University of Groningen

Value of neurohormonal and autonomic parameters for the assessment of the severity andprognosis in chronic heart failureTjeerdsma, Geert

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2001

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Tjeerdsma, G. (2001). Value of neurohormonal and autonomic parameters for the assessment of theseverity and prognosis in chronic heart failure. s.n.

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 24-04-2020

MIBEFRADIL AND AUTONOMIC DYSFUNCTION IN CHF

77

PART 4

Intervention in chronic heart failure

APPENDIX 8

78

MIBEFRADIL AND AUTONOMIC DYSFUNCTION IN CHF

79

APPENDIX 7

Effect of mibefradil on heart rate variability

in patients with chronic heart failure

G. Tjeerdsma, J. Brouwer, P.J. de Kam, J. Haaksma,H.J.G.M. Crijns, D.J. van Veldhuisen

Department of Cardiology/Thoraxcenter, University Hospital Groningen, Groningen

The Netherlands

Int J Cardiol 2000;73:55-60.

APPENDIX 8

80

ABSTRACT

Background Mibefradil was recently withdrawn from the marketbecause of an unfavorable clinical profile in patients with CHF.Although drug interactions appear to play a role, other mechanismssuch as proarrhythmia and autonomic deterioration could also berelevant. CHF is accompanied by autonomic impairment and analysisof HRV can be used to examine autonomic modulation of heart rate.

Methods and Results We studied 18 CHF patients (age 63.2 ± 10.1years (mean ± SD), ejection fraction 0.21 ± 0.07) treated withmibefradil or placebo, who participated in the MACH-I (MortalityAssessment in CHF) trial in our center, and compared them with 18healthy matched controls. HRV analysis was performed at baselineand after 7 months of treatment. At baseline, HRV parameters wereimpaired in patients with CHF compared to healthy controls (p<0.05).After seven months of treatment a reduction in (24-hour) heart ratewas observed (p=0.02, versus placebo). Apart from the effect on meanNN, no significant differences were observed for the remaining HRVparameters.

Conclusions Mibefradil does not impair autonomic balance and infact reduces heart rate in patients with CHF. These findings suggestthat autonomic activation did not contribute to the adverse effects ofmibefradil.

INTRODUCTION

Autonomic dysfunction is one of the hallmarks of CHF and has beenshown to relate to severity of disease and prognosis in CHF. 1

Measurement of HRV is a non-invasive method, which providesreliable information on autonomic modulation of heart rate, and ithas been shown to be a valuable tool in this syndrome. 2, 3 In addition,measurement of HRV has been shown useful in assessing drug-inducedchanges in CHF. 4 Calcium channel blockers are used by a substantialproportion of patients with CHF, although their value remainscontroversial. 5 One of the suggested mechanisms, by which the firstgeneration calcium channel blockers could have a negative effect onoutcome, was their propensity to increase neurohormonal activation.6 Mibefradil (Ro 40-5967) is a recently developed T-channel selectivecalcium channel blocker, which differs structurally from other availablecompounds. 7 Mibefradil was shown to be effective in patients withhypertension and angina pectoris 8, 9, and in patients with CHF earlydata suggested that it might cause a decrease in heart rate. 10 Since itis important to establish the autonomic effect of mibefradil, wemeasured HRV in a subset of patients with CHF, who participated ina recent large-scale CHF trial, the Mortality Assessment in CongestiveCHF trial (MACH). 11

METHODS

Study PopulationPatients aged 20 years or older were included in MACH if they fulfilledthe following criteria: (1) clinical evidence of symptomatic CHF ofNew York Heart Association Class II-IV within the last three months;(2) left ventricular ejection fraction, by radionuclide ventriculography< 0.35; (3) aged ≥ 20 years; (4) clinically stable on optimal therapyfor four weeks comprising loop diuretics and ACE inhibitors with orwithout other vasodilators and/or digitalis. Patients were excluded incase of acute CHF/recent myocardial infarction, significant valvularor congenital heart disease, hypertrophic obstructive cardiomyopathy,second or third degree AV block or other generally accepted exclusioncriteria. In addition to these criteria, patients in the present HRVanalysis had to be in regular sinus rhythm.

Study ProtocolAll patients in the present study participated in the MACH trial,which was a randomized, placebo controlled parallel group com-parison of mibefradil (Ro 40-5967, 50mg once or twice a day) andplacebo in patients with mild to severe CHF. 11 After a 2-week placebo

MIBEFRADIL AND AUTONOMIC DYSFUNCTION IN CHF

81

run-in period patients received 1 tablet (50 mg) placebo or mibefradil(double-blind phase) during 1 month. After one month double blindmedication was titrated up to 100 mg once daily.

The present pre-defined HRV substudy was only performed inpatients who were randomized at the University Hospital of Gronin-gen. Holter recordings were obtained at the end of the 2-week placeborun-in period, and after seven months of double-blind treatment(standard protocol visit). The study was approved by the Local EthicsCommittee and was conducted in accordance with the revisedDeclaration of Helsinki. Before entry in the study all patients gavewritten informed consent.

For comparison of HRV in these patients with CHF, healthyage and sex matched controls were also studied for HRV. The age ofthe healthy controls had to be between 20 and 75 years. Noabnormalities were allowed on routine physical examination, standardelectrocardiography, echocardiography and exercise testing. Within14 days before Holter recording, use of drugs was not allowed.

HRV AnalysisBefore entry in the study and during treatment all patients in thepresent substudy underwent 24-hours Holter recording. Ambulatory

monitoring was performed using 3 channel Holter recorders(Marquette series 8500). Modified leads aVF, V1 and V5 were used.The ECG was analyzed on Marquette XP analyzer (MarquetteElectronics, Inc., Milwaukee, WI, USA) and extensively reviewed byan experienced analyst, thus identifying all noise and ectopic beats.).All Holter analysis was performed without knowledge of the clinicalcharacteristics of the patients. Recordings with more than 15 % noiseor ectopic beats in total were excluded. For the calculations of HRVparameters, the data-file of RR intervals was transferred to a post-processor, developed at our institute. 12 From the series of RR intervals,time and frequency domain parameters were calculated. Time domainanalysis was performed over the entire 24-hour period segment. Asadvised in the guidelines, frequency domain analysis was performedusing consecutive 5-minute segments using discrete Fouriertransformation. 13, 14 Only data segments containing less than 5%noise or ectopy were accepted for analysis. This is a percentage basedon time, not number of beats. Finally a check was performed to ensurethat stationarity of the data segments was within 15% limits. Thefollowing time domain parameters were measured: the mean heartrate during 24 hours Holter recording (beats per minute), the meanof all normal to normal RR intervals (mean NN, ms), standard

Table 1. Characteristics of CHF patients (at baseline) and healthy controls

Heart Failure Patients

Parameters Healthy controls Placebo Mibefradil

( n = 18) (n = 9) (n = 9)

Mean age (years) 55 ± 10 62 ± 12 64 ± 8

Sex (male / female) 12 ± 6 5 / 4 7 / 2

Etiology of chronic heart failureIschemic heart disease 9 0Dilated cardiomyopathy 8 1

Mean blood pressure (mmHg) 91 ± 4 91 ± 7 91 ± 7

Heart rate (beats.min -1) 72 ± 11 76 ± 14 81 ± 14

Mean left ventricular ejection fraction 0.23 ± 0.07 0.19 ± 0.06

Duration heart failure (years) 3.2 ± 2.7 4.7 ± 3.8

NYHA II / III 6 / 3 4 / 5

MedicationDiuretics 9 9ACE inhibitors 9 9Digoxin 7 3

Data expressed as mean ± SD. CHF; chronic heart failure.

APPENDIX 8

82

Table 2. Heart rate variability parameters for healthy controls and patients with chronic heart failure at baseline.

Heart Failure patients *

Healthy controls Mibefradil Placebo

(n = 18) (n = 9) (n = 9)

Time domain parametersMean NN (ms) 846.2 ± 79.5 768.3 ± 88.3 770.8 ± 100.7SDNN (ms) 160.7 ± 23.2 102.8 ± 30.8 120.4 ± 32.3SDANN (ms) 135.5 ± 25.3 99.2 ± 24.4 112.8 ± 35.3RMSSD (ms) 32.3 ± 11.2 29.4 ± 17.5 26.4 ± 9.3

Frequency domain parametersLow frequency power (ms2) 947.5 ± 573.5 219.0 ± 47.6 174.2 ± 24.0High frequency power (ms2) 311.2 ± 215.4 77.4 ± 18.5 81.0 ± 9.0Total power (ms2) 4118.8 ± 1788.1 620.0 ± 104.0 571.2 ± 56.3

Mean NN, mean of all normal RR intervals; SDNN, standard deviation of heart period in milliseconds; SDANN, standard deviation of the averages of normal R-Rintervals in all 5-minute segments; rMSSD, root-mean-square successive difference of NN intervals; * All heart rate variability parameters were significantly different(p < 0.05) versus healthy controls.

deviation of heart period in milliseconds (SDNN, ms), standarddeviation of the averages of normal R-R intervals in all 5-minutesegments of a complete 24 hour recording (SDANN, ms) and root-mean-square successive difference of NN intervals (rMSSD, ms). Thelatter represents fast, beat-to-beat changes in heart rate, and has beenshown to be a measure of cardiac vagal activity. 13 Frequency domainanalysis was performed by using discrete Fourier Transformation. Lowfrequencies (LF, ms2; 0.04-0.15 Hz), high frequencies (HF, ms2; 0.15-0.4 Hz) and total power (TP, ms2; 0.0033-0.4 Hz) were calculated.The low frequency component is influenced by sympathetic andparasympathetic control mechanisms, whereas the high frequencycomponent is almost exclusively under vagal control. 13

Statistical AnalysisThe statistical analysis was performed using SAS-PC version 6.12(Cary, North Carolina, USA). Data are expressed as mean ± SD.Difference in baseline characteristics is analyzed by using the FisherExact test for categorical variables, and by the Wilcoxon 2-sampletest for continuous variables. To evaluate differences from baseline toseven months of the HRV parameters between placebo and mibefradil,a random effect model for longitudinal data was used, differentiatingwithin and between patient variation. Differences were consideredstatistically significant when a 2-tailed p value < 0.05 was found.

RESULTS

Of the original study population of twenty-eight patients in ourinstitution, seven patients had atrial fibrillation and in three patientsHolter recordings could not be analyzed for HRV since ectopic beatsor noise accounted for ≥ 15% of the signal. The remaining eighteenpatients formed the present study population. Of this group ninepatients received placebo and nine patients received mibefradil. Theirmean age was 63 ± 11 year (range 35-80) at the time of screening. Ofthese patients, twelve were men and six were women. There were nostatistically significant differences between the 2 groups. Forcomparison of HRV in the patients with CHF, HRV was also analyzedin 18 healthy controls. Their mean age was 55 ± 10 years (range 42 to65 year). The clinical characteristics of both patients and healthycontrols at baseline are summarized in Table 1.

At baseline, all HRV parameters were significantly lower in the CHFpatients when compared to the healthy controls (p < 0.05). Withinthe CHF group, there was no significant difference for any of theHRV parameters between the placebo and the mibefradil group atbaseline (Table 2).

MIBEFRADIL AND AUTONOMIC DYSFUNCTION IN CHF

83

Figure 1. Change in heart rate (beats per minute(BPM), mean ± SD) for patients with chronic heartfailure after seven months of treatment with studymedication. Mib-base, mean heart rate at baselinefor mibefradil group; Mib-Tx, mean heart rate afterseven months of treatment with mibefradil; Plac-base, mean heart rate at baseline for the placebogroup; Plac-Tx, mean heart rate after seven monthsof treatment with placebo.

Figure 2. Change in time domain parameters ofheart rate variability at baseline (black) and after 7months of treatment (white) with mibefradil orplacebo for patients with chronic heart failure.Mean NN, mean of all normal RR intervals; SDNN,standard deviation of heart period in milliseconds;SDANN, standard deviation of the averages ofnormal R-R intervals in all 5-minute segments;rMSSD, root-mean-square successive difference ofNN intervals, *p < 0.02

APPENDIX 8

84

After seven months of treatment, the use of mibefradil was associatedwith a significant decrease in heart rate compared with the placebogroup (p = 0.02), an effect that was primarily observed during the day(figure 1).

The relation of the treatment effects to the HRV parameters are shownin figure 2 for mean NN, SDNN, SDANN and rMSSD. Apart fromthe effect on mean NN, no significant differences were observed forSDNN, SDANN or rMSSD between the placebo and the mibefradilgroup. With regard to the frequency domain parameters, similar resultswere obtained, and there were no significant differences between themibefradil and placebo group.

DISCUSSION

Most trials in the past in patients with CHF, examining calciumchannel blockers have demonstrated disappointing results. Recently,mibefradil (Ro 40-5967), a compound with T-type calcium channelselectivity, was developed, and early data suggested that this agentwas not associated with significant negative inotropic and positivechronotropic effects on the heart in patients with ischemic heart diseaseand hypertension. 8, 9 In addition, no significant increases in plasmanorepinephrine levels were observed in patients with left ventriculardysfunction. 15 The use of this new calcium channel blocker wastherefore expected to be of potential benefit in patients with CHF,which was further supported by rather favorable observations inexperimental studies. 14 Nevertheless, mibefradil was recentlywithdrawn from the market because of interactions with a largenumber of drugs, which might lead to an increased mortality. 16, 17

In the present study, mibefradil did not affect HRV in patientswith CHF, who were found to have markedly impaired autonomicbalance at baseline. Moreover, the drug was found to cause a smallbut significant decrease in heart rate in these patients. Since a decreasein HRV parameters is related to poorer prognosis in patients withCHF 2, further deterioration by mibefradil would have beendisadvantageous, but this was not observed in the present study. Heartrate itself also is a well-known prognostic marker in CHF, and reducingheart rate during drug-therapy may be associated with improvementin survival. 18, 19 In the present study heart rate was reduced bymibefradil, which confirms earlier observations in CHF. 10, 15 Thiseffect could therefore be expected to be of benefit in heart failure, butdata on this issue from the MACH Study are not (yet) available.

LimitationsThe present study therefore provides no information regarding a

possible mechanism by which mibefradil could have an adverse effectin patients with CHF. Although the small number of patients limitsthe present findings, they do not support an unfavorable autonomiceffect of mibefradil in heart failure. Whether newer compounds ofthis group of T-type calcium channel blockers may be of more clinicalvalue, requires further study.

References

1. Packer M: The neurohormonal hypothesis: a theory to explain the mechanismof disease progression in heart failure. Am J Coll Cardiol 1992;20:248-54

2. Szabo BM, Van Veldhuisen DJ, Van der Veer N et al: Prognostic value of heartrate variability in chronic congestive heart failure secondary to idiopathic orischemic dilated cardiomyopathy. Am J Cardiol 1997;978-80

3. Nolan J, Batin PD, Andrews R et al: Prospective study of heart rate variabilityand mortality in CHF : results of the united kingdom heart failure evaluationand assessment of risk trial (UK-heart). Circulation 1998;98: 1510-6

4. Tuininga YS, Van Veldhuisen DJ, Brouwer J et al: Heart rate variability in leftventricular dysfunction and heart failure: effects and implications of drugtreatment. Br Heart J 1994;72:509-13

5. Gheorghiade M, Cody RJ, Francis GS et al: Current medical therapy for advancedheart failure. Am Heart J 1998;135:S231-48

6. Packer M: Pathophysiological mechanisms underlying the adverse effects of cal-cium channel-blocking drugs in patients with chronic heart failure. Circulation1989;80:IV59-67

7. Mishra SK, Hermsmeyer K: Selective inhibition of T-type Ca2+ channels by Ro40-5967. Circ Res 1994;75:144-8

8. Schmitt R, Kleinbloesem CH, Belz GG et al: Hemodynamic and humoral effectsof the novel calcium antagonist Ro 40-5967 in patients with hypertension. ClinPharmacol Ther 1992;52:314-23

9. Bakx AL, van der Wall EE, Braun S et al: Effects of the new calcium antagonistmibefradil (Ro 40-5967) on exercise duration in patients with chronic stableangina pectoris: a multicenter, placebo-controlled study. Ro 40-5967 Interna-tional Study Group. Am Heart J 1995;130:748-57

10. Van der Vring JA, Bernink PJ, Van der Wall EE et al: Evaluating the safety ofmibefradil, a selective T-type calcium antagonist, in patients with chroniccongestive heart failure. Clin Ther 1996;18:1191-1206

11. Levine TB: The design of the Mortality Assessment in Congestive Heart FailureTrial (MACH-1, mibefradil). Clin Cardiol 1997;20:320-6

12. Haaksma J, Brouwer J, Mulder LJM et al: Heart rate dependant changes inspectral analysis. IEEE Proc Comp in Cardiol 1994;45-8

13. Task Force of the European Society of Cardiology and the North AmericanSociety of Pacing and Electrophysiology: Heart rate variability: standards ofmeasurement, physiological interpretation and clinical use. Circulation 1996;93:1043-65

14. DeBoer RW, Karemaker JM, Strackee J: Comparing spectra of a series of pointevents particularly for heart rate variability data. IEEE Trans Biomed Eng1984;31:384-87

15. Rousseau MF, Hayashida W, van EC et al: Hemodynamic and cardiac effects ofthe selective T-type and L-type calcium channel blocking agent mibefradil inpatients with varying degrees of left ventricular systolic dysfunction. J Am CollCardiol 1996;28:972-9

MIBEFRADIL AND AUTONOMIC DYSFUNCTION IN CHF

85

16. Verheugt FW: Hotline sessions of the 21st European Congress of Cardiology.Eur Heart J 1999;20:1603-6

17. SoRelle R: Withdrawal of Posicor from market. Circulation 1998;98:831-32

18. Nul DR, Doval HC, Grancelli HO et al: Heart rate is a marker of amiodaronemortality reduction in severe heart failure. The GESICA-GEMA Investigators.Grupo de Estudio de la Sobrevida en la Insuficiencia Cardiaca en Argentina-Grupo de Estudios Multicentricos en Argentina. J Am Coll Cardiol 1997;29:1199-1205

19. Lechat P, Escolano S, Golmard JL et al: Prognostic value of bisoprolol-inducedhemodynamic effects in heart failure during the Cardiac Insufficiency BIsoprololStudy (CIBIS). Circulation 1997;96:2197-2205

APPENDIX 8

86

BETABLOCKERS AND AUTONOMIC IMPAIRMENT IN CHF

87

APPENDIX 8

Autonomic dysfunction in patients with mild

heart failure and coronary artery disease

and the effects of add-on b-blockade

G. Tjeerdsma *, MD, B.M. Szabó *, MD, L.M. van Wijk **, MD, J. Brouwer *,R.A. Tio *, H.J.G.M. Crijns *, D.J. van Veldhuisen *.

*Department of Cardiology/Thoraxcenter, University Hospital Groningen, Groningen

**Department of Cardiology, Refaja Hospital, Stadskanaal, The Netherlands

Eur J Heart Failure 2000;3:33-39

APPENDIX 9

88

ABSTRACT

Aims Autonomic impairment is related to the incidence of suddendeath in chronic heart failure (CHF). Our objective was to studyautonomic profiles in patients with mild CHF due to coronary arterydisease, and to investigate the value of add-on ß-blockade.

Methods Measures of autonomic function (plasma norepinephrine,heart rate [HR] variability, autonomic function testing), and exercisecapacity, were compared between 24 patients with mild CHF, and24 healthy controls. In this mechanistic study, we assessed the effectof 26 weeks metoprolol treatment in a double-blind, randomized,placebo-controlled design. All patients received metoprolol sustainedrelease (200 mg; n=12) or placebo (n=12). Assessments were made atbaseline and after 10 and 26 weeks treatment.

Results At baseline, norepinephrine levels were elevated, while HRvariability parameters were decreased in patients vs. controls (bothp<0.05). Autonomic function testing showed only small differences,although significant alterations were observed with deep breathingand head up tilting (both p<0.05). After 26 weeks, metoprolol didnot affect exercise capacity or norepinephrine concentrations. Incontrast, HR variability was markedly improved in metoprolol-treatedpatients versus placebo-treated patients (p<0.05). In particular, a shifttoward normal in the sympathovagal balance was observed (p<0.05).Autonomic function testing showed only small, and generally non-significant trends after metoprolol.

Conclusions Marked autonomic abnormalities are already present inmild CHF, which may be (partially) reversed by metoprolol. Theseobservations support the reported reduction of sudden death by β-blockade in patients with CHF.

INTRODUCTION

Neurohormonal activation and autonomic dysfunction are hallmarksof chronic heart failure (CHF) and the severity of these changes iscorrelated with the severity and prognosis of this syndrome. 1, 2

Experimental 3 and clinical studies 4 suggest that autonomicabnormalities occur already in the early stages of CHF, but only fewdata are available on this issue. These autonomic abnormalities areclinically relevant, as they have been associated with the occurrenceof sudden cardiac death. 5, 6 Angiotensin converting enzyme (ACE)inhibitors have become a cornerstone in the treatment of CHF, butdespite their success, total mortality and the incidence of suddencardiac death remain high. In the last few years, several large-scaletrials with β-blockers have shown that these drugs may reduce suddencardiac death and total mortality in CHF. 7-9 It has been suggestedthat this beneficial effect may be due to an improvement in autonomictone. 10 The purpose of the present mechanistic study was therefore,to examine autonomic function in patients with mild CHF andcoronary artery disease, and to compare them to healthy controls. Inaddition, we examined the effects of add-on β-blockade withmetoprolol during prolonged (26 weeks) treatment.

METHODS

Study DesignThe present study consisted of 2 parts: in the first we comparedautonomic function in patients with mild (NYHA functional classII) CHF, with their age- and sex-matched controls. In the secondpart, we conducted a randomized, double-blind, parallel groupcomparison of metoprolol and placebo. The drug intervention studyconsisted of a single-blind placebo run-in period of 1 week, and adouble-blind treatment period of 26 weeks. The study was approvedby the local ethics committee and conforms with the principlesoutlined in the Declaration of Helsinki (British Medical Journal1964;ii:177).

Prior to the start of the double-blind treatment protocol, severityof CHF and clinical stability were confirmed by evaluation of CHFsymptoms, and by a screening treadmill exercise test, including peakVO

2 measurement. At baseline, peak VO

2 testing was repeated, and

this test was taken as baseline value. At baseline, blood samples forplasma norepinephrine concentrations were taken for patients andhealthy controls. In addition, a 24-hour ambulatory electrocardiogramwas recorded for analysis of HR variability, and non-invasiveautonomic function testing (Ewing battery) 11 was performed in both

BETABLOCKERS AND AUTONOMIC IMPAIRMENT IN CHF

89

groups. Subsequent visits were planned at 2, 4, 6, 10, 18, and 26weeks. Patients were started on 25 mg metoprolol sustained release(SR); the first dose was given in the out-patient clinic, and HR andblood pressure were monitored for 2 hours. After 2 weeks, the dosewas doubled to 50 mg SR, followed by an increase to 100 mg at 4weeks, and to 200 mg metoprolol after 6 weeks. The dose was notincreased if systolic blood pressure was < 90 mm Hg, heart rate was< 50 bpm, or clinical signs and symptoms did not allow furtheruptitration.

After 10 weeks of treatment and at the end of the study (26weeks) all measurements of the baseline visit were repeated. Progressionof CHF requiring hospitalization, symptomatic hypotension requiringdiscontinuation, or intolerable side effects were predefined withdrawalcriteria during the study period.

PatientsPatients > 18 years, with clinically stable, mild CHF (New York HeartAssociation functional class II), due to old myocardial infarction (> 3months ago) were eligible for the study. Patients had to be clinicallystable on oral CHF medication, which had to include an ACEinhibitor, for at least 3 months. Concomitant use of digoxin or calciumantagonists was not allowed. Left ventricular ejection fraction had tobe < 0.40, and peak VO

2 < 80% of the age predicted value 12 with a

maximal upper limit of 25 ml/min/kg. Exercise tolerance had to belimited by fatigue or dyspnea, and patients had to be in sinus rhythm.Exercise limiting angina pectoris, and hemodynamically significantvalvular dysfunction, or any other contraindication for β-blockertreatment (including chronic obstructive pulmonary disease, brady-cardia [< 60 bpm], sick sinus syndrome, 2nd or 3d grade AV-block,hypotension [systolic pressure < 110 mmHg]) were exclusion criteria.

For comparison of autonomic function in the patients with CHF,healthy subjects were also studied. No abnormalities were allowed onroutine physical examination, standard electrocardiography andexercise testing. Within 1 month before participation, use of drugswas not permitted.

Exercise TestingPeak VO

2 was determined during symptom limited treadmill exercise

testing, using the modified Naughton protocol, as previously describedin detail. 13 Oxygen consumption, carbon dioxide production andthe respiratory exchange ratio were measured continuously duringthe test using an automated gas exchange measuring system(Sensormedics system 2900, Sensormedics Corp., Anaheim,California). Patients were familiar with treadmill exercise testing andwere encouraged to continue exercise until symptoms forced them tostop, and the gas exchange anaerobic threshold and a respiratoryexchange ratio > 1.0 were reached. Peak VO

2 was calculated as the

mean of the oxygen consumption values obtained during the lastminute of exercise.

Plasma NorepinephrineFor the purpose of neurohormonal blood sampling, an intravenous,indwelling antecubital cannula was inserted. After 30 minutes ofsupine rest blood samples for resting plasma norepinephrine weredrawn. Blood specimens were centrifuged immediately and the plasmawas separated. Plasma norepinephrine was measured using high-performance liquid chromatography with electrochemical detection.14

HRV AnalysisAmbulatory 24 hours ECGs were recorded using Marquette 3 channelAM recorders (8500 series, Laser System, Marquette Electronics Inc.,Milwaukee, Wisconsin) HR variability was analyzed with a Holteranalysis system (Marquette Series 8000) by one single analyst.Recordings with more than 15% of noise or ectopic beats wereexcluded from the HR variability analysis. After classification of theQRS morphology, both time domain and frequency domain HRvariability parameters were calculated, employing only normal tonormal intervals. Time domain HR variability parameters includedmean RR interval (mean NN), standard deviation of mean RR interval(SDNN) and the root mean square of successive difference (rMSSD).These parameters are considered to be mainly under vagal control. 15

The average value of the interval series was subtracted before spectralanalysis was performed using a discrete Fourier transformation algo-rithm. Frequency domain parameters included total power (TP), lowfrequency power (LF) and high frequency power (HF). The lowfrequency component is influenced by sympathetic and para-sympathetic control mechanisms, whereas the high frequencycomponent is almost exclusively under vagal control. The power ofLF and HF was computed in both absolute (ms2) and normalizedunits (nu). Finally, LF/HF ratio, was calculated, which is considereda measure of sympathovagal balance. 15

Autonomic Function TestingA series of autonomic function tests (sympathetic, parasympatheticor combined), based on cardiovascular reflexes, the so-called Ewingbattery 11 was performed as described earlier in detail. 16 Before thetests no caffeine-containing beverages/foods or tobacco was permitted.

Deep breathing (parasympathetic): The maximum heart rate duringeach breath during deep inspiration and expiration six times in oneminute was recorded. The mean of six breaths was used to calculatethe ratio between in- and expiratory heart rate.

Valsalva manoeuvre (parasympathetic): The ratio between thehighest heart rate during 15 seconds Valsalva manoeuvre, and thelowest heart rate afterwards during normal breathing. This manoeu-

APPENDIX 9

90

Table 1 Baseline characteristics of healthy controls and patients with mild heart failure (metoprolol vs. placebo).

Healthy controls CHF group

Metoprolol Placebo

Age (years) 57 (8) 61 (8) 54 (8)

Sex (male/female) 18/6 9/3 9/3

Localization prior myocardial infarctionAnterior/inferior-posterior 9/3 6/6Time since myocardial infarction [months] 41 (15) 49 (14)

Concomitant medicationAngiotensin converting enzyme inhibitors 12 12Diuretics 6 7Vasodilators 4 2

Blood pressure (mmHg)Systolic 133 (13) 134 (13)Diastolic 77(6) 84 (5)

Heart rate (beats per minute) 78 (11) 74 (11)

Left ventricular ejection fraction 0.28 (0.06) 0.27 (0.06)

Exercise time (seconds) 1128 (186) 887 (202) 857 (239)

Peak VO2Absolute value(ml/min/kg) 37.5 (8.1) 19.5 (3.6) 21.9 (2.9)% of normal 103.3 (6) 73 (13) 73 (10)

Resting plasma norepinephrine (pg/ml) 223 (97) 452 (193) 492 (293)

Values are presented as mean (SD), or number of subjects, as appropriate.

Table 2 HRV parameters for patients (at baseline and after 26 weeks treatment with metoprolol or placebo) and healthy controls.

Healthy Controls CHF- all patients

6 months treatment

1

Parameter Baseline Placebo Metoprolol

Time Domain ParametersMean NN [ms] 856 (18) 786 (21) ** + 24 + 139 *SDNN [ms] 157 (8) 137 (6.5) * + 5 -2RMSSD [ms] 31.3 (2.6) 8.0 (2.2) * + 0.6 + 2.9 **

Frequency Domain ParametersTP [ms2] 4107 (494) 2726 (257) +415 + 991LF [ms2] 957 (101) 508 (76) ** -36 + 112 **HF [ms2] 365 (66) 173 (21) ** -25 + 48 **LFnu [%] 75 (2) 69 (2) ** + 1 - 4 **HFnu [%] 25 (8) 30 (3) ** 0 + 4 **LF/HF ratio 4.6 (0.4) 3.6 (0.4) ** - 0.1 - 1.0 **

Values expressed as mean (SD). 1 changes vs. baseline. * p < 0.005, ** p < 0.05, both CHF vs. healthy controls. Mean NN; mean RR interval, SDNN; standarddeviation of mean RR interval, rMSSD; root mean square of successive difference, TP; total power, LF; low frequency power, HF; high frequency power.

BETABLOCKERS AND AUTONOMIC IMPAIRMENT IN CHF

91

vre was performed twice, and the mean of the two Valsalva ratios wascalculated.

Cold pressor test (sympathetic): The hand of the patient was putin ice water for three minutes. The increase of heart rate and meanarterial blood pressure during this period was calculated.

Mental stress test (sympathetic): Increase of heart frequency andblood pressure during repeated difficult subtractions, keeping thepatient under constant strain.

Isometric handgrip test (combined parasympathetic and sympathetic):The increase of heart rate and diastolic blood pressure during threeminutes handgrip held at 30% of the previously established maxi-mum force of the patient.

Head up tilting (combined parasympathetic and sympathetic):decrease of blood pressure after tilting of the bed to 80 grade uprightposition. The tilting was performed twice, first for two minutes, andthe second time for 10 minutes.

Standing up (combined parasympathetic and sympathetic): Patientswere instructed to stand up from the bed. The 30/15 ratio wascomputed (ratio between the highest and lowest heart rate after stan-ding up), and the difference in systolic blood pressure between stan-ding up and the supine position immediately preceding standing wascomputed.

Statistical AnalysisTo achieve a study power of 80 % power analysis was performed onprimary endpoints (changes in peak VO

2, exercise time, plasma

(nor)epinephrine, HR variability and Ewing battery score). Differencesin the autonomic parameters between patients with mild CHF andthe control group were compared by unpaired t-test. The distributionof all variables at baseline and at the endpoint evaluation was comparedwith analysis of variance and t-test. Statistical differences of p<0.05were considered statistically significant.

RESULTS

Study PopulationBaseline characteristics of the 2 CHF groups (placebo and metoprolol)are presented in Table 1; the 2 groups were generally well-matched,although age was slightly higher in the metoprolol group (p=NS).The age and gender of the overall CHF group and the healthy controlswas identical. Of the 24 CHF patients, who entered the treatmentprotocol, 18 finished the study (metoprolol: n=8, placebo: n=10).One patient in the metoprolol arm died suddenly 24 weeks afterstudy inclusion (no autopsy). Of the other 5 patients, 3 dropped outdue to progressive CHF (metoprolol n=2, both after >10 weeks

treatment, placebo n=1, after 8 weeks study treatment), and onepatient developed skin allergy to the study drug (metoprolol). Inaddition, one patient in the placebo group used prohibited medicationat baseline (haloperidol). All these exclusions were made prior tounblinding of the data. The maximum tolerated dose for all 8 patientstreated with metoprolol was 200 mg/day. After 26 weeks of treatmentheart rate decreased in the metoprolol group (78 (11) vs. 67 (8),p<0.05) without any effect in the placebo group (74(11) vs. 74(9)).Finally, both systolic and diastolic blood pressure were not effectedafter 26 weeks of treatment, although a slight but not significantdecrease was observed in the metoprolol group (133 (13) vs. 126(13)).

Exercise ParametersCompared to the healthy control group, patients had lower exercisetime (882 (221) vs. 1128 (186) seconds) and lower peak VO

2 (17.4

(3.5) vs. 37.5 (8.1)) at baseline (both p<0.05). Metoprolol neitheraffected exercise time, nor peak VO

2 after 10 and 26 weeks of

treatment, compared to placebo (Figure 1). Peak heart rate duringexercise was calculated before and after treatment. Before treatmentno differences for peak heart rate during exercise were observedbetween the placebo and the metoprolol group (140 (18) and 148(23)beats per minute, respectively).

In contrast, peak heart rate was decreased in the metoprolol group(125(21) beats per minute) compared to the placebo group (146(22)beats per minute) after starting treatment, (p<0.05 between bothgroups).

Plasma NorepinephrineAt baseline, resting plasma norepinephrine and epinephrine levelswere elevated in CHF patients, as compared to the healthy controlgroup (472 (230) pg/ml vs. 223 (97) and 189 (45) pg/ml vs. 38 (6),both p <0.05). After 10 and 26 weeks, metoprolol did not affectplasma neurohormonal levels.

HRV AnalysisResults are presented in Table 2 and figure 2. At baseline, HRvariability parameters were generally depressed (p<0.05), as comparedto the healthy control group. Within the CHF group, baseline HRVvariables were well matched between the two treatment groups.

HRV AnalysisAfter 26 weeks of metoprolol treatment, all HR variability parameterswere affected, although this effect did not reach statistical significancefor SDNN and TP. In general, a shift toward higher parasympatheticand lower sympathetic activity was observed.

APPENDIX 9

92

Figure 2. Change in time domain and frequencydomain para-meters of heart rate variability after26 weeks of treatment, expressed as percent ofbaseline values. * p < 0.05, ** p < 0.005, meto-prolol versus placebo group

Figure 1. Changes of peak VO2 (ml/min/kg) andexercise time (minutes) after 26 weeks studytreatment (per protocol analysis). Data are pre-sented as mean ± 95% CI.

Autonomic Function Testing (Ewing Battery)At baseline, of the parasympathetically mediated autonomic tests,heart rate response to deep breathing was significantly depressed inthe CHF group (p<0.05 vs. controls), while of the combinedsympathetically and parasympathetically affected parameters bloodpressure response to head up tilt testing was strongly elevated (p<0.05vs. controls). In the treatment (CHF) study, there were no statisticallysignificant baseline differences between the metoprolol and the placebogroups.

Deep breathing, Valsalva manoeuvre, Mental arrhythmic test,Head up tilting and Standing up test were all unaffected by 26 weeksmetoprolol treatment. Cold pressor test: In metoprolol treatedpatients, the HR increase after 3 minutes was reduced, as compared

to placebo (-12 vs. +1 bpm, p<0.05), after 26 weeks treatment.Isometric handgrip test: After 26 weeks treatment, metoprolol causeda stronger reduction of HR response than placebo (-12 vs. +1 bpm,p<0.05), while blood pressure remained unaffected (-6 vs. -2 mmHg).

DISCUSSION

There is increasing awareness, that drug treatment in patients withCHF must have a favorable, or at least a neutral, effect on autonomicfunction, since autonomic dysfunction is significantly related to theincidence of sudden death. 5, 6, 17 The main findings of the present

BETABLOCKERS AND AUTONOMIC IMPAIRMENT IN CHF

93

study are that, even in patients with only mild CHF, autonomicfunction is already markedly disturbed, but that treatment with theβ-blocker metoprolol improves this disturbance of autonomic tone.Given the fact, that recent large-scale trials with β-blockers in patientswith CHF have shown a marked reduction of all-cause mortality, butalso of the incidence of sudden death, 7-9 the present small mechanisticstudy may provide further insight into possible mechanisms involved.As such, these findings support recent other data, which have examinedthe mechanistic effects of β-blockers in CHF. 10

ACE inhibitors have been shown to markedly reduce morbidityand total mortality in patients with CHF, 18 but the incidence ofsudden death was unaffected. Still, sudden death accounts for a largeproportion of deaths in patients with early CHF. 8 Since sudden deathin these patients is particularly devastating, identification of suchpatients is important. 19 Left ventricular dysfunction, caused bycoronary artery disease remains the most common underlying dis-order in CHF. Therefore, in the present study the effects of metoprololon exercise parameters and measures of autonomic function (serumnorepinephrine, HR variability and autonomic function tests) wereinvestigated in homogeneous population of post-myocardial patientswith stable mild CHF. At baseline, plasma norepinephrine concen-trations and HR variability parameters showed marked signs ofdisturbed autonomic balance, with signs of sympathetic activationand parasympathetic depression. Plasma norepinephrine was nothelpful in this small population to evaluate drug effects. This may berelated to the fact, that a single measure-ment is subject to considerablebias, particularly in patients with mild CHF, and for this reason maybe less sensitive in these patients than the use of a more stable para-meter like HR variability. 6 HR variability parameters were significantlyaffected by metoprolol in this study, and may thus be used to examinedrug effects in these patients. Metoprolol caused a shift in autonomicbalance, as it increased parasympathetic power, and reducedsympathetic power. Both components have been related to a vulnerableautonomic profile, and the changes induced by metoprolol in thisstudy may thus be interpreted as a protective mechanism against theincidence of sudden death. Further, they are in line with earlier stu-dies in CHF, in which similar autonomic effects of β-blockade inpatients with ischemic heart disease were observed. 20-22 Autonomicfunction testing, using the Ewing battery did not provide additionalinformation in this study. Given the fact, that this technique is rathertime-consuming and expensive, the clinical value appears rather smallto assess autonomic changes.

Exercise time and peak VO2 were not improved after 10 and 26

weeks in the present study. This may be related to the fact that thenumber of patients in this study was rather small, and follow-up wasonly 26 weeks. Still, several other studies have also shown less pronoun-ced results of exercise parameters by β-blockade in patients with mild

CHF. 23 However, during long-term treatment, a favorable effect onthe progression of disease, as shown in several studies, will most likelyalso translate into a beneficial effect on clinical parameters, such asexercise capacity. It may thus be speculated, that in the first periodafter treatment initiation, β-blockade is more effective in reducingthe risk profile, than by affecting quality of life in patients with CHF.

ACKNOWLEDGEMENTS

We would like to thank Jaap Haaksma for technical assistance in theanalysis of HR variability data and Pieter J. de Kam for statisticalassistance and Andries Smit for supervising the autonomic functiontest.

References

1. Richards AM, Nicholls MG, Yandle TG et al: Neuroendocrine prediction of leftventricular function and heart failure after acute myocardial infarction. TheChristchurch Cardioendocrine Research Group. Heart 1999;81:114-20

2. Nolan J, Batin PD, Andrews R et al: Prospective study of heart rate variabilityand mortality in chronic heart failure : results of the united kingdom heartfailure evaluation and assessment of risk trial (UK-heart). Circulation1998;98:1510-6

3. Eaton GM, Cody RJ, Nunziata E et al: Early left ventricular dysfunction elicitsactivation of sympathetic drive and attenuation of parasympathetic tone in thepaced canine model of congestive heart failure. Circulation 1995;92:555-61

4. Grassi G, Seravalle G, Cattaneo BM et al: Sympathetic activation and loss ofreflex sympathetic control in mild congestive heart failure. Circulation1995;92:3206-11

5. Barr CS, Naas A, Freeman M et al: QT dispersion and sudden unexpecteddeath in chronic heart failure. Lancet 1994;343:327-9

6. Brouwer J, Van Veldhuisen DJ, Man in ‘t Veld AJ et al: Prognostic value of heartrate variability during long-term follow-up in patients with mild to moderateheart failure. The Dutch Ibopamine Multicenter Trial Study Group. J Am CollCardiol 1996;28:1183-9

7. CIBIS Investigators and Committees: The Cardiac Insufficiency Bisoprolol StudyII (CIBIS-II): a randomised trial. Lancet 1999;353:9-13

8. MERIT-HF study group: Effect of metoprolol CR/XL in chronic heart failure:Metoprolol CR/XL Randomised Intervention Trial in Congestive Heart Failure(MERIT-HF). Lancet 1999;353:2001-7

9. Spargias KS, Hall AS, Greenwood DC et al: beta blocker treatment and otherprognostic variables in patients with clinical evidence of heart failure after acutemyocardial infarction: evidence from the AIRE study. Heart 1999;81:25-32

10. Sanderson JE, Yeung LY, Chan S et al: Effect of beta-blockade on baroreceptorand autonomic function in heart failure. Clin Sci 1999;96:137-46

11. Rodrigues EA, Ewing DJ: Immediate heart rate response to lying down: simpletest for cardiac parasympathetic damage in diabetics. BMJ 1983;287:800

APPENDIX 9

94

12. Bruce RA, Kusumi F, Hosmer D: Maximal oxygen intake and nomographicassessment of functional aerobic impairment in cardiovascular disease. Am HeartJ 1973;85:546-62

13. Patterson JA, Naughton J, Pietras RJ et al: Treadmill exercise in assessment ofthe functional capacity of patients with cardiac disease. Am J Cardiol1972;30:757-62

14. Smedes F, Kraak JC, Poppe H: Simple and fast solvent extraction system forselective and quantitative isolation of adrenaline, noradrenaline and dopaminefrom plasma and urine. J Chromatogr 1982;231:25-39

15. Tjeerdsma G, Meinardi MT, van den Berg MP et al: Early detection ofanthracycline-induced cardiotoxicity in asymptomatic patients with normal leftventricular systolic function; autonomic versus echocardiographic parameters.Heart 1999;81:419-23

16. Imai Y, Abe K, Munakata M et al: Circadian blood pressure variations underdifferent pathophysiological conditions. J Hypertens Suppl 1990;8:S125-S132

17. Packer M: The neurohormonal hypothesis: a theory to explain the mechanismof disease progression in heart failure. J Am Coll Cardiol 1992;20:248-54

18. The SOLVD investigators: Effect of enalapril on mortality and the developmentof heart failure in asymptomatic patients with reduced left ventricular ejectionfractions. N Engl J Med 1992;327:685-91

19. Goldman S, Johnson G, Cohn JN et al: Mechanism of death in heart failure.The Vasodilator-Heart Failure Trials. The V-HeFT VA Cooperative StudiesGroup. Circulation 1993;87:VI24-VI31

20. Tuininga YS, Van Veldhuisen DJ, Brouwer J et al: Heart rate variability in leftventricular dysfunction and heart failure: effects and implications of drugtreatment. Br Heart J 1994;72:509-13

21. Pousset F, Copie X, Lechat P et al: Effects of bisoprolol on heart rate variabilityin heart failure. Am J Cardiol 1996;77:612-7

22. Copie X, Pousset F, Lechat P et al: Effects of beta-blockade with bisoprolol onheart rate variability in advanced heart failure: analysis of scatterplots of R-Rintervals at selected heart rates. Am Heart J 1996;132:369-75

23. MacMahon S, Sharpe N, Doughty R: Randomised,placebo-controlled trial ofcarvedilol in patients with congestive heart failure due to ischaemic heart failure.Lancet 1997;349:375-80

SELECTIVE DOPAMINE AGONISTS IN CHF

95

APPENDIX 9

Autonomic and hemodynamic effects of

a new selective dopamine agonist, CHF1035,

in patients with chronic heart failure

G. Tjeerdsma *, MD, L.M. van Wijk **, MD, G.P. Molhoek ***, MD,F. Boomsma ****, PhD, J. Haaksma *,MSc, D.J. van Veldhuisen *, MD

* Department of Cardiology/Thoraxcenter, University Hospital Groningen, Groningen

**Rejafa Hospital Stadskanaal, Stadskanaal

***Medisch Spectrum Twente Enschede, Enschede

****COEUR/Department of Internal Medicine, University Hospital Dijkzigt, Rotterdam

The Netherlands

Cardiovasc Drugs Ther 2001, in press

APPENDIX 10

96

ABSTRACT

Background Dopamine agonists have been studied in chronic heartfailure, but earlier reports with non-selective compounds demonstratedunfavourable long-term effects. CHF1035 is an orally active, newselective dopamine agonist, primarily activating DA

2- and α

2 receptors,

thereby inhibiting norepinephrine release, which may be beneficialin heart failure.

Methods We conducted a double-blind, placebo-controlledcomparison of CHF1035 (10 mg/day, n=20) and placebo (n=9) inpatients with mild to moderate chronic heart failure (left ventricularejection fraction < 0.45). Patients were clinically stable on diureticsand angiotensin converting enzyme inhibitors. Both acute and chronicassessments were made, including plasma neurohormones and 24-hrHolter monitoring for heart rate variability analysis.

Results CHF1035 was generally well tolerated during the study. After10 days, there were no significant changes between the groupsregarding heart rate and blood pressure. Compared to placebo, plasmanorepinephrine levels decreased on CHF1035, both in the first 4hours and after 10 days (p<0.05 between groups). Other neuro-hormones (natriuretic peptides, renin, aldosteron and endothelin)were not significantly affected. Heart rate variability parametersgenerally increased on CHF1035, but were unaffected by placebo(p<0.05 between groups).

Conclusions Short-term treatment with the selective dopaminergicagonist CHF1035 is well tolerated, reduces plasma norepinephrineconcentrations and increases heart rate variability in mild chronicheart failure.

INTRODUCTION

Neurohormonal activation is one of the hallmarks of chronic heartfailure and it is strongly related to prognosis. 1, 2 Inhibiting neuro-hormonal activation has been suggested to be of benefit in heart failure,3 and this may be accomplished by several classes of drugs, includingbeta-blockers and angiotensin converting enzyme (ACE) inhibitors,but also by dopaminergic agonists. 4-6 In the last 30 years, severaldopaminergic agonists have been developed for the treatment of heartfailure, of which ibopamine was the first to be studied on a largerscale. Ibopamine primarily affects dopamine (DA) receptors (DA

1and DA

2), but at higher doses α- and β-receptors may also get

activated. 7 Although earlier studies suggested that the drug could bebeneficial in heart failure, 7-9 a large mortality study showed thatibopamine was associated with an increased mortality. 10 Althoughthe reason for this adverse long-term effect has remained unknown,it is likely, that an inotropic effect of ibopamine, related to its β-receptor activating properties, has played a role. 11 For this reason,more selective compounds were developed, which more or less ex-clusivly affect DA receptors. 6, 12, 13

CHF1035 is a recently developed, orally active DA agonist, whichprimarily acts by activation of (prejunctional) DA

2- and α

2- receptors.

14-16 After ingestion, the drug is rapidly hydrolysed into its activemetabolite, CHF 1024. Earlier experimental studies showed thatCHF1035 reduces plasma norepinephrine levels and heart rate, 14

which would make it an attractive agent in left ventricular dysfunctionand heart failure. Only limited data are available regarding the clinicalvalue of CHF1035 in heart failure. 17, 18 While only few patients wereinvestigated in these studies, the results showed that CHF1035improved hemodynamics without affecting arrhythmias, but thesedata were only reported in abstract form.

The present study was therefore designed to evaluate the efficacyand safety of CHF1035 in patients with heart failure, who wereclinically stable on treatment with ACE inhibitors and diuretics forheart failure.

METHODS

Study designThis study was a randomised, placebo-controlled parallel-groupcomparison of CHF1035 on neurohormonal activation in patientswith mild to moderate heart failure. The institutional review boardapproved the protocol, and all patients gave written informed consent.After a placebo run-in period of 5 days, patients received double-

SELECTIVE DOPAMINE AGONISTS IN CHF

97

blind treatment for 10 days. The primary objective was to study theeffects of CHF1035 on neurohormonal activation as assessed by severalplasma neurohormones and heart rate variability (HRV) analysis. Thesecondary objective was to study hemodynamic responses after thefirst dose and after repeated administration of CHF1035.

Study populationPatients aged 18-75 years were eligible for enrolment if they had NewYork Heart Association (NYHA) functional class II or III within thethree months before enrolment and were stable for at least 3 weekspreceding the study. Patients received baseline therapy for heart failureincluding ACE inhibitors and a loop diuretic, and had a left ventricularejection fraction < 0.45 documented within the previous 2 months.Treatment with β-blockers, digitalis, anti-arrhythmic agents, α-agonists, angiotensin II antagonists, calcium antagonists and the useof monoamine oxidase inhibitors or dopamine agonists were notallowed. Patients were excluded from the study if they had primarycorrectable valvular heart disease, congenital heart disease, atrialfibrillation/flutter, severe hypo- or hypertension, unstable angina,chronic obstructive pulmonary disease and primary hepatic failure.Other criteria for exclusion were a serum creatinine concentrationhigher than 2.5 mg per decilitre (221 µmol per litre) and a serumpotassium concentration of more than 5.5 mmol per litre. Afterfulfilling all criteria, a single-blind placebo run-in period was startedfor 5 days. After this period patients were randomly assigned to eitherCHF1035 (5 mg twice a day) or placebo for a period of 10 days.Measurements for study outcomes were made before randomisation(baseline) and after 10 days of treatment.

Neurohormonal levelsPlasma levels of norepinephrine, atrial natriuretic peptide, brainnatriuretic peptide, aldosteron, endothelin, renin activity andCHF1035 pharmacokinetic parameters were determined at baselineand 2 and 4 hours after the administration of the study drug (the“acute effects”). This procedure was repeated after 10 days of treatment(the “chronic effects”). All venous blood for determination of plasmaneurohormonal levels was drawn in the morning after 30 minutes ofsupine rest. Transportation and measurements of plasma neuro-hormonal samples were performed, as described earlier in detail.19

Ambulatory electrocardiographic monitoring (HRV and arrhythmias)Ambulatory 24-hour electrocardiographic monitoring was performedusing 3 channel Holter recorders (Marquette series 8500). The ECGwas analysed on Marquette XP analyser (Marquette Electronics, Inc.,Milwaukee, WI, USA) and reviewed by an experienced analyst.Recordings with more than 15 % noise or ectopic beats in total wereexcluded. From the series of RR intervals, time and frequency domain

parameters were calculated according to the guidelines. 20 Time domainparameters included: Mean of all normal to normal RR intervals(AVGNN, ms), standard deviation of all normal to normal RRintervals (SDNN, ms), standard deviation of the averages of normalR-R intervals in all 5-minute segments of a complete 24 hour recording(SDANN, ms), and root-mean-square of successive differences of NNintervals (rMSSD, ms). The latter represents fast, beat-to-beat changesin heart rate, and has been shown to be a measure of cardiac vagalactivity. 20 Frequency domain parameters included: Low frequencies(LF, ms2; 0.04-0.15 Hz), high frequencies (HF, ms2; 0.15- 0.4 Hz)and total power (TP, ms2; 0.0033-0.4 Hz). The low frequencycomponent is influenced by sympathetic and parasympathetic controlmechanisms, whereas the high frequency component is almost ex-clusively under vagal control. 20 Finally, the ratio of LF to HF wascalculated, which is considered a measure of sympathovagal balance. 21

Statistical analysisData are presented as means ± standard deviation (SD). Thirty patientswere to be randomised and the ratio of patients with active medicationto placebo was 2:1. Treatment groups were compared at baseline (endof placebo run-in) using either analysis of variance (ANOVA), or theKruskall-Wallis test or the Cockran-Mantel-Haenszel test, whenappropriate. Statistical analysis was performed using SAS system(version 6.12 [Cary, NC, U.S.A.]). Differences between treatmentgroups were analysed using analysis of variance (ANOVA) and theStudent t test. For not normally distributed variables, the Kruskall-Wallis test and Wilcoxon Sum of Rank test were used. All p valueswere reported for two-tail tests, and an alpha < 0.05 was consideredstatistically significant. Mean values of heart rate variability werecalculated for the entire 24-hour period. Linear regression analysiswas used to test the relation between values of variables of HRV atbaseline and after treatment. Safety analysis included tabulation bytreatment group of type and frequency of all adverse events. To studythe effects on arrhythmias as assessed by 24-hour Holter monitoring,the chi-square test was used to study the effects in each treatmentgroup for categorical variables.

RESULTS

Study population and hemodynamic effectsOf the 30 patients who were found eligible, 1 patient withdrewconsent during the placebo run-in phase. The remaining 29 patientsform the present study population and their baseline characteristicsare shown in table 1. At baseline, there were no significant differencesbetween the two treatment groups, with respect to clinical charac-

APPENDIX 10

98

teristics, plasma neurohormones and variables of HRV. All patientscompleted the study and no serious side effects were noted. Onepatient had some complaints of nausea and dizziness early after drugadministration, possibly related to the study drug due to hypotension.After 10 days, no effects were observed on heart rate (from 70(7 to74(12 in the CHF1035 group and from 68(9 to 67(8 in the placebogroup) and blood pressure between both groups, although a smalldecrease was observed for systolic blood pressure in the CHF1035group (from 123 ± 11 to 119 ± 14 mmHg, p=NS between bothgroups).

Plasma neurohormonesThe chronic effects of CHF1035 on plasma norepinephrine con-centrations are demonstrated in figure 1. During the 10-day studyperiod, plasma norepinephrine concentrations decreased in theCHF1035 group (from 415 ± 180 to 360 ± 151 pg/ml) but not inthe placebo group (from 326 ± 140 to 368 ± 141 pg/ml; p<0.05,between groups). With regard to the acute effects on plasma norepine-phrine levels in the morning (figure 2); an increase was observed inthe placebo group during 4 hours (from 368 ± 141 to 565 ± 234 pg/ml), but this was not observed in the CHF1035 group (from 360 ±151 to 408 ± 186 pg/ml; p<0.05 between groups).

Table 1 Baseline characteristics for patients with heart failure; Placebo vs. CHF1035.

Placebo CHF1035

(n=9) (n=20)

Age (years) 63±9 64±10

Sex (%) male/female 89/11 80/20

Left ventricular ejection fraction 0.34±0.11 0.30±0.10

Duration of heart failure 48±58 45±37

New York Heart Association (%)II / III 75/ 25 85/15

Hemodynamic dataSystolic blood pressure (mmHg) 130±14 123±11Diastolic blood pressure (mmHg) 77±9 77±7Heart rate (beats/min) 68±9 70±7

Plasma NeurohormonesNorepinephrine (pg/ml) 326±140 436± 201Aldosteron ( pg/ml) 81±57 185± 229Renin (ng AngI/ml per h) 40±39 123± 173Atrial natriuretic peptide (pmol/l) 28±19 22± 20Brain natriuretic peptide (pmol/l) 49±74 33± 43Endothelin (pg/ml) 5± 4 6± 5

Heart rate variability parametersAVGNN (ms) 855±27 775±20SDNN (ms) 47±4 44±3SDANN (ms) 115±9 105±6rMSSD (ms) 35±5 27±5TP (ms2) 2848±1195 2249±1245LF (ms2) 413±104 351±65HF (ms2) 198±52 252±60Lfnu (%) 66±5 68±4Hfnu (%) 34±5 32±4LF/HF ratio 3.5±0.7 3.0±0.4

SELECTIVE DOPAMINE AGONISTS IN CHF

99

Figure 1. Change in plasma nore-pinephrine levelsat baseline and after 10 days of treatment inpatients with heart failure treated with placebo orCHF1035. * p<0.05 between treatments.

Figure 2. Change in plasma norepinephrine levelsin the morning during the first 4 hours afterreceiving placebo or CHF1035 in patients withheart failure treated with placebo or CHF1035. *p<0.01 within groups.

Figure 3. Change in time domain parameters inpatients with heart failure treated with CHF1035(White blocks) and placebo (black blocks),expressed as percent of baseline values. AVGNN,the mean of all normal to normal R-R intervalsduring 24 hours; rMSSD, root mean square ofsuccessive difference; SDANN, standard deviationof the averages of normal R-R intervals in all 5-minute segments of a complete 24 hour Holterrecording; SDNN, standard deviation of all normalR-R intervals. *p<0.05 between treatments.

APPENDIX 10

100

With regard to other plasma neurohormones (brain natriureticpeptide, renin, aldosteron and endothelin) no differences between orwithin both treatment groups were observed. Atrial natriuretic peptidewas not significantly affected after 10 days, but decreased after 4 hoursin the CHF1035 group (from 23 ± 25 to 19 ± 17 pmol/l) without aneffect in the placebo group (from 23 ± 23 to 23 ± 20 pmol/l; p<0.05between both groups).

Ambulatory electrocardiographic monitoring (HRV and arrhythmias)HRV analysis. Ambulatory 24-hour electrocardiographic Holterrecordings, were available at baseline and after 10 days in 28 patients(CHF1035 n=20, placebo n=8). The effects of CHF1035 on HRVparameters are shown in figure 3 and 4 for all time and frequencydomain parameters, respectively. Compared to the placebo group,the CHF1035 group showed an increase for all time domainparameters (p<0.05 between groups). Spectral analysis of HRV showedan increase for all frequency domain variables in the CHF1035 groupafter 10 days with an opposite effect in the placebo group (all p<0.05between groups). Further, compared to the placebo group the ratioof low to high frequency power was oppositely affected in theCHF1035 group (figure 4, p<0.05 between groups). Finally, meanheart rate in the first 4 hours after drug intake was unchanged in theplacebo group, but slightly decreased in the CHF1035 group (from80(8 to 72(6, p<0.01, p=NS between groups).

Electrocardiographic monitoring. During 24-hour Holtermonitoring there were large differences between patients. At baselineno significant differences were observed between both study groups.The median (range) number of ventricular premature complexes per24 hours was 586 (3-9687) in the CHF1035 group and 166 (18-

1393) in the placebo group. The median number of ventriculartachycardias was 2 (1-139) in the CHF1035 group and 2 (1-2) in theplacebo group and the median rate of ventricular tachycardia was156 (124-232) beats per minute for the CHF1035 group and 154(70-172) beats per minute for the placebo group. Finally, after 10days no significant differences were observed for all parameters bothbetween and within the CHF1035 and the placebo group.

DISCUSSION

The main finding of the present, short-term exploratory study is,that the selective oral dopamine agonist, CHF1035, reduces plasmanorepinephrine concentrations and increases autonomic function, asassessed by HRV analysis in patients with heart failure. In addition,the drug was well tolerated during short-term use in this study.

It is generally accepted that the activation of neurohormonalsystems plays a key role in heart failure, 22 and drugs that possesneurohormonal inhibiting properties have gained increasing attention.Dopamine agonists might theoretically have a role in this respect,but so far none of these compounds has gained a place in the treatmentof heart failure. There are no previous publications on the effects ofCHF1035 on autonomic parameters like HRV and plasmaneurohormones in patients with heart failure. In the present study,treatment with CHF1035 improved HRV parameters in patients with

Figure 4. Change in frequency domain parametersin patients with heart failure treated with CHF1035(White blocks) and placebo (black blocks),expressed as percent of baseline values. HF, Highfrequency power; LF, Low frequency power; TP,Total power. *p<0.05 between treatments.

SELECTIVE DOPAMINE AGONISTS IN CHF

101

mild to moderate heart failure, already treated with ACE inhibitors.More specifically, vagal activity was improved and sympathetic activitywas decreased, as reflected by an increased rMSSD and high frequencypower and a reduction of the ratio of low to high frequency power.This effect resembles that of other neurohormonal antagonists, likeACE-inhibitors and β-blockers, which have also been shown toimprove autonomic imbalance. 23-25 Earlier clinical studies with otherdopamine agonists, in particular ibopamine, showed that plasmanorepinephrine may decrease, 26-28 while small increases in HRV wereobserved advanced heart failure, 29 but not in those with mild heartfailure. 30 In the present study, plasma norepinephrine decreased after10 days on CHF1035. Also, the well-known increase of plasmanorepinephrine in the morning hours, 31 as was observed in the placebogroup in the present study, was blunted by CHF1035, again suggestinga sympatholytic effect.

CHF1035 did not consistently affect other neurohormonesduring the 10 day study period in the present study. Atrial natriureticpeptide levels were not significantly affected after 10 days, but theydecreased 2 after 4 hours after CHF1035, suggesting a directunloading effect of the drug. In an earlier study, Fontana et al 32 showedthat after intravenous dopamine infusion, atrial natriuretic peptidelevels increased in healthy volunteers, while they decreased in patientswith heart failure, suggesting a more direct (receptor-mediated?)relation between DA receptor activation and release of atrial natriureticpeptides.

Dopamine and its related compounds have long been suggestedto be of potential value in heart failure. 27, 33 While the intravenousmother drug dopamine is still used on a large scale in the intensivecare setting, none of the oral agonists has been licensed for chronicuse in heart failure. In contrast to ibopamine, CHF 1035 does nothave β-agonistic or α-1 activity, and it may thus be assumed that ithas no (potential) inotropic properties. For this reason, it may have amore promising pharmacological profile in patients with heart failure.

The number of patients included in this study was small, andconclusions should therefore be drawn with caution. Furthermore,in the present study the number of patients receiving active drug orplacebo was not equal (but 2:1, respectively), which could effect theresults in the present small-size study. Plasma norepinephrine levelsdecreased after 10 days of treatment with CHF1035, as compared tothe placebo group. However, it should be emphasised that plasmanorepinephrine levels are susceptible to spontaneous fluctuations,which might have affected the results of the present study. Althoughit is generally accepted that chronically elevated stimulation of thecardiac β-adrenergic system (by catecholamines) is toxic to the heartand that such stimulation may contribute to the pathogenesis of heartfailure. Whether blockade of this neurohormonal inhibition leads toa long-term beneficial effect is, however, uncertain and further

controlled studies with CHF1035 on this subject will be needed.Indeed, improvement of plasma norepinephrine and HRV, as observedin this study, is no guarantee for long-term success in heart failure, aswas recently shown for moxonidine. 34

In conclusion, the present exploratory study shows that the newselective new dopaminergic agonist CHF1035 reduces plasmanorepinephrine concentrations and increases HRV in patients withmild to moderate heart failure, while it was generally well-tolerated.Whether this may translate into a potentially beneficial long-termeffect needs further study.

References

1. Cohn JN, Levine TB, Olivari MT et al: Plasma norepinephrine as a guide toprognosis in patients with chronic congestive heart failure. N Engl J Med1984;311:819-23.

2. Swedberg K, Kjekshus J, Snapinn S: Long-term survival in severe heart failurein patients treated with enalapril. Ten year follow-up of CONSENSUS I. EurHeart J 1999;20:136-9.

3. Packer M: The neurohormonal hypothesis: a theory to explain the mechanismof disease progression in heart failure. J Am Coll Cardiol 1992;20:248-54.

4. Goldberg LI: Cardiovascular and renal actions of dopamine: potential clinicalapplications. Pharmacol Rev 1972;24:1-29.

5. Frishman WH, Hotchkiss H: Selective and nonselective dopamine receptoragonists: an innovative approach to cardiovascular disease treatment. Am HeartJ 1996;132:861-70.

6. Pouleur H, Raigoso J, Rousseau MF: Dopaminergic drugs in the managementof chronic heart failure. Eur Heart J 1991;12:C29-C34

7. Henwood JM, Todd PA: Ibopamine. A preliminary review of itspharmacodynamic and pharmacokinetic properties and therapeutic efficacy.Drugs 1988;36:11-31.

8. Dohmen HJ, Dunselman PH, Poole-Wilson PA: Comparison of captopril andibopamine in mild to moderate heart failure. Heart 1997;78:285-90.

9. Van der Ent M, van den Heuvel AFM, Remme WJ: Neurohumoral response tocarmoxirole, a selective dopamine (D2) receptor agonist, in patients with chronicmoderate heart failure. Cardiovasc Drugs Ther 1998;12:387-94.

10. Hampton JR, Van Veldhuisen DJ, Kleber FX et al: Randomised study of effectof ibopamine on survival in patients with advanced severe heart failure. SecondProspective Randomised Study of Ibopamine on Mortality and Efficacy (PRIMEII) Investigators. Lancet 1997;349:971-7.

11. Niebauer J, Coats AJ: Treating chronic heart failure: time to take stock. Lancet1997;349:966-7.

12. Meyer W, Buhring KU, Steiner K et al: Pharmacokinetics and first clinicalexperiences with an antihypertensive dopamine (DA2) agonist. Eur Heart J1992;13:D121-8

13. Teisman AC, Buikema H, Van Veldhuisen DJ et al: Direct vasodilating effectsof the new dopaminergic agonist Z1046 in human arteries. J CardiovascPharmacol 2000;35:581-5.

14. Latini R, Masson S, Jeremic G et al: Comparative efficacy of a DA2/alpha2agonist and a beta-blocker in reducing adrenergic drive and cardiac fibrosis inan experimental model of left ventricular dysfunction after coronary arteryocclusion. J Cardiovasc Pharmacol 1998;31:601-8.

APPENDIX 10

102

15. Masson S, Masseroli M, Fiordaliso F et al: Effects of a DA2/alpha2 agonist and abeta1-blocker in com-bination with an ACE inhibitor on adrenergic activity andleft ventricular remodeling in an experi-mental model of left ventricular dysfunctionafter coronary artery occlusion. J Cardiovasc Pharmacol 1999;34:321-6.

16. Puccini P, Zanelli U, Spinabelli D: In vivo metabolism of CHF1035, chiralaminotetraline, in the rat. Pharmacol Rev 1997; 35:209 Abstract.

17. Morisco C, Ricciardelli B, Argenziano L et al: Hemodynamic effects of gradedoral doses of a new dopaminergic analogue CHF 1035 in patients with congestiveheart failure. J Am Coll Cardiol 1995; 25:Abstract.

18. Crippa G, Umile A, Carrara GC et al: First clinical evidence with the new oraldopaminergic CHF 1035 in congestive heart failure. J Heart Failure 1995; 2:590Abstract.

19. Van Veldhuisen DJ, Genth-Zotz S, Brouwer J et al: High- versus low-dose ACEinhibition in chronic heart failure: a double- blind, placebo-controlled study ofimidapril. J Am Coll Cardiol 1998;32:1811-8.

20. Heart rate variability. Standards of measurement, physiological interpretation,and clinical use. Task Force of the European Society of Cardiology and theNorth American Society of Pacing and Electro-physiology. Eur Heart J1996;17:354-81.

21. Pagani M, Lombardi F, Guzzetti S et al: Power spectral analysis of heart rate andarterial pressure variabilities as a marker of sympatho-vagal interaction in manand conscious dog. Circ Res 1986;59:178-93.

22. Francis GS: Neurohumoral activation and progression of heart failure:hypothetical and clinical considerations. J Cardiovasc Pharmacol 1998;32:S16-S21

23. Binkley PF, Haas GJ, Starling RC et al: Sustained augmentation ofparasympathetic tone with angiotensin-converting enzyme inhibition in patientswith congestive heart failure. J Am Coll Cardiol 1993;21:655-61.

24. Copie X, Pousset F, Lechat P et al: Effects of beta-blockade with bisoprolol onheart rate variability in advanced heart failure: analysis of scatterplots of R-Rintervals at selected heart rates. Am Heart J 1996;132:369-75.

25. Coumel P, Hermida JS, Wennerblom B et al: Heart rate variability in leftventricular hypertrophy and heart failure, and the effects of beta-blockade. Anon-spectral analysis of heart rate variability in the frequency domain and in thetime domain. Eur Heart J 1991;12:412-22.

26. Rousseau MF, Raigoso J, van Eyll C et al: Effects of intravenous epinineadministration on left ventricular systolic performance, coronary hemodynamics,and circulating catecholamines in patients with heart failure. J CardiovascPharmacol 1992;19:155-62.

27. Rajfer SI, Davis FR: Role of dopamine receptors and the utility of dopamineagonists in heart failure. Circulation 1990;82:I97-202.

28. Van Veldhuisen DJ, Brouwer J, Man in ‘t Veld AJ et al: Progression of milduntreated heart failure during six months follow-up and clinical andneurohumoral effects of ibopamine and digoxin as monotherapy. DIMT StudyGroup. Dutch Ibopamine Multicenter Trial. Am J Cardiol 1995;75:796-800.

29. Szabo BM, Van Veldhuisen DJ, van der Burgh PH, et al. Clinical and autonomiceffects of ibopamine as adjunct to angiotensin-converting enzyme inhibitors inchronic heart failure. J Card Fail 1996;2:185-92.

30. Brouwer J, Van Veldhuisen DJ, Man in ‘t Veld AJ, et al. Prognostic value ofheart rate variability during long-term follow-up in patients with mild tomoderate heart failure. The Dutch Ibopamine Multicenter Trial Study Group. JAm Coll Cardiol 1996;28:1183-9.

31. Bleske BE, Chung HS, DiMagno M et al: Diurnal variation in plasmanorepinephrine in patients with heart failure. Pharmacotherapy 1999;19:984-8.

32. Fontana F, Bernardi P, Ruffini M et al: Atrial natriuretic factor after dopamineinfusion in healthy subjects and in congestive heart failure. Eur Heart J1991;12:803-6.

33. Horn PT, Murphy MB: New dopamine receptor agonists in heart failure andhypertension. Implications for future therapy. Drugs 1990;40:487-92.

34. Coats AJ: Heart Failure 99 -- the MOXCON story. Int J Cardiol 1999;71:109-11.