Br Heartj1 1995;74:140-148

Effects of receptor antagonists in patients withclinical evidence of heart failure after myocardialinfarction: double blind comparison of metoprololand xamoterol

Hans Persson, Eva Rythe'n-Alder, Anders Melcher, Leif Erhardt

AbstractObjective-To evaluate whether xam-oterol, a partial agonist, would improveexercise time more than metoprolol inpatients with mild to moderate heart fail-ure after a myocardial infarction.Design-Single-centre double blind ran-domised parallel group comparison ofmetoprolol 50-100 mg and xamoterol100-200 mg twice daily.Patients-210 patients aged 40-80 years(173 men) with clinical evidence of heartfailure early after a myocardial infarc-tion. 106 were given metoprolol and 104xamoterol.Main outcome measures-Exercise testresults and performance at three months;the exercise test, quality of life, and clini-cal assessments at baseline (5-7 daysafter the infarction) and after 3, 6, and 12months.Results-Exercise time increased atthree months by 22% in the metoprololgroup and 29% in the xamoterol group,but with no significant difference betweenthe groups. Patients taking xamoterolshowed overall non-significantly highermean values of exercise time achievedwith higher heart rates at rest and exer-cise. Improvements in quality of life,clinical signs of heart failure, and NewYork Heart Association functional classwere seen in both treatment groups overone year, with minor benefits of xam-oterol on breathlessness, peripheraloedema, and functional class. Eighteenpatients taking metoprolol and 22 takingxamoterol withdrew from the study dur-ing one year, with a low mortality, rein-farction rate, and progress of heartfailure in both treatment groups. Meandose from baseline to 3 months was 135mg metoprolol and 347 mg xamoterol.Conclusion-fl, Receptor antagonists withor without partial agonist activity are safeto use in mild to moderate heart failureafter a myocardial infarction. Exercisetolerance, quality of life, and clinicalsigns and functional class of heart failureimprove, and few patients show deterio-ration in their condition. Exercise toler-ance is no better with xamoterol thanmetoprolol.

(Br Heart J7 1995;74:140-148)

Keywords: heart failure; myocardial infarction; meto-prolol; xamoterol

Heart failure after a myocardial infarction is acommon and serious complication,' con-tributing appreciably to future morbidity andmortality2 irrespective of whether it is evalu-ated by clinical methods or objective measuresof left ventricular function.3 Digitalis anddiuretics have long been the mainstay of treat-ment in heart failure. Large scale studies haveshown that angiotensin converting enzymeinhibitors improve prognosis and symptomsand exercise tolerance in patients with chronicheart failure,4 with better effects on exercisetolerance than is obtained with digitalis.4These inhibitors reduce mortality and mor-bidity in patients after myocardial infarctionin those with asymptomatic left ventriculardysfunction6 and those with clinical signs ofheart failure.7 ,B Receptor antagonists substan-tially reduce mortality and morbidity aftermyocardial infarction especially in high riskpatients with enlarged hearts or with clinicalheart failure.89 The effect seems in addition tothat of angiotensin converting enzymeinhibitors.7

It has long been proposed that ,B receptorantagonists can be functionally beneficial inpatients with heart failure if used with cau-tion.'0 Prospective, randomised trials specifi-cally of ,B antagonists in patients with heartfailure after infarction are lacking.Vasodilating properties or partial agonistactivity have been suggested to reduce someof the negative inotropic or chronotropiceffects of ,B receptor antagonists, althoughconcerns have been raised that the agonistactivity could abolish the beneficial effect onmortality especially after myocardial infarc-tion."1 Xamoterol, a /3, receptor antagonistwith high partial agonist activity improvessymptoms and exercise tolerance in patientswith mild to moderate chronic heart failure'2without deleterious effects on prognosis, 3

although doubts have been raised in severeheart failure.'4 The effect of xamoterol onexercise tolerance was superior to that of digi-talis.'2 Metoprolol, a /i3 receptor antagonistdevoid of agonist activity, has been widelyused after myocardial infarction and alsoimproves exercise tolerance in patients withdilated cardiomyopathy.'0 15

This prospective study was undertaken tocompare the effects of these two types ofantagonists on exercise tolerance, clinicalassessment, and quality of life during one yearin patients with mild to moderate heart failuresoon after myocardial infarction. At the timeof planning the study in 1987 it was hypothe-sised that xamoterol would improve exercise

Section of Cardiology,Division of InternalMedicine, KarolinskaInstitute, DanderydHospital, SwedenH PerssonDivision of ClinicalPhysiology,Karolinska Institute,Danderyd Hospital,SwedenE Rythe'n-AlderA MelcherDepartment ofCardiology,University of Lund,Malmo UniversityHospital, SwedenL ErhardtCorrespondence to:Dr H Persson, Section ofCardiology, Division ofInternal Medicine,Danderyd Hospital, S-18288 Danderyd, Sweden.Accepted for publication4 January 1995

140

,B receptor antagonists in patients with clinical evidence of heartfailure after myocardial infarction

tolerance more than metoprolol because of itspartial agonist activity. It was consideredunethical to include a placebo group becauseof the routine use of fi receptor antagonists assecondary prevention at our institution inpatients with stabilised heart failure. This is,to our knowledge, the first prospective ran-domised trial to assess the functional effects offi receptor antagonists in heart failure aftermyocardial infarction.

Patients and methodsDESIGNThe study was a double blind randomisedparallel group single centre comparison ofmetoprolol (50-100 mg twice daily) and xam-oterol (100-200 mg twice daily) over a year.One cardiologist (HP) screened, randomised,and followed up all patients throughout thestudy. Patients were randomly allocatedmetoprolol or xamoterol at discharge 5-7 daysafter the index infarction. Stratification wasbased on enlargement of the left ventricle atthe time of randomisation, with left ventricu-lar end diastolic diameter of > 28 mm/m2body surface area being considered to reflectan enlarged left ventricle.'6 The primary effi-cacy variable in the study was exercise toler-ance at three months. Secondary end pointswere exercise tolerance at six and 12 months,clinical assessment of heart failure, and qual-ity of life during follow up.

Results of a previous study with xamoterolindicated that a total of 200 patients withcompleted exercise tests would be needed togive a 90% chance of detecting a 25 (10%)seconds difference between the two groups atthree months at the 5% level of significance.'2The protocol was approved by the ethics com-mittee at Karolinska Hospital, and all patientsgave their informed consent to take part. Thestudy was conducted in accordance with theDeclaration of Helsinki.

PATIENTSPatients aged 40-80 with one or more speci-fied clinical or radiological signs of left ven-tricular heart failure (table 1) at any timeduring the 5-7 days in the coronary care unitafter an acute myocardial infarction wereincluded after their condition had stabilised.Patients with severe heart failure (New YorkHeart Association class IV) were excluded, aswere patients with pulmonary disease, aorticstenosis, hypertrophic obstructive cardiomy-

Table 1 Occurrence of signs of heartfailure in the twotreatment groups. Values are numbers ofpatients

Metoprolol Xamoterol(n = 106) (n = 104)

Bilateral pulmonary rales 97 97Third heart sound 51 58Congestion on x ray film 41 43Type of congestion:Upper lobe diversion 36 40Interstitial oedema 19 15Alveolar oedema 1 2Pleural effusion 9 8

Sinus tachycardia 51 47Respiratory rate > 27/minute 17 18

opathy, unstable angina, drug misuse, otherdisabling diseases, or inability to carry out atwo minute exercise test. Concurrent treat-ment was kept as stable as possible through-out the study, but diuretics (mainlyfrusemide), nitrates, angiotensin convertingenzyme inhibitors, and digitalis were allowed.Angiotensin converting enzyme inhibitorswere given only to patients whose heart failuredeteriorated and who showed no responseafter adjustment of diuretics. Calcium antago-nists were not allowed at baseline.

METHODSAll patients had a clinical assessment, chestradiography, echocardiography, and a symp-tom limited exercise test before randomisationand discharge. A questionnaire about symp-toms and activities in their daily lives wasfilled in by the cardiologist. Randomisedtreatment was then started. Metoprolol 50 mgor xamoterol 100 mg was given twice daily onthe first day. Heart rate and blood pressurewere checked while patients were supine andstanding before and two hours after eachdose. On the second day the dose was dou-bled if no adverse reactions occurred. Ifadverse symptoms or signs of intolerancedeveloped the patient was discharged taking areduced trial dose. The lowest given dosebefore withdrawal was 50 mg metoprolol or100 mg xamoterol once daily. A visit to checksafety of treatment was made after four weeksof randomised treatment. Full assessmentswith an exercise test, quality of life question-naire, and clinical assessment were made afterthree, six, and 12 months.

Effort tolerance was assessed by a symptomlimited exercise test on a bicycle ergometer(Siemens Elema, Solna, Sweden) starting at30 W, with a continuous increase of 10 Weach minute. Heart rate was measured everyminute and blood pressure and respiratoryrate every third minute and at peak exercise.ST depression 60 ms after the J point wasmeasured at the end of exercise as the meanvalue from 3-5 consecutive beats in the lead(V,-V6) showing the maximal ST deviation.A capillary blood sample for estimation of

blood lactate concentration was taken threeminutes after exercise. The samples wereanalysed with the enzymatic method based onoxidation of L-lactate to pyruvate with aLactate Analyzer 640 (Roche Bioelectrics,Basel, Switzerland). The patients' perceptionof chest pain, dyspnoea, and leg fatigue wasassessed every minute and at peak exercise byusing Borg scales (0-10).'7Symptoms during daily life were assessed

by Likert questionnaires, which includedquestions about breathlessness, tiredness,chest pain, palpitations, speed of walking anddaily tasks, difficulty with walking and dailytasks, confidence, sleeping, and mood. Thephysician rated the answers given by thepatient on a scale of four or five points.A clinical assessment was performed noting

presence of rales, a third heart sound, respira-tory rate at rest, peripheral oedema, jugularvenous filling, hepatojugular reflex, and

141

Persson, Rythe'n-Alder, Melcher, Erhardt

hepatomegaly. The patients were assigned aNew York Heart Association functional classat each visit.

Echocardiography was performed 3-5 daysafter the index infarction with M moderecording of left ventricular end diastolicdiameter from a short axis parastemal viewbelow the level of the mitral valve16; left ven-tricular end diastolic diameter was related tobody surface area. An Interspec XT,equipped with a 2-5 MHz transducer, wasused. Wall motion score according to Berninget al'8 was determined at baseline in allpatients with an acceptable cross sectionalecho window.

Chest radiography was usually performed

Table 2 Characteristics ofpatients at baseline. Values are numbers ofpatients unlessstated othervise

Men:womenMedian age (range) (years)Mean (SD) weight (kg)New York Heart Association class:

I

II

III

Medical history:Myocardial infarctionLeft ventricular heart failureHypertensionAngina pectorisDiabetes mellitus

Index infarction:TransmuralAnterior/lateralPeak infarct size (mean (SD) ,ukat/l)

Aspartate aminotransferaseCreatine phosphokinaseCreatine kinase BLactate dehydrogenase

At hospital:Ventricular fibrillationVentricular tachycardiaAtrioventricular block

IIIIII

Treatment before admission to hospital:DiureticsDigitalis,B BlockersLong acting nitratesAngiotensin converting enzyme (ACE) inhibitorsCalcium antagonists

Medication in CCU:FrusemideDigitalis,B BlockersLong-acting nitratesACE inhibitorsCalcium antagonistsPotassium sparersThrombolysis

Medication at discharge:FrusemideDigitalisLong-acting nitratesACE inhibitorsPotassium sparersOral anticoagulantsAspirin

Mean (SD) heart rate (beats/min):SupineStanding

Mean (SD) blood pressure (mm Hg):Supine:

SystolicDiastolic

Standing:SystolicDiastolic

Mean (SD) ventricularend diastolic diameter (mm)

Mean (SD) wall motion score*Mean (SD) radiographic heart size (ml)t:

TotalRelativet

*According to Berning, et al."821 tAccording to Jonse

Metoprolol(n = 106)

82:2466 (46-80)77-5 (12-0)

27826

1514333117

9154

5-00 (3-57)37-5 (33-7)1-57 (1-51)

22-2 (13-5)

547

5

3

182

17728

10510

10090130

8766

1057

8111861189

69-3 (10-5)78-5 (13-1)

Xamoterol(n = 104)

91:1367 (40-80)76-5 (11-1)

46535

2016284019

8056

4-74 (3-17)36-5 (33-9)1-57 (1-33)

26-3 (16-5)

944

324

191323191

13

10413

10091102

9258

1039

827

911776

70-8 (11-6)79-2 (14-6)

119-0 (17 2) 121-3 (16-0)75-5 (8-8) 76-2 (8-9)

115 5 (17 4) 115 3 (18-6)79 4 (10-8) 78-0 (10-7)

58-3 (8-4) 58-1 (7-5)1-2 (0-4) 1-2 (0-4)

1040 (271) 1032 (212)543 (122) 542 (90)

ell.'9 tTo body surface area.

before discharge on days 3-5. Heart size(absolute and relative to body surface area)was determined according to the method ofJonsell.'9 The degree of pulmonary congestionwas judged.

Routine haematology and serum bio-chemistry tests were performed and compli-ance was checked by counting tablets at eachreturn visit.

All laboratory analyses (exercise test,echocardiography, radiography) were doneblind by independent physicians and techni-cians.

STATISTICAL ANALYSESAnalysis of covariance was used for exercisetolerance data. Adjusted means wereobtained, which allowed for the effect of thebaseline exercise tolerance and also the effectof heart size (large or small). The estimatedjugular venous pressure and respiratory rate atrest were analysed by the same methods.Differences between adjusted means and 95%confidence intervals for the differences wereestimated. A Mantel-Haenszel x2 test wasused to assess if the treatments differed in fre-quency of pulmonary rales, third heart sound,peripheral oedema, hepatomegaly, and hepa-tojugular reflex at follow up. A check for treat-ment by heart size interaction was performedusing the Breslow-Day test. For quality of lifequestions 95% confidence intervals were esti-mated according to Gardner and Altman forthe median changes from baseline, a withintreatment comparison. (Statistics withconfidence: confidence intervals and statisticalguidelines. BMJ 1989;299:690.) A betweentreatment comparison was made using theWilcoxon rank sum test. A P value <0 05 wasconsidered significant.

Adjustments of the P values were per-formed with respect to measurements at mul-tiple time points by a non-Bonferroni method.Statistical testing was performed at months 3,6, and 12 and at the last known value. Thelast known value was obtained to overcomethe potential bias of patients withdrawing atdifferent times from the two treatmentgroups. If a patient withdrew from testsbetween visits at three and six months theassessment value at three months was enteredat six and 12 months as the last known value.All patients with assessments at any follow upvisits were entered in the analysis whether ornot they were still taking the allocated treat-ment (intention to treat analysis). Data arepresented as means (SD), or as medians whenappropriate.

ResultsCHARACTERISTICS AT BASELINETwo hundred and ten patients entered thestudy (table 2). Their median age was 67years and 173 were men. Class I heart failureaccording to the New York Heart Associationwas present in six patients, class II heart fail-ure in 143, and class III in 61. A third heartsound was noted in 109 and rales in 194.Transmural infarction was diagnosed in 171

142

,B receptor antagonists in patients with clinical evidence of heartfailure after myocardial infarction

Table 3 Reasons for exclusion from study, 1988-90

No ofpatients(n = 312)

Echocardiography not possible 27No heart failure 106Severe heart failure 6Unstable angina 18Valvar heart disease 5Pulmonary disease 16Intermittent claudication 12Neurological disease 16Rheumatoid disease 2Haematological or malignant disease 10Psychiatric disease 6Other diseases 28Known intolerance to ,B blockade 16Other catchment area 19Alcohol misuse 4Other research project 3Language difficulties/consent not given 9Not classified 9

and anterior or lateral infarction in 110.Previous infarction was present in 35 and a

history of left ventricular heart failure in 30.Thrombolytic treatment was given to 124.The two treatment groups were similar

with respect to age, weight, sex, class of heartfailure, medical history, infarct size and loca-tion, left ventricular end diastolic diameter,wall motion score, vital signs, and cardiovas-cular drugs taken during the stay in the coro-

nary care unit and at discharge. Beforeadmission to hospital 13 patients randomlyallocated xamoterol had received digoxincompared with two patients randomly allo-cated metoprolol. This difference was not pre-sent in the coronary care unit or at discharge.

Receptor antagonists were given to 200patients before randomisation. Most patientsreceived treatment with frusemide, amiloride,long acting nitrates, and aspirin on discharge.The mean frusemide dose was 71 mg (range20-200) in both treatment groups at dis-charge. Angiotensin converting enzymeinhibitors were given to 18 patients and digi-talis to 16 at discharge.

PATIENT ELIGIBILITYA complete log book was kept during the firsttwo years of inclusion20; 632 patients surviving

Table 4 Reasons for withdrawalfrom study

Metroprolol Xamoterol(n = 18) (n = 22)

Death 5 6Deterioration in condition 8 6Atrial fibrillation 1 1Ventricular tachycardia 0 3Possible adverse reaction 2 2Patient defaulted 0 1Study closed down 2 1Other 0 2

Table 5 Adverse reactions reported in >5% of allpatients

Metroprolol Xamoterol(n = 77) (n = 44)

Vertigo 19 9Cold extremities 15 8Flatulence 9 10Myocardial infarction 7 6Angina pectoris 8 4Impotence 6 4Disturbed sleep 7 2Diarrhoea 6 1

3-5 days after a myocardial infarction wereregistered. Two hundred and fifteen patientswere not entered because they were older than75. An amendment to the protocol was madeafter two years to include patients older than75 in the study. Of the remaining 417patients, 105 patients were entered in thestudy. One hundred and six patients wereexcluded because of lack of heart failure signsin the coronary care unit and six patients wereexcluded because of severe heart failure. Ofthe 312 excluded patients (table 3), 256received fi blockade at the time of screening.Sixteen patients over 75 were entered duringthe remaining two years of inclusion.

CONCURRENT DRUG TREATMENTSDuring follow up the median number of otherdrugs was 4 in both groups on all visits, excepton the 12 month visit, when three other drugswere taken. At three months the dose offrusemide was increased in 10 patients takingmetoprolol and in three taking xamoterol (P <0 05). Mean doses of frusemide were not dif-ferent between the two trial groups after three,six, and 12 months of follow up. At 12months frusemide had been withdrawn in 17patients taking metoprolol compared with 15patients taking xamoterol. Angiotensin con-verting enzyme inhibitors were given to sixpatients taking metoprolol and nine takingxamoterol during follow up.

TRIAL TREATMENTSMean daily dose from baseline to threemonths was 135 (55) mg for metoprolol and347 (92) mg for xamoterol. From three to sixmonths the mean dose was 123 (63) mg formetoprolol and 324 (119) mg for xamoterol.From six to 12 months the dose was 116 (66)mg and 294 (145) mg, respectively. Mediandose was 100 mg for metoprolol and 400 mgfor xamoterol on all visits. The trial dose wasreduced because of symptoms of heart failurein 14 patients taking metoprolol and in 10taking xamoterol (P = 0.41).

ECHOCARDIOGRAPHY AND LEFT VENTRICULARFUNCTIONThe left ventricle was enlarged (left ventricu-lar end diastolic diameter > 28 mm/m2 bodysurface area) in 143 patients. A mean wallmotion score of 1 2 (0 4) was found at base-line in 189 patients, corresponding to a meanejection fraction2' of 36%, which was similarin both treatment groups (table 2). The meanwall motion score was 1-0 (0 4) in patientswith enlarged left ventricles, corresponding toa mean baseline ejection fraction of 30%. Themean wall motion score was significantly betterin patients with normal left ventricular size(1-4 (0 3) v 1-0 (0 4), P < 0.001).

WITHDRAWALSForty patients were withdrawn from ran-domised treatment during the one year followup (table 4). Five patients receiving metoprololand six receiving xamoterol died; three of thedeaths in the metoprolol group occurredwithin eight days of entry. Disease deteriora-

143

Persson, Rythe'n-Alder, Melcher, Erhardt

Table 6 Exercise variables. Values are means (SD)

Metoprolol Xamoterol

Exercise time (s):Baseline 391-0 (161-8) 386-7 (138-7)3 Months 482-4 (207 9) 494 3 (189-9)12 Months 508-3 (209.3) 528-0 (198.6)

Peak workload (W):Baseline 94 9 (26 9) 94 3 (23-1)3 Months 110 1 (347) 112-1 (31-6)12 Months 114-3 (35 0) 117-7 (33 2)

Heart rate at rest supine (beats/min):Baseline 69-3 (10-5) 70-8 (11 6)3 Months 57-5 (8-0) 70 3 (9*7)12 Months 59-1 (9 1) 71-2 (93)

Systolic blood pressure at rest supine (mm Hg):Baseline 119-0 (17-2) 121-3 (16-0)3 Months 125-8 (17-0) 134-2 (17-7)12 Months 131-6 (17-1) 136-1 (17-4)

Heart rate at peak exercise (beats/minute):Baseline 113-6 (15-7) 115-9 (20 6)3 Months 110 1 (17-4) 119-4 (18-2)12 Months 113-0 (17-2) 124-4 (17-9)

Systolic blood pressure at peak exercise (mm Hg):Baseline 158-5 (28 5) 156-9 (28 8)3 Months 159-7 (29 4) 163 1 (24 3)12 Months 167-8 (27 8) 170-7 (21-4)

Peak respiratory rate (breaths/minute):Baseline 29-4 (5-7) 29-2 (5 3)3 Months 31-0 (5 6) 30-6 (5 6)12 Months 32-6 (5 8) 31-6 (5 8)

Maximum ST depression at peak exercise (mm):Baseline 0-67 (0-96) 0-85 (1 01)3 Months 0-61 (0 84) 0 70 (0 93)12 Months 0 59 (0 79) 0-64 (0 89)

Blood lactate (mmol/l):Baseline 4-8 (1 5) 4*9 (1-6)3Months 6-0 (1 9) 6-3 (2-1)12 Months 6-4 (2-4) 7-0 (2 0)

Peak Borg dyspnoea:Baseline3 Months12 Months

Peak Borg leg fatigue:Baseline3 Months12 Months

Peak Borg chest pain:Baseline3 Months12 Months

5-1 (2 0)6-0 (2 2)6-4 (2-1)5-6 (2 2)6-7 (2-1)7-4 (1-9)

0-4 (1 0)0 7 (1-7)0.5 (1 4)

5 0 (2 0)5-7 (2 0)5-6 (1 9)

5-5 (2 1)6-6 (2 0)6-8 (17)

0.7 (1-5)0-8 (1-6)0-6 (1-5)

tion was the cause for withdrawal in 14patients, eight taking metoprolol and sixxamoterol. Five patients taking metoprololand four xamoterol were withdrawn from thestudy because of worsening heart failure.Eight of these nine patients had enlarged leftventricular end diastolic diameter at baseline.Two patients taking metoprolol and threexamoterol were withdrawn because of un-stable angina. All five had a normal left ven-tricular end diastolic diameter at baseline.Symptoms of heart failure improved in fivepatients after withdrawal. Eight patients who

Figure 1 Exercisetolerance in seconds atbaseline and after three,six, and 12 months offollow up and at the lastknown value. Thefigureshows adjusted means and95% confidence intervalsfor the difference in secondsbetween the treatmentgroups.

600 r550

500

450

400

350

XamoteMetopri95% Cointerval

- n= 100 n=95 n=94

n= 101 n=98 n=92

* Xamoterol

n = 101.

0

n = 102

n= 104o Metoprolol n= 106

Baseline 3 6 12 Last knownvalue

-rol 389 501 515 532 524olol 395 481 494 502 488nfidence -17.1 to -16.8 to -12.5 to -4.2 to

57.5 59.2 72.6 76.2

were withdrawn showed improvement afterrevascularisation; two patients died four and12 months after withdrawal. Seven of the 14patients whose condition deteriorated weretreated with open label ,B blockade afterwithdrawal. Three patients taking xamoterolwere withdrawn within one month because ofventricular tachycardia.

EXERCISE TESTINGExercise tests were continued until eithersymptoms or criteria determining cardiacsafety prevented continuation. Exercise testswere performed at three months in 201patients, at six months in 193 patients, and at12 months, the final visit, in 186 patients. Thegroups were comparable in baseline exercisetolerance, ST depression, blood lactate con-centration after exercise, and ratings of symp-toms on the Borg scale (table 6). Exercisetime increased at three months comparedwith baseline values by 22% in the metoprololgroup and 29% in the xamoterol group. Therewas no significant difference between the twogroups at any of the follow up visits (figure 1).The mean increase in exercise time waslargest at 12 months (27% increase for meto-prolol and 35% increase for xamoterol). Thelargest difference between the two treatmentgroups was 36 s, or 6 W (9% of baseline exer-cise time and 6% of baseline exercise toler-ance), at the last known value. The resultswere consistent in patients with normal andenlarged left ventricles.

Borg scale assessments showed fatigue tobe the major limiting symptom at baseline andall follow up visits. Breathlessness and fatiguewere the cause for stopping exercise in >95%,chest pain in <17%, arrhythmias <5%, and afall in blood pressure in <10% of all visits.The Borg scale ratings and blood lactate con-centration increased on follow up (P <0 001), showing that some of the improve-ment in exercise tolerance could be explainedby a higher degree of physical exertion. Ananalysis of the workload obtained at the maxi-mal Borg scale rating of dyspnoea on entrycompared with the workload obtained at thesame Borg scale rating at 12 months showedan average 12% increase in exercise tolerancein both treatment groups (P = 0 002) on fol-low up.Mean heart rate at rest (P < 0 001) and at

peak exercise (P < 0 001), and mean systolic(P < 0 05) and diastolic (P < 0 05) bloodpressures at rest were significantly higher inpatients receiving xamoterol at all follow upvisits. There was a mean decrease of heart rateof 9 9-11 4 beats/minute at rest and 3-6-0-8beats/minute at peak exercise with metoprololfrom baseline to follow up visits and a meandifference of -0 5-0-7 beats/minute at restand a mean increase of 3 4-7 9 beats/minuteon exercise with xamoterol.

QUALITY OF LIFE ASSESSMENTImprovements for both trial groups were seenin tiredness, chest pain, speed of walking, dif-ficulty with daily tasks, speed of daily tasks(table 7). Breathlessness improved only with

Co

a).0E

a,

xw

144

fi receptor antagonists in patients with clinical evidence of heartfailure after myocardial infarction

Table 7 Quality of life assessments duringfollow up

Change from baseline P value for(95% confidence interval) * difference

between metoprololMetoprolol Xamoterol and xamoterol

Tiredness:0-3 -1 to-0-5 -1 to-0-5 0-920-6 -1 to-0-5 -1 to-0-5 0-280-12 -1 to -0-5 -1 to -0-5 0-90

Breathlessness:0-3 OtoO -0-5toO 0-0030-6 0 to 0 -0-5 to 0 0-0460-12 -0-5to0 -0.5toO 0-48

Chest pain:0-3 -1 to -0-5 -0-5 to -0-5 0-530-6 -1 to -0-5 -1 to -0-5 0-930-12 -1 to -0-5 -1 to -0-5 0-99

Palpitations:0-3 OtoO OtoO 0.990-6 OtoO OtoO 0-380-12 -0.5toO OtoO 0-44

Difficulty with walking:0-3 0toO -0.5to0 0-390-6 OtoO -0.5toO 0-490-12 OtoO -0-5toO 0-41

Speed of walking:0-3 -1 to -0-5 -0-5 to -0-5 0-920-6 -0.5 to-0-5 -1 to-0-5 1-000-12 -1 to -0.5 -1 to -0-5 0-99

Difficulty with daily tasks:0-3 -1 to-0.5 -1 to-0-5 1-000-6 -1 to -0-5 -1 to -0-5 0-960-12 -1 to -1 -1 to -0-5 0-99

Confidence:0-3 -0.5 toO -0-5 toO 0-990-6 -0.5toO -0-5to0 0-830-12 -0-5toO -0-5toO 1-00

Quality of sleep:0-3 OtoO -0-5toO 0-300-6 -0-5 toO -0-5 to0 0-960-12 -0-5toO -0.5toO 0-94

*Answers were given on a 4 or 5 point scale. Minus signs mean improvements compared withbaseline scores. 95% confidence intervals were calculated according to Gardner and Altman.

xamoterol on follow up (95% confidenceinterval -05 to -05 at the last known value)and the improvement was better than that inthe metoprolol group at three (P = 0 003) andsix months (P = 0 046) but not at 12 months(P = 0-48). There was no improvement inpalpitations, difficulty with walking, mood,

Table 8 Clinical signs of heart failure at baseline andfollow up. Values are numbers(percentages) ofpatients unless stated otherwise

Sign Metoprolol Xamoterol P value

Gallop rhythm:Baseline 33 (31) 43 (41)3Months 14 (14) 12 (12) 0-636 Months 19 (19) 21 (21) 0-5012Months 17 (18) 16 (17) 0-82

Rales:Baseline 34 (32) 46 (44)3 Months 21 (21) 18 (18) 0-726 Months - -12Months 18 (19) 11 (11) 0-15

Mean (SD) respiratory rate at rest (beats/minute):Baseline 16-8 (3-9) 16-5 (3-7)3 Months 15-0 (3-2) 15-1 (3-3) 0-926Months 15-5 (3-1) 15-5 (3-3) 1-0012Months 15-3(3-4) 15-3(3-5) 1-00

Mean (SD) jugular venous pressure (cm H20):Baseline 9-2 (1-6) 9-5 (1-5)3 Months 9-1 (1-3) 8-8 (1-0) 0-136 Months 9-2 (1-3) 8-8 (1-0) 0-0912 Months 9-0 (1-1) 8-8 (1-0) 0-90

Peripheral oedema:Baseline 7 (7) 6 (6)3Months 11 (11) 4 (4) 0-156 Months 17 (17) 4 (4) 0-0212 Months 13(14) 6 (6) 0 27

Hepatomegaly:Baseline 0 1 (1)3 Months 0 1 (1) 0-626 Months 1 (1) 2 (2) 0-9912 Months 1 (1) 1 (1) 0-85

Positive hepatojugular reflex:Baseline 30 (31) 34 (34)3Months 26 (27) 17 (17) 0-246Months 27 (28) 19 (20) 0-6712 Months 22 (24) 18 (20) 0-92

100 r

0

0-

0._

0

0~cQt

75 H

50 H

25 H

o Bsi 3Baseline 3

IV 1 1III 26 15 1111 78 53 521 2 34 33

Metoprolol

12 Baseline 3 12

35 10 765 46 364 46 53Xamoterol

Figure 2 New York Heart Association (NYHA)fiunctional class at baseline and after three and 12 monthsoffollow up. Numbers ofpatients in each class are givenbelow the histogram.

confidence, and quality of sleep. No other dif-ferences between the treatment groups werefound.

CLINICAL ASSESSMENTClinical signs improved on follow up, but nodifferences were seen between the two treat-ment groups in frequency of rales, third heartsound, respiratory rate at rest, hepatomegaly,or hepatojugular reflex (table 8). Fewerpatients taking xamoterol had peripheraloedema on follow up, which was significant atsix months (4% and 17%, respectively; P =0-02). Estimated jugular venous pressure wasslightly lower with xamoterol; at six monthsthe adjusted mean was 8-7 cm H2O and 9 1cm H2O, respectively (P = 0-09). Patients'weight increased by 1-8 kg with metoprololand declined by 0-8 kg with xamoterol. Themean New York Heart Association class ofheart failure (figure 2) improved from 2-2 atentry to 1-8 at 12 months with metoprolol andfrom 2 3 to 1-5 with xamoterol (P < 0 01, fordifference between groups at 12 months).Heart failure deteriorated by one functionalclass in seven patients in the metoprolol groupand four in the xamoterol group from baselineto 12 months; but heart failure did not deteri-orate by more than one class in any patient.

ADVERSE EXPERIENCES AND ADMISSIONS TOHOSPITALWithin 12 months of entry 72 patients takingxamoterol and 85 taking metoprolol (P <0 07) reported 166 and 186 adverse reactions,respectively. Vertigo, cold extremities, flatu-lence, angina pectoris, impotence, myocardialinfarction, disturbed sleep, and diarrhoeawere each reported by at least 5% of allpatients (table 5). More adverse reactions

145

Persson, Rythe'n-Alder, Melcher, Erhardt

except for flatulence were reported by patientstaking metoprolol. The reinfarction rate waslow in both treatment groups: 7% (sevenpatients) with metoprolol and 6% (sixpatients) with xamoterol. There was no differ-ence between the treatments in numbers ofvisits to hospital. The number of days spent inhospital was also similar in both groups (5.2(9 4) in the metoprolol group and 6-5 (12-0)in the xamoterol group) (P = 0 33).

DiscussionControlled studies on exercise tolerance with,B blockade have mainly been performed inpatients with chronic heart failure caused byidiopathic dilated cardiomyopathy. Duringthe first month of treatment reduced exercisetolerance22 or no improvement was seen.23During long term treatment improvement wasobserved,'0 15 indicating that a short term dete-rioration may be expected before improve-ment is obtained. With xamoterolimprovement has been reported after 1-3months.'2 24 25 No improvement in exercise tol-erance was found with xamoterol in patientswith severe heart failure.'4 /3 Receptor antago-nists with vasodilatory effects such aslabetalol,26 bucindolol,27 and carvedilol28 alsoimprove exercise tolerance, but there are nocontrolled comparisons between differenttypes of /3 blockade and no trials have beenreported in patients with heart failure aftermyocardial infarction.

Exercise tolerance improved both withmetoprolol and xamoterol in our study, withno significant differences between the treat-ment groups. The improvement with xam-oterol over one year was similar to that seen inthe German austrian xamoterol study,'2 a sim-ilar workload at a similar peak exercise heartrate being achieved. The 27% improvementover one year with metoprolol is greater thanthe 16% improvement seen with digitalis inthe German study but 8-9% less than theeffect of xamoterol. The improvement frombaseline must be interpreted with caution aswe did not use a placebo group. In a previousstudy with exercise testing after mainlyuncomplicated myocardial infarction a similarimprovement was found with both metoprololand placebo, indicating the natural course ofexercise tolerance during the year after infarc-tion.29 Another reason for caution when inter-preting the data is that the higher exercisetime during follow up is achieved at a higherdegree of physical exertion according to theBorg scale data, measurements of lactate con-centration, and respiratory rate. The addi-tional analysis of Borg scale data supports atrue 12% improvement in both treatmentgroups at 12 months, as previously noted.The improvement in exercise tolerance wassimilar in the subgroups of patients with nor-mal and enlarged left ventricles, which meansthat exercise time improves also in patientswith a more obvious depression of left ventric-ular systolic function. We found significantdifferences in heart rate and blood pressurebetween the trial drugs at rest and on exercise,

which shows the partial agonist effect of xam-oterol. One reason for the lack of additionaleffect on exercise tolerance with xamoterolcould be a higher oxygen demand shown byThierfelder et al with xamoterol'0 but disputedby Rousseau et al.31One clinical benefit of xamoterol is

reflected in fewer patients with peripheraloedema, and this might be related to the dif-ference in weight between treatment groupsduring follow up. Increased doses offrusemide were also more frequent in themetoprolol group at three months, althoughmean doses were similar in both treatmentgroups during follow up. Metoprolol causespatients with heart disease to gain weight.32It may affect lipolysis.33 No other significantdifferences on clinical assessments could bedetected, indicating similar effects of thedrugs.

Quality of life questionnaires indicate animprovement with time in both treatmentgroups. This finding must, however, be inter-preted with caution because of the lack of acontrol group. Only breathlessness was signif-icantly better in patients taking xamoterolduring the first six months, indicating someclinical benefit and in agreement with otherstudies. 12 Fluid retention in patients takingmetoprolol might be one reason why breath-lessness did not improve with metoprolol. Wepreviously found that ventilation increases inpatients with heart failure treated with /3'receptor antagonists devoid of agonist activitywithout reducing exercise tolerance, peakoxygen uptake, and anaerobic threshold.34The beneficial effects of xamoterol on oedemaand breathlessness may explain the differencein New York Heart Association functionalclass at 12 months.

Safety data show a low mortality and mor-bidity for xamoterol and metoprolol in com-parison with mortality in patients with acomparable wall motion score3 or similar clin-ical inclusion criteria.7 The low mortality inour study suggests a superior effect of3 recep-tor antagonists in secondary prevention aftermyocardial infarction in high risk patients.35Our figures are in agreement with one studywith a partial agonist, acebutolol, in high riskpatients36 but in conflict with other studiesshowing that prognosis is improved more with/3 receptor antagonists without partial agonistactivity." However, our study was notdesigned to look at effects on mortality andthe data must therefore be interpreted withcaution. The reduction in mortality aftermyocardial infarction with metoprolol is seenwith a dose of 200 mg37 and the lower dosesused in heart failure may be less effective forpurposes of affecting mortality. TheMetoprolol in Dilated Cardiomyopathy study,however, showed that there was some benefiton the combined end point of mortality andmorbidity in patients with dilated cardiomy-opathy with metoprolol at a dose comparablewith our mean dose.'0 The safety of xamoterolshown in our study in mild to moderate heartfailure is in contrast to the findings in thexamoterol in severe heart failure study, in

146

/1 receptor antagonists in patients with clinical evidence ofheartfailure after myocardial infarction

which mortality was increased. 14We have thus shown that /3 receptor antago-

nists with and without partial agonist activityare well tolerated by patients with mild tomoderate heart failure after a myocardialinfarction, as has previously been shown forpatients with uncomplicated infarctions.'9The frequency of adverse reactions and thewithdrawal rate were no higher in our studythan in other heart failure studies withangiotensin converting enzyme inhibitors,5-7and only 5% of our patients showed a deterio-ration. The fear of deterioration in heart fail-ure is in most cases not warranted, althoughwe used a fast titration protocol and standarddoses of metoprolol and xamoterol. A similarincidence of heart failure was found inpatients allocated to ,B blockade compared toplacebo in a pooling off all long term sec-ondary prevention studies with /3 blockadeafter myocardial infarction.38 A beneficialtreatment effect of /3 blockade in selectedpatients with clinical evidence of heart failure ispossible because there may be a different out-come in these patients compared with patientswith uncomplicated myocardial infarction, asshown by the differentiated haemodynamicresponse in the Metoprolol in AcuteMyocardial Infarction trial.39 The BetablockerHeart Attack Trial showed an increase in inci-dence of heart failure confined to patientswith normal ejection fraction and not inpatients with reduced ejection fraction.40

Nearly all our patients were treated withopen /3 blockade as part of our routine treat-ment of myocardial infarction before inclu-sion in the study. We therefore knew thatmost patients tolerated at least low doses of,Bblockade, a strategy used in studies withangiotensin converting enzyme inhibitors.6Most patients (82%) with exclusion criteriaaccording to the study protocol obviously tol-erated /3 blockade. Only 16 patients wereexcluded because they could not tolerate suchblockade.The mechanism behind the improvement

after treatment with ,B blockade in patientswith heart failure is unclear. The lack of addi-tional effect with xamoterol on exercise toler-ance in spite of its beneficial haemodynamicproperties indicates that the /3l receptor block-ing effect is the most important. ,B Receptorantagonists with and without selectivity, par-tial agonist activity, and vasolidatory effectsimprove exercise tolerance.'5 24-29 Suggestedmechanisms include protection against ,Breceptor downregulation,41 anti-ischemiceffects,42 restoration of the chronotropic-inotropic relation,43 protection against cate-cholamine toxicity,44 and inhibition ofneurohormonal stimulation of the renin-angiotensin system." Our study does notallow us to draw any conclusion about thesebeneficial mechanisms, but they are probablyall more obviously affected by a /3 receptorantagonist devoid of partial agonist activity./3 Receptor antagonists with vasodilatoryproperties may prove to be superior to meto-prolol or xamoterol. Further comparativestudies are needed to answer this question.

We conclude that both metoprolol andxamoterol can be used safely in patients withmild to moderate heart failure after myocar-dial infarction, with improvement of exercisetolerance, quality of life, and clinical assess-ment in the two treatment groups and onlyminor differences in functional outcome dur-ing one year of follow up. The use of xam-oterol in patients with acute myocardialinfarction may be disputable until a propermortality study is performed because of theadverse data on mortality in severe chronicheart failure.'4

We thank Zeneca Pharmaceuticals, Alderley Park, for financialhelp, and Kjell Pennert at Zeneca AB, Gothenburg, for statisti-cal support. The study was also supported by grants from theSwedish National Heart and Lung Foundation, the SwedishSociety of Medicine, and the Karolinska Institute Foundation.We also thank E Johansson, M-L O'Konor, and E Linder-Klingsell for invaluable help; SV Eriksson for cross sectionalechocardiographic evaluations; and A Boavida, R Ude'n, andT Osby for chest x ray evaluations.

1 Wolk MJ, Scheidt S, Killip T. Heart failure complicatingacute myocardial infarction. Circulation 1972;45:1125-38.

2 Dwyer EM, Greenberg HM, Steinberg G, MulticenterPostinfarction Research Group. Clinical characteristicsand natural history of pulmonary congestion duringacute myocardial infarction. Am J Cardiol 1989;63:1423-8.

3 Beining J, Steensgaard-Hansen F, Appleyard M. Relativeprognostic value of clinical heart failure and earlyechocardiographic parameters in acute myocardialinfarction. Cardiology 1991;79:64-72.

4 Captopril-Digoxin Multicenter Research Group.Comparative effects of therapy with captopril anddigoxin in patients with mild to moderate heart failure.JAMA 1988;259:539-44.

5 SOLVD Investigators. Effect of enalapril on survival inpatients with reduced left ventricular ejection fractionsand congestive heart failure. N Engl J7 Med 1991;325:293-302.

6 Pfeffer MA, Braunwald E, Moye' LA, Rutherford J, BostaL, Brown EJ, et al. Effects of captopril on mortality andmorbidity in patients with left ventricular dysfunctionafter myocardial infarction. Results of the Survival andVentricular Enlargement Trial. N Engl J7 Med 1992;327:667-77.

7 Acute Infarction Ramipril Efficacy (AIRE) StudyInvestigators. Effect of ramipril on mortality and mor-bidity of survivors of acute myocardial infarction withclinical evidence of heart failure. Lancet 1993;342:821-8.

8 Olsson G, Rehnqvist N. Effect of metoprolol in postinfarc-tion patients with increased heart size. Eur Heart _J1986;7:468-74.

9 Chadda K, Goldstein S, Byington R, Curb JD. Effect ofpropranolol after acute myocardial infarction in patientswith congestive heart failure. Circulation 1986;73:503-10.

10 Waagstein F, Bristow MR, Swedberg K, Camerini F,Fowler MB, Silver MA, et al. Beneficial effects of meto-prolol in idiopathic dilated cardiomyopathy. Lancet1993;342: 1441-6.

11 Kjekshus JK. Importance of heart rate in determiningbeta-blocker efficacy in acute and long-term acutemyocardial infarction trials. Am J7 Cardiol 1986;57:43-7F.

12 German and Austrian Xamoterol Study Group. Double-blind placebo controlled comparison of digoxin andxamoterol in chronic heart failure. Lancet 1988;i:489-93.

13 Marlow HF, Lewis JA. Effects of xamoterol on mortality inmild to moderate heart failure [abstract]. Eur Heart J71990;11:132.

14 Xamoterol In Severe Heart Failure Study Group.Xamoterol in severe heart failure. Lancet 1990;336: 1-6.

15 Engelmeier RS, O'Conell, JB, Walsh R, Rad N, ScanlonPF, Gunnar RM. Improvement in symptoms and exer-cise tolerance by metoprolol in patients with dilatedcardiomyopathy: a double-blind, randomized, placebo-controlled trial. Circulation 1985;72:536-46.

16 Gardin JM, Henry WI, Savage DD, Ware JH, Burn C,Borer JS. Echocardiographic measurements in normalsubjects: evaluation of an adult population without clini-cally apparent heart disease. J7 Clin Ultrasound 1979;7:439-47.

17 Borg GAV. Psychophysical bases of perceived exertion.Med Sci Sports Exerc 1982;14:377-81.

18 Berning J, Steensgaard-Hansen F. Early estimation of riskby echocardiographic determination of wall motionindex in an unselected population with acute myocardialinfarction. Am J Cardiol 1990;65:567-76.

19 Jonsell S. A method for determination of the heart size byteleroentgenographs (a heart volume index). Acta Radiol

147

Persson, Rythe'n-Alder, Mekher, Erhardt

1939;20:325-40.20 Persson H, Linder-Klingsell E, Eriksson SV, Erhardt L.

Heart failure after myocardial infarction: the importanceof diastolic dysfunction. A clinical and echocardio-graphic study. Eur HeartJ 1995;16:496-505.

21 Beming J, Rokkedal Nielsen J, Launbjerg J, Fogh J,Mickley H, Andersen PE. Rapid estimation of left ven-tricular ejection fraction in acute myocardial infarctionby echocardiographic wall motion analysis. Cardilogqy1992;80:257-66.

22 Ikram H, Fitzpatrick D. Double-blind trial of chronic oralbeta blockade in congestive cardiomyopathy. Lancet1981;ii:490-3.

23 Currie PF, Kelly MJ, McKenzie A, Harper RW, rim YL,Federman J, et al. Oral beta-adrenergic blockade withmetoprolol in chronic severe dilated cardiomyopathy. JfAm Coil Cardiol 1984;3:203-9.

24 Vigholt-Sorensen E, Faergeman 0, Snow HM. Effects ofxamoterol, a fi, adrenoceptor partial agonist, in patientswith ischaemic dysfunction of the left ventricle. Br HeartJf 1989;62:335-41.

25 Waller DG, Webster J, Sykes CA, Bhalla KK, Wray R.Clinical efficacy of xamoterol, a beta,-adrenoceptor par-tial agonist, in mild to moderate heart failure. Eur HeartJ1989;10: 1003-10.

26 Leung WH, Lau C-P, Wong C-K, Cheng CH, Tai YT,Lim SP. Improvement in exercise performance andhemodynamics by labetalol in patients with idiopathicdilated cardiomyopathy. Am HeartJ 1990;119:884-90.

27 Pollock SG, Lystash JC, Tedesco C, Craddock G,Smucker ML. Usefulness of bucindolol in congestiveheart failure. Am J7 Cardiol 1990;66:603-7.

28 Krum H, Schwartz B, Sackner-Bernstein J, Penn J,Goldsmith RL, Kukin ML, et al. Double-blind, placebo-controlled study of the long-term efficacy of carvedilol inpatients with severe heart failure treated with converting-enzyme inhibitors. JAm Coll Cardiol 1993;21:114A.

29 Olsson G, Rehnqvist N, Lundman T, Melcher A.Metoprolol treatment after acute myocardial infarction.Effects on ventricular arrhythmias and exercise tests dur-ing 6 months. Acta Med Scand 1981;210:59-65.

30 Thierfelder L, Holubarsch C, Hasenfuss G, Heiss HW,Just H. The effect of the partial beta-l-adrenoceptoragonist xamoterol on hemodynamics and myocardialenergetics in patients with dilated cardiomyopathy.JI Cardiovasc Pharmacol 1991;17:593-9.

31 Rousseau MF, Pouleur H, Vincent MF. Effects of cardio-selective fl,-partial agonist ('Corwin') on left ventricularfunction and metabolism in patients with previousmyocardial infarction. Am Jf Cardiol 1 983;51:1267-74.

32 Rossner S, Taylor C, Byington R, Furberg C. Weight gainafter therapy with beta-blockade. Swedish Society of

Medicine. Hygiea 1989;98:189.33 Meiler SE, Leier CV. Effects of beta-adrenergic stimula-

tion on skeletal-muscle conditioning. Heart Failure 1986;128-37.

34 Persson H, Erhardt L, Melcher A, Nygren A. Aerobicresponses in exercise testing without and with betablockingtherapy in patients with chronic heart failure [abstract].Symposium on exercise testing in congestive heart failure,European Society of Cardiology, Brussels, 1990.

35 Beta-Blocker Pooling Project Research Group. The beta-blocker pooling project (BBPP): subgroup findings fromrandomized trials in post infarction patients. Eur Heart J1988;9:8-16.

36 Boissel J-P, Leizorovicz A, Picolet H, Peyrieux J-C, for theAPSI Investigators. Secondary prevention after high-riskacute myocardial infarction with low-dose acebutolol.Am ICardiol 1990;66:251-60.

37 Hjalmarsson A, Elmfeldt D, Herlitz J, Holmberg S, MalekI, Nyberg G, et al. Effect on mortality of metoprolol inacute myocardial infarction: a double-blind randomisedtrial. Lancet 1981;ii:823-6.

38 Yusuf S, Peto R, Lewis J, Collins R, Sleight P. Beta-blockade during and after myocardial infarction: anoverview of the randomized trials. Prog Cardiovasc Dis1985;26:335-71.

39 Held P, Corbeij HMA, Dunselman P, Hjalmarsson A,Murray D, Swedberg K, for the Haemodynamic StudyGroup of the MIAMI trial. Hemodynamic effects ofmetoprolol in acute myocardial infarction. A random-ized, placebo-controlled multicenter study. Am J Cardiol1985;56:47-54G.

40 Jafri SM, Khaja F, McFarland T, Capone R, Dahdah S,Haywood J, et al. Efficacy of propranolol therapy afteracute myocardial infarction related to coronary arterialanatomy and left ventricular function. Am J Cardiol1987;60:976-80.

41 Fowler MB, Bristow MR, Laser JA, Ginsburg R, Scott LB,Schroeder JS. Beta blocker therapy in severe heartfailure: improvement related to beta,-adrenergicreceptor up regulation? [Abstract.] Circulation 1984;70:II-112.

42 Bostrom PA, Balldin M, Lilja B, Johansson B. The effectof atenolol on the left ventricular performance inpatients with angina pectoris measured with isotopetechnique. Acta Med Scand 1988;223:239-45.

43 Mulieri LA, Hasenfuss G, Leavitt B, Allen PD, Alpert NR.Altered myocardial force-frequency relation in humanheart failure. Circulation 1992;85:1743-50.

44 Shanes JG, Subramanian R, Thompson EB, O'Neill HK,Kondos GT, Shanks P, et al. Chronic beta blockade inthe myopathic Syrian hamster [abstract]. Clin Res 1985;33:818.

148