JACC Vol. 16. No. 3 September 1990:533-46

One of the major causes of morbidity in patients with acute myocardial infarction is stroke (I 2). Bn the prethrombolytic era, the dominant form of stroke was a bland infarct arising

From the Divisions of Cardiology and Neurology, Department of Medi- cine, Duke University Medical Center, Durham, North Carolina; *Division of Cardiology, Department of internal Medicisle. University of Michigan Medi- cal Center, Ann Arbor, Michigan: iRiverside Methodist Hospital, Columbus, Ohio and Khrist Hospital. Cincinnati, Ohio. This work was supported by Research Grunt HS-t&G5 from the National Center for Health Services Research, Rockville. Maryland; Research Grant ML-36587 from the National Heart, Lung. and Blood Institute. Bethesda. Maryland: Training Grant LM-07003 and Grant LM-03373 from the National Library of Medicine. Bethesda; and grants from Genentech, Inc.. San Francisco. California. Abbott Laboratories, Chicago, Illinois. Andrew W. Mellon Foundation, New York, New York and the Robert Wood Johnson Foundation. Princeton. New Jersey.

Manuscript received August 16. 1989: revised manuscript received March 14. 1990, accepted April 5. f990.

Address for reorin& Christopher M. O’Connor, MD. Box 3356. Duke University Medical Center, Durham, North Carolina 27710.

~31990 by the American College of Cardiology

from thromboembohc disease, often in patients with atrial arrhythmias or a large anterior wall infarction, or both (t-3).

ore recently (4, intracranial hemorrhage has emerged as an infrequent but catastrophic complication of acute infarc- tion treated with thrombo~ytic therapy.

To date, few studies have examined in detail the inci- dence, clinical correlates and outcomes of stroke in patients with acute myocardial infarction treated with t~rombolyt~c therapy. ltn recent large randomized trials of thrombolytic

, the total stroke rate has ranged rate of i~tracra~~a~ hemo~rb~g

routinely use computed tomographic scannirrg in stroke patients, making the distinction betw be~~~)rr~agic and

nonhcmorrhag~c stroke that clearly emerges from the randomized trial data is that strokes of both types continue to occur in patients with acute myocardial infarctkn despite careful pretherapy screening

u135-KJ97/9o/$3.50

534 O’CONNOR ET AL. STROKE AND HEART ATTACK IN THE THROMBOLYTIC ERA

JACC Vol. 16. No. 3 September lWS33-40

lower doses of fibrin-specific drug and a more conservative approach to concomitant antiplatelet therapy. The Throm- bolysis in Myocardial Infarction (TIMI) Phase 11 trial (10) carefully screened for standard risk factors and withheld aspirin for 48 h after recombinant tissue-type plasminogen activator @t-PA) therapy, but still reported a 0.5% incidence

of intracranial hemorrhage (documented by computed tomo-

graphic scans). Although standard pretherapy screening may eliminate

the patients at highest risk for intracranial hemorrhage (for example, those with severe hypertension or intracranial tumor), there may be other risk factors for stroke or intra-

cranial hemorrhage that, if recognized, would allow the clinician to improve the riskibenefit ratio of this therapy. The purposes of this study were I) to describe the incidence of hemorrhagic and nonhemorrhagic stroke in a large consec- utive series of patien!s with acute myocardial infarction treated with aggressive reperfusion therapy and 2) to exam- ine the clinical correlates and in-hospital and long-term

outcomes of these patients.

patients, the dose was I mglkg body wc~~~t for L st hour (maximum 90 mg), with the ~en~ai~~ng dose divided over a

5 h infusion for a total dose of 150 . In all patients, 10% of the 1st hour’s dose was administe s an intravenous boius injection over 6 min. In the TAM1 II trial (121, 146 patients received combination therapy with rt-PA and low molectilar weight urokinase. The first 34 patients received 25 r&PA in additive to ~ro~~~~sc (14 patients received 0.5 million U of urokinase and 20 patients received I rn~~~io~ of urokinase). The remaini 1 I2 patients received I mgl

of rt-PA in addition to urok U in 24 patients, I mill

35 paticats). In the T

rgently to the i~ter~er~tion~~~ catheterization re diagnostic cardiac catheteriza

er arterial access was obtained~ 5,

intravenous heparin was inistered. The fl

grade 0 or i flow) underwent e

Study patients. From December 1985 to June 1987, 708 patients were enrolled in the Thrombolysis and Angioplasty in Myocardial Infarction (TAM11 I, II and Ill trials evaluat- ing various reperfusion strategies in acute myocardial infarc- tion. The entry and exclusion criteria for these three trials were the same and have been described in detail previously (I l-13). In brief, entry criteria included clinical evidence of an acute myocardial infarction within 4 to 6 h of symptom onset and electrocardiographic (ECG) evidence of epicardial injury.

emergency angioplasty received an additional 5 heparin. If residual thrombus persisted after an patients were eligible to receive intracoronary rt-PA (30 mg) or intracoronary urokinase (300

Patients were excluded from the study if they were >75 years of age, had previous coronary bypass surgery, were in cardiogenic shock at presentation or had an advanced life- limiting illness (for example, terminal cancer). In addition, particular emphasis was placed on excluding patients who had any of the following potential risk factors for hemor- rhage: I)?auma or major surgery (head, thoracic, abdominal or orthopedic) within 3 months: 2) severe. bleeding diathesis; 3) prolonged cardiopulmonary resuscitation I >I0 min) within 2 weeks before admission: 4) gastrointestinal or

genitourinary bleeding (within I month); 5) uncontrollled hypertension (diastolic blood pressure > I20 mm Hg on three consecutive measurements); 6) known structural intracranial lesions (for example, tumor or arteriovenous malformation); and 7) stroke, reversible ischemic neurologic deficit or transient ischemic attack within the previous 6 months.

Coronary care. Patients rece a standardized medical regimen after leaving the catheter~~~tio~ laboratory, which

included intravenou Sidocaine for 24 h, 325 mg of enteric- coated aspirin daily, iltiazem 60 mg every 6 h for 7 days and

a constant infusion of intravenous heparin for t to 7 days. The heparin dose was adjusted to maintain an activated partial thromboplastin time at I.5 to 2.5 times the control value. Management of the heparin infusion after throm- bolytic therapy was specified in the TAM1 protocols and was reviewed in detail with each investigator, but specific values of the activated partial thromboplastin time during the infu- sion were not collected in the TAM1 data base. Late revas- cularization procedures, including coronary angiopiasty and coronary artery bypass grafting, were used at the discretion of the patient’s private cardiologist. At 5 to IO days after enrollment, repeat diagnostic catheterization was per- formed.

Thrombolytic therapy. As previoc, ._$ described (1 l-13), study patients were treated with single chain &PA, In the

~~nrologie assessment and care. Patients were examined

TAM1 I trial !II), the first 178 patients received 60 mg of at least daily and any cba~~e in ~e~r~~Q~ic status WLS

t+PA during the 1st hour, 20 mg/h during the 2nd and 3rd promptly evaluated by the attending physician and a neurol-

hours and IO mg for eac5 of ?he last 5 h. in the remaiaiiag TO8 ogist. An emergency cosrl?crted tomographic scan of t!re head was obtained when new neuroioglc findings were

on the initiah 3can.

transferred to rehabi~ita~~o~ facilities status was stabilized.

e patimts had ~1 &Me by a ~1~~4~~~~~i~t. Str

sudden onsci of a persistent neurologic deficit, usually focal. A stroke was classified as bcmor~~~agic if there was evi of intracranial hemorrhage on the computed tomog scan of the bead (Fig. I). Other strokes were classified as ooohemQrrtlagic. A transient ischemic attack was defined as a reversible neurologic deficit (~24 h) without evidence of

checked by an ini *lendent study monitor. medical records of all stroke cases were reviewed by the neurologist coinvestigator (E.W.MJ for this study.

Continuous data were summarized by the median and

rhagic stroke group. A

Fewer patients with hemo medications but more use

of patients with a ~emorrbagic stroke.

with a hemorrhagic stroke tended to have single vessef coronary arterv disease, whereas those w&h a ~~onberno?. rhagic stroke tended to have multivessel dhease (‘fable 3).

536 O’CONNOR ET AL. STROKE AND HEART AITACK IN THE THROMBOLYTIC ERA

JACC Vol. 16. No. 3 September 199&533-40

Table 2. Clinical Characteristics a! Presentation*

Hemorrhagic Nonhemorrhagic No Stroke

Stroke Group Stroke Group Group

tn = 4) (n = 9) (n = 695)

Mean (range1 tyr! age 68 (60-74) 65 155-68) 56 (48-64)

Men (s/o) 25 56 80

Weight (kg) 73 (59-82) 72 t63-83) 81 (72~9!!)

Time from thrombolytic therapy to stroke (h) 6 (3-12) 72 t48-IOOt

Risk factors (c/o)

Hypertension 25 67 41

Diabetes 25 33 I5

Hyperlipidemia 0 22 II

History of smoking 75 56 23

Previous MI !%I 0 40 I2

Physical examination

Carotid bruit (%o) 0 0 ! Systolic BP (mm Hgt 130 tizu-141) E50 (I?Y-177) 130 !I 14-150)

Diastolic BP (mm Hgt 80 (80-95) 94 (74-108) a3 (70-92)

Abnormal neurologic examination (r/n) 0 0 I

Home medications (%I

None 2s II 42

Beta-blocker 0 44 13

Calcium channel blocker 0 II 8

Aspirin 50 33 I9

Dipyridamole (Persantinet 0 0 3

Warfarin (Coumadin) 0 0 0.4

*Continuous variables displayed as median values (25th to 75th percentile). BP = blood pressure: MI = myocar&l ir&ction: % = percent of patients in each group.

The distribution of the baseline ejection fraction in the nonhemorrhagic stroke group was substantially lower than that in the hemorrhagic stroke group or the patients without

Table 3. Baseline Coronary Angiographic Findings

Hemorrheic Stroke Nonhemorrhagic No Stroke Group Stroke Group Group

tn = 4tL (n = 9) In = 695)

No. of diseased vessels (%I

O-I 75 33 64 2 0 22 23 3 0 44 I3

Infarct-related artery (k)

Left anterior 0 78 39 descending

Left circumffex 50 22 I3 Right 25 0 48 Bypass graft 0 0 0.4

MI location (%/c)

Anterior 25 78 41 Inferior 75 22 59

Ejection fraction (%)t 62 (40~64) 36 (30-41) 53 w-591 Acute angioplasty 0 56 43

performed !I)

*One patient was unable to undergo cardiac catheterization. *Data displayed as median values (25th IO 75th percentile). MI = myocardial infarction. % = percent of patients in each group.

stroke. No patient with a hemorrhagic stroke had undergone acute coronary angioplasty compared with 56% of those with a nonhemorrhagic stroke and 43% of those with no stroke.

Patients with a hemorrhagic stroke tended to have the lowest nadir fibrinogen levels, but the incidence of major bleeding episodes (requiring transfu-ion of r2 U of blood) was similar for all three patient groups (Table 4).

~eu~o~og~c findings” The neurologic and tomographic findings of the head in 13 stroke patients are shown in Table 5. Depressed mental status was much more common after hemorrhagic stroke than after nonhemorrhagic stroke. All

Table 4. Laboratory Data*

Hemorrhagic

Stroke Nonhemorrhagic No Stroke Group Stroke Group Group (n = ~‘2 tn = 91 !n = 695)

I”. _I Hematocrit (nadir) !%I 37 129-42) 28 (22-32) 32 (28-36) Platelet count (nadir) 232 (145-319) I60 t79-260) 187 (147-235)

(X lO’/liler)

Fibrinogen (nadir) 50 ( 18-76) 86 (51-l 13) 120 (51-170) !gldl)

Major bleeding 25 33 I9 (transfusion of

=2 !J of blood) !%)

*Continuous variables displayed as median values (25th to 75th percen- tile). % = percent of patients in each group.

JAW Vol. 16, No. 3 September 1990:533-40

O’CONNOR ET AL. STROKE AND HEART ATTACK SN THE THROMBOLYTlC ERA

537

aracterislics of I? Palienls

Neurologic exam~nalio~ (%,)

~i~bl-sided weakness

Left-sided weakness

Aphasia

Depressed mental status

Seizure

Posterior circulation symptoms

Cl fmdiags of the head (%)

Normal (initial)

Small lesion (<I ~0~~)

Large lesion (z I cm’) Anlerior distribution

Posterior distribution

Edema

Midline shifi

44

22

33

33

22

22

0

CT = computed ~orn~~~ap~~c~ % = percent of patients in each group.

with bemQrrha~~c st

gic stroke. In 75% of

rrhagic stroke in pati , three c~a~cterist~cs

no stroke gloams (0% versus ease was absent in the non-

hemorrhagic stroke group but was present in 3% of the no stroke group. However, anterior myocard~a~ infarction with an ejection fraction ~45% occurred in 75% of the nonhem- orrhagic stroke group compared with 18% of the no stroke group. Of these three clinical cllaracteristics, only anterior

Tabte 6. Prediction of Nonhemorrhagic Stroke: Logistic Regressnon Analvsis of Three Prespecified Independent Variables

Variable Chl-Square p Value

Atrial fibrillation 0.06 0.80

Cerebrovascular disease 0.34 0.56

Anterior myocardial infarction 10.43 0.001

and EF ~45%

EF = ejection fraction.

Hemorrhagic No Stroke Nonhemorrhagic Stroke Group Stroke Group Group (il = 4, (ll = 9) (n = 695)

In-hospital death (%I 25 II 6 1 year death (%)* 0 25 3 Nonfalal Ml (r/r )I: 0 0 6

cdian hospilal 14 (12-51) 14 (1 I-30) 9 U-11) days?

Neurologic deficit at

discharge (%I*

None 0 0 IOU Mild 33 75 0

Severe 67 2s 0

o&al survivors; tdata displayed as median values (251h to 35th

percentile). Ml = myocardial infarction; % = perceni of patients in each group.

tients with intracranial

ospital death rate in

pared with 9 days in the no stroke group. The neurologic deficit present at discharge in hospital survivors was severe enough to require further rehabilitation in 67% cif the hem- orrhagic stroke group compared with 25% of the nonhemor-

rhagic stroke group.

emorrhage is the major complication of modern throm-

bolytic therapy and intracranial hemorrhage is the most astrophic form of hemorrhage seen with these th ytic agents. This study demonstrates that ~o~be

rhagic stroke remains t ost common form ef stroke seen

in patients with acut cardial infarction treated with

thrombolytic therapy a that a large anterior myocardial

infarction remains the major clinical correlate of this for stroke. However, hemorrhagic stroke has emerge important but unpredictable comphl-.aticin that often has

catastrophic consequences.

538 O’CONNOR ET AL. STROKE AND HEART ATTACK IN THE THROMBOLYTIC ERA

IACC Vol. 16, No. 3 September 1990533-40

Incidence of stroke. In the prethrombolytic era, approxi- mately 3% of patients with acute myo<ardial infarction developed B Stroke either concurrently with the infarction or during the same hospitalization (1,2). Recent randomized clinical trials (5-10) have demonstrated that the stroke rate in patients treated with thrombolytic therapy is 1% to 2% and is unchanged from that seen with conventional treatment. However, caution must be used when comparing the figures from current trials with the data from the prethrombolytic era because many patients at high risk for embolic stroke are excluded from treatment with thrombolytic therapy. Despite this important caveat, the available results suggest that the risk of embolic stroke may be reduced after thrombolytic therapy, perhaps because of a reduction in infarct size and the prevention of clot formation in the left ventricle asspci- ated with this therapy (22). This benefit is offset to shine extent, however, by the increase in the risk of hemorrhagic stroke compared with conservative treatment (8,22).

Some controversy has arisen about the rate of stroke associated with different thrombolytic agents. Direct com- parisons of available trials have been hampered by differ- ences in the methods of data collection, quality control and dosages used. The reported stroke rates in the large “sim- ple” clinical trials have consistently been lower than the rates reported in smaller trials with detailed data collection and access to rapid use of computed tomographic scans of the head. Accurate classification of an individual stroke as hemorrhagic or nonhemorrhagic requires computed tomo- graphic scanning.

In the TIM1 Phase II trial (IO), patients treated with 150 mg rt-PA had an intracranial hemorrhage rate of >I%; however, the 100 mg dose was associated with a risk of intracranial hemorrhage of 0.5%. We did not find an in- creased risk of stroke when rt-PA was used in combination with urokinase at doses lower than those recommended for monotherapy (Table 1).

Predictors of stroke. Most trials during the thrombolytic era did not examine specific clinical factors associated with stroke. In the prethrombolytic era, stroke was sssociatec! with a large anterior myocardial infarction, atrial arrhythmia and prior cerebrovascular disease (l-3). Most of these events were thought to be embolic and the remainder were attributed to thrombotic events, hypotension and, rarely (<O.l%), intracranial hemorrhage. The association in this study between a large anterior myocardial infarction and nonhemorrhagic stroke is consistent with previous reports (l,2). In this analysis, eight of nine patients with nonhemor- rhagic stroke were believed by the consulting neurologist to have an embolic stroke.

Although some trends were evident, we could not identify

risk fRctors for hemorrhagic stroke that would be helpful in excluding patients from treatment with thrombolytic ther- aPY. Importantly, patients with the standard major contrain- dications to thrombolytic therapy were already excluded

from the three trials reviewed in this study. Althouse et al. (23) reported that patients with an initial elevation of drash tolic blood pressure had an increase in intracranial hernor- rhage when treated with rt-PA. Other investigators (3, have postulated that systemic disease leading to loss vascular integrity, such as diabetes or prior hypertension, would increase the risk of intracranial hemorrhage. In addi- tion, concern about the possible effect of aspirin on hemor- rhagic stroke led the TIM1 investigators (110) to withhol aspirin for 48 h after thrombolytic therapy was initiated.

We attempted to use these sitnpie clinical factors to identify patients at increased risk of intracranial ble Unfortunately, these characteristics did not distinguis risk from low risk patients accurately enough to be us clinical decision making. In fact, the initial systolic and diastolic blood pressures were lower in the hemorrhagic stroke than in the nonhemorrhagic stroke group and were only marginally higher than in the group without stroke. Patients with diabetes mellitus had no increased risk of hemorrhagic stroke and prior use of aspirin was not associ- ated with an increased risk.

Patients in this study were screened with use of a careful medical history to ensure that no cerebral ischemic event bad occurred in the preceding 6 months. The absence of a stroke among the 18 patients with prior cerebrovascular disease >6 months before treatment raises the question of whether patients with a remote history of stroke or transient ischemic attacks should be excluded from receiving throm-

his event rate of 0% has a 95% upper confidence limit of 16%. however, so these results shoul not be regarded as definitive.

In our exploratory analysis, several characteristics were more frequent in patients with hemorrhagic stroke. Women may have been at greater risk because of the fixed dose of rt-PA used in most of these studies. We previously (24) demonstrated that women, probably because of their smaller size, have a higher r&PA concentration during the infusion phase, leading to more fibrinogenolysis and possibly a greater risk of lysis of hemostatic plugs. Older patients may have more defects in cerebrovascular integrity. Previous studies from our group (24) and others (25) found age to be an independent risk factor for general bleeding complica- tions after thrombolyt ic therapy. However, the Second International Study of Infarct Survival (lS~§-2) (7) has recently demonstrated a markedly beneficial effect of throm- bolytic therapy on the overall mortality rate, even in the very old. The absence of prior use of a beta-adrenergic or calcium channel blocker in patients with stroke suggests that these agents may be protective against hemorrhagic stroke by lowering blood pressure, reducing the rate of rise of systolic pressure or preventing catecholamine-induced sudden hy- pertension, thereby lessening the disruption of hemostatic plugs.

The TIM1 II investigators observed an intracranial hem-

JACC Vol. 16. No. 3 September 1990:%3-40

edema.

The clinical syndr

therapy should also be reevaluated. A co

graphic scan should be obtained immediately.

evacuation may be indicated on rare occasions, depending

on iocation of the bleeding.

e study. Two important limitations of

dy should be noted. First, despite a large overall

up, we only observed four hemorrhagic and nine

nonhemorrhagic strokes. Thus, our statistical p

detecting clinical associations is limiied. Second, t does not include any conservatively treated

comparkons :vith previous experience should b

with appropriate caution.

Conclusions. The use of thrombolytic therapy in patients

with acute myocardial infarction is associated with a small

risk of stroke. A large anterior myocardial iRfarctio~ with a

depressed left ventricular ejection fraction conti

the major risk factor for a ~Q~hemorrhagic stroke

patients at increased risk for hemorrhagic stroke cannot be

detected by clinical characteristics alone with enough ACCU-

racy to alter patient selection criteria once the standard

exclusion criteria have been applied. The clinical course of

both types of stroke is characterized by substantial morbid-

ity and mortality.

We i&be ~~ib8d~?l 10 Renee Story and Alexandria Lubans for assistance with Ihe m;mll\cripl. -. ---

I.

2.

7 _ .

4.

4.

6.

I.

8.

9.

10.

II.

I’ _,

13.

14.

IS.

16.

17.

I$.

Komrad MS. Coffey CE. Coffey KS. McKinnis RA. Massey EW. Califf WM. Myocardiai infarction and stroke. Neurology 1984;34:1403-9.

Thompson PL. Robinsoo JS. Stroke after acute myocardial infarction: relation lo infarcl size. Br Med J 1978:2:457-9.

Califf RM. Massey EW. Myocardial infarction and stroke in the throm- holyticera. In: CalilTRM. rk DB. Wagner GS. eds. Acute Coronary Care in the Thrombolytic Era. cage: Year Book Medical. 1988539-47.

Mark DB. Hlalky MA. O’Connor CM. et al. Adm~nislrat~on of throm- bolytic therapy ir the community hosphal: established principles ant? unresolved issues. J Am Coil Cardiol 1988:12(suppl ,4):32A-43A.

Gruppo Italian0 per 1,. Studiodella Streprochinasi nell”lnfarto Miocardico GlSSI). Effectiveness of intravenous thrombolytic treatment in acute myocardial infarction. Lancet 1986:1:397-402.

Slody Grow p. A prospective lrial of Intrabe:.ous Streptokinase in Acute Myocardial infarction W.A.M.): morbidity. and infarct size a( 21 days. N Engl J

ISIS-? (Second lnrernational Study of Infarct SUrViViIl) Collaborative Group. Randomized trial of intravenous streptoltina;e. oral nspirin. both or neither among 17.187 cases of suspected acute myoca,.%al infarction: ISIS-?. Lance’ 1988:2:?49-60.

Van de Werf F. Arnold FER for the European Study Groq for Recom- binant Tissue-Type P~asmi~ogen Activalor. Intravenous ticjsue plasmino. gen activator and :ize of infarct, left venlricular function, and survival ip acute myocardial infarction. Br Med J 1988:297:1374-9.

Wilcox RG. von der Lippe 6. Qlsson CG. Jensen 6. Skw AM, Hampton JR. Trial of tissue plas&nogen activator for mortality reduction in acute mvocardial infarcti,on: Anglo-Scandinavian Studv of Earls Thrombolvsis (ASSET). Lancet 1988:!:?!5-30.

TlMl Study Group. Comparison of invasive and conservative strategies after ?ieatniSrri wsth inlravenous tissue plasminogen activator in acute myocardial infarclion: results of Thrombolysis in Myocardial infarction (TIMI) Phase II Irial. N Engl J Med 1989;320:618-27.

Top01 EJ. Califf RM. George BS. et al. A randomized trial of immediate versus delayed elective angioplasty after in&venous tissue plasminogen activator in acute myocardial infarction. N Engl J Med 1987;317:581-8.

Top01 EJ. Califf RM. George BS, et al. Coronary arterial thrombolysis with combined infusion of recombinant tissue-type plasminogen activator and urokinase in patients with acute myocardial infarction. Circulation 1988:77: I 100-7.

Top01 EJ. George BS. Kereiakes Dl, et al. A randomized controlled trial of intravenous tissue plasminogen activator and early intravenous heparin in acute myocardial infarction. Circulation 1989;79:281-6.

O’Connor CM. Mark DB. Hinohara T. et al. Rescue coronary angioplasty after failure of intravenous streptokinase in acute myocardial infarction: in-hospital and long-term outcomes. J Inv Cardiol 1989;1:85-95.

arrell FE Jr. The LOGIST procedure. In: SUGl Supplemental Library User’s Guide. Version 5 Edition. Cary. NC: SAS hdtute. 1986269-93.

Fisher CM. Adams RD. Observation of brain embolism with special reference lo the mechanisms of hemorrhagic infarction. J Neuropathol Exp Neurol 1951:10:92-3.

Farris AS. Harden CA. Pi? er CM. WU~ogenesis of hemorrhagic infarc- tion of th,: brain. Arch rJeuo’ ‘3&:9:468-76.

Cerebral Embolism Study Grooe- Immediate anticoagulation of embolic stroke: brain hemorrhage and management options. Stroke 1984:15:779- 89.

540 O’CONNOR ET AL. JACC Vol. 16. No. 3 STROKE AND HEART ATTACK IN THE THROMBOLYTIC ERA September 1990x533-4(3

19. Rarnirez-Lassepas M, Quinones MR. Heparin therapy for stroke: hemor- rhagic complisations and risk factors for intracerebral hemorrhage. Neu- rology 1984;34: 114-7.

20. Homing CR, Domdorf W, Agnoli AL. Hemorrhagic cerebral infarction: a prospective study. Stroke 1986;17: 179-85.

21. Brett T, Thalinger K, Hertzbefy V. Hypertension as a risk factor for spontaneous intracerebral hemorrhage. Stroke 1986:17:1078-83.

22. Eigler N. Maurer G. Shah PK. Effect of early systemic thrombolytic therapy on left ventricular mural thrombus formation in acute anterior myocardial infarction. Am J Cardiol 1984:54:261-3,

23. Althouse R. Maynard C. Oisufka M. Kennedy JW. Risk factor for hemorrhagic and ischemic stroke in myocardial infarct patients treated with tissue plasminogen activator (abstr). J Am Coil Cardiol 1989; I#suppl A): I53 A.

24. Califf RM. Top01 EJ. George BS, et al. Hemorrhagic complications associated with the use of intravenous tissue plasminogen activator in treatment of acute myocardial infarction. An! J Med 1988:85:353-9.

25. Lew AS. Hod H. Cercek B, Shah PK. Cianz W. Mortality and morbidity rates of patients older and younger than 75 years with acute myocardial infarction treated with intravenous streptokinase. Am J Cardiol 1987;591-5.