Br Heart J 1986; 55: 391-9

Afterload reduction: a comparison of captopril andnifedipine in dilated cardiomyopathy

PIER G AGOSTONI, NICOLETTA DE CESARE, ELISABETTA DORIA,ALVISE POLESE, GLORIA TAMBORINI, MAURIZIO D GUAZZI

From Istituto di Cardiologia, Centro di Studio Ricerche Cardiovascolari del Consiglio Nazionale delleRicerche, University of Milan, and Fondazione "I. Monzino" and Istituto Ricerche Cardiovascolari"G. Sisini", Milan, Italy

SUMMARY Nifedipine and captopril are potent vasodilators and may be expected to help leftventricular failure by reducing afterload. Nifedipine (20mg three times a day) and captopril(50mg three times a day) were added to an optimal regimen of digitalis and diuretics in a doubleblind crossover trial in 18 cases of dilated cardiomyopathy. New York Heart Association func-tional class rating symptoms and exercise tolerance times improved on captopril but not on

nifedipine. The reduction in pulmonary capillary wedge pressure and the increase of cardiacoutput on captopril indicated that the augmented functional capacity may have resulted in part

from an improved performance of the left ventricle. Although there were comparable decreasesin systemic vascular resistance and presumably in impedence to ejection by the left ventricle onboth drugs, the dimensions of the ventricular cavity were found to be reduced by captopril andaugmented by nifedipine, and only captopril reduced the afterload (wall stress). In addition, theforce-length relation (between left ventricular end systolic stress and end systolic diameter) was

shifted to the left of baseline by captopril and to the right by nifedipine, suggesting that musclecontractility was reduced by nifedipine and not by captopril.These results suggest that nifedipine and captopril have different effects on afterload and

contractility and these may account for the different effects of these drugs on the performance ofthe heart and clinical responses.

Most vasodilators may be beneficial in both hyper-tension and cardiac decompensation. The angio-tensin converting enzyme inhibitor, captopril, issuch an example.'`3 The antihypertensive proper-ties of the calcium channel blocking agents are wellestablished,4 and nifedipine has been used to reduceventricular afterload.5-8Treatment of cardiac insufficiency in the advanced

stage of dilated cardiomyopathy is still one of themost challenging tasks in clinical cardiology, and thepoor prognosis justifies a search for more efficientmethods of treatment. Although the combination ofcaptopril with conventional treatment has been ofbenefit in many series of patients with refractorychronic congestive heart failure,9 -12 these series did

Requests for reprints to Professor Maurizio D Guazzi, Istituto diCardiologia, Via Bonfadini 214, 20138 Milan, Italy.

Accepted for publication 3 December 1985

not include many cases of dilated cardiomyopathyand the specific response of such patients mayhave been obscured. Acute pulmonary oedema dueto dilated cardiomyopathy can be relieved bynifedipine"6 but long term studies are lacking.For these reasons we have studied the effects of

these two agents in dilated cardiomyopathy. We per-formed a detailed haemodynamic study of thedifferent effects of captopril and nifedipine on bloodvessels (in terms of arterial and venous muscle relax-ation) and on the myocardium (in terms of changesin heart rate and contractility) in a crossover trial.

Patients and methods

PATIENTSTwenty six patients with chronic congestive heartfailure caused by dilated cardiomyopathy of un-known cause were considered to be eligible, but

391

on July 13, 2022 by guest. Protected by copyright.

http://heart.bmj.com

/B

r Heart J: first published as 10.1136/hrt.55.4.391 on 1 A

pril 1986. Dow

nloaded from

Agostoni, De Cesare, Doria, Polese, Tamborini, Guazzi

Table 1 Clinical data on 18 patients with dilated cardiomyopathy

Case Age (yr) Sex BSA (m2) ECG Previous drug treatment

1 57 F 1-26 SR Digitalis, frusemide, spironolactone2 54 M 1-93 AF + LBBB Digitalis, ethacrynic acid3 46 M 1-84 SR Digitalis, nitrates, hydrochlorothiazide4 61 M 1-80 AF Digitalis, frusemide5 45 M 1-75 LBBB Digitalis, nitrates, frusemide6 58 M 2-00 SR+ Digitalis, ethacrynic acid

1 degree AVB7 68 M 1-57 SR Digitalis, frusemide8 58 M 2-12 SR Digitalis, frusemide9 55 M 1-83 AF Digitalis, nitrates, frusemide10 49 M 1-99 AF + LBBB Digitalis, frusemide, spironolactone11 58 M 1-82 SR Digitalis, nitrates, hydrochlorothiazide12 34 M 176 SR Digitalis, hydrochlorothiazide13 62 M 188 SR Digitalis, frusemide14 38 F 1-45 SR Digitalis, frusemide15 46 M 1-85 SR Digitalis, ethacrynic acid16 57 M 1-80 SR + Digitalis, frusemide, spironolactone

1 degree AVB17 52 F 1-59 LBBB Digitalis, hydrochlorothiazide18 49 M 1-82 SR Digitalis, frusemide, spironolactone

BSA, body surface area; ECG, electrocardiogram; SR, sinus rhythm; AF, atrial fibrillation; LBBB, left bundle branch block; AVB,atrioventricular block.

three were lost to follow up, one died, and four werewithdrawn because at excessive hypotension ontreatment. The remaining 18 patients (Table 1) com-pleted the study. All of them had evident cardiacdysfunction causing dyspnoea at rest and were easilyfatigued; six of them also had orthopnoea despitetreatment with digitalis and diuretics. Vasodilators(nitrates) were used in four cases and treatment wasstopped seven days before the beginning of thestudy. Eleven patients had had cardiac symptoms for> two years and the other seven had had dyspnoeafor at least 10 months. Eleven subjects were in func-tional class IV (New York Heart Association) andseven were in class III. We excluded those patientswho had supine systolic blood pressure<100mmHg, angina pectoris, a history of or elec-trocardiographic signs of myocardial infarction, andhepatic or renal impairment (serum bilirubin or glu-tamic oxalacetic transaminase or both more thantwice the upper limit of normal, and serum cre-atinine concentration > 2 mg/dl). Patients weredeemed to be eligible for the study if they were capa-ble of exercising for at least three minutes and nomore than 12 minutes on a treadmill following theprotocol used by the Captopril Multicenter ResearchGroup."The study protocol was approved by our ethical

research committee, and each patient gave informedconsent.

STUDY DESIGNPatients were admitted to the hospital, and duringthe first seven days physical examination, 12 leadelectrocardiogram, and blood and urine laboratorytests were performed; the dosages of digitalis and

diuretics were also adjusted until we judged that anoptimal regimen had been achieved. This regimenwas kept constant during the trial. At the end of thisperiod a standard 6 ft posteroanterior chest x ray wastaken to measure the cardiothoracic ratio, and echo-cardiographic and blood pressure (with random zerosphygmomanometer) measurements were obtainedwith the patient supine. An exercise test was per-formed to determine maximum physical capacity.After this, patients were randomly allocated to atreatment group and right heart catheterisation wascarried out. Baseline records were obtained 30 minafter catheter positioning when patients felt com-fortable and heart rate and blood pressure had sta-bilised. Then nifedipine (20 mg) or captopril (50 mg)was administered and the haemodynamic variableswere recorded again every 30 min for the next120 min. Baseline values were compared with thosedetected when the fall in blood pressure after thedrug was greatest.

Patients continued to take the test drug three timesa day as outpatients. Each day patients noted theirsymptoms in a diary and classified dyspnoea on ascale of 0 to 4 (0 when none occurred, 1 + with morethan ordinary activity, 2+ with ordinary physicalactivity, 3 + dyspnoea at rest, and 4+ orthopnoea).We used the patient's dyspnoea grades to give a serialclinical function classification (from class I to IV).For two months physical examination, exercise

tests, and echocardiographic examination were per-formed each week. After this, patients were read-mitted to hospital and non-invasive and invasiveevaluations were repeated 90-120 min after the finaldose of the agent. Then patients were switched to theother drug, and continued on it as outpatients. We

392

on July 13, 2022 by guest. Protected by copyright.

http://heart.bmj.com

/B

r Heart J: first published as 10.1136/hrt.55.4.391 on 1 A

pril 1986. Dow

nloaded from

Afterload reduction: a comparison of captopril and nifedipine in dilated cardiomyopathy

considered a washout period to be unnecessary inview of the long duration of treatment. The studycontinued as described earlier. No complicationswere noticed with repeated haemodynamic mon-itoring. The cardiothoracic ratio was measured onchest x rays obtained at the end of both treatmentperiods.Both nifedipine (20mg) and captopril (50mg)

were given in identical capsules in a double blindrandomised fashion in both the short term and longterm studies. Capsules containing an inactive prepa-ration were given during the seven day run in period.

SAFETYFor the first three days of each drug treatment pa-tients remained in hospital and blood pressure andsymptoms related to hypotension were monitored.Hypotension with faintness on standing was experi-enced by three subjects while on nifedipine (one onnifedipine as second drug) and two while on cap-topril (one on captopril as second drug). In one ofthese, however, symptomatic hypotension was tran-sient and he was able to continue on captopril; theother four patients could not and they were with-drawn from the trial. No other patients had excessivehypotension or related symptoms during treatment.

HAEMODYNAMIC ASSESSMENTCirculatory measurements were obtained in fastingpatients who were supine and had not received pre-medication. For right atrial, pulmonary arterial, andpulmonary capillary wedge pressure measurementsand cardiac output determinations, a number 7 triplelumen flow directed thermodilution catheter was in-serted percutaneously under local anaesthesia into anantecubital vein and floated under fluoroscopic guid-ance to the pulmonary artery and, when necessary, tothe wedge position. Intra-arterial blood pressure wasmeasured through a Teflon catheter needle intro-ducted into a brachial artery. We used these vari-ables to calculate systemic (SVR) and pulmonary(PAR) arteriolar resistance by the following formu-las: SVR = AP - rap x 1332/CO and PAR = PAP- WPP x 1332/CO, where CO is cardiac output(average of three thermodilution determinations),and AP, rap, PAP, and WPP are mean arterial, rightatrial, pulmonary arterial and pulmonary capillarywedge pressure respectively. Mean pressures wereobtained by electronic damping, and heart rate wasderived from the electrocardiogram. All these vari-ables were recorded on a Hewlett-Packard eightchannel direct writing recorder (model 7868/A).

ECHOCARDIOGRAPHIC EVALUATIONLeft ventricular end systolic circumferential wallstress was taken as a variable that reflects afterload

and was evaluated non-invasively on echo-cardiograms recorded in triplicate. We used the an-giographically validated equation of Grossman andcoworkers,1 3 as modified by Reichek et al 4 to calcu-late end systolic circumferential wall stress (esCWS):

es CWS = 0 334 x Ps/LVIDsPWTs ( + PWTsPWT51 +LVIDs

where Ps, LVIDs, and PWTs are systolic cuff arte-rial pressure, left ventricular minor dimension (cm),and posterior wall thickness (cm) respectively at endsystole. Pressure was measured after the patient hadbeen recumbent for 15 min. An independent ob-server used a random zero sphygmomanometer totake three consecutive measurements with 5 min be-tween measurements. Ultrasound and blood pres-sure readings were averaged to improve theiraccuracy; variablity was within 2% for ventriculardiameter and within 5% for wall thickness and sys-tolic blood pressure. Two patients with atrialfibrillation, two with left bundle branch block, andtwo with both were excluded from this analysisbecause accurate echocardiographic recordings aredifficult to obtain in the presence of these disorders.We used a Hewlett-Packard cross sectional unit

(model 77020 A) for echocardiography. The end sys-tolic dimension of the endocardial surface was mea-sured as the shortest distance between the septal andposterior walls during systole. End systolic posteriorwall thickness was measured by the leading edge toleading edge method as the point ofmaximal anteriormotion. We analysed the means of three to five con-secutive cycles. The end systolic stress-left ventricu-lar end systolic diamater relation was taken as anindex of the end systolic force length relation.

STATISTICAL EVALUATIONAll values after treatment were compared with thebaseline values by two way analysis of variance.Differences in means were assessed by Student's ttest.

Results

Eighteen patients completed the study; all of themhad had nifedipine and captopril for eight weeks eachin a crossover study (Fig. 1).Mean exercise tolerance times on captopril were

significantly better than baseline values or values onnifedipine (Fig. 2). Nifedipine treatment values werenever significantly different from baseline values.The mean functional class was significantly better

while patients were on captopril than at baseline and

393

on July 13, 2022 by guest. Protected by copyright.

http://heart.bmj.com

/B

r Heart J: first published as 10.1136/hrt.55.4.391 on 1 A

pril 1986. Dow

nloaded from

394

HypotensicLost to follow

8 weeks

Hypotensi4Sudden dea

16 weeksFig. 1 Patient treatment.

Agostoni, De Cesare, Doria, Polese, Tamborini, Guazzi

Evaluable patients_26

13/' \i13Captopril Nifedipinean 1 Hypotension 2up 1 Lost to follow up 1

11 10Nifedipine Captoprilon 1 | | Hypotension 1 (recovered)ith 1 | |Lost to follow up 1

9 9

on nifedipine treatment (Table 2). Changes in clini-cal class paralleled the changes in exercise tolerancein all but two patients.

Baseline values for mean arterial pressure aver-aged 83 mm Hg, and the mean cardiac index for thegroup was reduced (2-5 I/min/m2). Pulmonary capil-lary wedge pressure was <20mmHg in all exceptone of the patients, and the mean value (12mmHg)was lower than expected given the severity of con-

700.

-'600-

-,"500-

0-In

- 400-W

.9400-EDU

c

sS0

,, 200-

1A-j

gestive heart failure. This was probably because pa-tients were on diuretics and on a salt restrictedhospital diet. Mean systemic and pulmonary arte-riolar resistance was 1480 and 145 dyn s cm-5respectively.Peak response to oral nifedipine occurred about

1-5 hours after administration. At this time there wasa 20% reduction in mean arterial pressure that wasmediated through a 29% decrease in systemic vascu-

r-- Nifedipinel0- --0 CaptoprilI

A

)r/

0Baseire

2I i

4 6. . . I I I .a.. I I r

*A A

A~~~~~A

11 X>e

X fo

8Vkeks of tren

Fig. 2 Comparison of times of maximal physical capacity during nifedipine and captopril treatments. *Value significantlydifferent from baseline value (p < 001); A Value sigmdficantly different from nifedipine treatment value (p< 001).

I %#V

2a

4I i 6 i

on July 13, 2022 by guest. Protected by copyright.

http://heart.bmj.com

/B

r Heart J: first published as 10.1136/hrt.55.4.391 on 1 A

pril 1986. Dow

nloaded from

Afterloadi reduction: a comparison of captopril and nifedipine in dilated cardiomvopathv

le 2 (Cl/iical and humoral results zith treatments in 18 patients with dilated cardiomvopathv

.\'YIIA class Body weight kg Scrumi potLassiuln Seri4I creatinine Dependent oedema Svrmptoms that developedMmol / tmig dl'

1 N C B .N C B N C IN" CB N C B ' C

4 4 3 5835 58 57 2 38 3 5 4 1 1 8 2-0 2-2 + + + Headache,palpitation

4 3 3 77 79 2 77 6 4 2 3 9 4-5 1.1 1 2 1 2 Palpitation4 4 3 81 82 8035 4 3 3-9 4-8 1(3 1.1 1.1 + +t Headache

4 3 81 6 83 78 2 3 7 3 3 4-2 1.9 1 4 1 6 0 + (3 Palpitation -

3 2 74 4 75 4 72 2 3 5 3 3 4 0 1 7 1-5 2 2 0 t 0 Headache,palpitation

4 3 3 88 89 87 3.7 3 6 4 1 1 8 1 6 1.9 + + t4t +3 4 3 63 62 5 62 4 3 3 9 4 6 1.5 1i1 1-5 + 4 0 Palpitation1 4 4 79 81 79 4-5 4 3 4 4 1 9 2-0 2-8 + + + ++ Palpitation4 4 3 77-6 79-2 77 4 0 3 6 4 1 1 1 1 3 1-2 l + + -I'aste alteration4 4 4 85 8 85-6 85 7 4 1 4-1 4 5 1.2 1 3 1 1 + + tt+ +4 3 3 86 87 5 84 8 3 6 3 3 4 1 1.9 1 7 18 + 4 + + + Palpitation3 3 3 80 5 81 6 77-9 4 3 4 1 4 7 1 8 1 9 2 2 0 + 0 Palpitation4 4 3 76 78 8 76 5 4 4 4-1 4.7 1 6 1 8 2 2 + + + + 0 Headache TIaste alteration3 3 2 67 66 646 4(0 36 47 1-8 1 8 20 0 0 0 -4 4 4 88 89 87 5 3 7 3 3 4 2 1-8 1 4 1-7 t t +t + + Headache,

palpitation3 4 3 614 65 60-3 44 41 47 14 16 1 1 + 03 3 2 65 4 65 64 3 6 3 3 4 1 1.2 1-5 1-5 0 0 (3 Palpitation4 4 3 73 76 3 72 3-9 3 9 4 3 1-5 1 7 1-7 + + + + + + Palpitation

3s6h 3-6 3.1* 75 7 76-9* 74 7 4(0 3 7 4 4* 1 6 1 6 1-7o 12 0.12 0 14 2 2 2 3 2 3 0(08 0 08 0(06 0(07 0(07 0 11

fficantls different (p<001) from basal values.basl, N, niltdipine; C, captopril; NYHA, New York Heart Association.

lar resistance and was associated with a 12", rise incardiac index. Mean pulmonary arterial pressure,pulmonary arteriolar resistance, and the pulmonarycapillary wedge pressure tended to decrease, al-though these changes were not statisticallysignificant. After long term treatment the responsewas substantially different. In fact, changes frombaseline in mean arterial pressure, and cardiac indexwere less pronounced and were no longer statisticallysignihcant; heart rate and pulmonary capillarywedge pressure showed a clear tendency to increase.The response to oral captopril was also greatest

approximately 1 5 h after administration. In five pa-tients the first dose induced considerable increases incardiac index and stroke volume index and decreasesin mean arterial pressure, heart rate, systemic vascu-lar rcsistance, and pulmonary capillary wedge pres-sure, which were sustained after eight weeks ofcontinuous treatment. Four patients had minimalhaemodvnamic responses to the first 50 mg of thedrug and also showed small changes when up tothree times this dose was given. Despite the reducedinitial response they continued on the standard pro-tocol and were treated with the standard dosage.Repeat assessment after eight weeks of continuoustreatment showed a spontaneous improvement ofefficacy in three patients, in whom values becamesimilar to those of the early responders in the samegroup. Because of this pattern, which has beenreported by others, I 'I' overall haemodynamic

changes from baseline that were present in the shortterm study were much more pronounced at longterm re-evaluation (Fig. 3).

Figure 4 shows means for the circulatory variablesthat were recorded in the entire population at base-line and after long term treatment with both agents.The changes induced by captopril and nifedipinemoved in the same direction, as shown in Fig. 3. Thisindicates that the sequence of drug administrationdid not interfere with the quality of the response tothe treatments. Cardiothoracic ratios measured byindependent observers on standard chest x rays weresignificantly reduced after captopril. The mean con-trol ratio was 0 575 compared with a mean ratio of0 560 after two months of captopril treatment(p< 001).

Left ventricular end systolic circumferential wallstress (mean for the entire population with the ex-ception of six patients) was significantly reduced bycaptopril and to a lesser extent and not significantlyby nifedipine at the end of the treatment period; theforce-length relation was shifted to the left after cap-topril but not after nifedipine (Fig. 5).No patients had to stop taking captopril or nife-

dipine during the study because of adverse reactions(Fig. 1).While on nifedipine five patients complained of

headache and 11 reported palpitation that was de-scribed as disturbing by six. We do not knowwhether this symptom was associated with wors-

395 on July 13, 2022 by guest. P

rotected by copyright.http://heart.bm

j.com/

Br H

eart J: first published as 10.1136/hrt.55.4.391 on 1 April 1986. D

ownloaded from

Agostoni, De Cesare, Doria, Polese, Tamborini, Guazzi

90

I60

70

a60

C' 1.5 8Hours Wee

,1,1800

E 1600u

c 1400

X 1200

tn 1000

I

1.5 8Hours Weeks

r-- Nifedipine|0|on CaiptoprilI

II I

,4 1.5 8Hours Weeks

Fig. 3 Haemodynamic values (mean (SD)) at baseline, after 1 5 hours, and after eight weeks of treatmentin patients given captopril (nine cases) and nifedipine (nine cases) as the first drug. AP, mean arterialpressure; SVR, systemic vascular resistance; PAR, pulmonary arteriolar resistance; WPP, mean pulmonarycapillary wedge pressure; PA1P, mean pulmonary arterial pressure. *Value significantly different frombaseline value (p < 001); A Value significantly different from nifedipine treatment value (p < 001).

ening arrhythmia because we did not evaluate this.We also found an increase in body weight in 12patients and enhancement of dependent oedema in11 while they were on nifedipine. This latter effect,however, may not necessarily be due to cardiaccauses.17 Nifedipine also reduced mean serum po-tassium concentration and did not depress renalfunction (Table 2).

Captopril did not lead to the development of de-pendent oedema or increase body weight. Two pa-tients on this treatment complained of a slight andtransient taste alteration; in one patient serum cre-atinine concentration rose from 1-9 to 2-8 mg/dl;mean serum potassium rose from a baseline concen-

tration of 4.Ommol/I to 4A4mmol/l (p<O001); thiswas also significantly higher than the concentration(3 7mmol/1) on nifedipine (Table 2).

Discussion

In this study two potent vasodilators with differentmechanisms of action were used to treat heart failuredue to cardiomyopathy. Captopril or nifedipine weregiven with a fixed regimen of digitalis and diureticsin a double blind crossover trial. Captoprilsignificantly improved the New York Heart Associ-ation functional class and exercise tolerance times.The incorporation of captopril into conventional

396

c

.100'

3 90'

.3 8070

*W 70'

I

C~4Ec*! 3000

fE2800x

n 2600

a 2400'

au

'u 200E, 150

100

t 500<L

I9L

on July 13, 2022 by guest. Protected by copyright.

http://heart.bmj.com

/B

r Heart J: first published as 10.1136/hrt.55.4.391 on 1 A

pril 1986. Dow

nloaded from

Afterload reduction: a comparison of captopril and nifedipine in dilated cardiomyopathy

0-

I40 , ?x / A

8 30 £i 20U)

"

14

El

IAl %..-~12

a.A

ISSate ao #St *ge

Fig. 4 Haemodynamic values (mean (SD)) of the entire population at baseline and after eight weeks' treatmentwith captopril and nifedipine. AZ, mean arterial pressure; SVR, systemic vascular resistance; PAP, meanpulmonary arterial pressure; WPJ, mean pulmonary capillary wedge pressure. *Value significantly different frombaseline value (p < 0-01); A Value significantly different from nifedipine treatment value (p < 0 01).

6 5LV end systolic diameter (cm)

Fig. 5 Relation between left ventricular end systolic wallstress and left ventricular end systolic internal diameter atbaseline and after eight weeks' treatment with captopril(broken arrow) and after eight weeks' treatment withnifedipine (solid arrow). Circles represent the means andlines the SD. *p < 0 01 for differences in wall stress anddiameter between baseline and treatment.

7.0

L.

X 100-

RL 80

- 70-

In

X:x

IE 1600-U

c 1400-

z120O4/)

418-

0x

c 4 6-

n

U)-2

c

4 2 -

397

A

6 0

I

- C* .& -

14 2AC 04'b, J1*4;

on July 13, 2022 by guest. Protected by copyright.

http://heart.bmj.com

/B

r Heart J: first published as 10.1136/hrt.55.4.391 on 1 A

pril 1986. Dow

nloaded from

398 Agostoni, De Cesare, Doria, Polese, Tamborini, Guazzi

heart failure treatment permitted patients who wereincapacitated by symptoms to perform at least amoderate physical exertion. A training effect is un-likely to be responsible for this result because thephysical tests were of short duration and infrequent.Nifedipine did not relieve symptoms, and madesome patients worse. In the long term trial thesignificant reduction of the pulmonary capillarywedge pressure and the increase of cardiac output bycaptopril but not by nifedipine showed that at leastpart ofthe increased functional capacity might be theresult of an improved performance of the left ven-tricle.Did vasodilating potency or some other factor

influence responses to the two drugs? In the shortterm study systemic vascular resistance was consis-tently reduced by nifedipine, cardiac output wasraised, and the left ventricular filling pressure tendedto fall. This pattern was similar to the response tonifedipine reported in patients with dilated cardio-myopathy and acute pulmonary oedema.5 7 Duringprolonged treatment, however, cardiac output didnot improve and the pulmonary capillary wedgepressure rose substantially, indicating that thebenefits of the reduced impedence to left ventricularejection observed in the short term study were lost,even though the vasodilating action of nifedipinepersisted. Changes in systemic vascular resistanceafter prolonged administration of captopril resem-bled those produced by nifedipine, but captopril im-proved the performance of the left ventricle, asindicated by the filling pressure-cardiac output re-lation.Heart rate was increased (by an average of nine

beats/min) on nifedipine and reduced (by an averageof 11 beats/min) on captopril. This finding may berelevant to the different haemodynamic effects of thetwo agents since the greater the dysfunction of aventricle the higher its sensitivity to changes in after-load."8 Also when the cavity becomes enlarged andsystolic wall thickening is diminished, as in dilatedcardiomyopathy, ventricular dimension becomes animportant determinant of afterload. In such a situ-ation, if cardiac output remains constant, afterloadmay vary in parallel with heart rate, because theventricular cavity is reduced more during systolewhen heart rate is slow than when it is fast. Indeedcaptopril treatment was associated with an increasedstroke index.End systolic stress represents the critical afterload

value which terminates ventricular ejection"9 andmeasurement of end systolic stress is a good index ofthe role ofafterload in determining ventricular pumpfunction. We assessed end systolic left ventricularcircumferential wall stress by a non-invasivemethod, although we were aware that echo-

cardiographic measurements in patients with dilatedhearts are liable to error and that the reproducibilitymay be unsatisfactory. In order to improve the re-liability and reproducibility of such measurementswe used appropriate techniques and criteria ofrecording, coded the echocardiograms, read themblindly and in random order, and took the average oftriplicate measurements for each period. The corre-spondence in most cases between changes in ventric-ular cavity dimensions evaluated by ultrasound andchanges in cardiothoracic ratio, and the uniformityof the trends during treatment indicate that ourechocardiographic records were reliable.The minor axis of the left ventricle at end systole

was significantly reduced by captopril but not bynifedipine, and end systolic afterload was reduced bycaptopril and not by nifedipine (Fig. 5), even thoughthe reduction of systemic vascular resistance andpossibly of aortic impedence was similar with thetwo treatments.The leftward shift of the force-length relation

after captopril may indicate either a shift to adifferent point in the same basal stress-length line(which would indicate no change in contractility) ora displacement to the left of the force-length lineitself (which would indicate improved con-tractility).13 14 This relation tended to move right-ward when patients were treated with nifedipine,indicating that some depression in contractility mayhave occurred.The reduction in heart rate with captopril, which

has been reported in several previous studies,9 16 20may be due to the absence of a facilitating actionof angiotensin II on the sympathetic neuro-transmission,21 or to reduction of catecholamine re-lease,17 or to venodilatation.- Captopril causes bothvenodilatation and arteriolar dilatation and henceheart rate may remain constant.17 If the adrenergicdrive to the heart were really depressed by captoprilwhen there was a concomitant improvement in car-diac function, the hypothesis advanced by some22that catecholamines are deleterious in primary car-diomyopathies would be reinforced.

In conclusion, the therapeutic efficacy and lowfrequency of important adverse effects indicate thatcaptopril is useful in patients with decompensateddilated cardiomyopathies that are refractory to con-ventional treatment. The poor or adverse responsesto nifedipine that we have reported in this studydiscourage its use in such patients.

References

1 Vidt DG, Bravo EL, Fouad F. Captopril. NEnglJMed1982; 306: 214-9.

on July 13, 2022 by guest. Protected by copyright.

http://heart.bmj.com

/B

r Heart J: first published as 10.1136/hrt.55.4.391 on 1 A

pril 1986. Dow

nloaded from

Afterload reduction: a comparison of captopril and nifedipine in dilated cardiomyopathy 399

2 Dargie HJ, Ball SG, Atkinson AB, Robertson JIS. Con-verting enzyme inhibitors in hypertension and heartfailure. Br Heart J 1983; 49: 305-8.

3 Abrams WB, Davies RO, Ferguson RK. Overview: therole-of angiotensin converting enzyme inhibitors in car-diovascular therapy. Fed Proc 1984; 43: 1314-21.

4 Guazzi MD. Use of the calcium channel blockingagents in the treatment of systemic arterial hyper-tension. In: Stone PH, Antman EM, eds. Calcium chan-nel blocking agents in the treatment of cardiovasculardisorders. New York: Futura, 1983; 377-401.

5 Polese A, Fiorentini C, Olivari MT, Guazzi MD. Clin-ical use of a calcium antagonistic agent (nifedipine) inacute pulmonary edema. Am J Med 1979; 66: 825-30.

6 Matsumoto S, Ito T, Sada T, et al. Hemodynamiceffects of nifedipine in congestive heart failure. Am JCardiol 1980; 46: 476-80.

7 Ludbrook PA, Tiefenbrunn AJ, Reed FR, Sobel BE.Acute hemodynamic responses to sublingual nife-dipine; dependence of left ventricular function. Circu-,lation 1982; 65: 489-98.

8 Guazzi MD, Cipolla C, Della Bella P, Fabbiocchi F,Montorsi P, Sganzerla P. Disparate unloading efficacyof the calcium channel blockers, verapamil and nife-dipine, on the failing hypertensive left ventricle. AmHeart J 1984; 108: 116-23.

9 Davis R, Ribner HS, Keung E, Sonnenblick EH, LeJemtel TH. Treatment of chronic congestive heart fail-ure with captopril, an oral inhibitor of angiotensin-converting enzyme. N Engl J Med 1979; 301: 117-21.

10 Ader R, Chatterjee K, Ports T, Brundage B, HiramatsuB, Parmley W. Immediate and sustained hemodynamicand clinical improvement in chronic heart failure by anoral angiotensin-converting enzyme inhibitor. Circu-lation 1980; 61: 931-7.

11 Captopril Multicenter Research Group. A placebo-controlled trial of captopril in refractory chronic con-gestive heart failure. J Am Coll Cardiol 1983; 2: 755-63.

12 Massie BM, Kramer BL, Topic N. Long-term cap-

topril therapy for chronic congestive heart failure. AmJ Cardiol 1984; 53: 1316-20.

13 Grossman W, Jones DM, Laurin LP. Wall stress andpattern of hypertrophy in the human left ventricle. JClin Invest 1975; 56: 56-64.

14 Reichek N, Wilson J, Sutton MSJ, Plappert TA, Gold-berg S, Hirshfeld JW. Noninvasive determination ofleft ventricular end-systolic stress: validation of themethod and initial application. Circulation 1982; 65:99-108.

15 Levine TB, Franciosa JA, Cohn JN. Acute and long-term response to an oral converting enzyme inhibitor,captopril, in congestive heart failure. Circulation 1980;69: 35-41.

16 Packer M, Medina N, Yushak M, Meller J. Hemo-dynamic patterns of response during long-term cap-topril therapy for severe chronic heart failure.Circulation 1983; 68: 803-12.

17 Fouad FM, El-Tobgi S, Tarazi RC, etal. Captopril incongestive heart failure resistant to other vasodilators.Eur Heart J 1984; 5: 47-54.

18 Cohn JN, Franciosa JA. Vasodilator therapy of cardiacfailure. N EngJI Med 1977; 297: 27-31.

19 Weber KT, Janiki JJ. Instantaneous force-velocity-length relations: experimental findings and clinical cor-relates. Am J Cardiol 1977; 40: 740-7.

20 Guazzi MD, De Cesare N, Galli C, et al. Calcium-channel blockade with nifedipine and angiotensinconverting-enzyme inhibition with captopril in thetherapy of patients with severe primary hypertension.Circulation 1984; 70: 279-84.

21 Antonaccio MJ, Kerwin L. Pre- and postiunctional in-hibition of vascular sympathetic function by captoprilin SHR. Implication of vascular angiotensin II in hy-pertension and antihypertensive actions of captopril.Hypertension 1981; 3: I-54-62.

22 Swedberg K, Hjalmarson A, Waagstein F, Wallentin I.Beneficial effects of long-term beta-blockade in con-gestive cardiomyopathy. Br Heart J 1980; 44: 117-33.

on July 13, 2022 by guest. Protected by copyright.

http://heart.bmj.com

/B

r Heart J: first published as 10.1136/hrt.55.4.391 on 1 A

pril 1986. Dow

nloaded from