Br HeartJ7 1994;72:422-427

Neurohormonal activation in patients with mildor moderately severe congestive-heart failure., and,effects of ramipril

Axel Sigurdsson, Ole Amtorp, Torstein Gundersen, Berit Nilsson, Jouko Remes,Karl Swedberg for the Ramipril Trial Study Group

Division ofCardiology,Department ofMedicine, OstraHospital, Gothenburg,SwedenA Sigurdssono AmtorpT GundersenB NilssonJ RemesK SwedbergCorrespondence to:Dr A Sigurdsson, Divisionof Cardiology, Departnentof Medicine, Ostra Hospital,416 85 Gothenburg,Sweden.

Accepted for publication16 May 1994

AbstractObjectives-To describe neurohormonalactivation in patients with mild or mod-erate heart failure and how it may bemodified by treatment with ramipril.Setting-Cardiology departments at 24hospitals in Denmark, Finland, Norway,and Sweden.Patients-223 patients with mild or mod-erately severe congestive heart failurewho were taking diuretics with or withoutdigitalis.Design-Randomised, double bind, mul-ticentre, placebo controlled comparisonof ramipril and placebo. Venous bloodsamples were drawn at rest, before blindtreatment, and after 12 weeks of treat-ment with the study drug. A probabilityprediction score for mortality derived bystepwise linear discriminant from neuro-hormone data in the first cooperativenorth Scandinavian enalapril survivalstudy (CONSENSUS I) was used toassess combined activity of the differentneurohormonal systems.Results-Plasma concentrations of atrialnatriuretic peptide were raised at base-line but angiotensin II, aldosterone,and noradrenaline concentrations werewithin normal limits. There was, how-ever, a wide interindividual variation.Plasma noradrenaline concentration andprediction score were higher amongpatients with class III congestive heartfailure according to the New York HeartAssociation's classification than amongpatients with class II congestive heartfailure (P < 0.05). There was a modestbut significant inverse correlationbetween exercise duration at baseline andplasma noradrenaline concentration(r = - 0*21, P = 0.0023), aldosterone con-centration (r = - 0-14, P = 0.04), andprediction score (r = - 0-24, P = 0.0004).Prediction score at baseline was signifi-cantly higher among those who died(n = 10) than among survivors (P = 0.03).Angiotensin converting enzyme activitywas suppressed and plasma concentra-tions of aldosterone and atrial natriureticpeptide were reduced after 12 weeks oftreatment with ramipril compared withplacebo. In patients with the most pro-nounced neurohormonal activation atbaseline (highest third of noradrenalineconcentration or prediction score), nora-drenaline concentration and predictionscore were significantly lower after 12

weeks of taking ramipril compared withplacebo. Patients with a prediction scorein the highest third at baseline had ahigher heart rate than to those in thelowest third (P = 0.003).Conclusions-Neurohormonal activationis associated with the degree of symp-toms and the severity of disease in mildor moderately severe congestive heartfailure. Treatment with ramipril attenu-ates neurohormonal activation. Thiseffect is most pronounced among patientswith the highest circulating concentra-tions of neurohormones before the startoftreatment.

(Br HeartJ 1994;72:422-427)

The efficacy of treatment of congestive heartfailure may be assessed in terms of sympto-matic improvement, by invasively or non-invasively measured indices of cardiac perfor-mance, and by evaluating functional capacity.The use of these variables, however, seems tobe of limited value in evaluating the long termefficacy of pharmacological interventions inheart failure. An improvement in exercisecapacity or cardiac performance is notnecessarily reflected in less progression of thedisease.

Recently neurohormonal factors haveemerged as important determinants of diseaseprogression in congestive heart failure.' Astrong relation has been found between circu-lating neurohormone concentrations andmortality in severe heart failure,2 and plasmanoradrenaline concentration seems to be amore sensitive guide to prognosis than tradi-tionally measured haemodynamic variables.3In the first cooperative north Scandinavianenalapril survival study (CONSENSUS I) thereduction in mortality with enalapril primarilyoccurred among patients with high plasmaconcentrations of neurohormones at the startof treatment.2 Less is known about the associ-ation of neurohormonal activation and diseaseprogression in milder forms of heart failure,although recent data have indicated a relationbetween activation of neurohormonal systemsand mortality in heart failure of moderateseverity.4 Despite the widespread use ofangiotensin converting enzyme inhibitors inpatients with mild or moderately severe heartfailure, the long term effects of such treatmenton the renin-angiotensin system, the sympa-thetic nervous system, and atrial natriureticpeptide have not been exactly defined.

422 on D

ecember 11, 2021 by guest. P

rotected by copyright.http://heart.bm

j.com/

Br H

eart J: first published as 10.1136/hrt.72.5.422 on 1 Novem

ber 1994. Dow

nloaded from

Neurohornonal activation in patients with mild or moderately severe congestive heartfailure and effects of ramipril

The Nordic Ramipril Trial Study Groupevaluated the effects of the angiotensin con-verting enzyme inhibitor ramipril on exerciseduration in patients with mild or moderatecongestive heart failure. The addition oframipril to treatment with diuretics did notsignificantly increase exercise time in thismulticentre, randomised, placebo controlledstudy.5 We present neurohormonal data fromthis trial. The main objectives were to analyseneurohormonal activation in patients withmild or moderately severe heart failure in rela-tion to different clinical variables, and tostudy the effects of ramipril on circulatingneurohormone concentrations in this popula-tion. In addition, the use of a probability pre-diction score for mortality, derived fromhormonal data in the CONSENSUS I trial,was tested as a measure of the combinedactivity of different neurohormonal systems.

Patients and methodsPATIENTSA total of 223 patients with mild or moder-ately severe congestive heart failure from 24centres in Denmark, Finland, Norway, andSweden were randomly allocated in a doubleblind fashion to treatment with ramipril orplacebo for 12 weeks.5 Patients were eligiblefor randomisation if they were taking adiuretic with or without digitalis and hadan enlarged heart size in a chest radiograph() 600 Ml/m2 for men, > 550 ml/m2 forwomen) or left ventricular ejection fraction <40% estimated by echocardiography, or both.No vasodilators other than nitrates wereallowed. The patients had to be able to per-form a bicycle exercise test with exercisecapacity restricted to the range of 4-10 min-utes (corresponding to 60-120W respec-tively), starting at 30 W, with stepwiseincrements of 10 W each minute. For patientswith a body weight above 80 kg a maximumworkload of 1-5 W/kg body weight wasallowed. Blind treatment with the study drugwas started when two consecutive exercisetests with identical exercise time had beenperformed and did not differ by more than 60s. Reproducibility had to be verified in nomore than four exercise tests.The mean age of the patients was 64-5

years (range 39-79), 160 were men. One hun-dred and forty had class II congestive heartfailure according to the classification of theNew York Heart Association and 83 had classII disease. The primary cause of heart failurewas considered to be coronary heart disease in137 patients, idiopathic dilated cardiomyopa-thy in 49, arterial hypertension in 26, and val-var disorders in six. In four patients theprimary cause of heart failure was consideredto be chronic atrial fibrillation and in onepatient the cause was unknown. The initialdose of ramipril or placebo was 1 25 mg oncedaily, which was increased weekly in two stepsto 2-5 mg and 5 mg once daily. After fourweeks the dose could be further increased atthe decision of the investigator to 10 mg oncedaily. The mean (SD) daily dose of ramipril

after 12 weeks was 7 9 (3 0) mg, and placebo8&3 (2 9) mg. Of patients randomly allocatedto ramipril, 69 were taking 10 mg daily at theend of the study and 31 were taking 5 mg orless.

BLOOD SAMPLING AND HORMONE ANALYSISAngiotensin converting enzyme activity andaldosterone concentration were determined inserum and angiotensin II, atrial natriureticpeptide, and noradrenaline concentrations inplasma. Venous blood samples were drawnafter at least 30 minutes of rest in the supineposition, before the start of blind treatmentand after 12 weeks of treatment with the studydrug. For analysis of angiotensin convertingenzyme activity and aldosterone concentra-tion 10 ml of blood were collected into anevacuated plain glass tube. The tube was leftat room temperature for around 30 minutes,after which it was centrifuged for 5 minutes at3000 rpm. For the analysis of angiotensin IIand atrial natriuretic peptide concentration 10ml ofblood were collected in a tube containingan anticoagulant (EDTA), and for the analy-sis of noradrenaline concentration 5 ml werecollected in a tube containing an anti-coagulant (EDTA) and an antioxidant (glu-tathione). The tubes were placed in an icebath and centrifuged within 20 minutes fromsampling in a refrigerated centrifuge (4°C) for5 minutes at 3000 rpm. The serum andplasma was transferred into disposable tubesand frozen at - 20°C or lower. If sampleswere not analysed within one month theywere stored at - 70°C. Concentrations ofangiotensin II, aldosterone, and atrialnatriuretic peptide were analysed by radio-immunoassays at the Unit for AppliedBiochemistry, Huddinge Hospital, Stock-holm, Sweden. The methods have beendescribed previously.2 The mean (SD) refer-ence values at the laboratory are: angiotensinII, 19(7) pmol/l (median 15 pmol/l); aldos-terone, 404(208) pmol/l (median 363 pmol/l);atrial natriuretic peptide, 57(21) pg/ml(median 47 pg/ml). Angiotensin convertingenzyme activity and noradrenaline wereanalysed at the Departnent of BioanalyticalChemistry, Astra Hassle AB, M6lndal,Sweden. Angiotensin converting enzymeactivity was analysed by an automatedmethod using 3-(2-furylacryloyl)-L-phenylalanyl-glycyl-glycine as substrate.6Noradrenaline concentration in plasma-EDTA was analysed by liquid chromatogra-phy and electrochemical detection usingalumina adsorption.7 The investigators incharge of the neurohormone determinationswere blind to the treatment groups.

STATISTICAL METHODSThe neurohormone values are presented asmeans (SD) unless otherwise specified. Therelation between continuous variables wasevaluated by the Spearman correlation coeffi-cient. The Mann-Whitney U test was used forthe non-paired comparisons between groups.Within group changes were evaluated by theWilcoxon signed rank test. P values less than

423 on D

ecember 11, 2021 by guest. P

rotected by copyright.http://heart.bm

j.com/

Br H

eart J: first published as 10.1136/hrt.72.5.422 on 1 Novem

ber 1994. Dow

nloaded from

Sigurdsson, Amtorp, Gundersen, Nilsson, Remes, Swedberg

Table 1 Hormone values and prediction scores at baseline and after 12 weeks treatmentwith study drug

Placebo Ramipril P value *

Angiotensin conuerting enzvme (,ukatll}Baseline:No of patientsMean (SD)Median (range)

12 weeks:No of patientsMean (SD)Median (range)

Baseline:No of patientsMean (SD)Median (range)

12 weeks:No of patientsMean (SD)Median (range)

Baseline:No of patientsMean (SD)Median (range)

12 weeks:No of patientsMean (SD)Median (range)

Baseline:No of patientsMean (SD)Median (range)

12 weeks:No of patientsMean (SD)Median (range)

Baseline:No of patientsMean (SD)Median (range)

12 weeks:No of patientsMean (SD)Median (range)

Baseline:No of patientsMean (SD)

12 weeks:No of patientsMean (SD)

102 1111-51 (0-57) 1-45 (0-53)1-43 (0 40-3-42) 1-35 (0-44-295)

89 1001-55 (0 58) 0-41 (0-38)1-48 (0 42-3 07) 0-28 (0-02-1-97)Angiotensin II (pmol/l)

104 11013-4 (26.6) 14-4 (18-5)70 (24-2550) 5-4 (20-855)

8812-4 (28 4)5-8 (2 3-256 8)

Aldosterone (pmolll)

9811-7 (15-5)5-9 (24-104-2)

103 111418 (307) 387 (317)349 (67-2482) 393 (69-2576)

90 100459 (344) 274 (165)367 (65-2370) 236 (67-1000)

Atrial natriurenc peptide (pg/ml)

103 111256 (222) 241 (138)207 (21-9-2000) 214 (156-719)

89 100234 (150) 204 (123)193 (104-740) 180 (23-5-612)Noradrenaline (nmol/l)

104 1102-54 (1-41) 2-69 (1-52)2-18 (0-41-10 0) 2-42 (0-8-13-0)

89 1002-50 (1-27) 2-48 (0 98)2-07 (0 66-584) 2-33 (0 77-5 94)

Prediction score (%)

103 10921-3 (3 0) 21-6 (3 4)

86 9821-2 (2-9) 20-9 (2 0)

>0 25

0-0001

>0 25

>0-25

0-16

possible pairs of hormones (two way interac-tion).9 The stepwise procedure was based on afixed alpha-to-enter level of 0.15.10 The dis-criminant score calculated from this modelwas converted into a predicted probability ofmortality using the formula: P = ex/l + ex,where P is the predicted probability of mortal-ity at six months. Among the 10 possibleterms (four individual hormones and six pairsof products) two contributed significantly tothe discrimination: noradrenaline and theproduct of angiotensin II and atrial natriureticpeptide. The model selected was x = - 1X6181+ (angiotensin II x atrial natriuretic pep-tide/109290) + (noradrenaline/1587), whereall three hormones are measured in pg/ml.A prediction score was calculated for each

patient at baseline and after 12 weeks oftreatment.

0o0o01 ResultsOf the 223 patients who were randomised toblind treatment, 20 were withdrawn because

>025 of various adverse events (10 were takingplacebo, 10 ramipril). Ten patients died(seven were taking placebo, three ramipril),

0-25 two after being withdrawn from the trial (bothwere taking ramipril). Blood samples wereavailable for analysis from 215 patients atbaseline (104 taking placebo, 111 ramipril)

0-21 and from 190 patients at 12 weeks (90 takingplacebo, 100 ramipril).

>0-25

>0-25

>0-25

*For between group comparisons (Mann-Whitney U test).

0 05 were considered significant.A probability prediction score for mortality

was calculated from neurohormonal data inthe CONSENSUS I trial database.8 Datafrom four hormones were used for the analy-sis; angiotensin II, aldosterone, atrial natri-uretic peptide, and noradrenaline. Stepwiselinear discriminant analysis was used with a

model including the four hormones and all

Table 2 Correlation between plasma- concentrations ofhormones at baseline

Atrialnatniuretic

Angiotensin II Aldosterone peptide Noradrenaline

Angiotensin II:Sample size - 213 213 213r - 0 30 0 11 0-22Pvalue - 0 0001 0-12 0 001

Aldosterone:Sample size - - 213 213r - - 0-20 0-20P value - - 0 004 0-004

Atrial natriuretic peptide:Sample size - - - 213r - - - 0-27P value - - 0 0001

r, Spearman correlation coefficient.

NEUROHORMONAL ACTIVATION BEFORE BLINDTREATMENTTable 1 summarises mean and median valuesfor the different hormones and the predictionscore at baseline and after 12 weeks of treat-ment. The concentrations of angiotensin II,aldosterone, and noradrenaline at baselinewere within normal limits. However, therewas generally a wide variation in neurohor-monal concentrations and some patientsclearly had raised values. Plasma atrial natri-uretic peptide concentration at baseline wasraised. A modest, but significant correlationwas found between the different hormones atbaseline, with r values in the range of 0-2-0-3(table 2).

Circulating concentrations of angiotensinII, aldosterone, atrial natriuretic peptide, andnoradrenaline at baseline tended to be higheramong those who did not survive (n = 10)than among survivors, but this was not signifi-cant. However, prediction score at baselinewas significantly higher among those who didnot survive (26.1%(8-8%)) than among sur-vivors (21-2%(2-6%)) (P = 0 03).

Plasma noradrenaline concentration and pre-diction score were both significantly higheramong patients with class III congestive heartfailure than among those with class II (figure 1).There was a significant inverse correlation atbaseline between exercise time and plasmanoradrenaline concentration (r = - 0 2 1, P =0-0023), aldosterone concentration (r = - 0 14,P = 0 04), and prediction score (r = - 0-24, P =0-0004) (figure 2). No relation was found

-c,.4,

424 on D

ecember 11, 2021 by guest. P

rotected by copyright.http://heart.bm

j.com/

Br H

eart J: first published as 10.1136/hrt.72.5.422 on 1 Novem

ber 1994. Dow

nloaded from

Neurohormonal activation in patients with mild or moderately severe congestive heartfailure and effects of ramipril

0

X E 4E Sro 3

ca2

0

0Class Class

o- 25

_

00

11 IIIlFi 10 _ Pla

in patients it I

1 la

predcl tion se b se

the New York Heart

Association. *P§ < 0-05.

-E 500 T.,400

0D 300

.0 200

100

0

W 419

Plasma

aldosterone

(pmol/l)

E 500400

300II2

300.1 00

x <1.93 1.93- >2.66

2.66

Plasmanoradrenaline

(nmol/=)

"a0 600

E 500._ 400

0) 300.0 200Q 1000 0

x

wU<19.9 19.9- >21.5

21.5Prediction score

(%)

Figure 2 Total exercisetime at baseline in relationto thirds ofplasmanoradrenaline andaldosterone concentrationand prediction score.*P < 0 05, **P < 0-01.

Table 3 Baseline characteristics stratified by the lowestand highest third ofprediction score

Lowest Highest(n = 73) (n = 70) P value

Mean (SD) age (years) 63-0 (8 8) 65-3 (8-2) NSSystolic blood pressure 136 134 NS(mm Hg)

Diastolic blood pressure 80 82 NS(mm Hg)

Heart rate (beats/min) 69 75 0 003S-Sodium (mmol/l) 141 140 NSS-Creatinine (umol/l) 93 100 0 07

S, serum.

between exercise time and plasma concentra-tions of angiotensin II or atrial natriuretic pep-

tide. Table 3 shows the relation betweenprediction score and some clinical and labora-tory variables at baseline. There was a signifi-cant difference between patients withprediction score in the highest and lowestthird with respect to heart rate.

EFFECTS OF RAMIPRIL ON CIRCULATINGNEUROHORMONE CONCENTRATIONSAfter 12 weeks of treatment angiotensin con-

verting enzyme activity and aldosterone con-

centration were significantly lower in theramipril group compared with the placebogroup (table 1). Figure 3 shows the relativechanges in hormone concentrations betweenbaseline and 12 weeks. In contrast to the datain table 1, the statistical analysis shown in fig-ure 3 is based on paired data alone. Plasmaangiotensin converting enzyme activity andaldosterone and atrial natriuretic peptide con-

centrations were significantly reduced in theramipril group but unchanged in the placebogroup. There were no significant changes inplasma concentrations of angiotensin II andnoradrenaline or in prediction score in eithergroup. The effects of treatment among

patients with plasma noradrenaline concen-

tration in the highest third at baseline were

separately analysed. In this group plasmanoradrenaline concentration was significantlylower in the ramipril group compared with theplacebo group at 12 weeks (figure 4). Also,among patients with a prediction score in thehighest third at baseline, the score was signifi-cantly lower in the ramipril group at 12 weekscompared with the placebo group (figure 4).The same was not true for plasma concentra-tions of angiotensin II. No differencesbetween the treatment groups were found forplasma noradrenaline concentration and pre-diction score in the middle and lowest thirds.

DiscussionNeurohormonal mechanisms are often acti-vated in patients with heart failure."-'4 Otherstudies have shown circulating neurohormoneconcentrations to be highly related to mortal-ity.2 315 Neurohormonal activation is alsomore pronounced in acute, unstable heartfailure than in the more stable, chronic formof the disorder.'6 Thus, activation of neuro-

hormonal systems seems to be related to theseverity of the underlying disease, although inpatients with the same clinical severity ofheart failure, neurohormonal activation may

vary and still be related to mortality.2 Ourstudy describes neurohormonal activation in afairly large group of patients with mild ormoderately severe heart failure. There wassome selection in the trial as the patients had tobe able to exercise within a predefined rangeon a bicycle before inclusion in the study.Otherwise, the group is representative ofpatients with heart failure receiving diureticsand with symptoms predominantly at exer-cise.

HORMONE CONCENTRATIONS BEFORE BLINDTREATMENTThe mean concentrations of angiotensin II,aldosterone, and noradrenaline were withinnormal limits at baseline suggesting much lessactivation of the vasoconstrictor systems thanamong patients with more advanced symp-toms.2 We emphasise, however, that there wasa large variation in circulating concentrationsof these hormones, with some patients havingvery high values. Plasma concentrations ofatrial natriuretic peptide were increased,which may be a reflection of the haemody-namic state.1718 Prediction score at baselinewas significantly related to subsequent mor-tality, which is in accordance with the resultsfrom CONSENSUS I.8 These results must,however, be interpreted carefully as therewere only 10 deaths during the study periodand because the study was not designed toassess mortality. There was a relatively weak,but significant, inverse correlation betweenexercise duration at baseline and noradrena-line, and aldosterone concentration as well asprediction score. A similar relation has beendescribed for noradrenaline in a smaller studyby Francis et al.'9 A relation was foundbetween the severity of symptoms and plasmanoradrenaline concentration as well as predic-tion score. Moreover, there was a correlationbetween the different hormone concentrationsat baseline, indicating that the differentneurohormon-al systems tend to be activatedtogether in the same patient; we emphasise,however, that the correlation coefficients wererelatively low.

It has not been exactly defined when neuro-hormonal systems become activated inpatients with congestive heart failure. Atrialnatriuretic peptide and noradrenaline concen-trations may be raised in patients with leftventricular dysfunction without overt heartfailure.20 Remes et al found that activation ofthe renin-angiotensin system is infrequent inpatients with new onset heart failure withoutdiuretic treatment.2' Our study shows thatthere is large interindividual variation in neu-rohormonal concentrations among patientswith similar clinical symptoms. Thus, despitesome correlation between neurohormonalactivation and symptoms, patients with stablesymptomatic heart failure do not always haveactivated neurohormonal systems and patientswith asymptomatic left ventricular dysfunc-tion may have activated systems. This indi-cates that neurohormonal activation,measured as hormonal concentrations in theblood, is not a strong determinant of the

425

on Decem

ber 11, 2021 by guest. Protected by copyright.

http://heart.bmj.com

/B

r Heart J: first published as 10.1136/hrt.72.5.422 on 1 N

ovember 1994. D

ownloaded from

Sigurdsson, Amtorp, Gundersen, Nilsson, Remes, Swedberg

* Placebo i Ramipril

**

tt

Angiotensin Angiotensin Aldosterone Atrial Noradrenaline Predictionconverting 11 natriuretic scoreenzyme peptide

Figure 3 Percentage changes in angiotensin converting enzyme activity, concentrations ofangiotensin II, aldosterone, and atrial natniuretic peptide, and prediction score betweenbaseline and 12 weeks. Results of the statistical analysis for within group changes areshown (ft P < 001) as well as for the comparison of the changes between the twotreatment groups (**P < 0-01).

-o- Placebo

6 +- Ramipril5

4 S32

Baseline 12 WeeksMean plasma

noradrenaline (nmol/l)

Baseline 12 WeeksMean

prediction score (%)

Figure 4 Plasmanoradrenaline andprediction score at baselineand 12 weeks amongpatients with plasmanoradrenalineconcentration in the highestthird at baseline. Resultsfor the between groupscomparison at baseline and12 weeks are shown.*P < 0-05, ns, notsignificant.

severity of clinical symptoms in heart failure.On the other hand, there is evidence suggestingthat neurohormonal activation is stronglyrelated to the progressive nature of the dis-order. Neurohormonal activation may be abetter indicator of progressive disease than theseverity of symptoms in patients with heartfailure. l4 22

EFFECTS OF RAMIPRIL ON NEUROHORMONALACTIVATIONAngiotensin converting enzyme activity andaldosterone and atrial natriuretic peptide con-centrations were significantly reduced byramipril compared with placebo, indicatingcompliance with treatment, interference withthe renin-angiotensin aldosterone axis, andbeneficial haemodynamic effects. Plasmaangiotensin II concentration was not signifi-cantly reduced by ramipril treatment. In somestudies angiotensin converting enzymeinhibitors reduce angiotensin II concentra-tions. However, these studies have mostlybeen done in patients with more severe or

unstable heart failure in which angiotensin IIconcentrations before treatment were higherthan in our patients.2327 The relatively lowplasma concentrations of angiotensin II mayto some degree explain why this hormone wasnot suppressed by ramipril in our study.Another possibility is that the reactive rise incirculating renin and angiotensin I concentra-tions may have overcome the inhibition ofangiotensin converting enzyme activity, allow-ing angiotensin II concentrations to rise.Among patients taking ramipril, 69% were

taking 10 mg daily and the mean daily dosewas 7-9 mg. Therefore ineffective dosing is anunlikely explanation for the lack of effect on

circulating angiotensin II concentrationamong patients receiving ramipril. Plasmanoradrenaline concentration was also unaf-fected by ramipril. Angiotensin convertingenzyme inhibitors have been found to reducecirculating concentrations of noradrenaline inpatients with heart failure, although this hasnot been consistent and the responses may

not be the same during acute and long termtreatment.28-34 In patients with the most pro-nounced neurohormonal activation both pre-diction score and plasma noradrenalineconcentration were significantly lower in theramipril group at 12 weeks compared with theplacebo group. In this type of analysis aregression to the mean has to be considered,but this phenomenon cannot explain the dif-ferences between the two treatment groups.Thus, a reduction in plasma noradrenalineconcentration with angiotensin convertingenzyme inhibitors seems to be most easilyachieved when the pretreatment values arehigh. This could explain the discrepanciesbetween earlier studies.

NEUROHORMONAL PREDICTION SCOREAs several plasma hormones are related tomortality in congestive heart failure, an assess-ment of the combined activation of neurohor-monal systems may be a better predictor ofprogression of heart failure and mortality thanplasma concentrations of individual hor-mones. A probability prediction score formortality derived by multivariate analysis ofhormone data from CONSENSUS I was usedto assess combined neurohormonal activity inour study. The aim was to find out if a com-bined nominator for the different hormonalsystems can be used as a measure of severityin mild to moderate forms of congestive heartfailure, and if this nominator can be affectedby treatment with an angiotensin convertingenzyme inhibitor. Although the selectedmodel is the best based on data from CON-SENSUS I, other models might be better ingeneral. Therefore it is important to test suchstatistical models in different patient popula-tions. In our study noradrenaline contributedrelatively more to the prediction score thanangiotensin II and atrial natriuretic peptide. Apopulation of patients with high predictionscore at baseline was identified as havinghigher heart rate, higher serum creatinineconcentration, and less exercise duration,probably indicating more severe heart failure.

NEUROHORMONAL ACTIVATION AND HEARTFAILUREThe determinants of neurohormonal activa-tion in chronic heart failure are not entirelyknown. There seems to be clear variationin neurohormonal concentrations amongpatients with similar clinical symptoms; noclear relation has been found between neuro-hormonal activation and haemodynamic mea-

surements.3536 Furthermore, neurohormonalactivation may be partly related to pharmaco-logical treatment such as with diuretics.37AoWhether neurohormonal activation is a reflec-tion only of the severity of the underlying dis-ease or whether it is an important componentin the pathophysiology of chronic heart failureis not known. The positive effects ofangiotensin converting enzyme inhibitors on

mortality and progression of heart failure andthe efficacy of fi blockers in patients with heartfailure due to idiopathic dilated cardiomyo-pathy certainly indicate that suppression of

** **

100

80

g 600

C 40e(a 20Co-oE 00

-0c -40.-ccu -60

-80

-100I

426 on D

ecember 11, 2021 by guest. P

rotected by copyright.http://heart.bm

j.com/

Br H

eart J: first published as 10.1136/hrt.72.5.422 on 1 Novem

ber 1994. Dow

nloaded from

Neurohormonal activation in patients with mild or moderately severe congestive heartfailure and effects of ramipril

neurohormonal activity is important.4' Aneurohormonal hypothesis has been formu-lated stating that prolonged activation of thesympathetic nervous system and the renin-angiotensin system exerts a direct adverseeffect on the heart that is independent of thehaemodynamic action of these systems.'4243Although still unproved, this hypothesis mayhave important implications for future treat-ment of congestive heart failure, as persistentactivation of different autocrine, paracrine,and endocrine systems in itself becomes atherapeutic target.

We thank Dr Steven Snapinn for statistical advice and forderiving the prediction score equation from the CONSENSUSI database. We also thank Professor Peter Eneroth, Britt MarieEriksson, and Lennart Svensson for supervising and performingthe laboratory assays.

1 Packer M. The neurohormonal hypothesis: a theory toexplain the mechanism of disease progression in heartfailure. JAm Coll Cardiol 1992;20:248-54.

2 Swedberg K, Eneroth P, Kjekshus J, Wilhelmsen L, for theCONSENSUS Trial Study Group. Hormones regulat-ing cardiovascular function in patients with severe con-gestive heart failure and their relation to mortality.Circulation 1990;82: 1730-6.

3 Cohn JN, Levine TB, Olivari MT, Garberg V, Lura D,Francis GS, et al. Plasma norepinephrine as a guide toprognosis in patients with chronic congestive heart fail-ure. NEnglJMed 1984;311:819-23.

4 Francis GS, Cohn JN, Johnson G, Rector TS, Goldman S,Simon A, for the V-HeFT VA Cooperative StudiesGroup. Plasma norepinephrine, plasma renin activityand congestive heart failure: relations to survival and theeffects of therapy in V-HeFT II. Circulation 1993;87(suppl VI):40-8.

5 Gundersen T, Swedberg K, Amtorp 0, Remes J, NilssonB, for The Ramipril Trial Study Group. Symptomaticimprovement but no effect on exercise capacity after 12weeks' treatment with ramipril in patients with moderatecongestive heart failure. Eur HeartJ 1994;15 (in press).

6 Johansen K, Martein S, Aas P. Automated method for thedetermination of angiotensin convertingenzyme inserum. Scandy Clin Lab Invest 1987;47:411-4.

7 Eriksson B-M, Persson B-A. Determination of cate-cholamines in rat heart tissue and plasma samples by liq-uid chromatography with electrochemical detection. JChromatogr 1982;228: 143-54.

8 Swedberg K, Kjekshus J. Cardiovascular hornones in rela-tion to mortality in severe heart failure. JAm ColU Cardiol1989;13:246A.

9 Morrison D. Multivariate statistical methods, 2nd ed, NewYork: McGraw-Hill, 1976.

10 Constanza M, Afifi A. Comparison of stopping rules inforward stepwise discriminant analysis. J7ournal of theAmerican Statistical Association 1979;74:777-85.

11 Thomas J, Marks B. Plasma norepinephrine in congestiveheart failure. Am J Cardiol 1978;41:233-43.

12 Curtiss C, Cohn J, Vrobel T, Franciosa J. Role of therenin-angiotensin system in the systemic vasoconstric-tion of chronic congestive heart failure. Circulation 1978;58:763-70.

13 Cody R, Atlas S, Laragh J, Kubo SH, Covit AB, RymanKS, et al. Atrial natriuretic factor in normal subjects andheart failure patients. J Clin Invest 1986;78:1362-74.

14 Goldsmith S, Fracis G, Cowley A, Levine T, Cohn J.Increased plasma arginine vasopressin levels in patientswith congestive heart failure. J Am Coll Cardiol 1983;1:1385-90.

15 Gottlieb S, Kukin M, Ahern D, Packer M. Prognosticimportance of atrial natriuretic peptide in patients withchronic heart failure. J Am Coll Cardiol 1989;13:1534-9.

16 Dzau V, Colucci W, Hollenberg N, Williams G. Relationof the renin-angiotensin-aldosterone system to the clinicalstate in congestive heart failure. Circulation 1981;63:645-51.

17 Richards AM, Cleland JGF, Tonolo G, McIntyre GD,Leckie BJ, Dargie HJ, et al. Plasma alfa natriuretic peptidein cardiac impainnent. BMJ 1986;293:409-12.

18 Sigurdsson A, Held P, Swedberg K. Short and long termneurohormonal activation following acute myocardialinfarction. Am HeartJ 1993;126:1068-76.

19 Francis G, Goldsmith S, Cohn J. Relationship of exercise

capacity to resting left ventricular performance and basalnorepinephrine levels in patients with congestive heartfailure. Am HeartJ 1982;104:725-31.

20 Francis G, Benedict C, Johnstone D, Kirlin PC, NicklasLiang C, Kubo SH, et al, for the SOLVD Investigators.Comparison of neureondocrine activation in patientswith left ventricular dysfunction with and without con-gestive heart failure. Circulation 1990;82:1724-9.

21 Remes J, Tikkanen I, Fyhrquist F, Pyorala K. Neuro-endocrine activity in untreated heart failure. Br Heart J1991;65:249-55.

22 Packer M. Pathophysiology of chronic heart failure. Lancet1992;340:88-92.

23 Cleland J, Dargie H, Ball S, Gillen G, Hodsman GP,Morton, JJ, et al. Effects of enalapril in heart failure: adouble blind study of the effects on exercise perfor-mance, renal function, hormones and metabolic state. BrHeartJ 1985;54:305-12.

24 McGrath B, Arnolda L. Enalapril reduces the cate-cholamine response to exercise in patients with heartfailure. EurJ Clin Pharmacol 1986;30:485-7.

25 Crozier I, Ikram H, Nicholls G, Jans S. Acute hemody-namic, hormonal and electrolyte effects of ramipril insevere congestive heart failure. Am J Cardiol 1987;59:155-63D.

26 Nicholls G, Ikram H, Espiner E, Moaslowski A, ScandrettM, Penman T. Hemodynamic and hormonal responsesduring captopril therapy for heart failure: acute, chronicand withdrawal studies. Am J Cardiol 1982;49:1497-501.

27 McAlpine H, Morton J, Leckie B, Dargie H. Hemo-dynamic effects of captopril in acute left ventricular failurecomplicating myocardial infarction. J CardiovascPharmacol 1987;9(suppl 2):S25-30.

28 Cody R, Franklin K, Kluger J, Laragh J. Sympatheticresponsiveness and plasma norepinephrine during ther-apy of chronic congestive heart failure with captopril.AmJMed 1982;72:791-7.

29 Wenting G, Man i n't Veld AJ, Woittiez AJ, Boomsma F,Laird-Meeter K, Simoons ML, et al. Effects of captoprilin acute and chronic heart failure: correlations withplasma levels of noradrenaline, renin, and aldosterone.Br HeartJ 1983;49:65-76.

30 Riegger G, Kochisek K. Vasopressin, renin and norepine-phrine before and after captopril administration inpatients with congestive heart failure due to idiopathicdilated cardiomyopathy. Am J Cardiol 1986;58:300-3.

31 Kubo S, Cody R, Laragh J, Prida E, Atlas SA, Yuan Y, et al.Immediate converting enzyme inhibition with intra-venous enalapril in chronic congestive heart failure. Am JCardiol 1985;55:122-6.

32 Kromer E, Riegger G, Liebau G, Kochsiek K.Effectiveness of converting enzyme inhibition (enalapril)for mild congestive heart failure. Am J Cardiol 1986;57:459-62.

33 Manthey J, Osterziel K, Rohrig N, Dietz R, Hackenthal E,Schmidt-Gayk H, et al. Ramipril and captopril inpatients with heart failure: effects on hemodynamics andvasoconstrictor systems. Am J7 Cardiol 1987;59:171-5D.

34 deGraeff P, Kingma J, Dunseiman P, Wesseling H, Lie K.Acute hemodynamic and hormonal effects of ramipril inchronic congestive heart failure and comparison withcaptopril. Am J Cardiol 1987;59:164-70D.

35 Kubo S. Neurohormonal activation ar' the response toconverting enzyme inhibitors in congestive heart failure.Circulation 1990;81(suppl ):107-14.

36 Levine T, Francis G, Goldsmith S, Simon A, Cohn J.Activity of the sympathetic nervous system and renin-angiotensin system assessed by plasma hormone levelsand their relation to hemodynamic abnormalities in con-gestive heart failure. Am J Cardiol 1982;49: 1659-66.

37 Ikram H, Chan W, Espiner E, Nicholls M. Hemodynamicand hormone response to acute and chronic furosemidetherapy in CHF. Clin Sci 1980;59:443-9.

38 Francis G, Siegel R, Goldsmith S, Olivari M, Levine T,Cohn J. Acute vasoconstrictor response to intravenousfurosemide in patients with chronic congestive heart fail-ure. Ann Intern Med 1985;103:1-6.

39 Bayliss J, Norell M, Canepa-Anson R. Untreated heartfailure: clinical and neuroendocrine effects of introduc-ing diuretics. BrHeartJ 1987;57:17-22.

40 Broqvist M, Dahlstrom U, Karlberg B, Karlsson E,Marklund T. Neuroendocrine response in acute heartfailure and the influence of treatment. Eur Heart J1989;10: 1075-83.

41 Waagstein F, Bristow MR, Swedberg K, Camerini F,Fowler MB, Silver MA, et al. Beneficial effects of meto-prolol in idiopathic dilated cardiomyopathy. Lancet1993;342:1411.

42 Packer M, Lee W, Kessler P, Gottlieb S, Bernstein J,Kukin M. Role of neurohormonal mechanisms in deter-mining survival in patients with severe chronic heart fail-ure. Circulation 1987;75(suppl IV):80-92.

43 Packer M. Neurohormonal interactions and adaptions incongestive heart failure. Circulation 1988;77:721-30.

427

on Decem

ber 11, 2021 by guest. Protected by copyright.

http://heart.bmj.com

/B

r Heart J: first published as 10.1136/hrt.72.5.422 on 1 N

ovember 1994. D

ownloaded from