CHF MAY/JUNE 2001146 DOFETILIDE IN PATIENTS WITH CHF

Introduction. Atrial fibrillation is a frequent cause of worsen-ing of symptoms in patients with congestive heart failure. Thedrugs currently available for maintenance of sinus rhythm allhave major side effects. Methods. In 34 Danish coronary care units, 1518 patientswith congestive heart failure and reduced left ventricular sys-tolic function were randomized to receive either placebo or anew class III antiarrhythmic drug, dofetilide. The dose ofdofetilide was adjusted according to the presence of atrial fib-rillation, the length of the QT interval, and renal function.Patients were continuously monitored electrocardiographicallyfor the first 3 days of the study. The primary end point was all-cause mortality and follow-up was for at least 1 year. Results. In the dofetilide/placebo groups, 311/317 patientsdied (41%/42%). The hazard ratio for dofetilide treatmentwas 0.95 (95% confidence interval, 0.81–1.11). Treatmentwith dofetilide reduced worsening of heart failure significantly(hazard ratio, 0.75; 0.63–0.89). After 1 year, 61% of patients with atrial fibrillation at the start of the study had converted to sinus rhythm on dofetilide, vs. 33% in the placebogroup. After conversion to sinus rhythm, 78%/43% of patients in the dofetilide/placebo groups remained in sinusrhythm for at least 1 year. There were 25 instances (3%) oftorsade de pointes ventricular tachycardia in the dofetilidegroup and none in the placebo group. Conclusion. In patients with congestive heart failure, dofetilidecan effectively convert atrial fibrillation to sinus rhythm andmaintain sinus rhythm after conversion. Hospitalization forcongestive heart failure is reduced. Dofetilide does not affectmortality. (CHF. 2001;7:146–150,155) ©2001 by CHF, Inc.

Congestive heart failure is a serious condition, mostoften caused by reduced systolic function of the left ven-tricle. Atrial fibrillation is frequent in patients with con-gestive heart failure and is associated with highmortality, reduced exercise capacity, fluid retention,and thromboembolic episodes. Therefore, treatmentsthat can maintain normal sinus rhythm in patients withcongestive heart failure are potentially useful.

Drugs currently available for conversion of atrial fib-rillation to sinus rhythm and for maintenance of sinusrhythm after conversion all have significant drawbacks.Class I antiarrhythmic drugs are associated with athree-fold increased risk of death in patients with con-gestive heart failure.1,2 The only antiarrhythmic drugthat has been shown not to increase mortality is amio-darone;3,4 however, this drug has frequent and seriousside effects.

Dofetilide is a new class III antiarrhythmic drug thatselectively inhibits the rapid component of the latepotassium current through cell membranes and therebyincreases the refractory period.5 As a pure class IIIdrug, dofetilide has no negative inotropic effect and noeffect on atrioventricular conductance. Dofetilide canconvert atrial fibrillation to sinus rhythm and maintainsinus rhythm in patients without heart failure.6–8

Dofetilide was recently approved in the U.S. and Eu-rope for maintenance of sinus rhythm after convertionof atrial fibrillation.

The Danish Investigators of Arrhythmia and Mortal-ity on Dofetilide in Congestive Heart Failure (DIA-MOND-CHF) trial was designed to determine whetherdofetilide influences mortality and morbidity in pa-tients with congestive heart failure and reduced systolicfunction of the left ventricle. When the study wasplanned, we anticipated an effect on atrial as well asventricular arrhythmias. Since the result was thatdofetilide was predominantly effective in the treatmentof atrial fibrillation, this paper is focused on this use ofthe drug.

MethodsPatients. The design of the study has previouslybeen published.9 Patients with congestive heart fail-

Dofetilide in Patients with Congestive HeartFailure and Left Ventricular

Dysfunction: Safety Aspects and Effect on Atrial Fibrillation

Mogens Møller, MD, PhD; Christian T. Torp-Pedersen,MD, PhD; Lars Køber, MD, PhD, on behalf of the Danish Investigators of Arrhythmia and Mortality on Dofetilide (DIAMOND) Study GroupFrom the Department of Cardiology, Odense UniversityHospital, Odense, Denmark

Address for correspondence/reprint requests: Mogens Møller, MD, Department of Cardiology, Odense University Hospital, DK5000 Odense, Denmark Manuscript received June 19, 2000;accepted October 12, 2000

CHF MAY/JUNE 2001DOFETILIDE IN PATIENTS WITH CHF 147

ure consecutively admitted to 34 Danish coronarycare units were screened for participation. The diag-nosis of congestive heart failure was based on clini-cal judgment. It was further required that within theprevious month the patient had experienced dysp-nea on minimal exercise or at rest (New York HeartAssociation class III or IV). Function of the left ventricle was studied by echocardiography. Eachparticipating center recorded echocardiograms onvideotape, and these were evaluated at a core lab be-fore randomization. Patients could participate if thewall motion index10 was ≤1.2 (corresponding to anejection fraction of ≤35%). It was further requiredthat patients be at least 18 years old and, if female,not at risk of pregnancy. Written informed consentwas required. Patients with an acute myocardial in-farction within the previous 7 days were evaluatedfor participation in the DIAMOND-MI study (un-published data) and are not discussed here. Otherexclusion criteria were a heart rate below 50/minwhile awake, sinoatrial block, second- to third-de-gree atrioventricular block unless treated with apacemaker, previous drug-induced proarrhythmia,a frequency-corrected QT interval above 460 msec1/2

(500 msec1/2 in patients with bundle branch block),diastolic blood pressure above 115 mm Hg, diastolicblood pressure below 80 mm Hg, serum potassiumbelow 3.6 mmol/L or above 5.5 mmol/L, recent useof a class I or class III antiarrhythmic drug, a calcu-lated creatinine clearance below 20 ml/min,11 signif-icant liver disease, acute myocarditis, plannedcardiac surgery or angioplasty, aortic stenosis, car-diac surgery within the preceding 4 weeks, andtreatment with an implantable defibrillator.

Organization and Design. The Danish Board ofHealth and the Central Scientific Ethical Committeeapproved the study, which was conducted in accor-dance with the Helsinki declaration and Guidelines forGood Clinical Practice in the European Union. Thestudy was led by a steering committee. All deaths in thestudy were classified by an event committee accordingto The Cardiac Arrhythmia Suppression Trial (CAST)12

criteria—with the exception that resuscitated cardiac ar-rest was not classified as death. An arrhythmia commit-tee classified documented arrhythmias. Polymorphicventricular tachycardia was classified as torsade depointes, if one of the following criteria was present: aheart rate ≤50/min before the episode, a frequency-ad-justed QT interval of >450 msec1/2, abnormal andchanging T wave configuration, ventricular prematurebeats with a long-short coupling interval or a typical ro-tating QTS axis during the tachycardia.

Eligible patients were randomized to double-blind treatment. All patients in the study were con-

tinously monitored by electrocardiography (EKG)during hospitalization for the first 72 hours of treat-ment to ensure that all arrhythmic events wererecorded and treated.

When the study was started, all patients withoutatrial fibrillation recieved 500 µg dofetilide (orplacebo) twice daily, and patients with atrial fibrilla-tion 250 µg twice daily. After inclusion of 288 pa-tients, dosing was changed on the basis of data fromother studies. Thereafter, the initial dose was ad-justed according to calculated creatinine clear-ance.11 Based on the doses above, patients with acreatinine clearance of 40–60 ml/min received, atmaximum, 250 µg dofetilide twice daily and pa-tients with a clearance of 20–39 ml/min received amaximum of 250 µg twice daily. The smallest dosein the study was 250 µg once daily. The dose ofdofetilide could be reduced if the corrected QT in-terval increased more than 20% or increased tomore than 550 msec1/2. If a reduction in dose wasrequired in patients receiving the smallest dose al-lowed, the study medication was discontinued. Thedose of study drug could also be reduced because ofa side effect or at the discretion of the investigator.In the case of a proarrhythmic event, the studymedication was discontinued. Compliance waschecked by tablet counting.

Participants in the study were seen on an outpa-tient basis after 1 and 3 months in the study andthereafter every 3 months. The study was concludedas planned when the last randomized patient hadbeen followed for 1 year. Survival of patients wascontinuously checked by computerized rolls in theDanish Central Person Register. No patient was lostto follow-up.

An independent data and safety monitoring com-mittee evaluated the results of the study with fourplanned interim analyses, and each time recom-mended continuation of the study.

End Points. The primary end point was all-causemortality. Secondary end points were death from acardiac cause, arrhythmic death, cardiac death orresuscitated cardiac arrest, arrhythmia requiringtreatment, and recurrent myocardial infarction. Inpatients with atrial fibrillation at the start of thestudy, a combined end point of death, stroke, andsystemic embolism was applied. The final sec-ondary end point was worsening of heart failure,which was defined as a hospital admission for con-gestive heart failure during which medical treat-ment was intensified. Conversion from atrialfibrillation to sinus rhythm and maintenance ofsinus rhythm after conversion were analyzed as apredefined substudy.

CHF MAY/JUNE 2001148 DOFETILIDE IN PATIENTS WITH CHF

Statistical Analysis. Time to events was analyzedwith a two-sided log-rank test, which was stratified ac-cording to treatment center and wall motion index.Kaplan-Meier estimators were used for survivalgraphs. The hazard ratio was calculated from Cox’sproportional hazard models, adjusted for center andwall motion index.

ResultsA total of 5548 patients with congestive heart fail-ure were screened, and 2531 (46%) had a wall mo-tion index of ≤1.2. Sixty percent (1518) of thesepatients had reduced left ventricular function andwere randomized. Characteristics of the two patientpopulations are shown in Table I. After 1 year inthe study, 421/397 patients in the dofetilide/placebogroups, corresponding to 76%/74% of those stillalive, were still on study medication. The medianfollow-up time was 18 months and no patients werelost to follow-up.

Survival. In the dofetilide group, 311 patients died(41%) and in the placebo group, 317 patients died(42%). Survival in the two groups did not differ sig-nificantly (Fig. 1A and B). When the survival analysiswas performed only on patients who continued to re-ceive the study medication, 89% were alive in bothtreatment groups after 1 year, and an analysis “ontreatment” for the whole study yielded a p value of0.54. Similarly, there were no differences in survivalwhen the analysis was restricted to patients who wereenrolled after the change in dosing according to cre-atinine clearance (p=0.49).

Table II shows causes of death in the two treat-ment groups. A large number of subgroups weredefined in advance, based on demographic charac-teristics, disease characteristics, and treatments; inall of these subgroups the mortality analysis was asneutral as in the overall analysis. In particular,there were no differences with respect to kidneyfunction or presence of atrial fibrillation at random-ization (Fig. 1A and B).

Other Clinical Data. The risk of hospitalization forworsening of congestive heart failure was significantlyreduced in the dofetilide group (p<0.001; hazard ratio,0.75; 95% confidence interval, 0.63–0.89). This findingwas independent of the presence of atrial fibrillation atrandomization, as shown in Fig. 1C and D. No othersecondary end points showed significant differences be-tween the two treatment groups, with hazard ratiosranging from 0.66–1.62.

Patients with atrial fibrillation at randomizationmore often converted to sinus rhythm in the

dofetilide group than in the placebo group. After 1month, 22 of 190 dofetilide patients with atrial fibril-lation at randomization (12%) converted, as com-pared to three of 190 patients in the placebo group(2%). Conversion to sinus rhythm continued to occurin both treatment groups. After 1 year, 61% of pa-

TABLE I. DEMOGRAPHIC CHARACTERISTICSOF THE PATIENT POPULATION AT BASELINE,BY TREATMENT GROUP

DOFETILIDE PLACEBOCHARACTERISTIC (N=762) (N=756)

Median duration of heart failure, months 12 12

Mean age in years (range) 70 (26–94) 70 (32–92)Males, number (%) 546 (72) 568 (75)Current smokers,

number (%) 254 (33) 268 (35)History, number (%)

Myocardial infarction 389 (51) 390 (52)Ischemic heart disease 509 (67) 508 (67)Diabetes 152 (20) 140 (19)Hypertension 111 (15) 115 (15)

Mean renal clearance, mm/min (SD) 57 (23) 57 (25)

Atrial fibrillation at randomization, number (%) 190 (25) 201 (27)

Median wall motion index (range) 0.9 (0.3–1.2) 0.9 (0.3–1.2)

Treatment at randomization, number (%)ß blocker 72 (9) 80 (11)ACE inhibitor 552 (72) 571 (76)Calcium channel blocker 153 (20) 170 (23)

NYHA functional class, number (%)I 16 (2) 17 (2)II 268 (35) 297 (39)III 423 (56) 385 (51)IV 49 (6) 52 (7)Not available 6 (<1) 5 (<1)

ACE=angiotensin-converting enzyme; NYHA=NewYork Heart Association

TABLE II. CAUSES OF DEATH, BY TREATMENTGROUP

DOFETILIDE PLACEBOCHARACTERISTIC (N=762) (N=756)

Cardiac deaths 255 (33) 251 (33)Total arrhythmic deaths 156 (21) 151 (20)Presumed arrhythmic 111 (15) 101 (13) Documented arrhythmic 28 (4) 28 (4)Unclassified* 17 (2) 22 (3)Noncardiac deaths 56 (7) 66 (9)Total deaths 311 (41) 317 (42)

*Unclassified deaths were analyzed as cardiac deathsthat were presumed arrhythmic.

CHF MAY/JUNE 2001DOFETILIDE IN PATIENTS WITH CHF 149

tients in the dofetilide group had converted (44%spontaneously and 17% electrically), compared to37% in the placebo group (17% spontaneously and20% electrically). These results are statistically signif-icant (p<0.001) after both 1 month and 1 year. Afterconversion from atrial fibrillation to sinus rhythm,the chance of maintaining sinus rhythm was signifi-cantly higher (p<0.001) in the dofetilide than in theplacebo group—after 1 year, 78% and 43%, respec-tively (Fig. 2). Among patients with sinus rhythm atrandomization, fewer developed atrial fibrillation ondofetilide than on placebo (11 of 556 compared to 35of 534 patients, respectively; p<0.001). The chanceof reaching the combined end point of death, stroke,or arterial embolism in atrial fibrillation was not dif-ferent in the two treatment groups: 52 of 190 pa-tients in the dofetilide group and 54 of 201 patientsin the placebo group (p=0.85).

Adverse Events. The total number of adverseevents and rate of treatment withdrawal were similarin the two treatment groups. Discontinuation of

study treatment due to QT prolongation was morecommon in the dofetilide group (14 vs. three patients). The maximal increase in the frequency-corrected QT interval was seen after 2 days of treat-ment.

The arrhythmia committee classified 25 cases ofdocumented arrhythmia as being torsade depointes ventricular tachycardia in the dofetilidegroup, compared to none in the placebo group. Ofthe 25 episodes, 15 required electrical conversionand two resulted in death. Nineteen of the 25episodes (76%), including the two deaths, tookplace during the 3 days of continuous EKG moni-toring at the start of treatment. Among the 146 pa-tients who received dofetilide before the protocolwas modified, seven had torsade de pointes (4.8%);there were 18 cases among the remaining 616 pa-tients (2.9%). Adjustment of the dofetilide dose ac-cording to creatinine clearance reduced the risk oftorsade de pointes. Occurrence of other ventriculartachyarrhythmias was similar in the two treatmentgroups.

Figure 1. Kaplan-Meier time-to-event curves for patients with and without atrial fibrillation at study entry. Solidlines=dofetilide; dotted lines=placebo. Graphs A and B show survival in patients with and without atrial fibrillation, re-spectively; graphs C and D show freedom from rehospitalization for congestive heart failure in patients with and withoutatrial fibrillation, respectively.

continued on page 155

CHF MAY/JUNE 2001150 DOFETILIDE IN PATIENTS WITH CHF

DiscussionOur study demonstrates that long-term treatmentwith dofetilide in a dose adjusted to the calculatedcreatinine clearance, the length of the QT interval,and the presence of atrial fibrillation does not in-crease the risk of death in high-risk patients withcongestive heart failure and reduced left ventricularfunction. The result was consistent in all subgroupsexamined, including patients with atrial fibrillationat randomization. Dofetilide is effective in re-estab-lishing and maintaining sinus rhythm in patientswith atrial fibrillation. This finding is consistent withthe findings in two other double-blind, placebo-con-trolled studies,7,8 in which treatment with dofetilideresulted in conversion of atrial fibrillation in 29%and 30% of patients and the persistence of sinusrhythm after 6 months in 62% and 71% of those whoconverted. Dofetilide can therefore be used for thetreatment of atrial fibrillation in patients with con-gestive heart failure. A prerequisite for safe use isthat the dose be carefully adjusted to the length ofthe QT interval and to the calculated creatinineclearance,11 and that treatment is started with 3 daysof continuous EKG monitoring. The efficacy ofdofetilide in the conversion of atrial fibrillation andmaintenance of sinus rhythm seems to be of thesame order of magnitude as that of amiodarone.13

Among the secondary end points in the study, therisk of hospitalization because of worsening of con-gestive heart failure was reduced in dofetilide-treatedpatients, regardless of whether sinus rhythm or atrialfibrillation was present at the beginning of the study.We have no explanation for this finding, but an

effect of dofetilide on atrial fibrillation may be a par-tial explanation.

Seventy-six percent of all registered episodes of tor-sade de pointes ventricular tachycardia took placewithin the first 3 days of treatment with dofetilide,which corresponds to the maximal prolongation of thefrequency-corrected QT interval, which was foundwithin 2 days. This finding demonstrates the impor-tance of starting treatment in the hospital, with contin-uous EKG monitoring. Dofetilide did not increase therisk of other secondary end points, including deathdue to documented or presumed arrhythmia. This isin contrast to the findings with some class 1A, class1C, and class III antiarrhythmic drugs.14,15 Thus far,antiarrhythmic drugs have not convincingly reducedmortality. Two studies have demonstrated reducedmortality with amiodarone treatment,3,16 but this find-ing was not confirmed in other, larger studies.4,17,18

Selection of patients for enrollment in controlledtrials can influence the rate of events observed. TheTrandolapril Cardiac Evaluation (TRACE) study19

used systematic screening of consecutive patients tomaximize the number included and to ensure that thepatients enrolled were as close as possible to beingrepresentative. In the current study, we used the samestrategy to recruit a substantial number of patients in alimited number of centers. The mortality rate in ourstudy was, in general, higher than that in other conges-tive heart failure studies, but is similar to the mortalityin one study of amiodarone.3 The mortality in the Sur-vival with Oral D-sotalol (SWORD) trial15 was lowerthan in our study, and the difference can be explainedby differences in design. The patients in SWORD wereless acutely ill, were not selected consecutively, andwere not hospitalized at the start of treatment.

The differences in class III antiarrhythmic drugsmay be important. Dofetilide blocks a single potassi-um channel, which mediates the late potassium cur-rent during the action potential, whereas amiodaroneblocks several potassium channels, calcium channels,and sodium channels—and is, furthermore, a ßblocker.5 D-sotalol is also less selective thandofetilide, as it influences three ion channels, inhibitsacetylcholine esterases, and has a residual ß blockingeffect.5 The clinical significance of these differences isnot known. The specific effect of dofetilide is translat-ed in good long-term tolerability, so that even pa-tients with structural heart disease and reducedfunction of the left ventricle tolerate the treatment.

Our conclusion is that dofetilide does not increasethe risk of death among patients with congestiveheart failure and reduced left ventricular systolicfunction, effectively converts atrial fibrillation to sinusrhythm, and effectively maintains sinus rhythm. The

Figure 2. Kaplan-Meier time-to-event curve for relapse ofatrial fibrillation or atrial flutter in patients who were inatrial fibrillation or atrial flutter at baseline and weresuccessfully converted to sinus rhythm, either pharmaco-logically or electrically. Solid lines=dofetilide; dottedlines=placebo; vertical bars=95% confidence intervals.

CHF MAY/JUNE 2001USING THE TRANSTHEORETICAL APPROACH 155

risk of hospitalization for worsening of heart failure isreduced. The dose of dofetilide needs to be carefullyadjusted according to the length of the QT intervaland the calculated creatinine clearance. Furthermore,treatment must be started in the hospital, with 72hours of continuous monitoring of cardiac rhythm.

Acknowledgment: This article is based on a study first reported in TheNew England Journal of Medicine. (Torp-Pederson C, Møller M,Bloch-Thomsen PE, et al. Dofetilde in patients with congestive heart fail-ure and left ventricular dysfunction. N Engl J Med. 1999;341:857–865). The Diamond-CHF study was supported by a grant fromPfizer Central Research.

REFERENCES1 Flaker GC, Blackshear JL, McBride R, et al. Antiarrhythmic

drug therapy and cardiac mortality in atrial fibrillation. TheStroke Prevention in Atrial Fibrillation Investigators. J AmColl Cardiol. 1992;20:527–532.

2 Coplen SE, Antman EM, Berlin JA, et al. Efficacy and safetyof quinidine therapy for maintenance of sinus rhythm aftercardioversion. A meta-analysis of randomized controlled tri-als. Circulation. 1990;82:1106–1116.

3 Doval HC, Nul DR, Perrone SV, et al., for Grupo de Estudiode la Sobrevida en la Insufficiencia en Argentina (GESICA).Randomized trial of low dose amiodarone in severe conges-tive heart failure. Lancet. 1994;344:493–498.

4 Singh SN, Fletcher RD, Fisher SG, et al. Amiodarone in patientswith congestive heart failure and asymptomatic ventricular ar-rhythmia. Survival Trial of Antiarrhythmic Therapy in CongestiveHeart Failure [see comments]. N Engl J Med. 1995;333:77–82.

5 Friedrich T, Nichols DJ, Alabaster CT, et al. Dofetilide. In:Messerli F, ed. Cardiovascular Drug Therapy. Philidelphia, PA:W.B. Saunders Co.; 1996:1296–1303.

6 Green MS, Dorian P, Roy D, et al. A randomized, double-blind,placebo-controlled comparison of intravenous dofetilide and pro-cainamide in the acute conversion of atrial fibrillation/flutter [ab-stract]. Circulation. 1996;8(suppl):453.

7 Singh SN, Zohle RG, Yellen L, et al. Efficacy and safety of oraldofetilide in converting to and maintaining sinus rhythm in pa-tients with chronic atrial fibrillation or atrial flutter: The sympto-matic atrial fibrillation investigative research on dofetilide(SAFIRE-D) study. Circulation. 2000;102:2385–2390.

8 Greenbaum RA, Campbell TJ, Channer KS, et al. Conversionof atrial fibrillation and maintenance of sinus rhythm bydofetilide. The EMERALD (European and Australian Multi-

center Evaluative Research on Atrial Fibrillation Dofetilide)study. Circulation. 1998;17(suppl):1633.

9 The Diamond Study Group. Dofetilide in patients with left ventricular dysfunction and either heart failure or acute myocardial infarction: Rationale, design and patientcharacteristics of the DIAMOND studies. Clin Cardiol.1997;20:704–710.

10 Køber L, Torp-Pedersen C, Carlsen J, et al., on behalf of theTRACE study group. An echocardiographic method for se-lecting high risk patients shortly after acute myocardial in-farction, for inclusion in multi-centre studies (as used in theTRACE study). Eur Heart J. 1994;15:1616–1620.

11 Cockcroft DW, Gault MH. Prediction of creatinine clearancefrom serum creatinine. Nephron. 1976;16(1):31–41.

12 Greene HL, Richardson DW, Barker AH, et al. Classifica-tion of deaths after myocardial infarction as arrhythmic ornonarrhythmic (The Cardiac Arrhythmia pilot study). Am JCardiol. 1989;63:1–6.

13 Deedwania PC, Singh BN, Ellenbogen K, et al. Spontaneousconversion and maintenance of sinus rhythm by amio-darone in patients with heart failure and atrial fibrillation:Observations from the Veterans Affairs Congestive HeartFailure Survival Trial of Antiarrhythmic Therapy (CHF-STAT). The Department of Veterans Affairs CHF-STAT In-vestigators. Circulation. 1998;98:2574–2579.

14 Echt DS, Liebson PR, Mitchell LB, et al. Mortality and morbidityin patients receiving encainide, flecainide or placebo. The CardiacArrhythmia Suppression Trial. N Engl J Med. 1991;324:781–788.

15 Waldo AL, Camm AJ, deRuyter H, et al. Effect of d-sotalol onmortality in patients with left ventricular dysfunction after re-cent and remote myocardial infarction. Lancet. 1996;348:7–12.

16 Garguichevich JJ, Ramos JL, Gambarte A, et al. Effect ofamiodarone therapy on mortality in patients with left ventric-ular dysfunction and asymptomatic complex ventricular ar-rhythmias: Argentine Pilot Study of Sudden Death andAmiodarone (EPAMSA). Am Heart J. 1995;130:494–500.

17 Julian DG, Camm AJ, Frangin G, et al. Randomised trial ofeffect of amiodarone on mortality in patients with left-ven-tricular dysfunction after recent myocardial infarction:EMIAT. European Myocardial Infarct Amiodarone Trial In-vestigators. Lancet. 1997;349:667–674.

18 Cairns JA, Connolly SJ, Roberts R, et al. Randomised trial ofoutcome after myocardial infarction in patients with frequentor repetitive ventricular premature depolarisations: CAMI-AT. Canadian Amiodarone Myocardial Infarction Arrhyth-mia Trial Investigators. Lancet. 1997;349:675–682.

19 Køber L, Torp-Pedersen C, Carlsen JE, et al. A clinical trialof the angiotensin converting enzyme inhibitor trandolaprilin patients with left ventricular dysfunction after myocardialinfarction. N Engl J Med. 1995;333:1670–1676.

Moller continued from page 150