Download Relation between myocardial infarct location and stroke

ester and New York, New
York
f stroke is increased after myocardial infarction
close association between anterior myocardial
infarction and left ventricu
‘on that embo~i~ati~n
a
tant cause of stroke i
a
reviews (5-8) recommend obtaining an
in patients with an anterior myoca~djal
viding anticoagulant therapy if thrombus is
d, a~tbo~~b this procedure is not accepted universally
(9). Yet, prospective studies have reported varying degrees
of success with the use of anticoagulants to prevent tbrombus (W-14) or reduce the likelihood of stroke (12-18).
From the Division of Cardiology, The Harris Chasanoff Heart Ins&%
Department of Medicine, Long Island Jewish Medical Center, New Hyde
Park and Long Island Campus for the Albert Einstein College OF Medicine,
Bronx; *Department of Preventive Medicine and Community Health and
Department of Medicine. University of Rochester Medical Center, Rochester;
and tDepartment of Biostatistics, Columbia University, New York, New
York. This study was supported by a consortium grant from Godecke
Aktiengesellschaft, Freiburg, Germany: Laboratories Dr Esteve, SA, Barcelona, Spain; Marion Laboratories, Inc.. Kansas City, Missouri; Nordic
Laboratories, Inc., Laval, Quebec, Canada; Lars Synthelabo, Paris, France;
Tanabe Seiyaku Co. Ltd., Osaka. Japan: an4 Warner-Lambert International,
Morris Plains, New Jersey.
Manuscript received December 3. 1993; revised manuscript received
Fe
: Dr. Monty M. Bodenheimer, Chief of
Cardiology, The Harris Chasanoff Heart Institute, Room 2135, Long Island
Jewish Medical Center, New Hyde Park, New York 11042.
81994 by the American College of Cardiology
ardial infarction
~othes~zed that if tI
nterior infarction is
with anterior infarction
increased incidence of stroke.
sbou
trospective
review of the
in the Multicenter
from February 17, 1983 to June 30, 1986. The rec~~tment
procedure and inclusion and exclusion criteria have been
previously reported (19). In brief, men or women 25 to 75
years old who were admitted to the hospital and subsequently documented to have had an acute myocardial infarction were eligible. Patients were excluded if there was
evidence of ongoing cardiogenic shock, pulmonary h
e at~iovent~c~~ar block or other c
ce survival. randomization
took
during hospital days 3 to I5 (median day 8) from the ti
onset of the myocardial infarction.
ths (averAll patients were followed up for 12 to 52
age 25 months).
Patients were seen by the investigator or
or both, at periodic intervals throughout
the trial. The MDPIT follow-up schedule consisted of five
clinical
coordinator,
07351097/94/$7.aQ
62
JACCVol. 24. No. I
July 19%61-6
BODENHEIMER ET AL.
STROKE AND MYOCARDIAL INFARCT SITE
1. ClinicalCharacteristicsby MyocardialInfarct Site
Covariate (no. of patients with anterior,
nonanterior infarction)
Age 260 yr (724, 1,740)
Previous Ml (723. 1,739)
NYIIA class II-IV (724. 1,740)
History of hypertension (723, 1,739)
Insulindependent diabetes (724, 1,740)
Cigarette smoking (72 I, 1,727)
Cardiachndings
Systolic BP cl00 mm IIg (722, 1,734)
Bibasiiar pulmonary rales (721,1,733)
Cteatine kinase ~619 IJ (724, 1,739)
brillation (724, 1,740)
sis (at index MI) (724, 1,746)
Aspirin
adrenergic blocking agents
abs
ridamole
Other antipkitelet agents
Reinfatction (before CVA) (724. !,74ft!
Warfaiin (at discharge) (724, 1,740)
StroktiIA (724, 1.740)”
Site uf MI (%I
Anterior
Nonanterior
v:ue
0.795
0.002
a.553
0.886
0.243
0.055
50
25
19
38
9
46
51
20
18
38
8
50
2
4a
79
9
56
I2
5
36
62
8
17
7
0.
e 0.
< 0.
0.
< 0.
< 0.
22
49
23
8
9
0.28
19
3
24
56
IO
7
6
0.34
4
4
0.145
0.
==z0.
0.
0.003
0.785
0.381
‘Total t!Jmber of cerebrovascular events during follow-up (see text). BP = blood pressure; CVA = cerebrovascular accident; EF = eiection fraction: MI = myocardia! infarction; NYHA class = New York Heart Associatiun
functional class; TIA i transient ischemic attack.
visits in the first year (an initial visit at I month followed by
a visit every 3 months), then a visit every 4 months until
study closure. Bata retrievalat the clinic visit consisted of an
interval medical and cardiac history, review of concurrent
medications and physical examinations.
During these visits any adverse experiences or intervening medicalconditions, or both, were noted and described on
the clinical record forms. Stroke was not coded separately
but included as either an adverse event or an interval
medical event. As a result, 1,731 medical follow-up repcsrts
had to be reviewed. This review was performed by two of us
(D.S. and B.S.) without knowledge of the site of the index
myocardial infarction.
A stroke was defined as an acute focal cerebra) disorder’that resulted in localizing findings characterized by either the patient’s physician or the clinical investitot as a cerebrovascular accident. If the duration of
symptoms was ~24 h, the episode was definedas a transient
ischemic attack (20).Nonlocalizingsymptoms, such as dizziness, giddiness or wooziness, or involuntary movements,
such as limb shaking, were not classified as transient ischemit attacks (20).Because this study was not prospectively
designed, confirmationwith computed tomography,ancephalography and other procedures was not always available
and was generally not recorded in the patient data.
The diagnosisof acute myocardialinfarctionrequiredserum
enzyme confirmation:an MBisoenzymefraction>4% of total
Creatinekmase (CK)or a qualitativelypositive MB band; an
elevatiott
of total Iactic deh
reversalof the LD
or an elevationof 22 times
normal of CK, se
xaloa?etic transaminaseor
LDH in the coronary care unit with a clinically consistent
pattern. Acute anterior wail myocardialinfarctionwas defined
as Q waves appearing in leads V, to Vd; absence of new Q
waves in these leads identifiedpatients with nonanteriormyocatdial infarction(19).The latter group included patients with
inferioror posterior Q wave infarction(61%)and those with
non-Q wave infarction(39%),defined as elevated serum ent with infarctionwithout Q waves (19).
zY
Univariateanalysis was performed to
determine the relation between baseline clinical variables
and infarct site (Table I). A model was then constructed
using Cox regression analysis with the occurrence of stroke
or transient ischemic attack as the dependent variable (21).
All covariates with a p value I 0.10 were forced into the
model. The use of warfarin was added to the model to
determine its contribution, and it was maintained in the
model regardless of its p value. The final model was constructed by adding infarct site to the model (21).Results are
expressed as hazard ratio with 95% confidence limits. A
Roweranalysiswas performedto determine the possibilityof
a beta error (22). The effect of infarct site on time to end
point was examined with the method of Kaplan and Meier.
The log-rank statistic was used to compare Kaplan-Meier
curves. Summary odds ratios were determined with the
technique of Mantel and Haenszel(22).
Covariates
(95% Cl)
Value
Systolic BP*
(on admission to CCU)
1.014
(1.007 to 1.021)
< 0.0001
0.930
NS
(0.397 to 2.179)
0.F
NS
Warfarin
Anterior i~a~ctiom
(0.508 to 1.340)
The
hazard ratio for sys!olic blood pressure (BP) is per unit increase in
, when compared with a patient with a systoli
g, a patient with a systolic blood pressure of 150
would have a hazard ratio of 1.014” (=1.97). CCU = coronary care unit;
C1 = confidence interval.
viva!curvesfor likeli
fined RSstroke or
attack, in patients wi!la a~it~ri~r or ~Q~~~t~rior m
I) at the time of enro~lmc~t.
and transient isc
s with anterior ~~~a~cti
acebo group and 44 (
patients with an event.
level and lower
cardial infarction as we
of admission with the index myoto be treated with
infarction had a stroke or transient is
pared with 68 (3.9%) of 1,742 patients with a nonanterior
myocardial infarction (relative risk 0.82; 95% confidence
interval [CI] 0.55 to 1.30).Power analysis revealed that this
study had an 86%chance of detecting a relativerisk of stroke
of 1.5in patients with anterior versus nonanterior infarction.
When the patients with nonanterior infarctionwere classified
into those with inferior or posterolateral Q waves a
with nap-Q wave infarction, the likelih
3.5% and 4.%, respectively. Figure 1
Meier life table event rate for patients with anterior and
r myocardial infarction. As is evident, no signifinona
ence exists (p = 0.42 by log-rank test). When
cant
analyzed by infarct site and treatment with placebo or
diltiazem, the findingswere essentially the same (p - 0.35).
Multivariate analysis (Table 2). Stepwise multivariate
Cox analysis was used to determine variables that were
limitingthe screening process to those with anterior infarction (5-S).
us ~~rrnat~Q~(Ta
A~t~coa~w~~t~~~
an
Echocardiographicstudies of the relation between anterior
myocardial infarction, thrombus and stroke differ in methstudy oatients, timing of echocardiog~aphy~there including thrombolytic agents) and follow-up.
e limitations in mind, pooling these studies shows
that the likelihood of thrombus and embolizationin patients
with anterior infarction is high whether or not anticoagulants
are used (Table 3). In four studies, patients were randomized
to receive either anticoagulantsor placebo, althoughthrombolytic agents were sometimes used (1l-14). Two studies
(11,12) showed benefit, whereas two (13,l
effect on left ventricular thrombus formation.
coagulation was
studies (I 1-14) were combined, no
bus
of 1.86 (95%
associated with an increased risk of th
CI 1.22to 2.82).
Thrombolytic agents have been reported to reduce thrombus formation (27,31) or to have no effect (14,18,29,32).
However, because anticoagulants are often used concur-
64
JACC Vol. 24, No. I
July 1994:61-Q
BODENHEIMER ET AL.
STROKE AND MYOCARDIAL INFARCT SITE
Total No.
Study
No. With
Emboli
No. With
Thrombus
Total No.
No. With
Thrombus
No. With
Emboli
497
203 (41%)
24t (4.8%)
Norarandomized PooZ Studies
31* (4.7%)
169 (26%)
653
Ref. IO, 15-18,23-32
Randomized Trials of Anticoanulation and Left Ventricular Thrombus
Vecchio et al. (14)
SCAT) trial (12)
Gueret et al. (13)
hhnnesilcrnet al. (II)
Totals
86
93
25
21
225
26
34
13
7
80 (36%)
I
94
2
107
21
1
21
243
2
6 (2.6%)
25
19
0
0
8
I
1
0
52 (21%)
2 (0.8%)
‘21 strokes and IO nonccreblal emboli. K?Ostrokes and 4 noncerebral emboli. SCumulativc odds ratio of thrombus it patients without aflt~coag~~~tjo~
vm.us
patients with antic ulation is 1.86 (95% confidence interval 1.22 to 2.82) with a chi-square of 8.36 with I degree of freedom.
rently, any benefit may not reflect the effect of throm~o~ytic
ents (14).
The ability of anticoagulation to prevent stroke is also
difficult to assess because of inadequate controls and a
relative paucity of embolic events (Table 3). In the present
study, warfarin was used in only 1% of patients with
anterior infarction, raising the possibility of a type 2 error to
he lack of contribution of warfarin to the model.
Won on stroke in
mtion paof as
tlents, The role of anticoagulation in patients after acute
myocardial infarction has been the subject of numerous
studies. Aithouqh stroke has generally not been a primary
prospectively aefined end point, we used previously published analyses (33-35) and a search of MEDLINE to find
those randomized studies in which the incidence of stroke
could be determined (Table 4) (36-44). As assessed with use
ofa cumulative odds ratio and the mantel-~aenszel test (22).
the risk of stroke was 2.05 (95% CI I.5 to 2.8) in patients
randomized to placebo versus anticoagulation. Left unre-
solved is the role of aspirin relative to coMmadi~ in this
setting (45).
(Table 1).
Dexter et al. (4) found a higher than expected rate of
stroke. particularly in the 1st 2 months after myocardial infarction. Our study also shows a relatively higher rate within
the 1st 2 months, with I8 of 91 strokes occurring during this
period. However, the rates were similar for patients with
anterior and nonanterior myocardial infarction (Fig. I).
Table 4. Analysis of RandomizedStudiesof AnticoagulationAfter Myocardial Infarction:
Incidenceof Stroke
Control Group
Study
MRC (36)
Wasserman et al. (37)
Lovell et al. (38)
Ritlsnd et al. (39)
MRC (40)
Drapkin and Merskey (II)
VA Coop (42)
Sixty Plus Reinfarction (43)
Smith et al. (44)
Total*
Total No.
188
70
240
101
715
391
499
439
607
3,250
--
No. With
Stroke
I
I
13
2
I8
9
16
21
44
125
Anticoagulation Group
Total No.
195
77
172
97
712
145
500
439
No. With
Stroke
3
0
2
0
8
13
4
I3
607
20
3,544
63
*Summary odds ratio of stroke in patients in control groupscompared with patients treated with anticoagulation
is 2.05 (95% confidence interval I .50 to 2.80) with a chi-square of 20.6 with 1 degree of freedom. MRC = Medical
Research Council: VA Coop = Veterans AtEM cooperative study.
JACC Vol. 24, No
3uly 1994:61-6
1
5. Ezekowitz MD Acute ~~~~ctio~, left ventricular thrombus and systemic
embo~izat~o~. An approach to ma~a8em~nt. J Am Coil Cardiol 1986;6:
IZSJ-82.
6. Meltzer RS, Visser CA, Fuster V. ~at~aca~ia~ thro
ernb~~~zatio~. Ann Intern
7. Wei~tra~b WS, Ba’albaki
concerning the apphcation ofe~~oca~di~g~~by
to :he diagnosis and treatment ofmuraltbrombi
after anterior wall acme myocaedial infarction. Am J Cardiol1989;64:70816.
8. Editorial. Left ven . ular tbrombus and stroke following myocardial
infarction. Lancet 1 ;335:759-60.
9. Cerebral Embolism Task Force. Cardiogenic brain embolism. The second
report of the cerebral embolism task force. Arch Nemo1 1989:46:729-43.
IQ. T&pie AGG, Robinson JG, Doyle DJ, et al. Comparison of high-dose
with lOwdose subcutaneous heparin to prevent left ventricular mural
thrombosis in patients with acute transmural anterior myocardial infarction. N Eng! J Med 1989;320:352-7.
n Der Lippe 6. Left ventricular
ticoagulants in acute myocardial
12. The SCATS Group. Randomised controlled trial of subcutaneous calciumheparin in acute myocardial infarction. Lancet 1989;2:182-6.
13. Gueret P. Dubourg 0, Ferrier A, Farcot JC, Rigaud M, Bourdarias IF.
Effects of full-dose hepatin anticoagulation on the development of
ventricular thrombosis in acute t~nsmM~al myocardial infarction. 3
Colt Cardiol 1986;8:419-26.
prevent more than half of the cere~r~vasc~la~events seen inn
our patients. Clearly, further studies are needed to determine the mechanis of these events. U
diagnostic criteria exist
strokes from those with ot
vascular atherosclerosis(
review of the ~~~lis~ed
agulation even though o
therapy. The exact place of this strategy in relation to other
interventions remains to be determined with randomized
trials (45).
We thank Mark Andrews, BA, for help with the data analysis and Shirley
Eberly, MS, for assistance with the statistical analysis.
I. Fulton RM, Kucket K. Plasma fibrinogen and thromboemboli after
myocardial infarction. Lancet 1976;2:1161-4.
2. Thompson PL, Robinson JS. Stroke after acute myocardial infarction:
relation to infarct size. Br Med J 1978;2:457-9.
3. Komrad MS, Coffey CE, Coffey KS, McKinnis R, Massey EW, Califf
RM. Myocardial infarction and stroke. Neurology 1984;34: 1483-9.
4. Dexter DD, Whisnant JP, Connolly DC, O’Fallon WM. The association of
stroke and coronary heart disease: a population study. Mayo Clin Proc
1987:62:1077-83.
14. Vecchio C, Chiarella F, Lupi G, Bellotti P, Domenicucci S. Left ventrict,lar thrombus in anterior acute myocardial infarction tter thrombolysis.
Circulation 1991;84:512-9.
15. Keating EC, Gross S, Schlamowitz RA, et al. Mural thrombi in myocardial infarctions. Prospeciive evaluation by two-dimensional ecbocardiography. Am J Med 1983;74:989-95.
16. Weinreich DJ, Burke JF, Pauletto FJ. Left ventricular mural thrombi
complicating acute myocardial infarction. Ann Intern Med 1984;100:78994.
17. Johannessen KA, Nordrehaug JE, Bon Der Lippe 6. Stein EV. Risk
ems with left ventricular thrombi and acute
factors for embolisation in
myocardial infarctioo. Br
rt J 1988;ti304-10.
18. Keren A, Goldberg S, Gottheb S, et al. Natural history of left ventricular
thrombi: their appearance and resolution in the posthospitalization period
of acute myocardial infarction. J Am Colt Cardiol 1990;15:798-8
ulricenter Diltiazem Postinfarction Trial Research Croup. The
effect of dihiazem ou mortality and reinfarction after myccadial infarction. N Engl1 Med 1988;319:385-92.
20. Koudstaal PJ, Alma A, Pop G&M, Kappelle LJ, Van Latum JC, Van Gijn
J, for the Dolch-TJA Study Group. Risk of cardiac events in atypisai
transient ischemic attack or minor stroke. Lancet 1992;340:630-3.
21. BMDP Statistical Software Los Angeles CA, Version 1999. Vol I and 2.
22. Fleiss IL. Statistical Methods for Rates and Proportions. 2nd ed. New
York: Wiley, 1981:145, 173-5.
23. Friedman M, Carlson K, Marcus FI, Woolfenden JM. Clinical comlalions in patients with acute myocardial infarction and left ventricular
thrombus detected by two-dimensional echocardiography. Am J Med
1982:72:894-S.
24. Ezekowitz MD, Kellerman DJ. Smith EO, Streitz TM Detection of aclive
left ventricular thrombosis during acute myocardial infarction using
indium-11 I platelet scintinraphy. Chest 1984;86:35-9.
25. Johannessen #A. Nordrehag JE, Van Der Lippe G. Left ventricular
thrombosis and cerebrovascular accident in acute myocardial infarction.
Br Heart 5 1984;51:553-6.
26. Visser CA, Kan G, Meltzer RS, Lie KI. Durrer D. Long term follow-up
of left ventricular thrombus after acute myocardial infarction. A twodimensional echocardiographic study ir 96 patients. Chest 1984;86:532-6.
27, Eigler N, Maurer 6, Predimqn SK. Elect of early systemic thromboytic
therapy on left ventricular m;r? ::uombus formation in acute anterior
myocardial infarction. Am J Cardiol 1934;54:261-3.
28. Spirit0 P, Bellotti P, Chiarella F, Domenicucci S, Sementa A, Vecchio C.
Prognostic significance and natural history of left ventricular thrombi in
patients with acute anterior myocardial infarction: a two-dimensional
echocardiographic study. Circulation 1985;72:774-80.
29. Held AC, Gore JM, Paraskos J, et al. Impact of thrombolytic therapy on
66
BODENHEIMER ET AL.
STROKE AND MYOCARDIAL INFARCT SITE
left ventricular mural thrombi in acute myocardial infarction. Am J
cardiol19@$2310-!.
30. Niiyanno~mulos P, Smith GC, Maseri A, Foale RA. The natural history
of left ventricular thrombus in myocardkl i&r&n:
a rationale in
support of masterly inactivity. J Am Coil Cardiil 1,B9;14:903-11.
31. Lupi G, Domeniczcci S. Chiarella F. Rellotti P, Vecchio C. Jntluenceof
thromboylytic ueatme~t followed by full dose anticoagulation on the
~~~~~~v~t~~
tbrombi in acute myocardial infarc!ion. Am
::
-.
32. Kupper AF, Verheught FW, Peels CH. Galema I%‘, Roes JP. Left
ventricular ihrombus incidence and behavior studied by serial twoy in acute anterior myowdial infarstion:
dimenJionat echocard
stcmic embolism and oral anticoagulation.
J Am CoU Cardid 193!+13:514-20.
33. CbalaKn TC, Matte RJ, Smith H Jr, Kurt&r AM. Evidence favoring the
use ofanticoagutants in the hospital phase of acute myocardial infarclion.
J Med 1977;297:1091-5.
37. Wasserman AJ, Gutterman LA, Yoe KR, Kemp V
Anticoagulants in acute myocardii infarction. Th
lation to alter mortahty in a randomized series. A
38. Lovell RRH. Denborougb h4A, Nestel FJ, GobbleAJ. A co~~o~ledtrial of
long-term treatment with ant~coa$ula~~ after myocardiat infarction in 412
male patients. Med J Aust 1
cacy of 3 and 12 months’
39. Ritland S. Lygren T. Compa
anticoagulant therapy after myocardial infarction. A controlled clinical
trial. hcet l!X9;1:122-4.
admi~l~tmtion after cardiac in-
Ianl therapy after acute myocardia!
iofarcdor~. Results of a c
erapy in ~~~e~~ypatients after
infarction, Second report
in Coronary Thrombosis to
hllicino FM. Xuereb RI. Aqt
acute myocardhl infarction in