on Prac-tice Guidelines (Committee on Management of Acute Myo-cardial Infarction). J Am Coll Cardiol 1999; 34:890–911

159 Shojania AM, Tetreault J, Turnbull G. The variations be-tween heparin sensitivity of different lots of activated partialthromboplastin time reagent produced by the same manu-facturer. Am J Clin Pathol 1988; 89:19–23

160 Bain B, Forster T, Sleigh B. Heparin and the activatedpartial thromboplastin time: a difference between the in-vitro and in-vivo effects and implications for the therapeuticrange. Am J Clin Pathol 1980; 74:668–673

161 Baker BA, Adelman MD, Smith PA, et al. Inability of theactivated partial thromboplastin time to predict heparinlevels: time to reassess guidelines for heparin assays. ArchIntern Med 1997; 157:2475–2479

162 Brill-Edwards P, Ginsberg JS, Johnston M, et al. Establish-ing a therapeutic range for heparin therapy. Ann Intern Med1993; 119:104–109

163 Kitchen S, Preston FE. The therapeutic range for heparintherapy: relationship between six activated partial thrombo-plastin time reagents and two heparin assays. ThrombHaemost 1996; 75:734–739

164 Olson JD, Arkin CF, Brandt JT, et al. College of AmericanPathologists Conference XXXI on laboratory monitoring ofanticoagulant therapy: laboratory monitoring of unfraction-ated heparin therapy. Arch Pathol Lab Med 1998; 122:782–798

165 Hirsh J, Warkentin TE, Raschke R, et al. Heparin andlow-molecular-weight heparin: mechanisms of action, phar-

276S Sixth ACCP Consensus Conference on Antithrombotic Therapy

macokinetics, dosing considerations, monitoring, efficacy,and safety. Chest 1998; 114(suppl):489S–510S

166 Markwardt F. Hirudin and derivatives as anticoagulantagents. Thromb Haemost 1991; 66:141–152

167 Rydel TJ, Ravichandran KG, Tulinsky A, et al. The structureof a complex of recombinant hirudin and human-thrombin.Science 1990; 249:277–280

168 Vu TK, Hung DT, Wheaton VI, et al. Molecular cloning ofa functional thrombin receptor reveals a novel proteolyticmechanism of receptor activation. Cell 1991; 64:1057–1068

169 Stone SR, Hofsteenge J. Kinetics of the inhibition ofthrombin by hirudin. Biochemistry 1986; 25:4622–4628

170 Organisation to Assess Strategies for Ischemic Syndromes(OASIS-2) Investigators. Effects of recombinant hirudin(lepirudin) compared with heparin on death, myocardialinfarction, refractory angina, and revascularisation proce-dures in patients with acute myocardial ischemia without STelevation: a randomised trial. Lancet 1999; 353:429–438

171 Maraganore JM, Bourdon P, Jablonski J, et al. Design andcharacterization of hirulogs: a novel class of bivalent peptideinhibitors of thrombin. Biochemistry 1990; 29:7095–7101

172 Theroux P, Perez-Villa F, Waters D, et al. Randomizeddouble-blind comparison of two doses of Hirulog withheparin as adjunctive therapy to streptokinase to promoteearly patency of the infarct-related artery in acute myocar-dial infarction. Circulation 1995; 91:2132–2139

173 White HD, Aylward PE, Frey MJ, et al. Randomized,double-blind comparison of hirulog versus heparin in pa-tients receiving streptokinase and aspirin for acute myocar-dial infarction (HERO). Circulation 1997; 96:2155–2161

174 Jang IK, Giugliano RP, Massey TJ, et al. A randomized,blinded study of two doses of Novastan (brand of argatro-ban) versus heparin as adjunctive therapy to recombinanttissue plasminogen activator (accelerated administration) inacute myocardial infarction: rationale and design of theMyocardial Infarction using Novastan and T-PA (MINT)study. J Thromb Thrombol 1998; 5:49–52

175 Fung AY, Lorch G, Cambier PK, et al. Efegatran sulfate asan adjunct to streptokinase versus heparin as an adjunct totissue plasminogen activator in patients with acute myocar-dial infarction. Am Heart J 1999; 138:696–704

176 Thrombin Inhibition in Myocardial Ischemia (TRIM) studygroup. A low molecular weight, selective thrombin inhibitor,inogatran, vs heparin, in unstable coronary artery disease in1209 patients: a double-blind, randomized, dose-findingstudy. Eur Heart J 1997; 18:1416–1425

177 Talbot M. Biology of recombinant hirudin (CGP 39393): anew prospect in the treatment of thrombosis. SeminThromb Hemost 1989; 15:293–301

178 Fareed J, Walenga JM, Leya F, et al. Some objective

considerations for the use of heparins and recombinanthirudin in percutaneous transluminal coronary angioplasty.Semin Thromb Hemost 1991; 17:455–470

179 Lee LV. Initial experience with hirudin and streptokinase inacute myocardial infarction: results of the Thrombolysis inMyocardial Infarction (TIMI) 6 trial. Am J Cardiol 1995;75:7–13

180 Metz BK, White HD, Granger CB, et al. Randomizedcomparison of direct thrombin inhibition versus heparin inconjunction with fibrinolytic therapy for acute myocardialinfarction: results from the GUSTO-IIb trial. J Am CollCardiol 1998; 31:1493–1498

181 Neuhaus KL, von Essen R, Tebbe U, et al. Safety observa-tions from the pilot phase of the randomized r-Hirudin forImprovement of Thrombolysis (HIT-III) study: a study ofthe Arbeitsgemeinschaft Leitender Kardiologischer Kran-kenhausarzte (ALKK). Circulation 1994; 90:1638–1642

182 Neuhaus KL, Molhoek GP, Zeymer U, et al. Recombinanthirudin (lepirudin) for the improvement of thrombolysiswith streptokinase in patients with acute myocardial infarc-tion: results of the HIT-4 trial. J Am Coll Cardiol 1999;34:966–973

183 Gold HK, Torres FW, Garabedian HD, et al. Evidence fora rebound coagulation phenomenon after cessation of a4-hour infusion of a specific thrombin inhibitor in patientswith unstable angina pectoris. J Am Coll Cardiol 1993;21:1039–1047

184 Kong DF, Califf RM, Miller DP, et al. Clinical outcomes oftherapeutic agents that block the platelet glycoprotein IIb/IIIa integrin in ischemic heart disease. Circulation 1998;98:2829–2835

185 Kleiman NS, Ohman EM, Califf RM, et al. Profoundinhibition of platelet aggregation with monoclonal antibody7E3 fab after thrombolytic therapy: results of the TAMI 8pilot study. J Am Coll Cardiol 1993; 22:381–389

186 Ohman EM, Kleiman NS, Gacioch G, et al. Combinedaccelerated tissue-plasminogen activator and platelet glyco-protein IIb/IIIa integrin receptor blockade with Integrilin inacute myocardial infarction: results of a randomized, placebo-controlled, dose-ranging trial. Circulation 1997; 95:846–854

187 The PARADIGM Investigators. Combining thrombolysiswith the platelet glycoprotein IIb/IIIa inhibitor lamifiban:results of the Platelet Aggregation Receptor AntagonistDose Investigation and reperfusion Gain in Myocardialinfarction (PARADIGM) trial. J Am Coll Cardiol 1998;32:2003–2010

188 Topol EJ. Toward a new frontier in myocardial reperfusiontherapy: emerging platelet preeminence. Circulation 1998;97:211–218

CHEST / 119 / 1 / JANUARY, 2001 SUPPLEMENT 277S