MD. FACC. WOLFGANG KIOWSKI, MD.
MD, W,TH THE TECHNKAL ASSISTANCE OF
n&e investigated pmspeclively in pafiat& with persistiig
asymptanattc canptex arrhythmias oner myocardiat in-
farction. Eod points were total mortality and arrhytbmir
evell,s (suddo” d&b, ws,ai”ed “a”,r*“lsr
and vsn,rictdar fibrillation).
bUmT,bx,, 312 bad Lown class 3 or Ib arrbylbmia on 24 h
eledmcardiograpbii rmmiings before bmpilal discharge
and come”,ed ,a ,lle study. They were randomized to
Miv,d”sdized Pnlisrrbyibmie ,Tea,men, Group 1, ” =
IW), hearmpn, with low dose amicxtamne, 200 mg/day
Comp!ex ventricular arrhythmias penisting after myocardial
infarction have been identified as a predictor of mortality
(l-3) independent of and additive to the prognostic value of
left vent&ular ejection fraction (4.5). However. to date.
antiarrhvthmic drue trials (6-9) have not demonstrated B
reduction in mortally: in fact. an excess mortality has been
rewrted (IO) in patients treated with class Ic antiarrhvthmic
d&s. Earlier &dies of antiarrhyrhmic agents have been
criticized on several grounds: fixed dosage. single antiar-
rhythmic drug and no serum drug level measurernenb.
However, an individualized approach to optimize :berapy
could bc important for the success of antiarrhythmic dw
therapy. Alternatively. a simple regimen favoring good com-
plrance and tolerance based on once daily application of a
Ions_acting drug might reduce the high dropout rate ob-
s&d in previ& &dies (9).
We designed a pm~pecttve randomized tnal in patents at
high risk after myocardial infarction: only patients with
complex ventricular ectopic activity persisting up to hospiral
discharge were entered-into the study. They were then
randomly allocated to one of three treatment regimens: 1)
clars I agents selected on the basis of suppression of
ventricular rctapic activity and patient tolennce: 2) wan-
dardized antmrrhythmic treatment with low dose amio-
rhythmic therapy (control group). Over a I year follow-up
period. clinsal events, antianhythmic drug effects and side
effects were worded to assess s!lether one or both of the
chow, ant+hvlhmir dn? @m**s
cuuld r+w m&a!ity.
treatment. startin% with
a drug with a long half-life (I I): and 3) na anti=-
tively admttted to the coronary care unit of the Universl!y
Hospitel of Beset with acute mywardial infarction between
Joly !, 1981 and June 30, 1967 were considered for this
study. leczur~ F. planned interim analysis after 2.5 years of
oatient recroitment showed that the incidence of complex
;enhicular arrhythmias as defined in this study was lower
than predicted. the study was extended to two afflialed
hos&ls (ffintonsspit;ile~ Bruderholz and Liesml) to obtain
adequate patient numbers.
Before hospitai dkhar ge.
for persistent complex ventticulrr
Death in the hospital, symptomatic: arrhythmias or cardio-
vascular surgery &ring ho~pitalizt.tion. life-threatening CD
morbidity and unavailability for tallow-up evaluation were
reasons for primary exclusion. Only patients who showed
asymptomatic complex ventricular ectopic activity on a 24 h
electrocardiographic (ECG) recording before hospital dis-
charge were included in the rtudy. Each patient received a
consecutive number based on the time of entry into the
hospital that WBP later used for randomization. which was
performed separately for the three hospitals.
Study protwoi. All patients were admitted to a coronary
care unit, where they remained for 48 to 72 h doting the
acute phase of infarction. They were then transferred to a
general ward and mobilized according to a defined scheme
socb that omieots with uncomdicated infarction were dis-
charged a&r 7 days. those withminor complications&r
to 14 days and those with severe complications after 3 to 4
One to 3 days before hoApiral discharge. all patients
without primary exclusion criteria had an ambulatory 24 h
ECG recording after discontinuation of all antiarrhythmic
drugs for 224 h. Amiodarone was never used during the
initial hospital stay. The 24 h ECG recordings et all three
hospitals were made on a Dynegram model 5000 cassette
recorder and centrally analyzed (playback at 60 times real
time speed, lnstrttments for Cardiac Research, model W.MC
Holler electrocardioscanner). The entire 24 h ECG was
condensed beat by beat on fiberoptic p:~per. Arrhythmias
detected by trained personnel were recorded at normal paoer
speed and analyzed by a cardiologist according to the Lawn
classification (12) and the number of ventricular premature
beats per hour. If frequent multiform or repetitive ventricu-
lar arrhythmias (Low class 3 or 4h) were present in 22 of
24 h. patients were asked to consent to the study. They we
then iandomized to one of the three treatment .&upson the
basis of their hospital entry number and the randomization
list. Standerd mdionuclide angiogrephy was performed to
determine left ventricular ejection fraction et rest (13).
Fdi~Mv-uo VI.,:~ were scheduled for all patients after 3. 6
sod 12 months and included clinical examination, 24 h ECG
All patients <?I years of age consecu-
l,?Zf patients were screened
recordings and drug level determinations if applicable. Pa-
tients who were treated individually had additional visits as
required. Patients no longer willing to adhere to the treat-
ment regimen or to show up TOI follow-up eiaiwtion were
contacted by telephone at the end of the year to obtain
information about survival.
The study protocol WBS approved by the Ethical Commit-
tee of the University Hospital on Januery 29. 1981 and has
previously been described in detail (14).
Anliarrilyrhmic Trenwnenf Regimens
Individualized autbrrhythmic lreatment (Group 1). While
still in the hospital, patients in this group underwent oral
drug testing with wntinuous ECC monitoring in the cm+
nary care unit or with repeated (usually six times) 30 min
ECG trend scribing. Thereafter, drug therapy was monitored
cm an outpatient b%s by frequent 24 h ECG recordings and
serum drue. level measurements. In this t?mup, treatment
was startea either with quinidine or mexiletine and dosage
adjusted according to eficacy and side effects. Suppression
of arrhythmias was considered effective if Lawn class b
arrhythmias were eliminated and the number of ventricular
premature beats was reduced by >SU%. In patients with
ventricuiar tachvcardia. elimination of runs of three or more
ventricular pre&tore beets and YW% reduction of couplets
had to be achieved. Ifthe first drop, did not achieve this goal,
the second drug WLS tested. In c&e of failure of both d&s.
other aotierrhythmic drugs (ajmeline. disopyramide, Recain-
id& propafenone or soralol) were tried. If none of these drug
regimens was effective, amiodarone was given.
Staadardiad anti&-rh@mii treatment with tow dose ami-
odemite (Grow 2). Patients were given amiodarone (1,ooO
mg orally for 5 days followed by a daily dose of 200 mglday).
If symptomatic arrhythmias developed despite amiodarone
adminishatian. antiarrhythmic therapy was changed and
patients were considered to be treatment deviators. Patients
were specifically questioned about side effects and under-
went thyroid function tests, but no chest X-ray films or
pulmonary function tests were routinely performed.
Cootrol group without erdiarrhythnde therapy (Group 3).
In this group, patients received no specific antierrhythmic
treatment but were seen after 3,6 and I2 months when they
underwent 24 h ECC monitoring. If symptoms due to
arrhythmias developed, they were given antiarrhythmic
drugs and were considered to be treatment deviators.
Conwmitent treatmeal. Treatment other than admiois-
tration of antiarrhythmic drop was not fixed and was left to
tbe judgment of the responsible physician. Antithrombotic
and beta-adrenergic blocking drugs es welt as coroeery
artery bypass surgery were considered standard treatment of
ischemic heart disease that could not be withheld for ethical
reasonr. Such concomitant rreatment wa recorded as in an
End point evaluation.
edge of the treatment group awgnment; evaluation was based
on death certificates, physician and hospital records. au~p\y
repons and tnfonnation obtained fmm relatives and witnews.
Each death was asrimted to the treatment erou~) accordinr to
the intention to &principle (15.16).
Sudden dcorh was defined as death occurring without
previous symptoms or within I h after the on&t oi new
symptoms of heart disease (17). an unwitnecsed deaeuih wa\
icduded in this delnidon if Ihc pa&xl had been ubscrvcd w
be free of symptoms ~24 h before he or rhe was found dead.
Other csrdiuc deolh was defined as death GCCU~
after the onset al cardiac symptoms. Cardiac arrest due to
ventricular fibrillation with successful resusc~tatmn and OE-
currence of sustained symptomatic or asymptomatic vcntric-
ular tachycardia (210 consecutive beatc) were constdered
secondary end points. Sudden death and sustained vattic-
ular tachycardia or ventricular fibrillation were considered
EmhWiott of treatmat etTWs. Effectiveness of the dif-
ferent muiarrhythmic dlug regimens in suppressing ventncular
ectopic astivity was assessed with respect to freqaency of
ventricular memature beats and Lawn class. Fresuencv was
defined by the average number of venrricular pre&mre~beatr
per hour duing the entire 20 to 24 h ECG recording and the
presence or absence of ?I0 ventricular premature beats.
Slalii methods. Ott the basis of our initial expectn-
Lions, a sample size of IO8 patients per group was estimated
on the assumption of a I ycer mortality rate of 20% and a
reduction in the mortality rate of ~50% by either drug
Deaths were assessed without knowl-
ring >I h
rcguncn s n = 0.05. pvwuer = 80% ““e-tailed
randomired patients were included in ihs mortfhty aw.tyses
and all pauentr for whom data uere available wcri: mcluded
,n rhe other analyses. All valoes :qxcsent rnezn wlucs - 1
SD as an Index of dtspernon. Wdhtn treatme”, gioups.
chxtge: i!! arrhy:hmias were aiae~d
WC\ analyw of variance and between group differences
wt.? the cht-square test. The pmbabdity of survival and the
occurrence ofarrhy~hmic events were calculated by Kaplan-
Mcier acnxanal analysts (18.191. The Cox proportional haz-
ardc rcgxc+n model (201 was used for adjustment of
Fowblc baseline differences according to an intention to
treat approach. All the reported p values are twwtxled.
Calcuk~tmns were performed with the use of the Systat
with repeated mea-
Incidenrr and natural ECMM of persistent venlricular
eclopic activity. During the study period. I.220 consecutive
patients ~71 years of age were screened for persisting
complex ventricular ectoptc actwtty before hospi!al dis-
charge Of this cohon. 340 08%) had qualifymg asympto-
matx arrhythmias: 28 patients (!3.2%1 did not consent to the
<tudy. rewiring in a study group of 312 patients. Of these.
C7r/, had repetitive forms of ventricular premature beatr
(Low class 4). The average number of ventricular prema-
ture beats per hour was 48 + 107. with 49% of patients
having sID ventricular premature beats/h.
Group 2 (low dose amiodarone) and I14 to Group 3 (control)
(Table I). Tbrre were no significant differences among the
three groups with regard to age, gender, history of coronary
artery disease and hypertension, Q wave infarction, peak
creating kinase values or left ventricular ejection fraction.
The incidence and severity of ventricular ectopic activity
were similar in all three groups, as was the number of
patients with concomitant antithrombotic and beta-blocker
E&x& of treatment on looaality. Tbe incidence of death
in the three study grows is shown in Table 2. Overall. 30
patients died (I year mortality rate 9.6%). Tbcre were three
noncardiac deaths (two due to carcinoma and one to major
mxardiac surgery). Twenty-two of the 27 cardisc deaths
were sudden; P of these were unwitnessed.
Figure I s.navs
the probability of survive! for the threr
treatment groups. TIx cumulative mortality rates at I:
months were 13% in the control group. 10% in the individ-
uatly treated group and 5% in amiodarone-treated patients,
indicating that in this study amiodarone reduced tbe totat
mortality rate by 61% from that in the control group (p c:
0.05). This diierettce also held true after adjustments in the
CQ?. model for the minor diierences in baseline variables,
such as age, previous myoeardial infarrrion and left vearic-
alar ejection fraction. individual treatment reduced the
mortality rate by 24% from t&t in the control group. but this
di5erenc.e wrts not significant. A&r exclusion of noncardiac
deaths, the mortality rate was still 55% less for amiodarane-
treated patientsandZl%les~ for individually treated patients
@we I. Prcbability of I2 month survival in the three treatment
grows of patients
mias after mywardii infarction.
complex ventricular arrhyth-
3. Comparison of Survivors and Patients Who Died From
than for the control patients. Patients with cardiac death
d&red significantly from sutvivors only by a lower left
ventricular ejection fraction (Table 3).
EBeds of tmlmeot on u&ythmk evett!s. During the I
year follow-up period, eight patients had symptomatic ven-
tricular tachycwdia or fibrillation and were rucccsstidly
resuscitated. In five additional patients, sustained asympto-
matic ventricular tachycardia was recorded on routine 24 h
ECG recordings (Table 4). Sudden death and sustained
ventricular tachycardia or fibrillation warred
less o&en in patients treated with amicdamne than in the
control group (5 of 98 versus 19 of 114. nduction of 66%.
p < 0.01). whereas the etlbct on arrbythmic events in
individually treated patients was less marked (10 of 100.
reduction of 40%. p = NS). Likewise,
clinical course free of symptomatic wrbythmic events for the
three groups showed a significant bettdit only for amic-
darone-treated patients (p < 0.025) (Fig. 2).
the probability of a
Lawn class z-3 arrhythmias. the main the&eutie targei for
this study, were signbicantly reduced by bath treatment
regimens at each follow-up study (Table 5). A comparison of
the number of ventricular premature beats per hour in tbe
three groups at the 3, 6 and I2 month studies showed no
significant differences; however, there was a significant
treatment effect over time with amiodamne, which w.xz less
pronounced and not significant for individually treated pa-
Table 4. Incidence of Anhythmic Events in the Three
Flgwr 1. Robability of an arrbythmic evenl-free course during the
I year follow-up period in ,he three trtaunen, wup3 of pa,xn,r
infarclion. Anhythmic even& are sudden dearh. vemriculx fib+
lalion and symptomatic suslained ‘~enrricular uchgcardxa.
Anttarrbythmie Irealmmt given. Gf IW drug trials in ,he
98 individually treated patients, 70 were performed with
qtinidine and 79 with mexiletine in different doses. Respec-
tively, these dlugs proved to be effective in 3% and 32%.
ineffective in 47% and 42% and were discominued because
of side effects in 23% and 26%. ThiTtY-three patients in
whom both treatments failed were given pmpafenonr tn =
r7), Recainide (,, = 14). sot&l (n = 5). disopyramide In = 4).
prajmaline (n = 4) or drug combinations (some patients
recetved up 10 seven diEwent drugs). In addirion. nine
patients crossed over 10 amicdarone aRer an average of 3.3
previous drug trials. Palients in lhis group who survived ,he
I yea, study paiod had an average of 2.1 drug trials. 7.4
?4 h ECF recordiqr and 3.5 drug level dererminations. In
the xmiodaronr-treated group. four patienls received an
mcreased dose. whereas five had their dose reduced ,o 101)
Treatment de&x&x, and tdereoce. In Table 6 the inci-
dence cl side effecects. lack of compliance and change of
treatmen, are compared ior all rhree study gmoups. Twenty-
two patients changed their study medicarion: II patients in
the conlml group experienced symplomatic aatxylhmias and
were given antiarrhythmic medication. In nine individually
treated pa,ien,s (Group I). no treatment proved to be e&c-
five or wilhoo, side etTec,s and they received amiodamne. In
two patients from the amiodamne treatment group (Group
2). the trearmen, age”, was changed 10 so,alol because of
Sixty-nine patients (22%) disconlinued the study prema-
turely because they wilhdrew their cooper&on (n = 42) or
T&k 5. E&c, of Anfiarrhylhmic Trearmen, on Lawn Arrhylhmld Glade and Frequency of
Vcn,ricular Ecropic Ac,ivi,y
Lawn gr& ,!a p,icn,s
whh Low” @. 1 I,
experienced side e5ec;s (n = 27). Fourteen individually
treated patients and 13 amiodamne-treated patients tina!ly
stopped their medication because of side effects. Reasons for
discontinuation of amiodarone were ocular side ebcts (n =
3). hyperthyroidism or hypothyroidism (n = 4). bradycardia
or dizziness (n = 2). dermatologic effects (n = Z), gastric
symptoms (n = t) and hair loss (n = I). No pulmonary and
no irreversible toxicity were observed. Altogether. 59 pa-
tients (59%) tn the individually treated group (Group I)
finished the study with their assigned treatment, compared
with 69 patients (75%) in the amiodarone group (Group 2)
and 93 patients (82%) in the cnntrol group (Group 3).
mitt. This study showed a significant reduction in mortality
by amiodarone in a high risk group of patients with arymg
tomatic persistent complex ventricular arrhythmias after
myocardial infarction. The prevalence of ventricular prema-
ture beats 2 weeks after infarction was similar to that
recently described in the placebo group of the Beta-Blocker
Heart Attack Trial (BHAT) (22). Lnwn class r3 arrhythmias
were selected as target arrhythmias for the present study in
1980, long before the greater predictive power of higher
density and repetitive forms of ventricular premature bents
was realized. The same is true for the definition of arrhyth-
mia suppression. The observed I year mortality rate of 13%
in the control group confhms the selection of a high risk
group of patients for this study: it is much higher than the
control group mortdity rnte observed in the Cardiac Ar-
rhythmia Suppression Trial (CAST) (IO) control patients.
fatluanee of nntiarrhythmic therapy nn mwtality afler
mynawtial iufarction. Although ventricular ectopic activity
is assacia:ed with ?n increased mortality rate in srrvivors of
acute myocardial infarction and although this effe :: seems to
be ittdependent of other risk factors, there has been no
evidence to date that treatment of arrhythmias cm reduce
this risk (6-9). In retrospective s!udies (9,22’. oatieats
responding to antiarrhythmic therapy had a si::n8tcantly
better clinical course than nonrespnuders. indicating that
drug response might be a marker of a more benign undedy-
ing disease. The results of the present study with the
individualized antiarrhythmic treatment approach showed a
suppression in ventricular ectopic activity without a signift-
cant reduction in mortality. However, no ewess mortality
was observed in our patients treated with class 1 antiarrhyth-
mic drugs in contrast to what has been reported recenrly in
CAST (lo) for class Ic drugs. Fkcainide. which depresses
conductbn more than other class I antiarrhythmie agents
(23). was used infrequently in our study and encainide was
not used at all. Conversely, amiodarone significantly re-
duced not only the total mortality rate, but also all arrhyth-
lhtrapy of pmtiufaretlau ventrfeufar arrhylh-
Possible meebantsms uf n&n
arrhythmia suppression with amiodarone was similar to that
with individualized therapy, other mechanisms for mortality
reduction need to be discussed. The antifibrillatory et&t of
amiodarone (23) seems to be the most likely beneficial
mechanism. acting by slowing of ventricular tachycardia.
Other mechanisms seem mnre speculative; amiodarone has
antiischemic (24.25) and antiadrenergic (26) effects that
might be contributory in patients after myoeardial infarction,
as in the case of bera-blocking drugs. The special phamta-
cokinetics of amiodarone with its long half-We might be
particularly impnrtaut for a sustained e5ect. Sudden death
occurred more frequently during the loading phase of this
treatment arm (three of four sudden deaths within the first 2
manths ofamiodarnne therapy), whereas it occurred later in
the individually treated and cotttrol patients. One could
speculate that with a more aggressive loading dose at the
beginning oftreatment, the result could have been improved
even further. In addition, amiodarone exerts virtuatly no
negative inotmpic effects during long-term administration
(25). Finally, the minimal proarrhythmic properties of ami-
odarone are in contrast to those of other anliatrhythmic
drugs such as Aecainide and encainide (27).
dividualized antiarrhythmic treatment approach was chosen
for Group I to find an aptimal drug treatment regimen for
each patient. Because of the large spontaneous vatiability in
arrhythmia nceurrence, this demanded not only aeute drug
testing, but continuous adjustments of therapy throughout
the follow-up period. Only 39 patients were given just one
autiarrhythmic drug, whereas 35 patients received at least
three different drugs in various dosages. Despite the possi-
bility of avoiding side effects by changing treatment, 14% of
the patients in this group finally discontinued all antiarrhyth-
mic therapy for this reason. Interpretation of the results in
this group is difficult because arrhythmia control was rela.
lively pow and these patients bad many more follow-np
visits for treatment management. In addition, 50% of pa-
tients who died in this group did so after protocol deviation
(Table 6). Aitogether, individualized antiarrhythmic therapy
proved to be cumbersome and costly and did not reduce
Spenific us@s of the study design. This study was
performed prospectively and randomization to the three
treatment groups was balanced. However, because of its
principal design, the study could not be performed in a
double-blind fashion. Therefore, the main emphasis of this
study was placed on mortality (28). Because there was a
significant mortality reduction with one treatment regimen,
this result is vu80 despite the lack of a double-blind design.
To cnrrect for minor di5erences in baseline characteristics
that could have influenced survival, the Cox model was used
and confirmed the significant difference in mortality between
the amiodarone and control groups. Beta-blocking drugs
of smkdamm. Because
treatmeut sppfoaeh. An in-
were not withheld in the mesenl studv. althoueh thev arc
known to influence sur&al
(29-32). On one hand, it was iudeed cthxallv imoowble to
withhold beta-blocker therapy, and on the oiher. trzatment
with propranolol has own shown not to alter the ielation
between ventricular ectopic activity and mortality 133).
Finally. amiodarone is a drug with ootcntially severe \ldc
effects (34-36). Therefore. the low- dose re&nen of ?W
mp/day was used. Careful patient monitoring (36) chowed a
much lower incidence of sihe effects than that reported wth
higher dosages (35.36).
Rekvance and implications of the reulls. Despite the
relatively small number of patients
overall differences in mortality and arrhythmic events were
large enough to reach statistical significance. The obwvcd
reduction in mortality was of such a magnitude that it is
clinically relevanl even if only cardiac deaths were consid-
ered. This inlerpretstion is supported by the highly GJ ,IS-
cant di6erences in symptomatic and all arrhythmlc evwi~
between amicdarone-treated and conlrol patients. In addi-
tion, similar differences in favor ofamiodarone were noted m
the smaller Canadian Amiwlaronc After Myocardial Infarc-
lion pilot study (37) as well as in patients with hypertrophic
cardiomyopathy (38). Nevertheless, these findings will re-
quire confirmation in more extensive studies under way in
Europe and the United States.
In view of the large number of palients who experience
sudden cardiac death (9.39). improving survival in these
patients is ofgreat clinical importance. Our findings indicate
that a simple treatment regimen with low dose amiodarone
may reduce mortality in the high risk subgroup of patients
with asymptomatic complex venbicula: ectopic activily per-
sisting after myocardial infarction.
after &ocardiai mf&ion
enrolle:! in the study. ths
JACC Vol. 16. No. 7
Dcccmbcr I!wz 111 I-8 THERAPY AFTER ltwARCT,ON
33. Korlir JB. Wdran AC. Sanders MB. B>,ngtan R for the BHAT Study
Group. Pronnstic significance of ventricular eclopic activky in w’vivors
of acult myocardial infarction who received propranolal. Am I Cwdiol
31. Hegerll. Ryrlowrky
dosage. drug concentmlions. and adverse ride4fects.
EN. Ziper DP. Rclalionrhipr between amiodarone
Am Hean J
27. Keren A. Tzivonl D. Gotllteb S. BenbonnJ. Stem S. Alypica! venlricular
tarhycard~n (torsades drpaintcrl inducedby am~odamne. Chest 1982:8l:
28. Greene HL. Sudden anhylhmlc cardiac death: mechanism. rewcilation
and elarsificabow ibe SeaClc pcrrpeclwe. Am J Cardml 1990:65:48_12B.
29. Norwegian Mullsenter Study Group. Tlmolol-induced reduclion in m m -
talky and reinfaaion in palien& surviving acute myocardbl infarction.
?4 End J Med 1981.304:801-7.
30. Bela.Blockr~ NearI Attack Trial Research Group. A randomized trial of
prap~a~alol in patn;:z with acute myocardial infarclion: I. Monalily.
J AMA 1982.247.1707-Il.
31. Hja!manon A, Elmfeidl D. Herlitz 1. et al. Effect on moitalily of
m~mprolol in acute myocardial infarcdan. LkxeI
32. Yusuf S. Pet0 R. LewisJ. Collino R. Sleight P. Betablockade duringand
alter myacvdlal infarction: an overview of mndnnized
Cardmvasr Dii 1%%27:33g771.
35. Ha&,ee CL Love JC. AlpeR IS. Ardaurian CK. Sloan KC Efficacy and
safely of long-term amiodaxme in frcatmenl of cardiac anhylhmiar:
dosage experience. Am Heart J 19831~935-43.
36. Singh BN. Nadcmance K. Kannan R. lkeda N. The cknical rewI1s of
amiadamne in cardiac anhylhmiar: apbmal dosing. PACE 1984:7:109-
37. Cairns 1. Connol!y 5. Holder D. 4 al. Suppression of port-myocardial
infarction VPBs by amiodamne: randomized double-blind placebo-
controlled study labstr). Circulation W?WB(suppl 111:1141.
38. McKenna WI. Adams KM. Paloniecki JO. Ed al. Long lam survival with
amiodaroneinpalientr with hypcrlrophiccardiomyapathyand
tachycardia labslr). Circulation 1489BXsupp1 11):11-7.
39. Mlllar JE. Ludmer PL. Willich SN. et al. Circadian variation in the
frequency of sudden rard,ac death. Crculation 1987:75:131-8.