Download PDF Article

JACC Vd.23. Na 5
April 19949016-32
1016
Technetium-99m Sestamibi Myocardial Perfusion Imaging in the
Emergency Room Evaluation of Chest Pain
THOMAS C. HILTON, MD, FACC, RANDALL C . THOMPSON, MD, FACC,
HUGH J. WILLIAMS, MD . ROBERT SAYLORS, DO, HOLLY FULMER, RN .
STEPHEN
A . STOWERS, MD. FACC
kromile, Fh 1da
Olijeriiws. The prpew d Ws iiveuIIga&a woo toewAssk tie
practicality d eiortdam predictive value otwo lgecmdW
pevlliam he" with kdmedoaofm dmlbl in mcpsey
room patient' win typkd angles and a normal ar aenRgamtk
dedromrdbpm (EGG).
Bahglveed. Aemraq d merpaey swam tans PRIM amen
meat may be Improved when clinical and ECG vselhke an meal
h Conjunction wide Mite *d1mmj 1 mpesmdY peefhtiw
Imaging. Tedmedmrfpm amadhi k a see, nallobdoper tied Is
taros up by de myoarBam is pmpseka to Well flew, let
mpka tallmrd01, it raWhiwtm slangy dam lejeetla ..
TeehrethmJ9mamtdbi mm dm be Injected dmhp they pair,
d bop agahad I to 2 h hear (when peeress have been
diulenI r.bawa) will cedrm whether aha .einnl w ofpers
des were present at tier times of hidecdon .
dhdmde. One hie,d ed two a gall room patients win
typical mom (on the book d ∎ standardised lyha gnmehsmire) and ∎ normal er needhpedle ECG Wd a hdiadm *m
seeIMrill y)adkn durbrg gmphms sail wen fellewad ap fw
oeasresa of .dean cardiac a In (svdae dWk osnfaW
myeewdal iduedon, "now molopleer am" surgery or
B-7 dremin(yeY).
Read,. tralvalaet prod oss d cardiac crab metal d de
peaeam d three or mew cwan ry room Bean (p - 0.009, elk
ratio 3.g mad as damormal or ogahoml sec tdtm ilinumm
ramlbi nma (p =1.d01, risk ratio 131) . M.Blvmide wpwdm ambeis Identified r ahnmd peetedm range m lie 6dy
hdepmded pride dadvmn madiae crew (Illm 0Aflf). Of
78 patiar win a am" pflde
. roa, only 1 lad a madjust
eras ampmed web is prisms Wish equip d area a 17
pad" will alaonsid ova, wed a cardiac treat sets of 1390
sad 71%, respectively (p - L11111161).
Cam/ata. Initial nyecaril perfusion Im{hg with
amlmdk.Alm .rbmibi when qipile h emmpaq ram pr
tame we gpkd won d ∎ aoomeI or amdkpmWr Et G
missiles to be bb* aonnrate le dkffmuMbg'
law said
Wills IM sabjects.
G As Cai (adW 1f9dd3:1Nd-22)
The emergency room evaluation of chest pain is a frequent
and troublesome problem. Accurate evaluation is made
difficult by the low specificity of standard clinical and
electrocardiograpihic (ECG) variables (1-3) . The come
qucnccs of inappropriate emergency room discharge of
patients with true myocardial Infarction may be serious (4,5).
Thus, most physicians practice a conservative admission
policy that results in a high Ievel of intensive can admissions
of patients with noncardiac chest pain (6 .7).
Previous studies have used computer algorithms (8,9),
two-dlcleeaional echocardiography (10,11), thalium-201 (121s) and technodum-PPm sesteodbi (16-22) to improve diagnos.
to accuracy and risk assessment . However, most studies have
used broad inclusion criteria . Including patients with typical or
mypkelchest pain and with dkgnosdcor andisgnostic ECGs.
TechmduntA9m sedmlbi is a new rialkhosop that Is
take
. up by die myocardium in proportion to blood flow . It has
a prolonged retention in the myocardium and redkblbwes
minimally after injection (23). That, teclmdism-9m cam&
Can be i*cted wide patina are acutely symptomatic . and
losses acquked 1 to 2 h later (when pelican have been
c iniealy stabilised) Will mission whether Profusion ebnorms4
irks were posed at the time of itdection (16-18), In the
current Investigation, we sought to determine whether acute
myocardial peAbdon imaging with technetium-99m seaamibi could accurately distinguish between low and high risk
emergency room patients with typical aopm and a normal or
nondiagoaic I;CO. Additionally, we sought to assess what
if any incremental predictive inn emation was provided by
initial technetium-99m seitkmibi imaging compared with
combinations of clinical and ECG variables .
F. the Dammed d Nudew Cudiol y. SI . Luke's Hogdel, leekrov9Ia FWW . TN. eddy *as supposed in pays by g ma fine Do post
Mack Pbuuaeudcol Cmpmy. Npah BBWWWce. Msemcauatswd Syocw
Rad9.phaymar,med aee.oy, Chtoweab . Csetea
Mmsmpt received Apt 29 .1990; revised osnnaipl received NovemW
ber 22, 1M. accepted December 1 . 1993.
Add,.re her n..~.emd•"es : Dr. 7be,ew C. ohm, 4105 Bdfon Rod,
Suite 2065 . Jackmav0a. Florida 32216 .
01994 by the Apraicm Cauea orCwaiokw
Methods
Patient selection. The study group comprised 102 patients
seen in the emergency room for evaluation of ongoing
agirmaike chest pain who had a normal or nondiagrwstic
5735 .10978M.00
3ACC yd .23, N.. 5
As s tm:lcls-n
TE[m1ET70M.99M
12-lead ECG and underwent acute technetium-99m sestamihi imaging
. Technetium-tpm sestamibi imaging was performed when requested by the referring physician for clinical
reasons. Chest pain was semigimntilatirely analyzed using a
modification of a previously validated chest pain score (24)
(see Appendix) and only patients with typical angina pectoris
(prospectively defined as a chest pain score a4) were
included. Patients were not included if there was historical
or ECG evidence of previous myocardial idarction . Patients
with atypical chest pain, ECG evidence of ischemia (ST
segment elevation. >_ I-nor ST segment depression or new T
wave Inversion) or contmindicatlons to !sdionuclide imaging
were also excluded. Of screened patients who were considered eligible <Ill% were not carolled because of lack of
available isotope . These patients had baseline clinical and
ECG characteristics similar to carolled patients .
Technique. All patients had a 12-lead ECG and bolas
[gjection of 10 to 30 mCi of technetium-99m sestamibi while
they were acutely symptomatic . To ensure that radioisotope
was available for injection after ecrmal operational hours of
the Department of Nuclear Medicine. two pee nixed slats of
techaetena-99m sesmmtbi were available for emergency
room use between B AM and midnight (morning and afternoon viols). The morning vial was calibrated for 30 mCi at
noon and the afternoon vial far 30 mCi a! 4 PM. This allowed
10 mCi of technetium-"m restamibi to be available far
injection until stidnigla . Thus. at teax out premixed vial was
available for initial imaging between B AM and midnight
7 daystweek. Favergen^y room physicians completed a
course in radiation safety, and radioisotope administration
was prrfmmed miler supervision of the emergency room
physician to ensure ute(-.in, during symptoms . Subsequent
to radieisompe Injection, patients were given appropriate
ant[dsehemiclantiaoginal lento . After stabilization of
symptoms patients were transferred to the Department of
Nuclear Medicine for imaging . Inmost patients imaging was
. 8D] 54 a
performed <60 mm alter igjeetian (mean time [
17 min, range 30 to 110) . Perfitaan houses were obtained in
a IV arc over bit views using 20 aiprQaction from the left
posterior oblique to the right ante for oblique position and a
siilgk.pbatan emission computed lemographic (SPECT)
camas equipped with a low enmgy, high resolution, paralW-hole collimator. Reconstruction was performed using a
Buttmwoeth filter without attenuation correction and a 20%
symmetric energy widow centered on 140 keV . The images
were rend immediately by a Scented expert nuclear cardiologist Or Iadlaloejat, and preliminary results were reported to
the attending physician . Images were subsequently analyzed
and soared by a consensus of two readers who were unaware
of clinical chmaeteristics . ECG results and outcome . Studies
were categorized as normal (low risk), abnormal (sigh risk)
or equivocal (intermediate risk) by visual qualitative amlyis.
Qualifying ECGs were interpreted by one of the inveslgamrs who was unaware of clinical characteristics,
technetium-99m sestamibi results and outcome . The ECGs
SESTAsOat1N
HILTON ai AL
771E EMERGrmey ROOM
1017
were categorized es normal or abnormal (bat nordiagnostic) .
Most abnormal (hut mmdiagnostic) ECGs had nonspecific
ST-T wave charges or bundle branch block. Normal (low
risk) ECGs were compared with abnormal but nondiagnostic
(higher risk) ECGs for predictive value. Follow-up ECGs
were obtained in 97 patients and analyzed for additional
predictive value.
Clinical end paints. End points were prospectively de .
hiked as prehospital discharge cardiac death. nonfatal myocardial infarction or the need for immediate coronary intervention (coronary artery surgery . coronary angioplasty Or
coronary thrombolysts) . Seventy-nine patients were admitted to hospital. Twenty-three (23%) of 102 enrolled patients
were discharged from the emergency room . These patients
were younger (44 t 11 vs . 52 ± 12 years [mean ± SDi . p
O.OOB) and had fewer risk factors (1 .4 ± 1 .2 vs . 2 .1 ± 1 .1 .
p = 0.01) than patients admitted to hospital but otherwise
had similar baseline clinical and ECG characteristics . Late
follow-up of these patients (including a limited history (23
patients] and 12-lead ECG [15 patients] m >90 days) revealed no adverse cardiac events or readmission for recurrent symptoms .
Statistical methods Statistical associations between adverse cardiac events and discrete variables were tested by
chi-square analysis. Associations between cardiac events
and continuous variables were analyzed with independent
r tests and univasiate linear regression analysis. Univariate
variables significantly (p <_ 0 .05) associated with adverse
cardiac events were examined for eofimeority and entered
into a multivariate regnsioa equation.
Remits
Clinical dmratmistira . Clinical charactetiatia, including
age. chest pain score, most coronary risk factors and use of
certain cardiovascular medications, such as digitalis. anti. anticoagulant or antianginal therapy, were evenly
platelet
distributed between patients with and without cardiac events
(Table I). Similarly, a patient history of previous heart
disease, conmaty anpography or exercise test was not
predictive of cardiac events . Conversely, compared with
patients without cardiac events . patients with cardiac events
had a signifiandy higher mean number of cardiac risk
factors (2 .5 x 1 .1 vs. 1 .8 ± 1 .1, p = 0.04) . a greater
prevalence of abnormal (but eondis®wstic) ECGs, were
more often men and had a slightly higher chest pain
scale .
Chains] risk groups . i n an effort to replicate clinical
judgment, patients were stratified into low, intermediate and
high risk groups on the basis of coronary risk factors (<3 vs.
a3) and their ECGs (normal vs . abnormal but nondiagnostic)
(fable 2) . Low (33 patients), intermediate (46 patients) and
high (23 patients) risk patients by these clinical criteria had
respective cardiac event rates of 6%, I t% and 35% (p =
O.0D7) (Fig . I). Combining the intermediate and high risk
groups and comparing them with the low risk group revealed
n
s
a
R
GJ
m Rifil l
n :1
Is
;!"
at
a
O
m
w
l~
721
~
all
Jg
81
JIM ~
MOIR!
'
l d
oil
ni ls--a If;
i
z
A3
JACC VU . 23. Na . 5
Aps 199CIO16-zz
TECHNETIUM-99M SESTANIEI
I
I
I
1019
l
I
X
HILTON ET AL
IN THE SNEOrENCY I910M
1
1
1
A
C
Fig- 2. A.'[he electtaardiegam of this 60 year old man with
typical angina shows oonspecife ST-T wave changes but is not
diagnostic d tscheaia. E, Tahxtium-99m sestemibi scan in the
shonmxis pop), honkeeal tastg-asis patddte) and vertical tongaxis
lissome views in the emergency roam lilt rdamel and before
hwtd disdmsge flight mists) . C. Early -may aagtogaphy
shows 80% namwiog in the fiatA obhue mooted stay and a Oiling
deiea sugpoatwe of thrombus In the bit chmtdkx co may artery.
The patient udoweat Was" administration of intmemenary,
tnakitme ad sabscgsent mummfd cmaary atgiopluty. Repeat
9ahnetiatr99n teslaoid scan become hospital discharge was nor.
.
ma, (B, rot mama)
tllal inters teats (oval accuracy 71%. p = 0 .01). Thirty-two
patients underwent prehospital discharge comar9 andTophy . Founeeo (70%) of 20padeds with= abnormal scan .
ihkb 4. Comparison d Smddvhy. Specificity and Overall
Mmacy d Nigh m hstmndiale Clinical Risk Shims' Versus
Abnormal or Egdvxd Tahoetoa-99m Sesteen'6i Raubs in
Prediction d Cardiac Evens
Overall
Semidvhy SpedOelty A-y
Clinical risk same;
Tedeniae89msesuabinadt
pYdu
-so T" 2.
111%
94%
NS
379
a%
0.mm
43%
t0%
0.0001
compared wish 3 (25%) of 12 with a normal scan, had
significant (>70%) coronary artery disease (p = 0.01). Significant coronary artery disease was evenly distributed be,
tween the haft anterior descending (39%) . left circumflex
. Of the 32 patients
(32%) and right (32%) coronary arteries
who underwent cardiac catheterizat(on, 8 had anterior perfusion abnormalities, and of these 6 had significant left
anterior descendingemonary artery disease (p = 0.0007) . 0f
4 patients with lateral perfusion abnormalities, 3 had dgnir
ieant left circumflex coroary attay stea0sia (p - 0.03), and
of 13 patients with inferior perfusion defects, 8 had significant right coronary artery stenosis (p = 0.0005) .
Follow-up. Follow-up, including limited history (94 patients) and I2-lead EC0 (32 patients) was performed at least
9o days after initial imagigg . One patient with a positive scan
1020
HILTON Er AL .
TECHNEr1UM.99M SRSTAMIm IN THE EMERGENCY
ROOM
underwent coronary angioplasty 5 weeks afterdischarso . No
late events occurred in patients with normal soaps .
Ualvariate rapessleu analysis . Univariate regression
analysis ofclinical, ECG and nuclear scintigmphic variables
identified the number of coronary risk factors (p = 0 .04) and
tochnetiumA9m sesteadbi scan results (P = 0 .0004) as predictors of adverse cardiac events . Technetium scan results
(p = 0.001) and chest pain acme (p = 0 .05) were significantly
associated with myocardial infarction .
MWdnrite mucelen amiss. Significant (p s 0.05)
univariale predictors ofadverse cardiac events were entered
into a multivariate wipessiou equation to assess their inde
.
pendent predictive value . Multivariate regression analysis
identified an abnormal techmriurnA9m scetamibi scan as the
only independent predictor of adverse cardiac events (p 0 .009). Both techoetium-99m eesmmibi and age were predictive of myocardial infarction (p < 0 .05) .
IDiscnon)on
Appropriate triage of emergency room patients with
either atypical chest pain and a normal BOG or typical
angina and a diagnostic ECO is usually straightforward.
Appropriate triage is more difficult in the subset of patients
with typical angina and a normal or nondiagnostic ECG. Tine
current investigation was limited to such patients . Correct
early management of this patient subset requires a risk
stratification process that can accurately separate the majorfly of patients into low and high risk groups . In the current
investigation, initial myocardial perfusion imaging with
technetium-99m seslamibi distinguished between most
(85%) patients at very low and very high risk far short-tctm
cardiac events (<2% and >70%, respectively) . Tints, this
technique identified both a lower and higher risk group
than standard clinical and ECO variables (alone and in
combiatkm).
®etrurardlegasphle and clinical variables. Previous
studies have confirmed the limited diagoslic and short-tern
predictive value of the initial ECG in acute chest pain
syndromes (1-3). Patients with chest pain and a normal or
nonspecific ECG have a low risk of Srathreatanlng compli-,
atins but the risk of nonfatal myocardial infarction is 1098
to 20% (25,26) . Thus, conservative management of patients
with chest pain and a normal or nonspecific ECO may yield
a low mortality but may result in lost opportunity for initial
Intervention and myocardial salvage.
PrevW donor with dwmmo4Ol. Thallium-201 scintigraphy is highly sensitive i the detection of myocardial
infarction when Isotope Injection and Image acquisition we
performed early after the onset of symptoms (12). Quantita.
Live analysis of the ncitigrams an enhance identification of
patients with multivessel coronary artery disease (I5). The
usefulness of thallium-201 imaging in unstable angina has
also been well described (27-29) . Large transmaal perfusion
abnormalities are seen in patients with variant angina and ST
LACC VaL 23. No. 5
Ap11 1991101e-22
segment elevation at rest (27). Rest angina with ST segment
depression or normalization of negative T wares is associated with more localized subendocardial perfusion abnor
malflles (28). However, there are logistic problems associated with the use of thallium-201 in acute isehemic
syndromes . Because thallium-201 redistributes after bdea
tom. optimal diagnostic accuracy requires that isotope iojec
tion and image apttdsition be performed while patients are
symptomatic_ This may reduce the safety of the technique.
Some investigators have suggested that diagnostic accuracy
is not affected when thallium-201 injection is delayed after
chat pain (15) . Other studies of initial nuclear scintigraphy
have demonstrated decreased sensitivity when radioisotope
injection is performed after resolution of symptoms (16) .
Also. because thallium-201 requires a cyclotron for its production, it Is impractical to keep doses available for an
unpredictable number of emergency room patients. In the
current investigation, logistic problems associated with
Ihalliuna20l were overcome by using technetimo-99m set,
consist.
Tedtnelr .Am a ulaalW . IM newer isotope has pro
.
longed retention by the ntyocardlum and redistributes mini .
mally after Injection (23) . Patients can be given an Injection
dwig cardiac symptoms and them treated with anti-ischemic
therapy before cardiac imaging . We believe that the high
level of predictive accuracy achieved with teehnetiwo-%n
sestamibi i this study Is attributable in pan to the absence
of significant delay between the symptoms and radioisotope
Injection . Additional advantages o' technetltmf99m seats .
Erbi include a phoiaemim)on of 140 keV and more optimal
doemery, allowing higher dosage and superior SPECI'
image quality. The advantageous properties of this new
isotope have been recognized in ontaagement of patients
with unstable coronary syndromes . Injection of the isotope
In pants with acute myacarial b Monition be Pore reperfusio therapy can define an area of risk, and repast study
titer therapy quantil tea myocardial salvage (17 .18,30.31).
In patients with acute myocerdil infarction and a nondiagnostic ECG, tedusad m-99m seslamlbi allows early
dlagtlosis and accurate identification of the ifrerteitad
artery (21).
TcineUwasd9m dam161 Nor Ii1W haagtq i peMsass
with shunt pai. Other studies have used techtetlum-99m
senemibi scintigraphy to asses spontaneous chest pain
(1622) and to rule out myocardial infarction in the enter
8ency room (19). These studies have ban limited by small
sample size (<50 sutyects), loose inclusion criteria (enrolling
patients with chest pain "suggestive of myaeardii eeh .
emda" [16D and delays of isotope Injection until after resolution of spontaneous chest pain. Nonetheless, these studies
have demonstrated high accuracy i identification of parents
with myocardial infarction . Similar to our experience .
Varetto a al . (19) observed that in emergency room patients
with a normal scan, there i a very low incidence ofadverse
cardiac events during hospital admission or after 3-month
Mow-up. Indeed, in the current investigation, a significant
JACC VOL 25 . No. 5
April 1994:1016-22
TECHNETIUM-99M
cardiac event did not occur in the group of patients with
normal scans until nerdy 100 patients had been enrolled .
Sensitivity in the current investigation (94%) is highly coosister with the observations of Bilodeau et al . (16) and
Gre
.goire and Theroux (22) . Both groups reported sensitivities of 96% when initial technetium-99m sestamibi was used
to diagnose acute myocardial infarction, Thus, widespread
application of this technique could achieve a significant
reduction in inappropriate emergency room discharge of
patients with myocardial infarction. The significantly higher
specificity (85%) of this technique compared with standard
clinical and ECU variables will hopefully allow reduction of
costly, inappropriate intensive care admissions of patients
with normal scans.
Cosacaoslderatloos . The cost associated with acute radionuclide seintigraphy ($600.011 in our institution) should be
considered. This expense may be partially offset by decreases in inappropriate intensive care admissions and shortened length of hospital slay in patients with normal acute
perfusion scans. Such patients are candidates for expeditious cardiac evaluation, including early stress testing (usually in <12 to 24 h) . The initial chest pain seen in such
patients is used as the rest or baseline image . Patients with a
compatible history and an abnormal scan are at significant
risk for myocardial infarction and may be candidates for
initial coronary interventions. such as lhrombolysis or coronary revascularization . Further study will be needed to
determine whether such patients might benefit from intravenous thrombdysis .
Ltd of the atody. Because of limited sample size,
only one event occurred in patients with a normal scan .
Small cell size could result In statistical error . Also, quantitative analysis of the tomographie images may have improved predictive accuracy . but our results are very similar
to previous reports (16,22) in which quantitative SPECT
analysis was used. Moreover, qualitative analysis is commonly used for the interpretation of myocardial perfusion
studies. Finally, formal analysis of cost-effectiveness was
not a put of tte current investigation but would have
obvious relevance .
Comeldaes. The present observations suggest that acute
technetium-99m sestamibi imaging is a practical technique
that can accurately distinguish between very low (<20) and
very high (>70%) risk for adverse cardiac events in most
(85%) emergency, room patients with typical angina and a
noadtagnostlc ECO. Compared with clinical and ECO vadshles (almne and in combination), this new technique appears
to have significantly greater specificity and overall accuracy
in the prediction of adverse cardiac events . Patients with
normal scans are candidates for expeditious cardiac evaltmtiast and possible curly hospital discharge. An abnormal scan
is associated with a high risk of myocardial infarction, and
such patients may be candidates for initial coronary intervention.
HILTON ET AL.
SESrAMIBI IN THE EMERGENCY ROOM
1021
Appendix . Chest Pain Score
Pdsitire
Location
Suhsternal
Preeordial
Left chest, base of neck, lower jaw, epigastrium
Radiation
Either arm
Shoulder, back, neck, lower jaw (arm not
mentioned)
Charactcr
Crushing, pressing . squeezing
Heaviness or tightness
Severity
Severe
Moderate
Associated symptoms
Dyspnea
Diaphoresis
Nauseatvomiting
Previous angina history
(previous history of angina pectoris : associated
with exertion. relieved by rest or relieved by
nitrvvglycerin)
Nrgarire
Location
Localized to apex
Character
Sticking . stabbing, pinprick catching
+3
+2
+1
+2
+I
+3
+2
+2
+1
+2
+2
+3
-I
-t
References
I . sends JE. Band DA, A®mpum D. Chahner B, 1550110,5 FJ. L- : ai the
initial decoamdiaWam 19 predict is-hospital conplicdiae of roots
my-dial nfluction. N Engl J Med 198$312:1137-41.
2 . Vihanueva FS tuba PJ, Afrenkteh A. Pdlork SG, Hwang LI, Kaul S .
Value mod GpytallnpsafMSPdmotkadsufwahwlingpaliaas presealil8
m the emergency mum with a udiao-rtted symptom for detemdeay
lane-mm marimmh . Am J Cstdiol 19D2dv746-JB.
3 . Lee TH . Rouen GW, weuog MC . Bated DA, Cook EF, Acne- D.
GoMmm L . Sealtivay, of merit ddiet criteria far diagnoring myoean,
dial iafumim abide 24 hours of hmpimBzadaa, Ann liters Mad 1917:
106:181-6 .
4. Lee TH, Ram GW. Weaberg MC. m al. Cladml ehmrctemtics and
assured bone, of paomu with MUM myocordkd kdseoiwe sent home
from the emergeney mom. Am J Cerdlo11987:m:219-20
.
LeeTH .userLW.Brand DA .leanChmdYD,et .1.Palheawhhuum
chest pain who tune cancycsay room depettmenls agmmt mcdcst
advice . J Gm burn Mod 19119 :21-4.
6. Lee TH, Good-L. The .-my cue null trans 3: Maorial trends
and fat- direction . Ann totem Mod 1989;10807-91.
7. Hall AG, Thlbmdt GE, Hbghes RA, Samett GO . Reder VA, Sheen
IL. The emme or pilau. with mspacd myucudb infmctiuo; the
idenriticoian of towwksk pticma for easy sander from intensive care.
N End J Mad 1910:10!:947-8.
8. Goodman L. Cook EF. Brand DA, m al. A compuler peanocd to predict
myarmrsal infaretioe in edurueuy deportment patients with taut pain .
N EraoJ Mad 1988318 :797401.
9. Potty MW. D'Agonioo an . Seam HP, Sylkowekt PA, Hoad wa. A
5.
psedit ice ietrsntem to improve coronary core unit admislm pmmtees
in seat- hehaoia heart disuse ; a prospective mukiceider clinical trial .
N Erro I Mad 1984;7m127.1-8.
lo. Homwita as. Margaruotk J . Immedioe detection of early high-risk
pltnls with acne myocardial Infection win& two-0Imensimal electN
;s
gad
IC
INN
Id
IL
F
a
p
ELI
10
P!
P!
Sa
y
-ei!j so
Pi
a
rj
i
U,
RURIP14 111