Left ventricular diastolic function and cardiac diasease · diastolic left ventricular dysfunction...

Left ventricular diastolic function andcardiac diasease

A radionuclide angiography study

Rijksuniversiteit Groningen

LEFT VENTRICULAR DIASTOLIC FUNCTION ANDCARDIAC DISEASE

A radionuclide angiography study

PROEFSCHRIFT

ter verkrijging van het doctoraat in deMedische Wetenschappen

aan de Rijksuniversiteit Groningenop gezag van de

Rector Magnificus, Dr. D.F.J. Bosscherin het openbaar te verdedigen op

woensdag 10 mei 2000om 14.15 uur

door

HARM JANS MUNTINGA

geboren op 17 mei 1965te Groningen

Promotores: Prof. Dr. H.J.G.M. CrijnsProf. Dr. E.E. van der Wall

Co-promotor: Dr. P.K. Blanksma

Referent: Dr. M.G. Niemeyer

Beoordelingscommissie: Prof. Dr. W.H. Van GilstProf. Dr. F.L. MeijlerProf. Dr. T. Van der Werf

Paranimfen: G.T. Bult-MuntingaA.E. Tuinenburg

The work described in this thesis has been carried out at the departments ofCardiology and Nuclear Medicine of the Martini Hospital and the department ofCardiology of the University Hospital in Groningen.

© Copyright 2000 by H.J. Muntinga. All rights reserved. No part of this publicationmay be reproduced, stored in a retrieval system, or transmitted in any form or by anymeans, electronically, mechanically, by photocopying, recording, or otherwise,without the written permission of the author.

ISBN 90-367-1178-9

Druk: Drukkerij van Denderen B.V., Groningen, NL

Design and Layout: H.J. MuntingaFront cover showing a heart threatened with thistles by stupidity, with thorns bygenius and with blossoms by love (reproduced from Alfons Mucha) on a backgroundof RNA derived pictures of the heart showing one cardiac cycle.

Financial support for the publication and realisation of this thesis by WCN (WorkingGroup on Cardiovascular Research The Netherlands), Martini Ziekenhuis, StichtingEdu Cardio Groningen, Roche Nederland B.V., Sanofi~Synthelabo, Byk nederlandB.V., Merck Sharp & Dohme B.V., AstraZeneca, Servier Nederland B.V., andNovartis Pharma B.V. is gratefully acknowledged.

Ter nagedachtenis aan mijn vaderAan mijn moeder

Aan Monique, Sanne en Jacco

ContentsChapter 1 General introduction and aim of the thesis 1

Chapter 2 Diastolic function: physiology, methods and clinicalsignificance. Background of the present thesis

5

Chapter 3 Normal values and reproducibility of left ventricular fillingparameters by radionuclide angiographyInternational Journal of Cardiac Imaging 1997;13:165-171Editorial comment by M. PillayInternational Journal of Cardiac Imaging 1997;13:173

30

39

Chapter 4 Quantification of the atrial contribution to diastolic filling duringradionuclide angiographyNuclear Medicine Communications 1997;18:642-647

41

Chapter 5 Circadian rhythm in left ventricular relaxation of patients withcongestive heart failure: diagnostic and therapeuticimplicationsEuropean Journal of Internal Medicine 1998;9:91-97

50

Chapter 6 Effect of mibefradil on left ventricular diastolic function inpatients with congestive heart failureJournal of Cardiovascular Pharmacology 1996;27:652-656

60

Chapter 7 Left ventricular beat-to-beat performance in atrial fibrillation:dependence on contractility, preload and afterloadHeart 1999;82:575-580Letter to the editor by Professor Mark I.M. Noble and responseAccepted for publication in Heart

69

83

Chapter 8 Left ventricular diastolic function after cardioversion of chronicatrial fibrillationSubmitted

87

Chapter 9 Summary and conclusionsSamenvatting en conclusies

100

2 Chapter 1

often believed to be present in patients with heart failure in whom no systolic dysfunctioncan be found (Cohn 1990, Goldsmith 1993). Therapy depends on the type and phase ofthe disease and the underlying mechanisms (Brutsaert 1993). It is therefore of clinicalimportance to establish criteria for the diagnosis of diastolic heart failure.

To date, diastolic left ventricular heart failure is diagnosed in patients withevidence of congestive heart failure by excluding a systolic cause and where possibleincreasing the plausibility of presence of diastolic dysfunction by various diagnostictests. Many diagnostic tools are available for the diagnosis of diastolic dysfunction,whether primary or secondary. These tests can make the presence or absence ofdiastolic dysfunction more likely. None of the tests however can diagnose diastolicdysfunction unambiguously. The European study group on diastolic heart failure recentlyproposed guidelines for the diagnosis of diastolic heart failure (European study groupon diastolic heart failure 1998). In order to diagnose diastolic heart failure, signs orsymptoms of congestive heart failure together with normal or mildly abnormal systolicfunction, and evidence of abnormal left ventricular diastolic function must simultaneouslybe present. Independent predictive values of each technique and each index for thediagnosis of diastolic heart failure are however not yet available.

Radionuclide angiography of left ventricular systolic function at rest and duringexercise has proven to be of great value for diagnosis and prognosis in patients withcoronary artery disease, valvular heart disease, and congestive heart failure (Bonow1991). Although left ventricular ejection fraction is the most important variable derivedfrom radionuclide angiography, numerous variables describing diastolic left ventricularfunction may also be obtained. They may provide clinically relevant additionalinformation in selected patients, and give insight into pathophysiologic processes ofvarious cardiac diseases (Bonow 1991).

AIM OF THE THESISThe present thesis was conducted in order to outline the value of radionuclideangiography in diagnostic testing and follow-up of diastolic left ventricular function, andto provide insight in the role of diastolic dysfunction of the left ventricle in thepathophysiology of various cardiac disorders.

In Chapter 2 an introduction is given on “diastology” and its role in cardiacdisease. A variety of diagnostic tools which can be used to assess diastolic leftventricular dysfunction will be summarised. Also the therapeutic implications of diastolicdysfunction will be discussed.

Chapter 3 deals with the normal values and reproducibility of the radionuclideangiography derived diastolic function parameters which are used in the nuclearlaboratory of the Martini Hospital in Groningen. These findings are discussed and acomparison with other studies is made.

In chapter 4 a specific problem with regard to this technique, the quantification ofthe atrial contribution to diastolic filling is addressed. The moment of onset of atrial

General introduction and aim of the thesis 3

contraction is mostly derived from the left ventricular volume curve. No external referencepoint for the onset of atrial contraction, e.g. the P wave on the electrocardiogram, isused. We investigated whether the use of either the left ventricular volume curve or the Pwave on the electrocardiogram as starting-point for the atrial contribution phase led todifferent estimations of the atrial contribution to diastolic filling.

In chapter 5 the unexpected observation of diurnal variation of left ventriculardiastolic function in patients with congestive heart failure and decreased left ventricularejection fraction is discussed. This diurnal variation could have important implicationsfor the timing of diagnostic testing of diastolic left ventricular function in patients withheart failure and left ventricular systolic dysfunction. The diurnal variation of diastolic leftventricular function could play a role in the circadian rhythm of the onset of acutecardiogenic pulmonary oedema in these patients (Kitzis 1999). Treatment regimensshould account for increased risk of pulmonary oedema early in the morning.

In chapter 6 the effect of the calcium antagonist mibefradil on left ventriculardiastolic function is tested in patients with congestive heart failure and decreased leftventricular ejection fraction. Although the administration of calcium antagonists inpatients with decreased systolic function is potentially hazardous because of theirnegative inotropic action, some representatives of this group are (practically) safe in thisrespect. Because some investigators described a beneficial effect of calciumantagonists on diastolic function in patients with depressed systolic function, it was ofinterest to test this hypothesis with mibefradil. During this investigation mibefradil did notaffect diastolic function.

Our haemodynamic findings in patients with atrial fibrillation are discussed inchapters 7 and 8. The beat-to-beat variation of haemodynamics is the point of interestin chapter 7. In this study we used a nuclear stethoscope to collect data of the leftventricular time activity curve in order to assess independent determinants of beat-to-beat variations in left ventricular performance during atrial fibrillation. The short-term andlong-term left ventricular diastolic function after electrocardioversion in patients withchronic atrial fibrillation is discussed in chapter 8.

In chapter 9 the results of this thesis are summarised.

References

1. Kitzman DW, Higginbotham MB, Cobb FR, Sheikh KH, Sullivan MJ. Exercise intolerance inpatients with heart failure and preserved left ventricular systolic function: failure of the Frank-Starling mechanism. J Am Coll Cardiol 1991;17:1065-72.

2. Bonow RO, Udelson JE. Left ventricular diastolic dysfunction as a cause of congestive heartfailure. Mechanisms and management. Ann Intern Med 1992;117:502-510.

3. Vasan RS, Benjamin EJ, Levy D. Prevalence, clinical features and prognosis of diastolic heartfailure: an epidemiologic perspective. J Am Coll Cardiol 1995;26:1565-1574.

4. Grossman W. Diastolic dysfunction in congestive heart failure. N Engl J Med 1991;325:1557-64.

Chapter 1General introductionand aim of the thesis

he clinical syndrome of congestive heart failure has become a major cause ofmorbidity and mortality. Physiologists and clinicians distinguish many causes andmanifestations. In the past decades the importance of the diastolic function of the

heart was recognised in the genesis of this syndrome. Although the prevalence ofdiastolic left ventricular dysfunction as a manifestation of congestive heart failure varieswidely between the majority of studies, it is generally estimated between 30 and 40%(Kitzman 1991, Bonow 1992, Vasan 1995).

Diastolic dysfunction is the result of a diversity of structural and/or physiologicabnormalities of myocardial relaxation and/or ventricular compliance that increaseresistance to ventricular inflow, e.g. constrictive pericarditis, amyloidosis, hypertrophiccardiomyopathy and, more commonly hypertension and myocardial ischaemia(Grossman 1991). Similar to patients with systolic heart failure, signs and symptoms ofpatients with diastolic heart failure are related to increased pulmonary venous pressureand/or decreased cardiac output, and include (exertional) dyspnoea, fatigue, gallopsounds, lung crepitations and pulmonary oedema, (Goldsmith 1993, Vasan 1996). Inpatients with congestive heart failure and systolic left ventricular dysfunction, diastolicfunction parameters are consistently found to correlate significantly to symptom status(Franciosa 1985, Szlachcic 1985, Rihal 1994, Lapu-Bula 1999). This contrasts with thefinding that the severity of systolic dysfunction, i.e. left ventricular ejection fraction, onlycorrelates weakly with exercise capacity or symptom status (Franciosa 1981, Rihal1994).

Despite the similarity between the clinical symptoms of diastolic and systolicdysfunction, the reported mortality rate for patients with heart failure and predominantlydiastolic dysfunction is considerably lower (Cohn 1990, Vasan 1995). In addition to thestrong correlation between the severity of systolic dysfunction and prognosis in patientswith heart failure with left ventricular systolic dysfunction, the common finding of diastolicdysfunction, a restrictive filling pattern in particular, also appears to correlate withincreased cardiac mortality (Pinamonti 1993, Rihal 1994, Temporelli 1998).

Diastolic heart failure is a common clinical entity with similar signs and symptomsas systolic heart failure, but with different prognosis. Isolated diastolic dysfunction is

T

4 Chapter 1

5. Goldsmith SR, Dick C. Differentiating systolic from diastolic heart failure: pathophysiologic andtherapeutic considerations. Am J Med 1993;95:645-655.

6. Vasan RS, Benjamin EJ, Levy D. Congestive heart failure with normal left ventricular systolicfunction. Arch Intern Med 1996;156:146-157.

7. Franciosa JA, Baker BJ, Seth L. Pulmonary versus systemic hemodynamics in determiningexercise capacity of patients with chronic left ventricular failure. Am Heart J 1985;110:807-813.

8. Szlachcic J, Massie BM, Kramer BL, Topic N, Tubeau J. Correlates and prognostic implicationof exercise capacity in chronic congestive heart failure. Am J Cardiol 1985;55:1037-1042.

9. Rihal CS, Nishimura RA, Hatle LK, Bailey KR, Tajik AJ. Systolic and diastolic dysfunction inpatients with clinical diagnosis of dilated cardiomyopathy. Circulation 1994;90:2772-2779.

10. Lapu-Bula R, Robert A, De Kock M, D’Hondt AM, Detry JM, Melin JA, Vanoverschelde JL.Relation of exercise capacity to left ventricular systolic function and diastolic filling in idiopathicor ischemic dilated cardiomyopathy. Am J Cardiol 1999;83:728-734.

11. Franciosa JA, Park M, Levine TB. Lack of correlation between exercise capacity and indexes ofresting left ventricular performance in heart failure. Am J Cardiol 1981;47:33-39.

12. Cohn JN. Heart failure with normal ejection fraction. The V-HeFT Study. Circulation1990;81(suppl III):III-48-III-53.

13. PinamontiB, Di Lenarda A, Sinagra G, Camerini F, and the Heart Muscle Disease Study Group.Restrictive left ventricular filling pattern in dilated cardiomyopathy assessed by Dopplerechocardiography: clinical, echocardiographic and hemodynamic correlations and prognosticimplications. J Am Coll Cardiol 1993;22:808-815.

14. Temporelli PL, Corra U, Imparato A, Bosimini E, Scapellato F, Giannuzzi P. Reversiblerestrictive left ventricular diastolic filling with optimized oral therapy predicts a more favorableprognosis in patients with chronic heart failure. J Am Coll Cardiol 1998;31:1591-1597.

15. Brutsaert DL, Sys SU, Gillebert TC. Diastolic failure: Pathophysiology and therapeuticimplications. J Am coll Cardiol 1993;22:318-325.

16. European study group on diastolic heart failure. How to diagnose diastolic heart failure. Eur HeartJ 1998;19:990-1003.

17. Bonow RO. Radionuclide angiographic evaluation of left ventricular diastolic function. Circulation1991;84(suppl I):I-208-I-215.

18. Kitzis I, Zeltser D, Kassirere M, Itzcowich I, Weissman Y, Laniodo S, Keren G, Viskin S.Circadian rhythm of acute pulmonary edema. Am J Cardiol 1999,83:448-450.

6 Chapter 2

and Brutsaert 1978).Heart failure can be defined as "the pathophysiological state in which an

abnormality of cardiac function is responsible for failure of the heart to pump blood ata rate commensurate with the requirements of the metabolising tissues, or to do soonly from an elevated filling pressure" (Braunwald 1992). As a consequence, indiastolic failure increased resistance to ventricular filling leads to elevated ventricularfilling pressures or inadequate cardiac output. Accordingly, an increase in pulmonarywedge pressure may lead to symptoms of congestion. It is clear that within thisdefinition many cardiac diseases may result eventually in diastolic failure, includingreduced systolic performance, pericardial and valvular disease (Table 2.1). The maincauses of diastolic failure can be divided into relaxation abnormalities, decreasedcompliance and inappropriately high heart rates (Brutsaert 1993). All three causes cancontribute separately to diastolic failure, but in many conditions they act together.

FIGURE 2.1. Definition of diastoleaccording to the analogy betweenisolated cardiac muscle and intactleft ventricular function. Time tracesof force (f) and length (l) of anafterloaded twitch with physiologicrelaxation sequence aresynchronized with left ventricularpressure (P) and volume (V). In oneheart cycle systole (S) and diastole(D) alternate. Systole is the periodof contraction and relaxation.Diastole is the period between twocontraction-relaxation cycles(Gillebert 1994). In the clinicaldefinition diastole starts with theisovolumetric relaxation (IR) phase,includes the rapid filling phase(RFP), the period of diastasis, andthe atrial contraction phase, andends before the isovolumetriccontraction (IC) phase, (Brutsaert1984).

IC EJECTION IR RFPSUCTION

DIASTASISPASSIVE FILLING

ATRIALCONT

R

S D

S D CLINICAL

DEFINITION

MUSCLE-

PUMP

Diastolic function: focus on the present thesis 7

DIASTOLIC PHASESThe four phases into which diastole is divided (isovolumic relaxation, rapid filling,diastasis and atrial systole) will be discussed briefly (Figure 2.1). Relaxation of the heartis a dynamic process of isovolumic relaxation and early rapid filling. Rapid filling stillcontinues after relaxation has been completed. A relatively small volume portion isshifted into the left ventricle during diastasis. In the atrial contraction phaseintraventricular blood volume increases again.

Isovolumic relaxation. Relaxation is a catecholamine dependent energy consumingprocess in which large portions of adenosine triphosphate (ATP) are used. Thedissociation of actin-myosin crossbridges results from an allosteric action of ATP, i.e. areaction of ATP with a site at the myosin head other than the binding site of actincausing the actomyosin “rigor complex” to dissociate (Figueredo 1993, Apstein 1994).This process is called the "plasticising effect" of ATP. Further relaxation is assured byrapid resequestration of cytosolic calcium in the sarcoplasmic reticulum bysarcoplasmic reticulum calcium ATPase, which is activated by phosphorylation ofphospholamban, a regulatory subunit of the calcium pump of the sarcoplasmicreticulum. The affinity of the calcium receptor site on the troponin-tropomyosin complexfor calcium is decreased by phosphorylation, thereby increasing the rate ofdissociation of calcium from troponin C and enhancing relaxation (Morgan 1991). Thisprocess of relaxation starts in late systole and ends in mid-diastole, causing theintraventricular pressure to decline (Brutsaert 1984).

Left ventricular pressure first falls below the pressure in the aortic root, whichcauses the aortic valve to close. Pressure continues to decline until below left atrialpressure, so that the mitral valve will open, and rapid filling begins. The decline ofpressure in time approximates an exponential curve, but in the non-filling heart the leftventricular pressure frequently reaches a negative asymptote due to elastic recoil (Yellin1986, Yellin 1994). The rate of myocardial relaxation is influenced by severalindependent factors in the intact heart (Brutsaert 1980, Brutsaert 1984): (1) Pre- andafterload, (2) inactivation (which itself is influenced by neurohumoral factors, thecoronary circulation, and the use of drugs), and (3) regional nonuniformity (of load andinactivation). Furthermore, impaired relaxation, whether incomplete or slow, must bediscerned conceptually from prolonged systolic contraction, which is a physiologic andcompensatory situation merely leading to delayed or retarded relaxation (Brutsaert1993). In prolonged contraction, which can be seen in acute and chronic systolicpressure or volume loading, the early phase of hypertrophy and increased contractility,an upward shift of the diastolic part of the pressure-volume relation which is seen inimpaired relaxation, is not observed.

8 Chapter 2

TABLE 2.1. Summary of conditions in which left ventricular diastolic dysfunction may be involved, their

pathophysiology and mechanisms (Vasan 1996, Grossman 1991).

Mechanism Pathophysiology

Hypertension ↓ relaxation • ↑ afterload (↑ contraction load)

• nonuniformity (regional variation in

• ↓ coronary reserve

↓ chamber compliance • ↓ myocardial compliance (fibrosis)

• altered chamber geometry

• ↑ coronary turgor

Coronary artery disease

- Myocardial ischemia ↓ relaxation • retarded inactivation (diastolic calcium


↓ chamber compliance • ↓ myocardial compliance (altered

- Myocardial infarction ↓ relaxation • retarded inactivation

• ↑ contraction load


↓ chamber compliance • regional fibrosis

• ↓ regional myocardial compliance

Valvular heart disease

- Aortic stenosis ↓ relaxation • pressure overload

↓ chamber compliance • concentric hypertrophy

- Mitral stenosis ↑ atrial ventricular pressure • resistance to atrial emptying /

- Aortic regurgitation ↓ relaxation • volume overload

↑ chamber compliance • eccentric hypertrophy but decreased

• altered collagen matrix

- Mitral regurgitation idem idem

Hypertrophic cardiomyopathy ↓ relaxation • nonuniformity (regional variation in

• ↓ coronary reserve

• ↓ relaxation load

• ↑ contraction load (obstruction)

↓ chamber compliance • ↓ myocardial compliance

• altered chamber geometry

Restrictive cardiomyopathy ↓ chamber compliance • ↓ myocardial compliance (deposits,

• deposits (e.g. sarcoid)

• altered collagen matrix

Constrictive pericarditis ↓ end-diastolic chamber • ↓ diastolic capacity

Dilating cardiomyopathy ↓ relaxation • diastolic calcium overload

• retarded inactivation

↓ compliance • altered collagen matrix

• ↑ end-diastolic volume


Rapid filling. When left ventricular pressure falls below left atrial pressure the mitralvalve will open. The left ventricle will then be filled by the blood which is accumulated inthe left atrium in the previous systole. The rate of early left ventricular filling isdetermined uniquely by the atrioventricular pressure gradient and the impedance of themitral valve (Ishida 1986, Yellin 1990, Yellin 1992, Yellin 1994). The atrioventricularpressure difference, in turn, is determined by the active and passive properties(relaxation and compliance) of both left atrium and left ventricle. Thus, factors whichdetermine these properties, like e.g. loading conditions, contractility, and heart rateinfluence the early diastolic filling pattern by means of their influence on the pressuredifference.

In experimental canine models in which the left ventricle was withheld from fillingby end-systolic volume clamping, the left ventricular pressure frequently fell below zero(Yellin 1986, Sabbah 1981). This diastolic suction of the left ventricle is caused by thestorage of potential energy generated by preceding systolic contraction to below the“equilibrium volume”, i.e. the volume that the left ventricle exhibits when there is notransmural pressure (Brecher 1966, Yellin 1994). In an experimental setting in dogs thedegree of pressure negativity after volume clamping was increased when contractilitywas increased (Hori 1982). Evidence was provided by Udelson et al. that elastic recoiland restoring forces are also operative in the intact human heart during β-adrenergicstimulation (Udelson 1990). When, in this experiment, end-systolic volume was furtherreduced below the equilibrium volume, minimal diastolic pressure was also reduced.This probably resulted from an augmentation of internal restoring forces and elasticrecoil, which in turn may result in negative transmural pressure and diastolic suction. So,the fall in left ventricular pressure at the time of opening of the mitral valve is not onlyresulting from left ventricular relaxation, it is also caused by elastic recoil.

The atrioventricular pressure gradient, the driving force of left ventricular filling, isalso determined by left atrial pressure. In experiments by Ishida et al. increased leftatrial pressures by volume loading of mongrel dogs resulted in increased peak rapidfilling rates induced by increased atrioventricular pressure gradients, despitedecreased rate of left ventricular relaxation induced by increased loading conditions(Ishida 1986). Left atrial compliance will also affect the peak atrioventricular pressuredifference (Keren 1985, Suga 1974). In early diastolic filling, blood leaves the left atriumfaster than it is filled by the pulmonary veins, and its pressure will fall (y-descent).

At first, left ventricular pressure will continue to decline despite early leftventricular filling, thus accelerating blood into the left ventricle. With increasing leftventricular volume, left ventricular pressure will rise according to the passive fillingcharacteristics, that are determined by viscoelastic properties of the myocardium,myocardial thickness, and external constraints, e.g. pericardium, right ventricle andlungs (Yellin 1990, Little 1990, Janicki 1990, Little 1995). Once relaxation and elasticrecoil are completed, left ventricular filling will continue because of inertia, i.e. the massof flowing blood (Yellin 1990). Early left ventricular filling rate, i.e. the first derivative of

10 Chapter 2

volume-versus-time curve, will diminish when atrioventricular pressure is reversed(Courtois 1988). This "deceleration" of rapid left ventricular filling is greatly determinedby left ventricular chamber stiffness, so that early diastolic filling deceleration timedecreases as left ventricular chamber stiffness increases (Little 1995). Conversely,early diastolic filling deceleration time may increase in the presence of mitral stenosisas a consequence of increased impedance of the mitral valve with prolongedatrioventricular pressure difference associated with elevated left atrial pressure(Meisner 1991).

Diastasis. This stage of ventricular filling is generally addressed as being the phase ofpassive filling in which the filling rate is slow, the rise in ventricular pressure ismoderate, and the pressures of the left atrium and ventricle have reached an equilibrium(Arrighi 1995). Some authors however observed a mid diastolic inflow peak into the leftventricle, arising from a reestablishment of a positive atrioventricular gradient due to leftatrial filling via the pulmonary veins (Keren 1986, Yellin 1992, Biasucci 1990). Mid-diastolic inflow may contribute more than 20 % of total filling of the left ventricle in dogs,but is absent in the dilated heart and in mitral stenosis (Biasucci 1990, Yellin 1992).

Atrial systole. The last phase of ventricular filling is dominated by the contraction of theatria, which causes atrial pressure to rise again above ventricular pressure and inducea new blood flow into the left ventricle. This phase is influenced by left atrial function,loading, and heart rate (Wang 1995 , Courtois 1994, Udelson 1994). Atrial dysrhythmia,hypertrophy, and dilatation may alter this phase considerably (Rowlands 1967, Bonow1983a, Kono 1992).

ASSESSMENT OF DIASTOLIC FUNCTIONThe generally applied non invasive methods to assess diastolic left ventricular functionare radionuclide angiography (RNA), two-dimensional echocardiography and Dopplerechocardiography. The principles of measuring diastolic function with RNA will bediscussed below, whereas the methodology of the latter two techniques will bediscussed only briefly. Cardiac catheterisation with ventriculography is an invasivemethod of assessing diastolic properties of the heart and will be discussed as well. Thenewer methods of assessing diastolic function, including 2-D colour Doppler, colour M-mode Doppler, and cardiac magnetic resonance imaging, will be mentioned but are notdiscussed because they are not (yet) used in routine clinical practice for the purpose ofassessing diastolic left ventricular function. In all techniques, temporal resolution andfiltering strongly influence measurements of left ventricular volume change rate.

Cardiac catheterisation. With cardiac catheterisation it is possible to collectmeasures of left ventricular function invasively. When focusing on diastolic function, theparameters can be divided into those expressing (passive) compliance, and those


expressing (active) relaxation of the left ventricle (Gaasch 1994). Chamber compliance,the inverse of chamber stiffness can be defined as the instantaneous volume changeper unit change in pressure (dV/dP) and can be calculated when measures of diastolicleft ventricular volume and pressure have been gathered together (Grossman 1986, Vander Werf 1991). Relaxation can be described with help of the left ventricular pressurecurve during isovolumic relaxation. The first derivative of this curve describes the rate ofleft ventricular pressure decline (dP/dt). Although relaxation can be described by peak -dP/dt, this factor is influenced by changes in loading (Grossman 1986). A betterparameter to describe left ventricular relaxation is obtained when the time constant τ ofleft ventricular pressure decline is calculated (Grossman 1986, Van der Werf 1991,Shintani 1994). In slow myocardial relaxation τ may be prolonged and vice versa.

Radionuclide angiography. The assessment of global left ventricular diastolicfunction with radionuclide angiography is derived from the time activity curve of the leftventricle, which closely matches the left ventricular volume curve (Figure 2.2). It thereforerepresents relative volume changes throughout the cardiac cycle. This curve is alsoused for the assessment of global left ventricular ejection fraction. When appropriatedata acquisition has taken place and attention has been paid to the technicalconsiderations (see below) this volume curve may also be used to study left ventricularfilling, which, as outlined before, is dependent on diastolic left ventricular function.Parameters of diastolic function include those expressing the filling rate at a certainmoment (e.g. peak filling rate), the timing of this event (e.g. time to peak filling rate), andrelative filling fractions (e.g. early diastolic filling fraction and atrial contribution todiastolic filling, Udelson 1994).

In the interpretation of diastolic function parameters technical details as well asphysiological variations (the dependency of left ventricular filling parameters on

0

20

40

60

80

100

0,0 0,2 0,4 0,6 0,8 1,0

Time (sec)

Lef

t ve

ntr

icu

lar

volu

me

(% e

nd

-dia

sto

lic v

olu

me) AC

EDFPFR

TPFR

FIGURE 2.2. Time-activity curveobtained from radionuclideangiography after temporalsmoothing with 5 Fourierharmonics. Variables used todescribe the diastolic part of thecurve are the contributions of earlydiastolic filling (EDF) and atrialcontraction (AC) to the fillingvolume; peak filling rate (PFR),measured as the peakinstantaneous slope of earlydiastolic filling; time to peak fillingrate (TPFR) measured from minimalleft ventricular counts or frommaximal left ventricular counts.

)

12 Chapter 2

variables as age, heart rate and ejection fraction, see chapter 3) must be taken intoaccount (Bonow 1991, Udelson 1994). The acquisition of left ventricular gated bloodpool scintigraphy can lead to many technical errors (Wagner 1989). Cycle lengthfluctuations have a detrimental effect on the diastolic part of the left ventricular volumecurve (Hammermeister 1974). It is therefore essential to exclude extrasystolic andpostextrasystolic beats and use an appropriate cycle-length window (Juni 1988). In theMartini Hospital in Groningen count “drop-off” was minimised using a 5% cycle lengthwindow with forward gating (Juni 1988). Further, temporal resolution with a high framingrate is necessary to accurately reflect the instantaneous deflections in the time activitycurve (Udelson 1994). High temporal resolution in the multigated studies of the presentthesis was achieved with a framing rate of 50 frames per second (20 msec per frame).Obviously, high temporal resolution should not lead to a decrease of the count numberper frame, which might lead to an increasing degree of statistical imprecision. We useda minimum of 150,000 counts per frame in the present studies of this thesis. Temporalsmoothing could therefore be achieved with five Fourier harmonics, thereby keeping therisk of systematic underestimation of the filling parameters low (Bacharach 1983,Bonow 1989, Bonow 1991). Because of an unpredictable variation in red cell taggingwith technetium-99m pertechnetate, and a variable amount of absorption betweenpatients, measurements of count changes in the time-activity curve must be related(normalised) to a physiologic variable derived from the time activity curve, e.g. end-diastolic volume and stroke (or filling) volume. Therefore this “normalised” parameter isnot only influenced by the true filling rate (in milliliters/second) but also by thenormalisation parameter. It is therefore recommended to perform normalisation to morethan one parameter (Udelson 1994).

From the obtained time activity curve which closely resembles the left ventricularvolume curve, a number of determinants of left ventricular filling can be taken (Figure2.2). These parameters are used throughout the present thesis. The peak filling rate(PFR in end diastolic volume (EDV) or filling volume (FV) per second) is the mostwidely used parameter of left ventricular filling, and measures the maximuminstantaneous filling rate during early rapid filling. The normal value varies widely, and isdependent on age, heart rate, and ejection fraction (see Chapter 3). The time to peakfilling rate (TPFR) is the time interval from minimal left ventricular counts to the peakfilling rate expressed in milliseconds. In the present thesis this interval is used for TPFR,unless otherwise stated. Time to peak filling rate can alternatively be measured frommaximal left ventricular counts to peak filling rate (Friedman 1986). The duration ofsystole is then included. This parameter therefore concerns the entire contraction-relaxation cycle. This may be of importance since in many clinical conditions of diastolicfailure impaired relaxation and prolonged contraction are present in combinationdepending on aetiology and stage of the disease (Brutsaert 1993).

The atrial contribution to diastolic filling can be assessed by measuring theincrease in relative left ventricular volume that is due to atrial contraction. The moment of


onset of atrial contraction can either be determined from the time-activity curve or one ofits derivatives, or by using the onset of the P wave on the electrocardiogram and adding40 ms for assumed atrial electrical-mechanical delay as starting-point for the atrialcontribution phase in the time-activity curve (see Chapter 4). In chapter 8 the termadditional filling fraction (AFF) is introduced because passive diastolic left ventricularfilling is assumed to be partly responsible for left ventricular filling in late diastole.

Echocardiography. The dynamics of left ventricular filling is studied with Dopplerechocardiography of the mitral inflow pattern, which can be divided in an early (E) andan atrial (A) peak filling velocity (Thomas 1991, Quinones 1991). Deceleration time ofthe peak early transmitral Doppler recording and isovolumic relaxation time can bemeasured with this technique as well. These parameters have been successfullycorrelated with invasive measures of diastolic function, e.g. left ventricular fillingpressures, the time constant of relaxation τ, and pulmonary wedge pressure (Scalia1997, Störk 1989, Yamamoto 1997, Temporelli 1999). An important Dopplerechocardiographic pattern associated with impaired prognosis in patients withdepressed systolic left ventricular function is the restrictive filling pattern (Nijland 1997).This pattern has a high ratio of peak E and peak A, accompanied by a shortdeceleration time. Information on diastolic left ventricular function can be supplementedby Doppler pulmonary venous flow parameters (Gentile 1997). High temporal andvelocity resolution are important advantages of the Doppler echocardiographicexamination of left ventricular diastolic function (Thomas 1994). One of the draw-backsof Doppler echocardiographic measurement of left ventricular diastolic function ishowever its reproducibility (Heesen 1998). For repeated measurements in two-dimensional echocardiography it is therefore advised to have the patients followed byone single investigator (Otterstad 1997).

Studies which compared radionuclide angiographic parameters of left ventricularfilling with Doppler echocardiographic measurements show a good correlation,provided that the parameters are normalised correctly (Friedman 1986, Bowman1988). Many observations on left ventricular filling physiology and pathophysiology byradionuclide angiography are paralleled by similar observations on Dopplerechocardiography of the mitral inflow pattern, like e.g. the findings with ageing,hypertension, and myocardial ischaemia (Mantero 1995, Hoit 1994, Thomas 1994).More recent techniques like e.g. colour kinesis (Vignon 1998) and colour M-mode(Stugaard 1993) still lack comparison with radionuclide angiography but are promising.

Magnetic resonance imaging. With magnetic resonance (MR) imaging it is not onlypossible to collect anatomic data, but also functional data of the heart (Van Rossum1998). These functional data can be obtained using different MR techniques, includingcine MR imaging and myocardial tagging with radiofrequency pulses (Van Rossum1998). With phase velocity mapping it is possible to evaluate parameters of diastolic

14 Chapter 2

left ventricular function, i.e. left ventricular inflow propagation and vortex flow in the leftventricle (Yoganathan 1997). How these parameters relate to specific cardiac diseaseremains to be investigated.

DIASTOLIC DYSFUNCTION IN CARDIAC DISEASENormal systolic function. An increasing number of studies have reported on thepresence of congestive heart failure in patients with normal systolic function (Cohn1990, Bonow 1992, Vasan 1995). The prevalence of normal systolic function amongpatients with congestive heart failure varies widely between 13% to 74%. The majorityof studies suggest a prevalence of about 40% (Vasan 1995). The criteria used foridentification of congestive heart failure, age, and the ratio between acute and chroniccongestive heart failure in the study groups are responsible for the variability of thereported prevalences (Vasan 1995). Patients with congestive heart failure and normalsystolic function at rest are often presumed to have heart failure on the basis of leftventricular diastolic dysfunction (Vasan 1995). However, the presence of diastolicdysfunction needs to be substantiated where possible, although uniform criteria fordiastolic dysfunction are not available and have not been applied (Vasan 1995). Inorder to come to standardisation of diagnostic criteria for diastolic heart failure, theEuropean study group on diastolic heart failure proposed guidelines for the diagnosis ofdiastolic heart failure (European study group on heart failure 1998). According to theseguidelines, diagnosis of diastolic heart failure requires the presence of signs andsymptoms of congestive heart failure, presence of normal or only mildly abnormal leftventricular systolic function, and evidence of abnormal left ventricular relaxation, filling,diastolic distensibility or diastolic stiffness (European study group on heart failure1998). Radionuclide angiographic abnormalities are only described for PFR: < 2.0EDV/s (age < 30 year), PFR < 1.8 EDV/s (age 30 - 50 year), PFR < 1.6 EDV/s (age >50 year) based on a study performed by Bonow et al (Bonow 1988).

In isolated diastolic dysfunction the left ventricle is unable to fill adequately atnormal diastolic pressures (Grossman 1991). Exercise intolerance is one of theimportant early symptoms of patients with diastolic heart failure (Kitzman 1991). In thegenesis of acute pulmonary oedema diastolic dysfunction can play an important role(Grossman 1990). Identification of patients with diastolic left ventricular dysfunction isimportant because pathophysiology and therapy differ from that of patients withprimarily systolic dysfunction.

Congestive heart failure with normal systolic function may be seen in a variety ofdisorders (Table 1; Grossman 1991). These disorders include structural abnormalitiesof the pericardium, and of the myocardium (e.g. amyloidosis, and fibrosis; Kushwaha1997, Janicki 1994). Valvular heart disease (primarily mitral and tricuspid stenosis)may cause elevated atrial pressure without systolic dysfunction. In chronic left ventricularvolume overload, e.g. in aortic and mitral regurgitation, diastolic dysfunction expressedby increased slope of the end-diastolic pressure volume relation may be present without


systolic dysfunction (Grossman 1976). An acute increase of volume load howevercauses elevated diastolic filling pressures without primary myocardial dysfunction(Arrighi 1995). In the presence of right ventricular dilatation, left ventricular diastolicdysfunction may be present due to ventricular interdependence (Janicki 1990). In leftventricular hypertrophy diastolic dysfunction may be present as a result of increasedmyocardial fibrosis and altered myocardial relaxation. Finally, dynamic disordersresulting in relaxation abnormalities, e.g. pressure overload in hypertension with orwithout left ventricular hypertrophy, and myocardial ischaemia may result in diastolicdysfunction.

The prognosis of patients with congestive heart failure and normal systolicfunction is variable and depends on the underlying pathophysiologic mechanismresponsible for the diastolic filling abnormalities. Reported annual mortality rates arelower than that of systolic heart failure, and range between 1.3% and 17.5% dependingon aetiology of diastolic dysfunction and age (Vasan 1995).

The definite diagnosis of diastolic congestive heart failure is based on theevaluation of systolic and diastolic performance showing elevated left ventricular fillingpressures with normal systolic function. A diagnosis of diastolic heart failure is likelywhen non-invasive techniques show abnormalities of the left ventricular filling patternwith normal systolic function (Arrighi 1995).

Depressed systolic function. Signs and symptoms of cardiac failure in patients withdepressed systolic function may partly be due to concomitantly altered diastolicproperties of the myocardium (Grossman 1976). In patients with heart failure anddepressed systolic left ventricular function the severity of systolic dysfunction correlateswith prognosis, but not with exercise capacity or symptom status. (Franciosa 1981,Rihal 1994). Diastolic function parameters are found to correlate significantly tosymptom status in such patients (Franciosa 1985, Szlachcic 1985, Rihal 1994). Anearly sign of diastolic heart failure is diminished exercise tolerance (Packer 1990).Recent studies on chronic tachycardia-induced cardiomyopathy leading to markedsystolic dysfunction demonstrate that this is associated with an impairment of intrinsicmyocardial relaxation (Shinbane 1997, Zile 1996). In patients with left ventricular systolicdysfunction the common finding of diastolic dysfunction, a restrictive filling pattern inparticular, appears in addition to left ventricular ejection fraction to be correlated withincreased cardiac mortality (Pinamonti 1993, Rihal 1994, Nijland 1997).

Coronary artery disease. Myocardial ischaemia may cause transient abnormalities ofleft ventricular diastolic filling (Arrighi 1995). The majority of patients with coronary arterydisease have abnormal diastolic parameters at rest resulting in decreased earlydiastolic filling and increased atrial transport function (Mahmarian 1990). After coronaryartery bypass grafting and percutaneous transluminal coronary angioplastyabnormalities of diastolic filling appear to normalise (Bonow 1982, Lawson 1988).

16 Chapter 2

Impaired relaxation is caused by increased myocardial cytosolic calcium ionconcentrations arising from decreased calcium sequestration or increased calciumentry (Bonow 1990). In patients with coronary artery disease left ventricular asynchronydue to regional ischaemia affects global left ventricular filling (Perrone-Filardi 1992).Regional left ventricular nonuniformity can be present in ischaemia, after myocardialinfarction, but also in other cardiac conditions e.g. hypertrophic cardiomyopathy and thenormal ageing process (Bonow 1990). After myocardial infarction a restrictive fillingpattern is indicative of diastolic dysfunction due to initial myocardial stiffness ormyocardial failure (Algom 1995). Impaired diastolic filling is a constant pathologicalfinding in patients with previous myocardial infarction, and is more severe in patientswith concomitant heart failure (Bareiss 1990). In a recent study with Dopplerechocardiography, a restrictive left ventricular filling pattern was a good predictor ofcardiac death (Nijland 1997).

Tachycardia. Left ventricular diastolic dysfunction may emerge in the presence ofinappropriately high heart rates and chronic tachycardia (Brutsaert 1993, Zile 1996). Inthe presence of chronic tachycardia left ventricular diastolic dysfunction coincides withsystolic abnormalities (Shinbane 1997). In the experimental setting of chronic pacinginduced cardiomyopathy, the nature of diastolic dysfunction has not been made exactlyclear. Generally, in chronic tachycardiomyopathy impaired left ventricular relaxation,increased left ventricular diastolic wall stress, and decreased compliance have beenfound (Komamura 1992, Zile 1995, Ohno 1994). Administration of inotropic agentscould however normalise decreased early left ventricular relaxation, and decreasingloading conditions could normalise increased left ventricular end-diastolic wall stress byreduction of both increased myocardial stiffness and increased time constant ofrelaxation τ (Sasayama 1991, Komamura 1992). After cessation of pacing in chronicpacing induced tachycardiomyopathy the normalisation of left ventricular systolicfunction which may occur, is accompanied by development of left ventricular hypertrophywith persistent diastolic dysfunction, consisting of decreased relaxation and decreasedcompliance (Tomita 1991).

In humans with chronic atrial fibrillation with rapid ventricular response, intrinsictachycardiomypathy characterised by decreased systolic left ventricular functionnormalises after restoration of sinus rhythm or adequate rate control (Morris Jr 1965,Packer 1986, Grogan 1992, Rodriguez 1993). This tachycardiomyopathy may becaused by inadequate high heart rate at rest and/or with exercise (Van den Berg 1993).In chronic atrial fibrillation also the atria itself are dilated (Davies 1972). Most authorshowever agree on (partial) functional recovery of atrial function after chemical andelectrical cardioversion, and the maze procedure (Manning 1989, Jovic 1997, Yashima1997). The increase in Doppler echocardiographic A wave velocity, and decreased E/Aratio after cardioversion of atrial fibrillation may also indicate (transient) impairment ofleft ventricular relaxation (Xiong 1995). New data on this subject are further discussed in


Chapter 8.

Other cardiovascular disease. Impairment of diastolic function is the mostcharacteristic pathophysiologic abnormality in patients with hypertrophiccardiomyopathy, leading to decreased early diastolic filling and increased atrial fillingdue to diminished relaxation and increased chamber stiffness (Hess 1993, Posma1994). Also hypertension and secondary left ventricular hypertrophy are manifested bysuch diastolic abnormalities (Hoit 1994). Patients with cardiac syndrome X also haveimpaired resting left ventricular diastolic filling which improves after the betablockeratenolol (Fragasso 1997). The normal ageing process is generally associated withstructural and functional changes in the heart, leading to decreased early diastolic fillingand increased atrial filling (Nixon 1994). A recent study addressed the changes of leftventricular relaxation with age to presence of coronary artery disease, systemichypertension, left ventricular systolic dysfunction or hypertrophy (Yamakado 1997). Inaddition, restrictive cardiomyopathy including cardiac amyloidosis and sarcoidosis isassociated with impaired ventricular filling (Kushwaha 1997).

THERAPEUTIC CONSEQUENCESIn the choice of a therapeutic regimen for the treatment of congestive heart failure it is ofgreat importance to know whether it originates from systolic dysfunction or diastolicdysfunction or a combination (Bonow 1992). The identification of the underlyingmechanisms responsible for diastolic dysfunction has therapeutic significance as well.Diastolic left ventricular dysfunction may originate from structrural pericardial (e.g.constrictive pericarditis), myocardial changes, (e.g. myocardial fibrosis or amyloidosis),or valvular disease (e.g. mitral stenosis, Covell 1990, Kushwaha 1997). Theseconditions may require specific surgical or medical interventions. Diastolic dysfunctionmay originate from functional changes of the heart as well. Here, diastolic dysfunctionmay be reduced by treatment of the underlying disease, e.g. myocardial ischaemia,blood pressure in case of hypertension, or tachycardia in case of arrhythmia.

Calcium channel blockers in heart failure. The negative inotropic effects of calciumchannel blockers result from depression of calcium transmembrane transport (Iliceto1997, Packer 1989). In the intact hemodynamic system this negative inotropic effect ishowever counterbalanced by peripheral vasodilatation, which reduces left ventricularafterload, but also initiates neurohormonal activation. Despite these factors, verapamiland diltiazem reduce ventricular contractility (Packer 1989, Su 1994, Clozel 1989).Nifedipine however, being a dihydropyridine, increases all contractile indices, probablyresulting from reflex stimulation of the sympathetic system (Ferrari 1997, Iliceto 1997).Second generation dihydropyridines may also exert their positive hemodynamic effectby the same mechanism (Lambert 1990). The improvement in cardiac performanceafter dihydropyridine administration is however less marked than after the

18 Chapter 2

administration of vasodilators that do not depress cardiac contractility (Packer 1989).In ischaemic heart disease, regional systolic dysfunction may lead to global left

ventricular dysfunction by remodelling. Calcium channel blockers can play an importantrole here, by reducing the extent of dysfunctioning myocardium caused by ischaemia,and improving the functional recovery of post-ischaemic but viable myocardium (Iliceto1997). This effect is obtained by coronary vasodilation thereby improving themyocardial oxygen supply and removing harmful by-products of anaerobic glycolysis(Nayler 1994). Calcium channel blockers also have an energy saving effect on themyocardium by means of their peripheral vasodilative action, which reduces afterload,and by means of their direct negative inotropic myocardial effect. In addition,myocardial energy demand is depressed by heart rate reduction after administration ofverapamil or diltiazem (Iliceto 1997, Nayler 1994).

Despite beneficial haemodynamic effects of calcium channel blockers incoronary artery disease, the administration of these drugs in congestive heart failurecan be detrimental (Packer 1989, Elkayam 1993). The unfavourable effects of short-term administration have been attributed to the negative inotropic properties of calciumchannel blockers. In addition, the reflex augmentation of sympathetic activity innifedipine often fails to compensate its direct negative inotropic properties in patientswith congestive heart failure (Elkayam 1990). Long-term treatment with calcium channelblockers may also have deleterious effects which are probably related to the activationof neurohormonal systems, including the sympathetic nervous system and the renin-angiotensin system (Packer 1989, Elkayam 1993). Although second generationcalcium antagonists do not induce significant neurohormonal activation, no data areavailable justifying the use of these drugs in the treatment of heart failure (De Vries1997).

In the treatment of concomitant angina pectoris, the anti-ischaemic effects ofcalcium antagonists may however be useful (The DEFIANT II Research Group 1997).Several approaches have been suggested to improve the safety of calcium channelblockers in patients with heart failure including development of calcium channel blockersthat do not activate the neurohormonal system (Iliceto 1997, Elkayam 1993). In onestudy the short term treatment of patients with heart failure with the T-channel selectivecalcium antagonist mibefradil did not raise sympathetic activity (Van der Vring 1998).Because of serious drug interactions and arrhythmogenic effects the drug was howeverwithdrawn from the market by the manufacturer in 1998.

Calcium channel blockers and myocardial relaxation. In coronary artery diseaseregional and global left ventricular diastolic function may be impaired as a result of'silent' myocardial ischaemia (Mahmarian 1990, Bareiss 1990). Relaxation, being anenergy consuming process, may be improved by reversing myocardial ischaemia e.g.by bypass surgery or percutaneous transluminal coronary angioplasty (Lawson 1988,Bonow 1982). Calcium channel blockers are also capable of increasing coronary blood


flow significantly, either via natural vessels or collaterals or both. In the presence of leftventricular relaxation abnormalities due to myocardial ischaemia, they may thereforeimprove diastolic function (Dienstl 1992, The DEFIANT II Research Group 1997, Hanet1990). In addition, a reduction in afterload of the left ventricle by peripheralvasodilatation may also improve diastolic left ventricular function, as it decreasesmyocardial oxygen demand (The DEFIANT II Research Group 1997). The negativeinotropic effect of calcium channel blockers may decrease myocardial energy demanddirectly, thus decreasing ischaemia, and improving myocardial relaxation. The latterexplanation is more likely in the administration of phenylalkylamines orbenzothiazepines, but not in dihydropyridines for they may enhance myocardialcontractility (Lahiri 1990).

In hypertensive left ventricular hypertrophy and hypertrophic cardiomyopathy, leftventricular filling is often impaired as a result of reduced left ventricular relaxation,reduced subendocardial coronary reserve, a reduction in adenosine triphosphate, orincreased chamber stiffness due to increased myocardial collagen content (Smith1985, Vatner 1990). In addition, regional nonuniformity which arises fromnonhomogeneous distribution of these factors may further impair global left ventricularfilling (Bonow 1990). Indexes of left ventricular relaxation have been shown to beimproved after the administration of nifedipine, diltiazem, and verapamil (Bonow 1990).In addition, in these patients verapamil improved exercise tolerance and diastolicpressure volume relations despite its negative inotropic effect, normalised atrialcontribution to left ventricular filling, and reduced left ventricular asynchrony (Bonow1985, Bonow 1983b, Bonow 1983a, Bonow 1987).

Finally, the structural and functional changes in diastolic function in the normalageing process, may at least in part be reversed by the administration of verapamil(Nixon 1994, Arrighi 1994).

Diuretics. Diuretics effectively reduce pulmonary congestion and oedema. Inhaemodynamic studies with diuretics, filling pressures of both left and right ventriclesare reduced (Udelson 1993). Exercise tolerance can thus be improved with little changein systolic performance (Brutsaert 1993).

Nitrates. A favourable effect on diastolic performance is also noted with nitroprussideand nitroglycerin resulting in diminished filling pressures at smaller chamber volumes(Udelson 1993). However, in diastolic heart failure a drastic reduction of preload mayresult in decreased cardiac output with fatigue and exercise intolerance as aconsequence.

ß-Blocking agents. In systolic heart failure ß-blockers depress left ventricularcontractility and relaxation initially (Eichhorn 1992). Long term treatment of heart failurewith ß-blockers may eventually result in increased contractility and improved early

20 Chapter 2

diastolic relaxation (Udelson 1993, Eichhorn 1990). Although the direct effect of ß-blockers on the myocardium are unfavourable for diastolic function, the favourableeffects on diastolic function result from reduction of heart rate, blood pressure, leftventricular hypertrophy, and left ventricular ischaemia, (Bonow 1992, Fragasso 1997).

ß-Adrenergic drugs. Although the short-term infusion of ß-adrenergic drugs has afavourable effect on early diastolic relaxation, there is no influence on the end-diastolicpressure-volume relation (Udelson 1990). Because in patients with diastolicdysfunction, increased heart rate may be unfavourable, ß-adrenergic drugs must beprescribed with great caution (Smith 1992).

Angiotensin-converting enzyme inhibitors. Angiotensin converting enzymeinhibitors possibly have a favourable effect on diastolic function. They diminish fillingpressures of both left and right ventricles (Udelson 1990). In addition, left ventricularhypertrophy and blood pressure are decreased (Bonow 1992). In patients with dilatedright ventricles, the impeding effect of pericardial constraint and ventricularinterdependence on early diastolic filling is diminished by acute angiotensin convertingenzyme inhibition (Konstam 1990). In patients with severe systolic left ventriculardysfunction long-term therapy with the angiotensin converting enzyme inhibitor enalaprilprevented left ventricular dilatation accompanied by a decreased left ventricularchamber stiffness (Pouleur 1993).

Digitalis glycosides. Because digitalis glycosides increase intracellular calciumconcentration, they possibly have unfavourable effects in diastolic heart failure. Incontrast, possible indirect effects of digitalis including afterload reduction and lowerend-systolic volumes may provide an association with lowered filling pressures(Udelson 1993). In a recent study on the effects of chronic digitalisation in systolic heartfailure a decrease in the rate and degree of ventricular relaxation was confirmed(Hassapoyannes 1998). This effect did however not abolish favourable effects ofdigitalis on mortality.

A general approach of patients with diastolic heart failure in terms of drug treatmentdoes not (yet) belong to present-day practice. Although effects on heart rate,remodelling and relaxation must be taken into account, the aetiology of heart failure andthe type of diastolic dysfunction are often the key to treatment.

In clinical trials no internationally accepted uniform diagnostic criteria are usedfor diastolic heart failure. The lack of uniformity in diagnostic criteria explains whydiastolic heart failure in present day practice often is not recognised and poorlyunderstood. Prognosis of patients with heart failure and diastolic left ventriculardysfunction differs from patients with systolic heart failure. Because of the risingincidence of diastolic dysfunction and heart failure in ageing populations it is of


importance to define normal values and variability in cardiac disease of measures ofdiastolic left ventricular function. Then tailored treatment options for patients withdiastolic heart failure may be tested successfully.

References

1. Algom M, Schlesinger Z. Serial changes in left ventricular diastolic indexes derived from dopplerechocardiography after anterior wall acute myocardial infarction. Am J Cardiol 1995;75:1272-1273.

2. Apstein CS, Morgan JP. Cellular mechanisms underlying left ventricular diastolic dysfunction. In:Gaasch WH, LeWinter MM, ed. Left ventricular diastolic dysfunction and heart failure.Philadelphia: Lea & Febiger, 1994;3-24.

3. Arrighi JA, Dilsizian V, Perrone-Filardi P, Diodati JG, Bacharach SL, Bonow RO. Improvement ofthe age-related impairment in left ventricular diastolic filling with verapamil in the normal humanheart. Circulation 1994;90:213-219.

4. Arrighi JA, Soufer R. Left ventricular diastolic function: physiology, methods of assessment, andclinical significance. J Nucl Cardiol 1995;2:525-543.

5. Bacharach SL, Green MV, Vitale D, White G, Douglas MA, Bonow RO, Larson SM. Optimumfourier filtering of cardiac data: a minimum-error method: concise communication. J Nucl Med1983;24:1176-1184.

6. Bareiss P, Facello A, Constantinesco A, Demangeat JL, Brunot B, Arbogast R, Roul G.Alterations in left ventricular diastolic function in chronic ischemic heart failure. Assessment byradionuclide angiography. Circulation 1990;81(supp III):III-71-III-77.

7. Biasucci LM, Pajaro OE, Tanoue T. Mid-diastolic mitral flow rebound: mechanisms, importanceand relation to Doppler indexes of diastolic function. (Abstract). Circulation 1990;82(suppl III):III-127.

8. Bonow RO, Kent KM, Rosing DR, Lipson LC, Bacharach SL, Green MV, Epstein SE. Improvedleft ventricular diastolic filling in patients with coronary artery disease after percutaneoustransluminal coronary angioplasty. Circulation 1982;66:1159-1167.

9. Bonow RO, Frederick TM, Bacharach SL, Green MV, Goose PW, Maron BJ, Rosing DR. Atrialsystole and left ventricular filling in hypertrophic cardiomyopathy: effect of verapamil. Am JCardiol 1983a;51:1386-1391.

10. Bonow RO, Ostrow HG, Rosing DR, Cannon RO 3d, Lipson LC Maron BJ, Kent KM,BacharachSL, Green MV. Effects of verapamil on left ventricular systolic and diastolic function in patientswith hypertrophic cardiomyopathy: pressure-volume analysis with a nonimaging scintillationprobe. Circulation 1983b;68:1062-1073.

11. Bonow RO, Dilsizian V, Rosing DR, Maron BJ, Bacharach SL, Green MV. Verapamil-inducedimprovement in left ventricular diastolic filling and increased exercise tolerance in patients withhypertrophic cardiomyopathy: short- and long-term effects. Circulation 1985;72:853-864.

12. Bonow RO, Vitale DF, Maron BJ, Bacharach SL, Frederick TM, Green MV. Regional leftventricular asynchrony and impaired global left ventricular filling in hypertrophic cardiomyopathy:effect of verapamil. J Am Coll Cardiol 1987;9:1108-1116

13. Bonow RO, Vitale DF, Bacharach SL, Maron BJ, Green MV. Effects of aging on asynchronousleft ventricular regional function and global ventricular filling in normal human subjects. J Am CollCardiol 1988;11:50-58.

14. Bonow RO, Bacharach SL, Crawford-Green C, Green MV. Influence of temporal smoothing on

22 Chapter 2

quantitation of left ventricular function by gated blood pool scintigraphy. Am J Cardiol1989;64:921-925.

15. Bonow RO. Regional left ventricular nonuniformity. Effects on left ventricular diastolic function inischemic heart disease, hypertrophic cardiomyopathy, and the normal heart. Circulation1990;81(suppl III):III-54-III-65.

16. Bonow RO. Radionuclide Angiographic Evaluation of Left ventricular diastolic function.Circulation 1991;84(suppl I):I-208-I-215.


18. Bowman LK, Lee FA, Jaffe CC, Mattera J, Wackers FJTh, Zaret BL. Peak filling rate normalizedto mitral stroke volume: a new doppler echocardiographic filling index validated by radionuclideangiographic techniques. J Am Coll Cardiol 1988;12:937-943.

19. Braunwald E. Pathofysiology of Heart Failure. In: Braunwald E, ed. Heartdisease. Philadelphia:Saunders, 1992; 393-418.

20. Brecher GA, Kolder H, Horres AD. Ventricular volume of nonbeating excised dog hearts in thestate of equilibrium. Circ Res 1966;19:1080.

21. Brutsaert DL, Housmans PR, Goethals MA. Dual control of relaxation. Its role in the ventricularfunction in the mamalian heart. Circ Research 1980;47:637-652.

22. Brutsaert DL, Rademakers FE, Sys SU. Triple control of relaxation: implications in cardiacdisease. Circulation 1984;69:190-196.

23. Brutsaert DL, Sys SU, Gillebert TC. Diastolic failure: pathofysiology and therapeuticimplications. J Am Coll Cardiol 1993;22:318-325.

24. Brutsaert DL, Sys SU. Relaxation and diastole of the heart. Physiol Rev 1989;69:1228-1315.

25. Clozel JP, Banken L, Osterrieder W. Effects of Ro 40-5967, a novel calcium antagonist, onmyocardial function during ischemia induced by lowering coronary perfusion pressure in dogs:comparison with verapamil. J Cardiovasc Pharmacol 1989;14:713-721.

26. Cohn JN. Heart failure with normal ejection fraction. The V-HeFT Study. Circulation1990;81(suppl III):III-48-III-53.

27. Courtois M, Kovacs SJ, Ludbrook PA. Transmitral pressure-flow velocity relation. Importance ofregional pressure gradients in the left ventricle during diastole. Circulation 1988;78:661-671.

28. Courtois M, Mechem CJ, Barzilai B, Gutierrez F, Ludbrook PA. Delineation of determinants ofleft ventricular early filling. Saline versus blood transfusion. Circulation 1994;90:2041-2050.

29. Covell JW. Factors influencing diastolic function. Possible role of the extracellular matrix.Circulation 1990;81(suppl III):III-155-III158.

30. Davies MJ, Pomerance A. Pathology of atrial fibrillation in man. Br Heart J 1972;34:520-525.

31. De Vries RJ. Second generation calcium channel blockers in chronic heart failure. Thesis,University of Groningen, 1997.

32. Dienstl F, Motro M, Poole Wilson PA, Schartl M, Bassand JP, Van Dalen FJ, EmmottSN,Horowitz M, Hugenholtz PG. Improved diastolic function with the calcium antagonistnisoldipine (coat-core) in patients post myocardial infarction: results of the DEFIANT study. EurHeart J 1992;13:1496-1505.

33. Ebstein. Die diastole des herzens. Ergebnisse der Physiologie 1904;3 (2):123-194.

34. Eichhorn EJ, Bedotto JB, Malloy CR, Hatfield BA, Deitchman D, Brown M, Willard JE,GrayburnPA. Effect of ß-adrenergic blockade on myocardial function and energetics in congestive heartfailure. Improvement in hemodynamic, contractile, and diastolic performance with bucindolol.Circulation 1990;82:473-483.

35. Eichhorn EJ. The paradox of ß-adrenergic blockade for the management of congestive heart


failure. Am J Med 1992;92:527-538.

36. Elkayam U, Amin J, Anilkumar M, Vasquez J, Weber L, Rahimtoola SH. A prospective,randomized, double-blind, crossover study to compare the efficacy and safety of chronicnifedipine therapy with that of isosorbide dinitrate and their combination in the treatment ofchronic congestive heart failure. Circulation 1990;82:1954-1961.

37. Elkayam U, Shotan A, Mehra A, Ostrzega E. Calcium channel blockers in heart failure. J AmColl Cardiol 1993;22(suppl A):139A-144A.

38. European study group on diastolic heart failure. How to diagnose diastolic heart failure. EurHeart J 1998;19:990-1003.

39. Ferrari R. Major differences among the three classes of calcium antagonists. Eur Heart J1997;18 (suppl A):A56-A70.

40. Figueredo VM, Camacho SA. Basic mechanisms of myocardial dysfunction: cellularpathophysiology of heart filure. Curr Opin Cardiol 1993;8:369-375.

41. Fragasso G, Chierchia SL, Pizetti G, Rossetti E, Carlino M, Gerosa S, Carandente O,Fedele A,Cattaneo N. Impaired left ventricular filling dynamics in patients with angina and angiographicallynormal coronary arteries: effect of beta adrenergic blockade. Heart 1997;77:32-39.

42. Franciosa JA, Baker BJ, Seth L. Pulmonary versus systemic hemodynamics in determiningexercise capacity of patients with chronic left ventricular failure. Am Heart J 1985;110:807-813.

43. Franciosa JA, Park M, Levine TB. Lack of correlation between exercise capacity and indexes ofresting left ventricular performance in heart failure. Am J Cardiol 1981;47:33-39.

44. Friedman BJ, Drinkovic N, Miles H, Shih WJ, Mazzoleni A, De Maria AN. Assessment of leftventricular diastolic function: comparison of doppler echocardiography and gated blood poolscintigraphy. J Am Coll Cardiol 1986;8:1348-1354

45. Gaasch WH. Passive elastic properties of the left ventricle. In: Gaasch WH, LeWinter MM, ed.Left ventricular diastolic dysfunction and heart failure. Philadelphia: Lea & Febiger, 1994;143-149.

46. Gentile F, Mantero A, Lippolis A, Ornaghi M, Azzolini M, Barbier P, Beretta L,Casazza F,Cornot R, Faletra F, Giagnoni E, Gualtierotti C, Lombroso S, Mattioli R. Pulmonary venous flowvelocity patterns in 143 normal subjects aged 20 to 80 years old. An echo 2D colour Dopplercooperative study. Eur Heart J 1997;18:148-164.

47. Gillebert TC, Sys SU. Physiologic control of relaxation in isolated cardiac muscle and intact leftventricle. In: Gaasch WH, LeWinter MM, ed. Left ventricular diastolic dysfunction and heartfailure. Philadelphia: Lea & Febiger, 1994;25-42.

48. Grogan M, Smith HC, Gersh BJ, Wood DL. Left ventricular dysfunction due to atrial fibrillation inpatients initially believed to have idiopathic dilated cardiomyopathy. Am J Cardiol 1992;69:1570-1573.

49. Grossman W, McLaurin LP. Diastolic properties of the left ventricle. Ann Intern Med1976;84:316-326.


51. Grossman W. Evaluation of systolic and diastolic function of the myocardium. In: Grossman W,Baim DS, ed. Cardiac catheterization, angiography and intervention. Philadelphia: Lea &Febiger, 1990;333.

52. Grossman W. Evaluation of systolic and diastolic function of the myocardium In: Grossman W,Baim DS, ed. Cardiac catheterization, angiography and intervention. Philadelphia: Lea &Febiger, 1986;301-319.

53. Hammermeister KE, Warbasse JR. The rate of change of left ventricular volume in man: diastolicevents in health and disease. Circulation 1974;49:739-747.

54. Hanet C, Rousseau MF, Van Eyll C, Pouleur H. Effects of nicardipine on regional diastolic left

24 Chapter 2

ventricular function in patients with angina pectoris. Circulation 1990;81(suppl III):III-148-III-154.

55. Hassapoyannes CA, Bergh ME, Movahed MR, Easterling BM, Omoigui NA. Diastolic effects ofchronic digitalization in systolic heart failure. Am Heart J 1998;136:688-695.

56. Heesen WF. Isolated systolic hypertension. Pathophysiology and effects of treatment. Thesis.Groningen University, The Netherlands, 1998.

57. Hess OM, Krayenbuehl HP. Management of hypertrophic cardiomyopathy. Curr Opin Cardiol1993;8:434-440.

58. Hoit BD, Walsh RA. Diastolic dysfunctionin hypertensive heart disease. In: Gaasch WH,LeWinter MM, ed. Left ventricular diastolic dysfunction and heart failure. Philadelphia: Lea &Febiger, 1994;354-372.

59. Hori M, Yellin EL, Sonnenblick EH. Left ventricular diastolic suction as a mechanism ofventricular filling. Jpn Circ J 1982;46(1):124-129.

60. Iliceto S. Left ventricular dysfunction: Which role for calcium antagonists? Eur Heart J1997;18(suppl A):A87-A91.

61. Ishida Y, Meisner JS, Tsujioka K, Gallo JI, Yoran C, Frater RWM, Yellin EL. Left venricular fillingdynamics: influence of left ventricular relaxation and left atrial pressure. Circulation 1986;74:187-196.

62. Janicki JS, Matsubara BB. Myocardial collagen and left ventricular diastolic dysfunction. In:Gaasch WH, LeWinter MM ,ed. Left ventricular diastolic dysfunction and heart failure.Philadelphia: Lea & Febiger, 1994;125-140.

63. Janicki JS. Influence of the pericardium and ventricular interdependence on left ventriculardiastolic and systolic function in patients with heart failure. Circulation 1990;81(suppl III):III-15-III-20.

64. Jovic A, Troskot R. Recovery of atrial systolic function after pharmacological conversion ofchronic atrial fibrillation to sinus rhythm: a Doppler echocardiographic study. Heart 1997;77:46-49.

65. Juni JE, Chen CC. Effects of gating modes on the analysis of left ventricular function in thepresence of heart rate variation. J Nucl Med 1988;29:1272-1278.

66. Keren G, Meisner JS, Sherez J, Yellin EL, Laniado S. Interrelationship of mid-diastolic mitralvalve motion, pulmonary venous flow, and transmitral flow. Circulation 1986;74:36-44.

67. Keren G, Sherez J, Megidish R, Levitt B, Laniado S. Pulmonary venous flow pattern-itsrelationship to cardiac dynamics. Circulation 1985;71:1105-1112.

68. Kitzman DW, Higginbotham MB, Cobb FR, Sheikh KH, Sullivan MJ. Exercise intolerance inpatients with heart failure and preserved left ventricular systolic function: failure of the Frank-Starling mechanism. J Am Coll Cardiol 1991;17:1065-72.

69. Komamura K, Shannon RP, Pasipoularides A, Ihara T, Lader AS, Patrick TA, Bishop SP,VatnerSF. Alterations in left ventricular diastolic function in conscious dogs with pacing-induced heartfailure. J Clin Invest 1992;89:1825-1838.

70. Kono T, Sabbah HN, Rosman H, Alam M, Stein P, Goldstein S. Left atrial contribution toventricular filling during the course of evolving heart failure. Circulation 1992;86:1317-1322.

71. Konstam MA, Kronenberg MW, Udelson JE, Metherall J, Dolan N, Edens TR, Howe DM,YusufS, Youngblood M, Toltsis H, and the SOLVD Investigators. Effect of acute angiotensinconverting enzyme inhibition on left ventricular filling in patients with congestive heart failure.Relationto right ventricular volumes. Circulation 1990;81(suppl III):III-115-III-122.

72. Kushwaha SS, Fallon JT, Fuster V. Restrictive cardiomyopathy N Eng J Med 1997;336:267-276.

73. Lahiri A, Rodrigues EA, Carboni GP, Raftery EB. Effects of long-term treatment with calciumantagonists on left ventricular diastolic function in stable angina and heart failure. Circulation1990;81(suppl III):III-130-III-138.


74. Lambert CR, Buss DD, Pepine CJ. Effect of nicardipine on myocardial function in vitro and invivo. Circulation 1990;81 (suppl III):III-139-III-147.

75. Lawson WE, Seifert F, Anagnostopoulos C, Hills DJ, Swinford RD, Cohn PF. Effect of coronaryartery bypass grafting on left ventricular diastolic function. Am J Cardiol 1988;61:283-287.

76. Little WC, Downes TR. Clinical evaluation of left ventricular diastolic performance. ProgCardiovasc Dis 1990;32:273-290.

77. Little WC, Ohno M, Kitzman DW, Thomas JD, Cheng CP. Determination of left ventricularchamber stiffness from the time for deceleration of early left ventricular filling. Circulation1995;92:1933-1939.

78. Mackenzie J. Diseases of the heart. London: Oxford medical publications, 1914;148.

79. Mahmarian JJ, Pratt CM. Silent myocardial ischemia in patients with coronary artery disease:possible link with diastolic left ventricular dysfunction. Circulation 1990;81(suppl III):III-33-III-40

80. Manning WJ, Leeman DE, Gotch PJ, Come PC. Pulsed Doppler evaluation of atrial mechanicalfunction after electrical cardioversion of atrial fibrillation. J Am Coll Cardiol 1989;13:617-623.

81. Mantero A, Gentile F, Gualtierotti C, Azzolini M, Barbier P, Beretta L, Casazza F,Corno R,Giagnoni E, Lippolis A, Lombroso S, Mattioli R, Morabito A, Ornaghi M, Pepi M, Pezzano A.Left ventricular diastolic parameters in 288 normal subjects from 20 to 80 years old. Eur Heart J1995;16:94-105.

82. Meijler FL, Brutsaert DL. Relaxation and diastole. Precedings of the fourth workshop oncontractile behavior of the heart. Introduction. Eur J Cardiol 1978;7 (supplement 1):1.

83. Meisner JS, Keren G, Pajaro OE, Mani A, Strom JA, Frater RWM, Laniado S,Yellin EL. Atrialcontribution to ventricular filling in mitral stenosis. Circulation 1991;84:1469-1480.

84. Morgan JP. Abnormal intracellular modulation of calcium as a major cause of cardiac contractiledysfunction. N Engl J Med 1991;325:625-32.

85. Morris Jr JJ, Entman M, North WC, Kong Y, McIntosh H. The changes in cardiac output withreversion of atrial fibrillation to sinus rhythm. Circulation 1965;31:670-678.

86. Nayler WG. The ischemic myocardium and calcium antagonists. In: Opie LH, ed. Myocardialprotection by calcium antagonists. New York: Wiley-Liss, 1994;46-61.

87. Nijland F, Kamp O, Karreman AJP, Van Eenige MJ, Visser CA. Prognostic implications ofrestrictive left ventricular filling in acute myocardial infarction: a serial Doppler echocardiographicstudy. J Am Coll Cardiol 1997;30:1618-1624.

88. Nixon JV, Burns CA. Cardiac effects of aging and diastolic dysfunction in the elderly. In: GaaschWH, LeWinter MM, ed. Left ventricular diastolic dysfunction and heart failure. Philadelphia: Lea& Febiger, 1994;427-435.

89. Ohno M, Cheng CP, Little WC. Mechanism of altered patterns of left ventricular filling during thedvelopment of congestive heart failure. Circulation 1994;89:2241-2250.

90. Otterstad JE, Froeland G, St. John Sutton MS, Holme I. Accuracy and reproducibility of biplanetwo-dimensional echocardiographic measurements of left ventricular dimensions and function.Eur Heart J 1997;18:507-513.

91. Packer DL, Bardy GH, Worley SJ, Smith MS, Cobb FR, Coleman E, Gallaghar JJ,German LD.Tachycardia-induced cardiomyopathy: a reversible form of left ventricular dysfunction. Am JCardiol 1986;57:563-570.

92. Packer M. Abnormalities of diastolic function as a potential cause of exercise intolerance inchronic heart failure. Circulation 1990;81(suppl III):III-78-III-86.

93. Packer M. Pathophysiological mechanisms underlying the adverse effects of calcium channel-blocking drugs in patients with chronic heart failure. Circulation 1989;80(supplIV):IV-59-IV-67.

94. Perrone-Filardi P, Bacharach SL, Dilsizian V, Bonow RO. Effects of regional systolic

26 Chapter 2

asynchrony on left ventricular global diastolic function in patients with coronary artery disease. JAm Coll Cardiol 1992;19:739-744.

95. PinamontiB, Di Lenarda A, Sinagra G, Camerini F, and the Heart Muscle Disease Study Group.Restrictive left ventricular filling pattern in dilated cardiomyopathy assessed by Dopplerechocardiography: clinical, echocardiographic and hemodynamic correlations and prognosticimplications. J Am Coll Cardiol 1993;22:808-815.

96. Posma JL, Blanksma PK, Van der Wall E, Lie KI. Acute intravenous versus chronic oral drugeffects of verapamil on left ventricular diastolic function in patients with hypertrophiccardiomyopathy. J Cardiovasc Pharmacol 1994;24:969-973.

97. Pouleur H, Rousseau MF, Van Eyll C, Stoleru L, Hayashida W, Udelson JA, Dolan N,Kinan D,Galagher P, Ahn S, Benedict CR, Yusuf S, Konstam M, for the SOLVD Investigators. Effects oflong-term enalapril therapy on left ventricular diastolic properties in patients with depressedejection fraction. Circulation 1993;88:481-491.

98. Quinones MA. How to assess diastolic function by doppler echocardiography. In: Braunwald E,ed. Heartdisease update. Philadelphia: Saunders, 1991;15;351-358.

99. Rihal CS, Nishimura RA, Hatle LK, Bailey KR, Tajik AJ. Systolic and diastolic dysfunction inpatients with clinical diagnosis of dilated cardiomyopathy. Circulation 1994;90:2772-2779.

100. Rodriguez LM, Smeets JLRM Xie B, de Chillou C, Cheriex E, Pieters F, Metzgers J,den Dulk K,Wellens HJJ. Improvement in left ventricular function by ablation of atrioventrcular nodalconduction in selected patients with lone atrial fibrillation. Am J Cardiol 1993;72:1137-1141.

101. Rowlands DJ, Logan WFWE, Howitt G. Atrial function after cardioversion. Am Heart J1967;74:149-160.

102. Sabbah HN, Stein PD. Pressure-diameter relations during early diastole in dogs: incompatibilitywith the concept of passive left ventricular filling. Circ Res 1981;48:357-365.

103. Sasayama S, Asanoi H, Ishizaka S. Mechanics of contraction and relaxation of the ventricle inexperimental heart failure produced by rapid ventricular pacing in the conscious dog. Eur Heart J1991;12(Supplement C):35-41.

104. Scalia GM, Greenberg NL, McCarthy PM, Thomas JD, Vandrvoort PM. Noninvasive assessmentof the ventricular relaxation time constant (T) in humans by Doppler echocardiography.Circulation 1997;95:151-155.

105. Shinbane JS, Wood MA, Jensen DN, Ellenbogen KA, Fitzpatrick AP, Scheinman MM.Tachycardia-induced cardiomyopathy: a review of animal models and clinical studies. J Am CollCardiol 1997;29:709-715.

106. Shintani H, Glantz A. The left ventricular diastolic pressure-volume relation, relaxation, andfilling. In: GaaschWH, LeWinter MM, eds. Left ventricular diastolic dysfunction and heart failure.Philadelphia: Lea & Febiger, 1994;57-88.

107. Smith TW, Braunwald E, Kelly RA. The management of heart failure. In: Braunwald E, ed.Heartdisease. Philadelphia: Saunders, 1992; 464-519.

108. Smith VE, Schulman P, Karimeddini MK, White WB, Meeran MK, Katz AM. Rapid ventricularfilling in left ventricular hypertrophy: II. pathologic hypertrophy. J Am Coll Cardiol 1985;5:869-874

109. Störk TV, Müller RM, Piske GJ, Ewert CO, Hochrein H. Noninvasive measurement of leftventricular filling pressures by means of transmitral pulsed doppler ultrasound. Am J Cardiol1989;64:655-660.

110. Stugaard M, Smiseth OA, Risoe C, Ihlen H. Intraventricular early diastolic filling during acutemyocardial ischemia. Assessment by multigated color M-mode echocardiography. Circulation1993;88:2705-2713.

111. Su J, Renaud N, Carayon A, Crozatier B, Hittinger L, Laplace M. Effects of the calcium channelblockers, diltiazem and ro 40-5967, on systemic haemodynamics and plasma noradrenalinelevels in conscious dogs with pacing-induced heart failure. Br J Pharmacol 1994;113:395-402.


112. Suga H. Importance of atrial compliance in cardiac performance. Circ Res 1974;35:39-43.

113. Szlachcic J, Massie BM, Kramer BL, Topic N, Tubeau J. Correlates and prognostic implicationof exercise capacity in chronic congestive heart failure. Am J Cardiol 1985;55:1037-1042.

114. Temporelli PL, Scapellato F, Corra U, Eleuteri E, Imparato A, Giannuzzi P. Estimation ofpulmonary wedge pressure by transmitral doppler in patients with chronic heart failure and atrialfibrillation. Am J Cardiol 1999;83:724-727.

115. The DEFIANT-II Research group. Doppler flow and echocardiography in functional cardiacinsufficiency: assessment of nisoldipine therapy. Results of the DEFIANT-II study. Eur Heart J1997;18:31-40.

116. Thomas JD, Weyman AE. Echocardiographic doppler evaluation of left ventricular diastolicfunction. Physics and physiology. Circulation 1991;84:977-990.

117. Thomas JD. Dopller echocardiography and left ventricular diastolic function. In: GaaschWH,LeWinter MM, eds. Left ventricular diastolic dysfunction and heart failure. Philadelphia: Lea &Febiger, 1994;192-218.

118. Tomita M, Spinale FG, Crawford FA, Zile MR. Changes in left ventricular volume, mass, andfunction during the development and regression of supraventricular tachycardia-inducedcardiomyopathy: disparity between recovery of systolic versus diastolic function. Circulation1991;83:635-644.

119. Udelson JE, Bacharach SL, Cannon RO 3d, Bonow RO. Minimum left ventricular pressureduring ß-adrenergic stimulation in human subjects: Evidence for elastic recoil and diatolic"suction" in the normal heart. Circulation 1990;82:1174-1182.

120. Udelson JE, Bonow RO. Radionuclide angiographic evaluation of left ventricular diastolicfunction. In: Gaasch WH, LeWinter MM, ed. Left ventricular diastolic dysfunction and heartfailure. Philadelphia: Lea & Febiger, 1994;167-191.

121. Udelson JE, Surks H, Konstam MA. The importance of left ventricular diastolic dysfunction andright ventricular performance in the hart failure syndrome. Curr Opin Cardiol 1993;8:383-396.

122. Van den Berg MP, Crijns HJGM, Gosselink ATM, Van den Broek SAJ, Hillege HJ, VanVeldhuisen DJ, Lie KI. Chronotropic response to exercise in patients with atrial fibrillation:relation to functional state. Br Heart J 1993;70:150-153.

123. Van der Vring JAFM, Cleophas AJM, Van der Wall EE, Niemeyer MG, Bernink PJ, Knol HR,Van Veldhuisen DJ, Braun SJ, Kobrin I, Crijns HJGM. Effects of mibefradil, a T-channel-selectivecalcium channel blocker, on sympathetic activity in patients with congestive heart failure.Cardiologie 1998;;5:506-511.

124. Van der Werf T. Invasief onderzoek van de systolische en diastolische linkerventrikelfunctie. In:Werf T van der, red. Linkerventrikelfunctie tijdens systole en diastole. Utrecht: Bunge, 1991;56-73.

125. Van Rossum AC. Magnetic resonance imaging of cardiac function and flow: present and future.In: Van der Wall EE, Blanksma PK, Niemeyer MG, Vaalburg W, Crijns HJGM, ed. Advancedimaging in coronary artery disease. Dordrecht: Kluwer academic publishers, 1998;289-306.

126. Vasan RS, Benjamin EJ, Levy D. Prevalence, clinical features and prognosis of diastolic heartfailure: an epidemiologic perspective. J AM Coll Cardiol 1995;26:1565-1574.

127. Vasan RS, Benjamin EJ, Levy D. Congestive heart failure with normal left ventricular systolicfunction. Clinical approaches to the diagnosis and treatment of diastolic heart failure. Arch InternMed 1996;156:146-157.

128. Vatner SF, Shannon R, Hittinger L. Reduced subendocardial coronary reserve. A potentialmechanism for impaired diastolic function in the hypertrophied and failing heart. Circulation1990;81(suppl III):III-8-III-14.

129. Vignon P, Mor-Avi V, Weinert L, Koch R, Spencer KT, Lang RM. Quantitative evaluation ofglobal and regional left ventricular diastolic function with color kinesis. Circulation 1998;97:1053-

Chapter 2Diastolic function:

physiology, methodsand clinical significance

Background of the present thesis

rom a physiological point of view the heart is an integrated muscle-pump system.The term diastole is interpreted as a division, notch, or separation between twocontraction-relaxation cycles (Brutsaert 1984). In this interpretation its meaning isrestricted to the passive properties of the heart (Gillebert 1994). Diastole of the

left ventricle starts when active relaxation has been completed and includes thediastasis and the atrial contraction phase (Figure 2.1). In the English medical literaturediastole has however come to mean "the dilatation or period of dilatation of the heart,especially that of the ventricles, coinciding with the interval between the second andfirst heart sounds" (Brutsaert 1984). In this interpretation it is that part of the cardiaccycle which starts with the isovolumetric relaxation phase and ends with cessation ofmitral inflow (Arrighi 1995). In the present thesis the latter, clinical definition of diastolewill be used.

Ebstein extensively discusses the history of views on diastole in 1904(Ebstein 1904). Mackenzie devides the heartcycle in a presphygmic period with risingventricular pressure between closure of the atrio-ventricular valves and opening of theaortic valves, a sphygmic or pulse period with opened aortic valves, and apostsphygmic period after aortic valve closure and before atrio-ventricular valveopening where ventricular pressure is falling (Mackenzie 1914). Wiggers suddevidedthe cardiac cycle into smaller phases (Wiggers 1921). In his view, diastole is precededby protodiastole, the period between end of ventricular contraction and closure of thesemilunar valves. The period of diastole then begins with closure of the semilunar valvesand isometric relaxation. After opening of the atrio-ventricular valves a period ofdiastolic inflow begins. In case of a sufficiently long diastolic interval a period ofdiastasis is recognised in addition. Wiggers subdevided the period of auricular systolein a period in which auricular contracction exerts effect on ventricular filling or ventriculartension, and a period of auricular filling by venous return from the pulmonary circulation(Wiggers 1921).

In 1978 new interest arose on relaxation and diastole, since symptoms ofcardiac disease could be succesfully related to pathophysiology of diastole (Meijler

F

28 Chapter 2

1061.

130. Wagner RH, Halama JR, Henkin RE, Dillehay GL, Sobotka PA. Errors in determination of leftventricular functional parameters. J Nucl Med 1989;30:1870-1874.

131. Wang K, Gibson DG. Non-invasive detection of left atrial mechanical failure in patients with leftventricular disease. Br Heart J 1995;74:536-540.

132. WiggersCJ. Studies on the consecutive phases of the cardiac cycle. I. The duration of theconsecutive phases of the cardiac cycle and the criteria for their precise determination. Am JPhysiol 1921;56:415-438.

133. WiggersCJ. Studies on the consecutive phases of the cardiac cycle. II. The laws governing therelative duration of ventricular systole and diastole. Am J Physiol 1921;56:438-459.

134. Xiong C, Sonnhag C, Nylander E, Wranne B. Atrial and ventricular function after cardioversion ofatrial fibrillation. Br Heart J 1995;74:254-260.

135. Yamakado T, Takagi E, Okubo S, Imanaka-Yoshida K, Tarumi T, Nakamura M, Nakano T.Effects of aging on left ventricular relaxation in humans. Analysis of left ventricular isovolumicpressure decay. Circulation 1997;95:917-923.

136. Yamamoto K, Nishimura RA, Chaliki HP, Appleton CP, Holmes Jr DR, Redfield MM.Determination of left ventricular filling pressure by doppler echocardiography in patients withcoronary artery disease: critical role of left ventricular systolic function. J Am Coll Cardiol1997;30:1819-1826.

137. Yashima N. Nasu M, Kawazoe K, Hiramori K. Serial evaluation of atrial function by Dopplerechocardiography after the maze procedure for chronic atrial fibrillation. Eur Heart J1997;18:496-502.

138. Yellin EL, Hori M, Yoran C. Left ventricular relaxation in the filling and nonfilling intact canineheart. Am J Physiol 1986;250 Heart Circ Physiol 19):H620-H629.

139. Yellin EL, Meisner JS, Nikolic SD, Keren G. The scientific basis for the relations betweenpulsed-doppler transmitral velocity patterns and left heart chamber properties. Echocardiography1992;9:313-338.

140. Yellin EL, Nikolic S, Frater RWM. Left ventricular filling dynamics and diastolic function. ProgCardiovasc Dis 1990;32:247-271.

141. Yellin EL, Nikolic SD. Diastolic suction and the dynamics of left ventricular filling. In: GaaschWH, LeWinter MM, ed. Left ventricular diastolic dysfunction and heart failure. Philadelphia: Lea& Febiger, 1994;89-102.

142. Yoganathan AP. Assessment of diastolic function with velocity-encoded magnetic resonanceimaging. Heartforum 1997;10 (suppl1):50-56.

143. Zile MR, Mukherjee R, Clayton C, Kato S, Spinale FG. Effects of chronic supraventricular pacingtachycardia on relaxation rate in isolated cardiac muscle cells. Am J Physiol 1995;268 (HeartCirc Physiol 37):H2104-H2113.

144. Zile MR. The development of diastolic dysfunction in chronic tachycardia. In: Spinale FG, ed.Pathophysiology of tachycardia-induced heart failure. Armonk, NY: Futura Publishing Company,Inc.1996;25-44.

Chapter 3Normal values and reproducibility ofleft ventricular filling parameters by

radionuclide angiographyH.J. Muntinga1, F. van den Berg1, H.R. Knol2, M.G. Niemeyer1, P.K. Blanksma3, H. Louwes4 and E.E. van der Wall5

1 Department of Cardiology, Martini Hospital, Groningen, 2 Northern Centre for Health Care Research,University of Groningen, Groningen, 3 Department of Cardiology, University Hospital Groningen,Groningen, 4 Department of Nuclear Medicine, Martini Hospital, Groningen, 5 Department of Cardiology,University Hospital Leiden, Leiden, The Netherlands.

Summary

Background: In physiologic situations age, heart rate (HR) and left ventricular ejection fraction(EF) may influence left ventricular filling rate. In this study, we determined normal values forradionuclide angiography (RNA) derived diastolic filling parameters, the correlations with age,HR and EF and their reproducibility.Methods: The study was performed in 20 patients, 40-76 years old (mean 57), with normalfindings at coronary angiography and left ventriculography. The first RNA was performed atrest (RNA1). Then, five minutes bicycle ergometry was performed and the patients wereallowed five minutes rest before RNA was repeated (RNA2). From the left ventricular timeactivity curve we determined peak filling rate (PFR), time to peak filling rate (TPFR) and atrialcontribution (AC) to ventricular filling.Results: Values for PFR1 were 2.2 ± 0.6 EDV/s (PFR2 2.4 ± 0.7 EDV/sec, r=0.82), for TPFR1 198 ±

22 ms (TPFR2 203 ± 24 ms, r=0.45) and for AC1 31 ± 11 % (AC2 31 ± 10 %, r = 0.72). Thecorrelation’s of PFR and TPFR with age were statistically significant (respectively r=-0.68 andr=0.48, P<0.05). PFR was also influenced by HR and EF (respectively r=0.51 and r=0.50, P<0.05).TPFR however was not influenced by HR and EF, whereas AC was positively correlated withHR (r=0.79, P<0.01).Conclusions: Radionuclide angiography is a reliable and reproducible method to assessparameters of diastolic left ventricular filling in individual patients. It may therefore be used toserially follow diastolic function. When used for interindividual comparison the dependency ofRNA derived left ventricular filling parameters on age, HR and EF should however beconsidered.

International Journal of Cardiac Imaging 1997;13:165-171

Reproducibility of normal diastolic function parameters 31

or the quantitative interpretation of diagnostic test results used to evaluate heartfunction it is necessary to define limits of normal. Healthy volunteers,1,2,3,4,5

patients with a low probability of coronary artery disease,1,5,6,7 and patients with normalcoronary angiograms2 are commonly used to provide standards in assessing theaccuracy of diagnostic tests in cardiology.

In the context of tests used to evaluate diastolic heart function, the definition ofnormal is often based on findings in patients with a normal exercise electrocardiogramand a normal Doppler echocardiogram.1,8 Alternatively, clinically normal subjects withnormal findings at coronary arteriography and left ventriculography may also be used todefine the normal range.2 Until now several studies reported on radionuclideangiographic assessment of normal diastolic function.1,2,3,5,6,16 Unfortunately, only fewstudies address the reproducibility of the normal values for the patient group used.2 Inthe present study a description is made of normal values of diastolic functionparameters, together with the reproducibility of these parameters for patients withnormal findings in coronary arteriography. The parameters were assessed using acomputer program which automatically defines filling parameters derived byradionuclide angiography.

METHODSPatients. The present study was performed in 20 subjects, 8 male persons, mean age57 ± 11 years, with normal findings at coronary arteriography and left ventriculography.Coronary angiography was performed as a diagnostic test because of precordial painof unknown origin. All results of the cardiac angiography were reviewed by one person.Patients with significant coronary abnormalities were excluded, as were patients withdetectable valvar abnormalities. Nine patients were successfully treated forhypertension. Left ventricular hypertrophy was excluded with electrocardiography andwith echocardiography. Three patients had paroxysmal atrial fibrillation. In thirteenpatients all medication if any was used was withdrawn more than 48 hours beforeevaluation. The other seven patients remained on oral therapy, which included oralnitrates (1 patient), diuretics (1), calcium channel blockers (1), beta-blocking agents (5)and angiotensin-converting enzyme inhibitors (1).

Radionuclide angiography. Each patient's red blood cells were labelled withtechnetium 99m, after intravenous administration of pyrophosphate. The total dosagewas 550-740 MBq. Left ventricular function was evaluated by radionuclide angiographyusing a gamma camera (Siemens Orbiter) with an all-purpose parallel-hole collimatorinterfaced with a Pinnacle computer (Medasys Inc., Ann Arbor). Rest supine multigatedimages were obtained in left anterior oblique view with a caudal tilt, so that the left andright ventricles were entirely separated. Only a 5% cycle-length-window with forward

F

32 Chapter 3

gating was accepted.9 Acquisition was completed after 150,000 counts per frame of 20ms. Temporal smoothing was performed by 5 Fourier harmonics.10,11

Measurements. From the systolic part of the left ventricular time-activity curve (TAC)we measured the time to end of systole (TES) and the ejection fraction (EF). From thediastolic part of the curve we took measurements only when early diastolic filling couldbe separated from late diastolic filling by a diastasis period. The diastolic parametersincluded peak filling rate (PFR) and time to peak filling rate (TPFR). The PFR wasnormalised to end-diastolic volume (EDV/sec) and to stroke volume (SV/sec). TheTPFR was measured from the beginning of contraction (TPFR') and from the beginningof relaxation (Figure 3.1). Further diastolic filling was divided into rapid filling and atrialcontribution (AC) as percentage of stroke volume and end-diastolic volume.

Reproducibility data. Reproducibility was assessed by performing 2 differentradionuclide angiograms. They were separated by a supine bicycle ergometry of 5minutes and 5 minutes of rest, so that heart rate in both studies was not automaticallythe same. After a new adjustment of the collimator the 2nd data collection was started.

Statistical analysis. Measured data are presented as mean ± 1 SD. Linear regressionanalysis was used to test the relation between the radionuclide angiographic variableson the two occasions. This was also used to test the relation between age, HR and EFand the radionuclide angiographic data. Statistical analysis for paired samples wasperformed by a Student's t-test.

FIGURE 3.1. Time-activity curve of the left ventricle (Figure 3.1A) and its first derivative (Figure 3.1B).Ejection fraction is the ratio of stroke volume (SV) and end-diastolic volume (EDV). Atrial contribution(AC) is expressed as a fraction of SV or EDV. In Figure 3.1B PFR is the rapid instantaneous filling rateduring early rapid filling expressed as EDV per second and as SV per second. TPFR is measured fromthe beginning of relaxation (TPFR) and from the beginning of contraction (TPFR').

A B

TABLE 3.1. Radionuclide angiographic values of systolic and diastolic parameters in the studypopulation.

Mean ± SD RangeAge (years) 57 ± 11 40-76 HR (beats/min) 68 ± 12 52-105EF (%EDV) 62 ± 8 46-76 PFR (EDV/s) 2.2 ± 0.6 1.0-3.4 PFR (SV/s) 3.5 ± 0.8 2.0-4.9 TES (ms) 382 ± 42 301-464 TPFR (ms, from TES) 198 ± 22 144-233 TPFR' (ms, from t=0) 579 ± 47 444-650 AC (%SV) 31 ± 11 15-67 AC (%EDV) 19 ± 8 9-50 HR = heart rate; EF = ejection fraction; PFR = peak filling rate; EDV = end-diastolic volume; SV = strokevolume; TES = time to end of systole; TPFR = time to peak filling rate.


RESULTSNormal values. The mean values, standard deviations and ranges of systolic anddiastolic parameters measured by the first radionuclide angiography together with ageand heart rate are listed in Table 3.1. A description of the values of diastolic fillingparameters with radionuclide angiography with comparable techniques in normalindividuals is made in Table 3.2. Most authors had used different selection criteria fortheir study population, including a low probability for coronary artery disease, normalvolunteers and patients with normal findings in coronary angiography. Age was anotherdifference between these studies. Lee et al., Miller et al. and Iskandrian et al. hadselected patients of all age categories.2,6,16 Iskandrian et al. had divided theirpopulation according to age.16 Arora et al. had selected two groups of different ages.1

All authors except Iskandrian et al. had used 5 harmonics for temporal smoothing.Taking these differences into account, the values of PFR (normalised to both EDV andSV), TPFR (measured from the beginning of filling and from the beginning of emptying)and AC (%SV) were approximately the same as in our study.

Reproducibility. Two patients were not evaluable for the second RNA study, and theywere excluded from this part of the study. The reproducibility data are presented inTable 3.3. Compared to the first RNA no changes of the means of the variables werepresent in the second RNA. The average difference between the individual values of thetwo RNA studies ranged from 3 % of the initial value in TPFR' to 20 % of the initial value

TABLE 3.2. Values of diastolic filling parameters assessed with radionuclide angiography withcomparable techniques in normal individuals.

Lee Arora Miller IskandrianGroup I Group I I Group I Group I I

Ref. no. 6 1 1 2 16 16Screening low. vol. vol. vol. &

NCAGlow. low.

No. of subjects 64 10 13 30 29 36Age 17-77 20-33 68-86 22-80 17-50 50-75Harmonics 5 5 5 5 3 3PFR (EDV/s) 3.0 ± 0.8 3.5 ±0.5 2.3 ±0.5 2.67 ±0.95 3.1 ±0.6 2.6 ±0.6 Range 1.4 - 5.3 - - 1.15 - 4.84 - -PFR (SV/s) 4.5 ±1.0 - - - - - Range 2.5 - 6.9 - - - - -TPFR (ms) - - - 180 ±40 165 ± 41 175 ±46 Range - - - 96 - 257 - -TPFR' (ms) - 471 ± 33 543 ± 52 - - - Range - - - - - -AC (%SV) - 16 ± 4 31 ± 10 - - - Range - - - - - -ref.no. = reference number; low. = low probability for coronary artery disease; vol. = normal volunteers;NCAG = normal coronary angiogram; PFR = peak filling rate; EDV = end diastolic volume; SV = strokevolume; TPFR = time to peak filling rate measured from the beginning of relaxation; TPFR' = time to peakfilling rate measured from the beginning of contraction; AC = atrial contribution to diastolic filling.

34 Chapter 3

in AC (%EDV). The values of the parameters of the first and the second RNA correlatedwell. Only the correlation of TPFR1 to TPFR2 was relatively low (r=0.45).Correlations of Age, HR and EF to parameters of the TAC. The results of theregression analysis of Age, HR and EF in relation to the mentioned diastolicparameters of the TAC are listed in Table 3.4. The negative relation of age with leftventricular filling was reflected in a statistically significant correlation with PFRexpressed as EDV/sec and SV/sec (P<0.01, Figure 3.2A) and both TPFR and TPFR'P<0.05, Figure 3.2B). Age was however not correlated to AC (r=0.15 and r=-0.08 forAC expressed as %SV and as %EDV respectively). The positive influence of heart rateon diastolic filling was expressed in the relation with PFR normalised to EDV (P<0.05),TPFR' (P<0.01) and AC expressed as fractions of EDV and of SV (P<0.01). PFRnormalised to EDV, TPFR' and AC (%EDV) were also correlated to EF.

DISCUSSIONNormal values. In the radionuclide angiographic evaluation of diastolic left ventricularfunction, the matter of interest is the left ventricular filling pattern. In the absence of mitralvalve abnormalities, this filling pattern is the result of the diastolic atrial ventricular

TABLE 3.3. Reproducibility data of 18 patients who underwent two sequential RNA's.

RNA 1(mean ± SD)

RNA 2(mean ± SD)

Averagedifference

between thestudies

(mean ± SD)

MaximumDifference

Between thestudies r

HR (beats/min) 69 ± 12 71 ± 9 4 ± 4 15 0.89EF (%EDV) 63 ± 8 62 ± 8 5 ± 4 18 0.71PFR (EDV/s) 2.22 ± 0.61 2.37 ± 0.66 0.32 ± 0.26 1.02 0.82PFR (SV/s) 3.53 ± 0.85 3.80 ± 0.89 0.56 ± 0.47 1.84 0.70TES (ms) 380 ± 39 373 ± 37 16 ± 12 45 0.89TPFR (ms) 198 ± 22 203 ± 24 20 ± 15 51 0.45TPFR' (ms) 579 ± 47 576 ± 36 18 ± 13 45 0.89AC (%SV) 31 ± 11 31 ± 10 5 ± 6 28 0.72AC (%EDV) 20 ± 9 19 ± 7 4 ± 4 20 0.74HR = heartrate (beats/min); EF = ejection fraction of the left ventricle; EDV = end-diastolic volume; PFR= peak filling rate; SV = stroke volume; TES = time to end of systole; TPFR = time to peak filling ratemeasured from the beginning of relaxation; TPFR' = time to peak filling rate measured from the beginningof contraction; AC = atrial contribution; SD = standard deviation; r = correlation coefficient relating thefirst and second study.

TABLE 3.4. Correlations between age, heart rate and ejection fraction and diastolic parameters of theTAC.

PFR(EDV/s)

PFR(SV/s)

TES(ms)

TPFR(ms)

TPFR'(ms)

AC(%SV)

AC(%EDV)

Age -0.68* -0.61* 0.17+ 0.48+ 0.45+ 0.15 -0.08HR (beats/min) 0.51+ 0.33 -0.68* -0.33 -0.81* 0.79* 0.82*

EF (%EDV) 0.50+ 0.039 -0.50+ -0.16 0.54+ 0.28 0.50+

P<0.01; + P<0.05; All corelations not listed are not significant (P>0.05).HR = heartrate (beats/min); EF = ejection fraction of the left ventricle; EDV = end-diastolic volume; PFR= peak filling rate; SV = stroke volume; TES = time to end of systole; TPFR = time to peak filling ratemeasured from the beginning of relaxation; TPFR' = time to peak filling rate measured from the beginningof contraction; AC = atrial contribution.

FIGURE 3.2A The relation between peak filling rate (PFR) and age in the 20 patients with normalcoronary angiograms. EDV/s = end-diastolic volume per second. B The relation of time to peak filling rate(TPFR) to age in the 20 patients with normal coronary angiograms.

A B


pressure gradient,12 which in addition is influenced by many factors including leftventricular relaxation and compliance.13,14,15 Some filling parameters therefore show aclear relation with physiological (e.g. age and exercise) and pathological (e.g.ischaemia and hypertension) conditions that influence these factors.11 It is therefore ofgreat importance to describe the normal values and the reliability of the test in individualcases. The comparison of the normal test results with the findings of other investigators,must however be made with caution, for also a comparison of the tests has to be made.Previously, other investigators have described the values of PFR, TPFR and AC inhealthy volunteers and patients with a low probability of coronary artery disease.1,2,3,5,16

The measured values of PFR (normalised to end diastolic volume and to strokevolume), TPFR (both from beginning of contraction and from beginning of relaxation)and AC in these studies roughly coincided with our mean values. Also, the ranges wereapproximately the same.

Relationship with age. With the normal ageing process, the heart is subjected to anumber of anatomic changes which include a decreased amount of myocytes, fibroustissue proliferation, loss of elastic fibres and increased left ventricular wall mass.8 Withthese findings, a decreased left ventricular compliance and a decreased relaxation rateare likely to be present with increasing age. The described relation between age andPFR and TPFR was found earlier by other authors.1,2,3,5,6,16 The increasing AC withrising age which was earlier described, is thought to be a compensation of decreasedearly diastolic filling.1 This is however not supported by the present study.

Relationship with ejection fraction. In the present study the measured fillingparameters were also related to the ejection fraction. As was described before, PFRnormalised to end-diastolic volume, but not PFR normalised to stroke volume, isinfluenced by EF.11 This relation expresses the influence of systolic performance onventricular relaxation. PFR normalised to stroke volume and AC expressed as apercentage of stroke volume are however relatively independent of EF.

Relationship with heart rate. Heart rate is also an independent variable influencingthe timing and maximum rate of early rapid diastolic filling.3,6,11 Its positive effect onPFR is believed to be resulting from the disappearance of the diastasis period intachycardia. In this situation PFR is highly heart rate dependent.11 In every TAC in ourstudy however a diastasis period was present. Despite this, the relation between HRand PFR was still statistically significant, as was reported before.2 An increased leftventricular relaxation in higher heart rates may be the explanation for this. Also negativecorrelations between HR and TPFR' (measured from the beginning of contraction) andTES were measured. They indicate a decreased duration of the contraction-relaxationcycle in higher heart rates mainly due to a reduction of the duration of ejection.

36 Chapter 3

Reproducibility. On PFR and TPFR excellent reproducibility data were reportedbefore by Miller et al.2 In this study the second data collection was not, as in our study,preceded by a new adjustment of the collimator. Also, the patients had been lyingquietly between the measurements while in our study they had a short period ofexercise. The release of catecholamines during exercise has a potential effect ondiastolic function. In view of the redressment of heart rate in the following period of rest,this effect was probably not long-lasting in our study. Although these factors may havebeen of influence on the results, the present study shows good and consequentindividual reproducibility of the values of the measured variables.

Limitations. As mentioned, the methodological limitations of radionuclide angiographyshould not be underestimated. Not only technical considerations, as temporalsmoothing and resolution are important in the assessment of diastolic function, alsomethodological considerations as normalisation, influence the outcome and thereliability of the measurements. In addition, in the present study only 20 patients wereinvestigated, which decreases the power of describing normal values and referencevalues for other investigators.

Conclusions. We demonstrated the values of filling parameters with radionuclideangiography in patients with normal findings in coronary angiography. A good individualreproducibility of the values was found. RNA may therefore be used to serially followdiastolic function. Unfortunately, the diagnostic use of radionuclide derived leftventricular filling parameters remains limited because of the wide ranges and therelatively high standard deviations of the normal values. Also, the dependency on age,heart rate and ejection fraction make the interpretation of the diagnostic test resultsdifficult.

References

1. Arora RR, Machac J, Goldman ME, Butler RN, Gorlin R Horowitz SF. Atrial kinetics and leftventricular diastolic filling in the healthy elderly. J Am Coll Cardiol 1987;9:1255-1260.

2. Miller TR, Grossman SJ, Schectman KB, Biello DR, Ludbrook PA, Ehsani AA. Left ventriculardiastolic filling and its association with age. Am J Cardiol 1986;58:531-535.

3. Bonow RO, Vitale DF, Bacharach SL, Maron BJ, Green MV. Effects of aging on asynchronousleft ventricular regional function and global ventricular filling in normal human subjects. J Am CollCardiol 1988;11:50-58.

4. Kuo LC, Quinones MA, Rokey R, Sartori M, Abinader EG, Zoghbi WA. Quantification of atrialcontribution to left ventricular filling by pulsed doppler echocardiography and the effect of age innormal and diseased hearts. Am J Cardiol 1987;59:1174-1178.

5. Bowman LK, Lee FA, Jaffe CC, Mattera J, Wackers FJTh, Zaret BL. Peak filling rate normalizedto mitral stroke volume: a new doppler echocardiographic filling index validated by radionuclideangiographic techniques. J Am Coll Cardiol 1988;12:937-943.


6. Lee KJ, Southee AE, Bautovich GJ, Freedman B, McLaughlin AF, Rossleigh MA, HuttonBF,Morris JG. Normalised radionuclide measures of left ventricular diastolic function. Eur J NuclMed 1989;15:123-127.

7. Miyatake K, Okamoto M, Kinoshita N, Owa M, Nakasone I, Sakakibara H, NimuraY.Augmentation of atrial contribution to left ventricular inflow with aging as assessed byintracardiac doppler flowmetry. Am J Cardiol 1984;53:586-589.

8. Nixon JV, Burns CA. Cardiac effects of aging and diastolic dysfunction in the elderly. In: GaaschWH, LeWinter MM, ed. Left ventricular diastolic dysfunction and heart failure. Philadelphia: Lea& Febiger, 1994;427-435.


10. Bacharach SL, Green MV, Vitale D, White G, Douglas MA, Bonow RO, Larson SM.Optimumfourier filtering of cardiac data: a minimum-error method: concise communication. J Nucl Med1983;24:1176-1184.

11. Udelson JE, Bonow RO. Radionuclide angiographic evaluation of left ventricular diastolicfunction. In: Gaasch WH, LeWinter MM, ed. Left ventricular diastolic dysfunction and heartfailure. Philadelphia: Lea & Febiger, 1994;167-191.

12. Ishida Y, Meisner JS, Tsujioka K, Gallo JI, Yoran C, Frater RWM, Yellin EL. Left ventricularfilling dynamics: influence of left ventricular relaxation and left atrial pressure. Circulation1986;74:187-196.



15. Shintani H, Glantz A. The left ventricular diastolic pressure-volume relation, relaxation, andfilling. In: GaaschWH, LeWinter MM, eds. Left ventricular diastolic dysfunction and heart failure.Philadelphia: Lea & Febiger, 1994;57-88.

16. Iskandrian AS, Hakki AH. Age-related changes in left ventricular diastolic performance. AmHeart J 1986;112:75-78.

Editorial comment on chapter 3When is a visually or mathematically diagnosed

scan abnormality clinically important?International Journal of Cardiac Imaging 1997;13:173

M. PillayDr. Daniel den Hoed Clinic, University Hospital Rotterdam

This question is frequently addressed as more and more sensitive and sophisticateddiagnostic tools become available.

The first step in setting up a diagnostic test is to evaluate its reliability in terms ofsensitivity, specificity and reproducibility under a number of conditions. In a diagnosticimaging department, this is usually achieved by repeat imaging and analyses.

In this respect the authors have satisfied the basic requirement necessary tointroduce the test. The next question, of course, is the test being utilised on a regularbasis? If so, is there an expectation to supplement the results of this paper withadditional data (preferably a combination of diseased and normal subjects) so thatvalidity of the repeated measurements will be tested to determine the specificity,sensitivity and accuracy of the method. Without this additional data the reproducibilityfigures derived from repeated measurements in small studies can surely only be ofimportance as a quality control procedure. It is also well recognised that the normality inparameters derived by radionuclide cardiac angiography are geographically andpopulation biased so that multicentre trials to establish normal values are essential. Thepooling of data using meta-analyses has limitations and inherent deficiencies so thatchoice of analysis should be avoided where possible.

Presenting p-values alone can lead to their being given more merit than theydeserve. There is a tendency to equate statistical significance with medical importanceor biological relevance. Small differences of no real interest can be statisticallysignificant with large sample sizes, whereas clinically important effects may bestatistically non-significant only because the number of subsets studied was small.

This phenomenon is clearly evident in this study where a number of p-values areshown to indicate significant differences and where correlation coefficients demonstratelarge spread of individual data. Including data on diseased subjects will test thereliability of the hypotheses. The estimation and use of confidence intervals would havenbeen more appropriate.

In larger studies on impaired left ventricular function and diastolic filling Bonow etal. and Pace et al. have concluded that “the differences between the diastolic fillingcharacteristics of patients with coronary artery disease without prior infarction and thoseof normal volunteers could not be explained on the basis of any differences in age, inheart rate or in left ventricular end-diastolic size…”. This finding is not supported by this

Editorial comment on chapter 3 40

study and may be a result of the geographic, population sample and methodologicalbias.

The evaluation of reproducibility is surely of much importance provided thatstudies are designed to go beyond laboratory based quality control and include data totest hypotheses on a broader base, i.e. sensitivity, specificity and accuracy.

References

1. Bonow RO et al. Impaired left ventricular diastolic filling in patients with coronary artery disease:assessment with radionuclide angiography. Circulation 1981;64:315-323.

2. Pace L et al. Diagnosis of coronary artery disease by radionuclide angiography: effect ofcombining indices of left ventricular function. J Nucl Med 1989;30:1966-1971.

Chapter 4Quantification of

the atrial contribution to diastolic fillingduring radionuclide angiography

H.J. Muntinga1, F. van den Berg1, H.R. Knol2, J.J. Schuurman3 and E.E. van derWall4

1 Department of Cardiology, Martini Hospital, Groningen; 2 Northern Centre for Healthcare Research,University of Groningen, Groningen; 3 Department of Nuclear Medicine, Martini Hospital, Groningen;4Department of Cardiology, Leiden University Hospital, Leiden, The Netherlands

Summary

In order to define exactly the onset of late diastolic filling with respect to atrial contraction,atrial contribution (AC) to left ventricular filling was quantified in 34 patients with a variety ofdiseases using radionuclide angiography. From the time-activity curve and its first derivative, aflow-volume loop was constructed. Using the flow-volume loop, the period between minimalflow and the moment of maximal end-diastolic counts was defined as AC-interval andcorrelated with the PQ-interval on the electrocardiogram. The relative filling volumes withinthese time periods were very closely related in all patients (r=0.99, p<0.0001). Also, thecorrelation between PQ-interval and AC-interval was statistically significant (r=0.82, p<0.0001).In a subset of patients PQ-interval and AC-interval were not exactly the same. In thesepatients, AC-interval was always longer than PQ-interval, indicating the existence of passivediastasis flow before the onset of atrial contraction. In patients with low heart rates this wasmore apparent than at higher heart rates. Despite the close relation of the PQ-interval on theelectrocardiogram and the AC-interval of the flow-volume loop, they may represent differententities. In radionuclide angiography, the PQ-interval is therefore a better parameter to definethe moment of onset of atrial activity than the AC-interval.

Nuclear Medicine Communications 1997;18:642-647

42 Chapter 4

n the non-invasive assessment of diastolic left ventricular function, thequantification of the parameters of both early and late diastolic filling play animportant role.1,2,3,4,5 The parameters of late diastolic filling are not only related to

atrial contraction, but also to left ventricular relaxation and compliance. Variousradionuclide angiographic and echocardiographic methods can be used to measurethe volume displacement during atrial contraction.2,3,6,7,8 The moment of onset of atrialcontraction is mostly determined by the shape of the measured curves.4,7,8,10,11,12,13 Theuse of an external reference point has not been applied to verify the exact duration ofatrial contraction.9 We incorporated the PQ-interval of the electrocardiogram into theflow-volume loop as derived from the radionuclide time-activity curve of the left ventricleto define the onset of atrial contraction and separate this from late diastolic passivefilling. The atrial contribution to left ventricular filling was thus quantified.

METHODSPatients. Thirty-four patients participated in the study. Baseline characteristics of thesepatients are presented in Table 4.1. Twelve patients (35%) were normal, 4 patients(12%) had coronary artery disease, and 18 patients (53%) had other underlyingdisease. No patients had clinically significant mitral valve disease or aorticregurgitation. All patients were in regular sinus rhythm.Radionuclide Angiography. In vivo labeling of red blood cells was performed byintravenous injection of 99mTc-pertechnetate (20-30 mCi), 30 minutes after the

intravenous administration of pyrophosphate. Image acquisition was obtained by agamma camera (Siemens Orbiter) with an all-purpose parallel-hole collimatorinterfaced with a Pinnacle computer (Medasys Inc, Ann Arbor, Mich.). The gammacamera was positioned in the left anterior oblique view with a caudal tilt to optimallyobtain separation of the left and right ventricle. The acquisition was gated to the QRS-complex of the electrocardiogram. To minimize count fall-off at the end of the cardiaccycle in the forward gated data, only a 5% cycle-length-window was accepted.14

Acquisition was completed after 150,000 counts per frame of 20 msec duration. Theregion of background correction was selected manually.15 After temporal smoothingwith a five-harmonic Fourier fit of the raw data radionuclide time-activity curves of theleft ventricle were generated from the cardiac images.16

I

TABLE 4.1. Baseline characteristics of the 34 patients.Age (years) 57 ± 13 (mean ± SD)Sex (male/female) 17/17 (50/50%)Previous history:

No CAD (normal CAG) 12CAD 4Hypertension 6Others 12

CAD = coronary artery disease; CAG = coronary angiogram.

Quantification of atrial contribution to diastolic filling 43

Radionuclide image analysis. From the time-activity curve of the left ventricle the firstderivative curve (time-flow curve) and then a flow-volume loop were constructed(Figures 4.1A and 4.1B). Both flow and volume are expressed as fractions of theirmaximal value during one cycle. The flow-volume loop is thus largely independent of theabsolute flow and volume parameters and ejection fraction.2,17,18 The plot of flow againstvolume describes their relationship throughout the cardiac cycle. It can therefore beused to study the relative volume portions of early and late diastolic filling. Electrico-mechanical delays as well as gating delays were accounted for by moving the time-activity curves so that end-diastolic counts coincided with the beginning of the QRS-complex on the electrocardiogram.14,15

Measurements. Left ventricular ejection fraction (LVEF) was automatically determinedwith help of the time-activity curve. Atrial contraction time (AC-interval) was defined asthe time interval between the diastasis point and the moment of maximal counts (beingend-diastole). The diastasis point was found graphically in the flow-volume loop by aidof the computer, as it is represented by the lowest flow point between early and latediastolic filling. The moment of maximal counts was found analytically.

The PQ-interval was measured during data acquisition from anelectrocardiographic rhythm strip by two observers with aid of a magnifying-glass andcalibration scale. The PQ-interval was then plotted into the flow-volume loop. This wasused to define two subgroups: A) patients in which AC-interval and PQ-interval wereapproximately the same (difference of < 10 msec) and B) patients in which AC-intervaland PQ-interval differed > 10 msec (Figure 1B).

FIGURE 4.1A. Left ventricular time-activity curve and first derivative (time-flow curve) of study no. 31without compensation for gating and electrico-mechanical delays. After early diastolic filling (EDF), theflow diminishes in the diastasis (D). Late diastolic filling (LDF) is a composition of passive flow and atrialcontraction. EDV = end diastolic volume. B. Flow-volume loop of study no. 31. The PQ-interval (PQ) ismeasured on the electrocardiogram and the AC-interval (AC) is measured in the flow-volume loop. Here,AC-interval is longer than PQ-interval, probably due to passive filling before the onset of atrial contraction.

A B

44 Chapter 4

The relative contribution of late diastolic filling to total diastolic filling was thendetermined by using both AC-interval and PQ-interval as a reference point (AC% andPQ%). These relative volume portions were automatically calculated by the computer byintegration of the Fourier function.

Statistical analysis. Measured data are presented as mean ± 1 standard deviation(SD) unless otherwise stated. Comparison of the baseline characteristics and theradionuclide angiographic results between the subgroups was performed with Student'st test. For the correlations between the parameters of relative contribution of latediastolic filling to total diastolic filling and atrial contraction duration the Pearsoncorrelation was used. To evaluate that AC-interval was always longer than PQ-interval,and never shorter, a sign-test was used. A test for declining probabilities was used toevaluate the chance of AC-interval to exceed PQ-interval in relation with heart rate.

RESULTSAnalysis of all patients. A scatter diagram for the relation between PQ-interval andAC-interval is presented in Figure 4.2 (r=0.82, p<0.0001). AC% and PQ% were also

FIGURE 4.2. Scatter diagram showingthe relation of the PQ-interval on theelectrocardiogram and the AC-intervalin the flow-volume loop. Since a subsetof 11 patients had a longer AC-intervalthan PQ-interval, the linear regressionline for all the 34 patients was shiftedupward from the line: AC-interval = PQ-interval.

FIGURE 4.3. Scatter diagram showingthe relation of the late diastolic fillingfraction measured with two differentmethods. In PQ% the moment of onsetof late diastolic filling is defined withthe PQ-interval on theelectrocardiogram and in AC% with theAC-interval in the flow-volume loop.


very closely related (Figure 4.3: r=0.99, p<0.0001). Weak but statistically significantcorrelations were found between all measured late diastolic filling parameters (PQ-interval, AC-interval, PQ% and AC%) and heart rate (r=0.43 to 0.66, p<0.05). A positivecorrelation existed also between the atrial filling fractions, AC% and PQ%, and age(r=0.54 and 0.53 respectively, p<0.05). No correlations were found with LVEF or withthe other baseline characteristics.Subgroup analysis. In a subset of patients, AC-interval was longer than PQ-interval,but never shorter (sign-test p<0.01). In Table 4.2 the baseline characteristics and theradionuclide angiographic data of this subgroup are compared to the subgroup in whichAC-interval and PQ-interval were equal. Heart rate was the only statistically significant

TABLE 4.2. Comparison of the baseline characteristics and the findings between the subgroups atradionuclide angiography (mean ± SD).

PQ-interval=

AC-interval(Group A)

n=23

AC-interval>

PQ-interval(Group B)

n=11

p-value

Male/female 43/57% 64/36% NSAge (years) 55 ± 14 61 ± 10 NSLVEF 66 ± 12 62 ± 13 NSHR 67 ± 10 60 ± 8 <0.05Underlying disease 15 (65%) 8 (73%) NSMedication: ß-blockers 6 (26%) 4 (36%) NS Ca2+ channel blockers 3 (13%) 1 (9%) NSPQ-interval 177 ± 30 164 ± 18 NSAC-interval 177 ± 30 189 ± 20 NSPQ% 31 ± 11 27 ± 9 NSAC% 31 ± 11 28 ± 8 NS

LVEF = left ventricular ejection fraction; HR = heart rate; PQ-interval = PQ-interval on theelectrocardiogram; AC-interval = atrial contraction time measured in the flow-volume loop; PQ% = volumedisplacement during atrial contraction using the PQ -time as reference point; AC%= id. using AC-intervalas reference point.

FIGURE 4.4. Histogram showing thenumber of patients in each heart raterange of 10 bpm. The patients withdifferent values for AC-interval and PQ-interval had relatively slow heart rates(p<0.05).

46 Chapter 4

parameter in which the subgroups differed. It ranged from 51 to 85 beats per minute (35to 59 frames) in group A, and from 43 to 78 beats per minute (38 to 69 frames) in groupB (p<0.05). In group B there was no correlation between heart rate and the magnitudeof the difference between AC-interval and PQ-interval. In Figure 4.4 a frequencydistribution of the patients of both subgroups is presented for each heart rate interval of10 bpm. The probability that AC-interval exceeded PQ-interval increased with loweringheart rates (p<0.01). In 55 percent of the patients with heart rates < 60 beats per minuteAC-interval exceeded PQ-interval. In 82 percent of the patients with AC-intervalexceeding PQ-interval, heart rate was < 65 beats per minute.

DISCUSSIONIn this study we compared two radionuclide angiographic methods to quantify atrialcontribution to diastolic filling using flow-volume loops. Both the PQ-interval on theelectrocardiogram and the AC-interval on the flow-volume loop were used to define theonset of atrial contraction. Using the flow-volume loop, the relative volume portions oflate diastolic filling (AC% and PQ%) rather than the absolute filling rates werecalculated.2,17,18 Both these parameters correlated well. In a subset of patients withlower heart rates AC-interval however exceeded PQ-interval.

Previous measurements of atrial contribution to ventricular filling byradionuclide angiographiy. The assessment of late diastolic filling parameters byradionuclide blood pool scintigraphy has been established previously.2,4,9,17,18 Green etal. described a method using flow-volume loops.17,18 They assumed that the temporalcomponent of electrical atrial activity to ventricular activity equals that of the subsequentmechanical activity. Bonow et al. corrected their time-activity curve for gating delays.4,9

They also used the PQ interval to define the moment of onset of atrial contraction.Electrico-mechanical delay was standardized to 40 ms. Arora et al. on the other handused the diastasis point in the flow-volume loop to define the moment of onset of atrialactivity.2

Echocardiographic measurement of atrial function. Late diastolic fillingparameters can also be obtained with Doppler echocardiography.10,11,19,20,21,22 Withthis approach, the volume displacement during atrial contraction can indirectly bemeasured by integration of the early and late transmitral flow velocity curve andcalculation of this ratio.11 However, a filter is used to eliminate the low frequency signal.The measurement of the duration of late diastolic filling may then become inaccuratesince the presence of a diastasis flow might be overlooked.5,10,20,21

The correlation of AC-interval and PQ-interval. A strong correlation was foundbetween AC-interval and PQ-interval in the present study.2,4,9,17,18 This correlation


represents atrial electrico-mechanical coupling. However, in a subset of 11 of 34 (32%)patients AC-interval and PQ-interval differed. This may partly be related to the low heartrates of our patients. In 32 percent of the patients heart rate was < 60 beats per minute.With lowering heart rates, the diastasis period may become longer.1 The accuracy ofthe measurement of the diastasis point in the flow-volume loop may then decrease. Thedifference between AC-interval and PQ interval may also result from the correction wemade for electrico-mechanical delay of ventricular activation. The electrico-mechanicaldelay of atrial activation does not need to be exactly the same. However, in respect ofthe exact match of PQ-interval and AC-interval in the majority of the performed studiesthis explanation is rather unlikely.

In all patients in which AC-interval and PQ-interval differed, the former alwaysexceeded the latter, but never vice versa. In other words, the onset of an earlier atrial-ventricular flow can sometimes be detected before the beginning of electrical atrialactivity. This passive diastasis flow has been described before and can be expected tooccur in patients with low heart rates, increased ventricular stiffness, and increasedventricular relaxation rate.6,19,20,21,22 It may therefore be assumed that the AC-intervalwithin the flow-volume loop represents the beginning of a combination of diastasis flowand atrial contraction flow. The PQ-interval on the electrocardiogram can be considereda better parameter of atrial contraction time than AC-interval.

Although it is conceivable to define the atrial contraction flow as beginning afteratrial electrical activity, the measured flow in late diastole may remain a composition ofpassive diastasis flow and atrial contraction flow. One must therefore be cautious todefine atrial contractile function on the basis of late diastolic flow only. This drawback isshared by other radionuclide angiography derived parameters on late diastolic fillingsuch as peak late filling rate.

Limitations. There are two potential limitations to the present study. First, themethodological limitations of radionuclide angiography should not be underestimated.The assessment of global left ventricular filling requires a high temporal resolution andsmall cycle length fluctuations. Secondly, in this study different patient groups wereincluded, which may impose on the accuracy of the assessment of diastolic fillingpatterns. However, no differences between these parameters were found between thevarious subgroups.

Conclusions. A close correlation was observed between the PQ-interval on theelectrocardiogram and the atrial contraction time (AC-interval), a derivative of the shapeof the left ventricular time-activity curve. In addition, a very close correlation was presentbetween the subsequently calculated relative contributions of late diastolic filling to totaldiastolic filling. In patients with low heart rates however, the probability of AC-intervalexceeding PQ-interval increased. This is likely due to a relatively small diastasis flow. Itappears therefore better to use only the PQ-interval on the electrocardiogram, after

48 Chapter 4

correction for electrico-mechanical delay, to define the beginning of atrial contractionrather than the AC-interval.

ACKNOWLEDGMENTSWe wish to thank Oebele Dijkstra and Henk Louwes for their technical assistence withthe image analysis.

References

1. Bonow RO. Radionuclide Angiographic Evaluation of Left ventricular diastolic function.Circulation 1991;84(supplI):I-208-I-215.

2. Arora RR, Machac J, Goldman ME, Butler RN, Gorlin R, Horowitz SF. Atrial kinetics and leftventricular diastolic filling in the healthy elderly. J Am Coll Cardiol 1987;9:1255-1260.

3. Thomas JD, Weyman AE. Echocardiographic doppler evaluation of left ventricular diastolicfunction. Physics and physiology. Circulation 1991;84:977-990.

4. Bonow RO, Vitale DF, Bacharach SL, Maron BJ, Green MV. Effects of aging on asynchronousleft ventricular regional function and global ventrcular filling in normal human subjects. J Am CollCardiol 1988;11:50-58.

5. Störk TV, Müller RM, Piske GJ, Ewert CO, Hochrein H. Noninvasive measurement of leftventricular filling pressures by means of transmitral pulsed doppler ultrasound. Am J Cardiol1989;64:655-660.

6. Yellin EL, Nikolic SD. Diastolic suction and the dynamics of left ventricular filling. In: GaaschWH, LeWinter MM, eds. Left ventricular diastolic dysfunction and heart failure. Philadelphia: Lea& Febiger, 1994:89-102

7. Kuo LC, Quinones MA, Rokey R, Sartori M, Abinader EG, Zoghbi WA. Quantification of atrialcontribution to left ventricular filling by pulsed doppler echocardiography and the effect of age innormal and diseased hearts. Am J Cardiol 1987;59:1174-1178.

8. Miyatake K, Okamoto M, Kinoshita N, Owa M, Nakasone I, Sakakibara H, Nimura Y.Augmentation of atrial contribution to left ventricular inflow with aging as assessed byintracardiac doppler flowmetry. Am J Cardiol 1984;53:586-589.

9. Bonow RO, Frederick TM, Bacharach SL, Green MV, Goose PW, Maron BJ, Rosing DR. Atrialsystole and left ventricular filling in hypertrophic cardiomyopathy: effect of verapamil. Am JCardiol 1983;51:1386-1391.

10. Shapiro EP, Effron MB, Lima S, Ouyang P, Siu CO, Bush D. Transient atrial dysfunction afterconversion of chronic atrial fibrillation to sinus rhythm. Am J Cardiol 1988;62:1202-1207.

11. Mantero A, Gentile F, Gualtierotti C, Azzolini M, Barbier P, Beretta L, Casazza F, Corno R,Giagnoni E, Lippolis A, Lombroso S, Mattioli R, Morabito A, Ornaghi M, Pepi M, Pezzano A.Left ventricular diastolic parameters in 288 normal subjects from 20 to 80 years old. Eur Heart J1995;16:94-105.

12. Bonow RO, Vitale DF, Maron BJ, Bacharach SL, Frederick TM, Green MV. Regional leftventricular asynchrony and impaired global left ventricular filling in hypertrophic cardiomyopathy:effect of verapamil. J Am Coll Cardiol 1987;9:1108-1116.

13. Friedman BJ, Drinkovic N, Miles H, Shih WJ, Mazzoleni A, De Maria AN. Assessment of leftventricular diastolic function: comparison of doppler echocardiography and gated blood poolscintigraphy. J Am Coll Cardiol 1986;8:1348-1354.



15. Wagner RH, Halama JR, Henkin RE, Dillehay GL, Sobotka PA. Errors in determination of leftventricular functional parameters. J Nucl Med 1989;30:1870-1874.

16. Bacharach SL, Green MV, Vitale D, White G, Douglas MA, Bonow RO, Larson SM. Optimumfourier filtering of cardial data: a minimum-error method: concise communication. J Nucl Med1983;24:1176-1184.

17. Green MV, Juni JE, Goose PW, Bacharach SL, Bonow RO. A method for surveying patientpopulations for differences in global left ventricular function. In: Ripley K, Ostrow H, eds.Computers in cardiology 1982. Silver Springs, MD: IEEE Computer Society Press, 1982:307-310.

18. Green MV, Findley SL, Bonow RO, Bacharach SL, Juni JE. Left ventricular flow-volume relationsin normal subjects and patients with heart disease. In: Ripley K, Ostrow H, eds. Computers incardiology 1983. Silver Springs, MD: IEEE Computer Society Press 1983:295-298.

19. Yellin EL, Nikolic S, Frater RWM. Left ventricular filling dynamics and diastolic function. ProgCardiovasc Dis 1990;32:247-271.

20. Yellin EL, Meisner JS, Nikolic SD, Keren G. The scientific basis for the relations betweenpulsed-doppler transmitral velocity patters and left heart chamber properties. Echocardiography1992;9:313-338.

21. Keren G, Meisner JS, Sherez J, Yellin EL, Laniado S. Interrelationship of mid-diastolic mitralvalve motion, pulmonary venous flow, and transmitral flow. Circulation 1986;74:36-44.

22. Meisner JS, Keren G, Pajaro OE, Mani A, Strom JA, Frater RWM, Laniado S, Yellin EL. Atrialcontribution to ventricular filling in mitral stenosis. Circulation 1991;84:1469-1480.

Chapter 5Circadian rhythm in

left ventricular relaxation ofpatients with congestive heart failure:

diagnostic and therapeutic implicationsHJ Muntinga1, F van den Berg2, HR Knol3, MG Niemeyer2, AJM Cleophas4, andEE van der Wall5

1 Department of Cardiology, Medisch Spectrum Twente, Enschede, 2 Department of Cardiology, MartiniHospital, Groningen, 3 Northern Centre for Health Care Research, University of Groningen, Groningen, 4

Department of Internal Medicine, Merwede Hospital, Dordrecht, 5 Department of Cardiology, UniversityHospital Leiden, Leiden, the Netherlands.

Summary

Background: Since afterload is an important determinant of left ventricular (LV) relaxation, thephysiological circadian variation of arterial blood pressure may induce diurnal variability of LVfilling parameters in patients with congestive heart failure (CHF).Methods: In 15 patients with chronic stable CHF (NYHA class II or III) and reduced LV ejectionfraction (<40%) due to previous myocardial infarction, treated with either a calcium channelblocker or placebo, radionuclide angiography was performed at 10 AM and at 1 PM to assesswhether or not LV filling indices follow a diurnal rhythm. Blood pressure was measuredsimultaneously. This procedure was repeated after one week.Results: At 10 AM compared to 1 PM a higher systolic blood pressure (SBP, 138 ± 16 vs 131 ± 15mmHg, P<0.01) and diastolic blood pressure (DBP, 75 ± 10 vs 71 ± 11 mmHg, P<0.01), a lowerpeak filling rate (PFR, 0.88 ± 0.30 vs 1.02 ± 0.29 end diastolic volume/second, P<0.01), and alonger time to peak filling rate (TPFR, 598 ± 74 vs 574 ± 61 millisecond, P<0.05) were found.Using repeated measures analysis of variance, the changes of PFR were related to the dailychanges of SBP and DBP (each P<0.05).Conclusion: These data suggest that LV relaxation as measured by PFR and TPFR is inverselycorrelated with a circadian fall in blood pressure in patients with CHF. This could have theimportant diagnostic implication that the timing should be taken into account when assessingdiastolic LV function in patients with CHF, as well as the therapeutic implication that treatmentregimens should account for an increased risk of pulmonary edema in these patients early inthe morning.

European Journal of Internal Medicine 1998;9:91-97

Circadian Rhythm in Left Ventricular Relaxation 51

circadian rhythm exists in physiological mechanisms such as blood pressure,vagal and sympathetic action, and heart rate,1,2,3,4 and has been associatedwith the circadian occurrence of cardiac events such as myocardialinfarction,5,6,7 transient myocardial ischemia,8,9 and sudden cardiac death.10

Recently, a circadian variation of left ventricular (LV) performance as measured bydiastolic function in healthy voluntary subjects has been observed.11 The existence of adiurnal variation of LV performance in patients with congestive heart failure (CHF) couldhave important implications for the appropriate timing of diagnostic assessments. Inaddition, it may have therapeutic consequences for patients with imminent pulmonaryedema due to cardiac failure. However, no such data are available to date. Accordingly,the aim of the present study was to investigate the presence of such a diurnal variationof LV filling in patients with stable CHF. For that purpose we analyzed the data of aradionuclide angiography (RNA) study in patients with stable cardiac failure treated witheither placebo or incremental dosages of calcium channel blocker. The results of thisstudy on the effects of calcium channel blockade on diastolic function compared toplacebo have been published,12 and did not reveal important differences. For thepurpose of the current paper the data of the two groups were taken together andanalyzed for circadian differences in diastolic function as measured by radionuclideangiography. Since arterial blood pressure is an important determinant of LVrelaxation,13 and in addition, the load dependence of relaxation in heart failure isincreased,14 we emphasized the relation between the levels of blood pressures and LVfilling.

METHODSPatients. Fifteen patients with chronic CHF, NYHA class II or III, were included in thestudy. Baseline demographic characteristics of the study population are previouslypublished.12 All patients had sustained a myocardial infarction > 3 months prior to theinvestigation. The LV ejection fraction (EF) was < 40%. Exclusion criteria were unstableangina pectoris, clinically significant valvular disease including mitral regurgitation,hypertrophic obstructive or dilated cardiomyopathy, moderate or uncontrolledhypertension, clinically significant arrhythmia’s including atrial fibrillation, bradycardia,and atrioventricular block of any degree. All calcium channel blockers and ß-blockingagents, were withdrawn at least 5 half-life’s before the start of trial. All other oralmedication was continued.

Study design. The patients underwent RNA at approximately 10 AM and 1 PM. DuringRNA resting brachial artery cuff systolic and diastolic blood pressure (SBP and DBPrespectively) were measured and mean arterial pressure (MAP) was calculated. Heartrates and heart rate pressure products were also determined. After the first RNA, five ofthe patients were treated with placebo and ten of them with mibefradil in doses rangingfrom 6.25 to 100 mg once daily for one week. All of the patients were measured again

A

52 Chapter 5

at the completion of the week of treatment.

Radionuclide Angiography. To evaluate LV systolic and diastolic function rest supinemultigated RNA images were obtained according to the protocol we described in ourprevious work.12 Shortly, the studies were made with forward gating and a 5% cycle-length-window. After 150,000 counts per 20 ms frame acquisition was completed.Using five Fourier harmonics temporal smoothing was achieved.

Analysis of data. From the LV time-activity curve and the first derivative of this curvewe derived the parameters of systolic and diastolic function (Figure 5.1). EF wasdefined as the ratio of stroke volume (SV) and end diastolic volume (EDV). Atrialcontribution (AC) was expressed as a fraction of SV. The first derivative of the volumecurve was depicted as a flow curve. In this, peak filling rate (PFR) was defined themaximum instantaneous flow during early rapid filling, and peak ejection rate (PER) asthe maximum flow during LV emptying. Time to peak filling rate (TPFR) was measuredfrom the beginning of emptying, and was divided into two parts: time to end of systole(TES), and time from end of systole to peak filling rate (TPFR').

84,0

100,0

-0,08 0,12 0,32 0,52 0,72

10 AM

LV

vo

lum

e (%

ED

V)

SV

AC

EDV

84

100

0 0,2 0,4 0,6 0,8

1 PM

0

-0,80

0,00

0,80

0 0,5

Time (s)

Fill

ing

rat

e (E

DV

/s)

TPFR

PFR

PER

TES TPFR'

-0,8

0

0,8

0 0,5

Time (s)

FIGURE 5.1. Redrawn time-activity curves of the left ventricle obtained from one of the study patients(top) and their first derivatives (bottom) at 10 AM (left) and at 1 PM (right). Left and right time axis andvertical axis are identical. AC indicates atrial contribution to diastolic filling; EDV, end diastolic volume;PER, peak ejection rate; PFR, peak filling rate; sec, seconds; SV, stroke volume; TES, time to end ofsystole; TPFR, time to peak filling rate measured from the beginning of emptying; TPFR’ time from end ofsystole to peak filling rate. At 1 PM compared to 10 AM left ventricular filling is enhanced as shown byincreased PFR and decreased TPFR, despite unchanged left ventricular ejection fraction (SV/EDV).


Statistical analysis. Since the same variable was obtained on several occasionsstatistical analysis was performed using repeated measures analysis of variance.Circadian rhythm of a parameter with similar changes on both study days was testedwith a cubic trend. With this trend similarly directed variability is tested on repeatedoccasions. To test whether the change patterns of the parameters equaled the variationof blood pressure we added mean blood pressure as a co-variate to the repeatedmeasures model. Because of the use of co-variation of differences we tested in thissituation with a quadratic trend. In a sequence of three differences between followingmeasurements this trend tests opposite parameter directions. Also, the measurementsof 10 AM from both study days, and those of 1 PM were taken together. Differenceswere tested by paired t tests. For the correlations between the changes between 10 AMand 1 PM the Pearson correlation was used (r). All of the data were expressed as mean± 1 SD unless otherwise stated. A P value <0.05 was considered as statisticallysignificant. Values of 0.05 to 0.10 were also stated, since they may indicate a trend ofthe measured variable.

RESULTSOf the fifteen patients who entered the study mean EF was low (25 ± 7%). As therewere no significant differences between the effects of mibefradil and placebo on bloodpressure or LV function, the data of the two groups were analyzed both separately andtogether (Table 5.1).

Radionuclide angiographic variables. The repeated measures analysis of the

TABLE 5.1. Hemodynamic and radionuclide angiographic measurements in 15 patients with heart failure.The measurements at study start on 10 AM and those on 1 PM were taken together with themeasurements after one week of treatment with placebo (n=5) or a calcium channel blocker (n=10).

placebo mibefradil all measurements10 AM 1 PM 10 AM 1 PM 10 AM 1 PM

SBP (mmHg) 130 ± 12 124 ± 11 141 ± 16 135 ± 15 138 ± 16 131 ± 15**DBP (mmHg) 76 ± 8 71 ± 6** 75 ± 11 71 ± 12 75 ± 10 71 ± 11**MAP (mmHg) 94 ± 8 89 ± 7* 97 ± 10 92 ± 10 96 ± 10 91 ± 9**HR (beats/min) 71 ± 7 75 ± 5* 70 ± 14 72 ± 13 70 ± 12 73 ± 11*RPP (mmHg.bpm/1000) 9.3 ± 1.5 9.3 ± 1.3 10.0 ± 2.3 9.7 ± 1.8 9.7 ± 2.1 9.5 ± 1.7EF (%EDV) 27 ± 13 27 ± 12 30 ± 7 31 ± 6 29 ± 9 30 ± 8PER (EDV/s) -1.2 ± 0.5 -1.3 ± 0.5 -1.4 ± 0.3 -1.5 ± 0.3 -1.3 ± 0.4 -1.5 ± 0.4**PFR (EDV/s) 0.79 ± 0.32 1.03 ± 0.33 0.92 ± 0.28 1.01 ± 0.27 0.88 ± 0.30 1.02 ± 0.29**TPFR (ms) 589 ± 29 570 ± 53 602 ± 88 576 ± 64 598 ± 74 574 ± 61*TES (ms) 422 ± 55 379 ± 69* 387 ± 55 385 ± 58 399 ± 57 383 ± 62TPFR' (ms) 167 ± 41 192 ± 31 215 ± 86 191 ± 60 199 ± 78 191 ± 52AC (%SV) 40 ± 17 40 ± 19 42 ± 13 43 ± 15 41 ± 15 42 ± 17p<0.05 vs. 10 AM; ** p< 0.01 vs. 10 AMAC = atrial contribution; bpm = beats per minute; DBP = diastolic blood pressure; EDV = end diastolicvolume; EF = ejection fraction; HR = heart rate; MAP = mean arterial pressure; PER = peak ejectionrate; PFR = peak filling rate; RPP = heart rate pressure product; SBP = systolic blood pressure; SV =stroke volume; TES = time to end of systole; TPFR = time to peak filling rate measured from thebeginning of emptying; TPFR' = TPFR measured from the beginning of filling.

54 Chapter 5

radionuclide angiographic data showed no diurnal change of EF. In contrast, astatistically significant cubic trend of the mean value of PER was demonstrated(P<0.01), indicating an equal variation of this parameters on both study days. For thePFR (P<0.06) and TPFR (P<0.08) a tendency towards a cubic trend was obtained. Thediurnal change of these variables was consistently and equally present on both studydays. However, there was no diurnal variation present in the TES, TPFR', and AC. Themeasurements of both study days were then evaluated together. At 10 AM compared to1 PM a clear difference was showed with respect to PER, PFR and TPFR (Table 5.1).At 10 AM PER and PFR were lower, and TPFR was longer. There were no differencesbetween 10 AM and 1 PM with respect to EF, TES, TPFR', and AC.

Blood pressure, heart rate, and heart rate pressure product. The repeatedmeasures analysis revealed a statistically significant cubic trend for the means of thevalues of SBP (P<0.05), DBP (P<0.05), MAP (P<0.05), and heart rate (P<0.05). Thechange of value of these variables was equally directed on both study days, butpredominated on the first day. At 10 AM compared to 1 PM SBP, DBP, and MAP werehigher, and heart rate was slightly but statistically significant lower. Heart rate pressureproduct did not show a diurnal variation because of opposite variations of heart rateand blood pressure.

Blood pressure and heart rate vs RNA. The relation between variations of bloodpressure values and variations of PER and PFR was examined by adding mean valuesof SBP, DBP, and MAP as co-variates to the model of repeated measures of variance.In the change pattern of PFR, SBP and DBP were recognized as co-variates (both:P<0.05). Also MAP was recognized for co-variation (P<0.01). In the change pattern ofPER no co-variation for this parameters was demonstrated (all parameters: P>0.10).Weak but statistically significant correlations were found between the change from 10AM to 1 PM of the mean PFR, and the change of the mean DBP (r=-0.44, P<0.05),MAP (r=-0.38, P<0.05), and heart rate (r=0.37, P<0.05), but not with SBP (r=-0.169,P>0.10).

DISCUSSIONThe aim of the present study was to investigate the possible existence of a day-timevariability of LV filling in patients with CHF. For this purpose we used RNA which is areliable noninvasive technique of measuring LV filling and may be used to serially followdiastolic LV function.15 The parameters of early diastolic filling, PFR and TPFR’, arecorrelates of left ventricular relaxation, i.e. of early diastolic LV function.16 In manyclinical conditions of LV diastolic failure, prolonged contraction and impaired relaxationare, although conceptually different, difficult to distinct, and may lead to prolongation ofthe contraction-relaxation cycle.17 In the present study TPFR is used to measure this.The analysis of the data resulted in the observation of a diurnal variation of LV filling


parameters. The observed changes in PFR and TPFR suggest the presence of adecreased LV relaxation in the morning when compared with early afternoon. Alsodiurnal fluctuations of blood pressure and heart rate, but not of heart rate pressureproduct, were found. The diurnal changes of LV filling and the diurnal changes of bloodpressure and heart rate were interrelated.

To our knowledge no previous studies on diurnal physiological variations ofdiastolic cardiac function have been performed in patients with CHF. In healthyvolunteers, however, a recent study showed diurnal changes of LV relaxation suggestiveof a circadian rhythm.11 The proposed underlying mechanism might have been the day-night cycle in sympathoadrenal activity. The circadian changes in blood pressure inthese volunteers were too small to explain the diurnal variation of LV filling. Althoughearlier investigators have reported a significant decrease in LV relaxation due to anincrease in afterload,18,19 early LV filling results in variable responses to blood pressureelevations in subjects with normal LV systolic function without mitral regurgitation.20 Inheart failure, however, the load dependence of relaxation may be enhanced.14,21

Because patients with heart failure without mitral regurgitation were concerned in ourstudy, the variations of blood pressure may have had a direct effect on the diastoliccardiac function parameters. The variations in PFR were not only related to bloodpressure but to heart rate as well. Earlier reports on a correlation between heart rateand PFR were based on observations made during exercise in patients with heart ratesabove 90 beats per minute.22,23 In these high heart rate patients, loss of diastasismakes the measurement of PFR more sensitive to heart rate changes.22 In our study noexercise testing was performed. So, the heart rates were well below 90 beats perminute, and every time-activity curve showed clear signs of diastasis. Because of thesemethodological limitations in our material, our results do not allow to indicate the natureof the relation between heart rate and PFR. Besides the interrelation of heart rate andPFR, associations of high heart rate and low PFR with increased AC are oftendescribed.15,24. In the present study however AC was not associated with heart rate orPFR because these parameters lead to opposite effects on AC. The day-timefluctuations of blood pressure and heart rate as demonstrated studies of healthysubjects, are probably related to a circadian variation in vagal and sympathetic cardiaccontrol, and are modified by physiological actions, such as exercise and food intake.1,2

Compared to healthy volunteers, patients with heart failure due to coronary arterydisease may be subject to attenuation of circadian variability of blood pressure andheart rate.25 Nevertheless, the present data show a diurnal variation of blood pressureand heart rate and a trend towards diurnal variability in LV filling in patients with severelyimpaired LV function. During ß-adrenergic stimulation in human subjects LV earlydiastolic pressure is reduced by changes in contractility and LV relaxation, causing anincreased left atrial-ventricular pressure gradient.26 This in turn augments the rate andmagnitude of early diastolic filling.16,26 Augmentation of LV filling parameters, like PFRand TPFR’ at 1 PM in the present study, are therefore expected in the presence of ß-

56 Chapter 5

adrenergic stimulation. Because ß-adrenergic stimulation also enhances contractility,the entire contraction-relaxation cycle may be shortened,17 thereby decreasing TPFR inthe present study as well. Circadian variability of sympathoadrenal activity, although notmeasured, may be directly responsible for the observed changes in LV filling. Theincreased PFR and shortened TPFR of 1 PM however were associated with andcorrelated to decreased blood pressure. Therefore the decreased PFR and TPFR at10 AM compared to 1 PM may indicate the presence of diastolic failure with systoliccompensation in response to an increased afterload, instead of a directsympathoadrenal effect on diastolic LV relaxation. In this situation increased end-diastolic LV pressure increases risk of pulmonary edema.17 The absence of a circadianrhythm in TPFR’ in the present population may be of methodological nature. Because inpatients with coronary artery disease increased regional systolic asynchrony isassociated with prolonged isovolumic relaxation 27 or early segmental relaxation,24 thiscondition may increase the inaccuracy of estimation of TES. Although the presentglobal assessment of LV filling does not allow conclusions on regional asynchrony, thehigher standard deviation of TES compared to the standard deviation of TES in anormal population measured in the same laboratory,15 may indicate the presence ofprolonged isovolumic relaxation time.

The occurrence of transient myocardial ischemia also shows a circadianvariation which may be related to variations in LV filling.28,29,30 Although in ourpopulation no patients complained of chest pain or other symptoms of myocardialischemia during the follow-up period, transient myocardial ischemia may remainunnoticed. In earlier studies a circadian variation in the onset of silent and symptomaticischemia was described with peaks between 8 and 10 AM, and 4 and 5 PM.28

Myocardial ischemia decreases relaxation causing a reduction of PFR and an increaseof TPFR,29,30 thereby giving a possible explanation for our findings. However, nochanges of rate pressure product, which is a measure of myocardial oxygen demand,were found between 10 AM and 1 PM. In addition, if the observed variability of bloodpressure and LV filling parameters was due to silent myocardial ischemia, theadministration of mibefradil, a potent anti-ischemic calcium channel blocker with asignificant effect on silent myocardial ischemia,31 would have resulted in differences ofPFR and TPFR between placebo and drug treated patients.

Although it cannot be completely excluded that small variations of bloodpressure, heart rate, and LV function may be related to the administration of the studydrug, the influence of this mechanism is probably negligible. The slight reduction ofheart rate during mibefradil (p<0.05), can definitely not explain all of the circadianchanges observed in our material. The fact that the differences between the variousdose groups and placebo with regard to blood pressure and RNA data were small andunsignificant,12 suggests that the measured variations are at least partly due to effectson the measured parameters in patients with mibefradil coinciding with parallel effectsin the placebo group (Table 5.1). Although all other calcium channel blockers and ß-


blocking agents were withdrawn before the start of the trial, other administered drugs,including angiotensin-converting enzyme inhibitors, diuretics, long-acting nitrates, anddigoxin, might have had a confounding influence on the study result. In view of the factthat both type of medication and moment of intake were diverse, effects on bloodpressure, heart rate and left ventricular filling were probably heterogeneous. This, in fact,supports that the variation of blood pressure and LV filling was a circadian phenomenonrather than a drug associated effect. We therefore assume that a circadian fall of bloodpressure and a rise of diastolic performance reflect a true circadian phenomenon ratherthan an epiphenomenon of drug treatment.

The conclusions of the present study regarding circadian variability were basedon observations on diastolic LV filling and blood pressure and heart rate on 2 differentoccasions each study day. The detailed assessment of circadian variability of LVdiastolic function in patients with stable congestive heart failure however requires morefrequent measurements at regular intervals on 1 study day, and not only during day-time.To confirm the findings of the present study, future studies of circadian rhythm inpatients with congestive heart failure should meet this criteria. A greater number ofpatients entering the study would possibly increase the statistical power. Conclusions. In the present study we found a circadian rhythm of RNA parameters ofLV filling in patients with CHF due to previous myocardial infarction. This is probablycaused by a physiological diurnal variation of blood pressure, as demonstrated in ourpatients, producing an increased afterload for the LV early in the morning. This mayhave the important implication that the timing should be taken into account whenassessing diastolic LV function in patients with CHF. Also, it may have the therapeuticconsequence that treatment regimens should account for an increased risk ofpulmonary edema in these patients early in the morning.

ACKNOWLEDGMENTSThe authors wish to thank Oebele Dijkstra, Hans Schuurman and Henk Louwes from thedepartment of nuclear medicine in the Martini Hospital Groningen for their advice andtechnical assistance.

58 Chapter 5

References

1. Millar-Craig MW, Bishop CN, Raftery EB. Circadian variation of blood pressure. Lancet1978;1:795-797.

2. Hayano J, Sakakibara Y, Yamada M, Kamiya T, Fujinami T, Yokoyama K, Watanabe Y, TakataK. Diurnal variations in vagal and sympathetic cardiac control. Am J Physiol 1990;258 (HeartCirc Physiol 27):H642-H646.

3. Weber MA, Drayer JI, Nakamura DK, Wyle FA. The circadian blood pressure pattern inambulatory normal subjects. Am J Cardiol 1984;54:115-119.

4. Furlan R, Guzzetti S, Crivellaro W, Dassi S, Tinelli M, Baselli G, CeruttiS, Lombardi F, PaganiM, Malliani A. Continuous 24-hour assessment of the neural regulation of systemic arterialpressure and RR variabilities in ambulant subjects. Circulation 1990;81:537-547.

5. Muller JE, Stone PH, Turi ZG, Rutherford JD, Czeisler CA, Parker C, Poole WK, Passamani E,Roberts R, Robertson T, Sobel BE, Willerson JT, Braunwald E, and the MILIS Study Group.Circadian variation in the frequency of onset of acute myocardial infarction. N Engl J Med1985;313:1315-1322.

6. Hansen O, Johansson BW, Gullberg B. Circadian distribution of onset of acute myocardialinfarction in subgroups from analysis of 10,791 patients treated in a single center. Am J Cardiol1992;69:1003-1008.

7. Gilpin EA, Hjalmarson A, Ross Jr J. Subgroups of patients with atypical circadian patterns ofsymptom onset in myocardial infarction. Am J Cardiol 1990;66:7G-11G.

8. Rocco MB. Timing and triggers of transient myocardial ischemia. Am J Cardiol 1990;66:18G-21G.

9. Stone PH. Triggers of transient myocardial ischemia: circadian variation and relation to plaquerupture and coronary thrombosis in stable coronary artery disease. Am J Cardiol 1990;66:32G-36G.

10. Aronow WS, Ahn C. Circadian variation of primary cardiac arrest or sudden cardiac death inpatients aged 62 to 100 years (mean 82). Am J Cardiol 1993;71:1455-1456

11. Voutilainen S, Kupari M, Hippelainen M, Karppinen K, Ventila M. Circadian variation of leftventricular diastolic function in healthy people. Heart 1996;75:35-39.

12. Muntinga HJ, Van der Vring JAFM, Niemeyer MG, Van den Berg F, Knol HR, Bernink PJLM,Van der Wall EE, Blanksma PK, Lie KI. Effect of mibefradil on left ventricular diastolic functionin patients with congestive heart failure. J Cardiovasc Pharmacol 1996;27:652-656.

13. Brutsaert DL, Sys SU. Relaxation and diastole of the heart. Physiol Rev 1989;69:1228-1315.

14. Eichhorn EJ, Willard JE, Alvarez L, Kim AS, Glamann DB, Risser RC, Grayburn PA. Arecontraction and relaxation coupled in patients with and without congestive heart failure?Circulation 1992;85:2132-2139.

15. Muntinga HJ, Van den Berg F, Knol HR, Niemeyer MG, Blanksma PK, Louwes H, Van der WallEE. Normal values and reproducibility of left ventricular filling parameters by radionuclideangiography. Int J Cardiac Imaging 1997;13:165-171.

16. Ishida Y, Meisner JS, Tsujioka K, Gallo JI, Yoran C, Frater RWM, Yellin EL. Left venricular fillingdynamics: influence of left ventricular relaxation and left atrial pressure. Circulation 1986;74:187-196.

17. Brutsaert DL, Sys SU, Gillebert TC. Diastolic failure: pathofysiology and therapeuticimplications. J Am Coll Cardiol 1993;22:318-325.




20. Takenaka K, Shiota T, Sakamoto T, Hasegawa I, Suzuki J, Amano W, Sugimoto T. Effect ofacute systemic blood pressure elevation on left ventricular filling with and without mitralregurgitation. Am J Cardiol 1989;63:623-625.

21. Little WC. Enhanced load dependence of relaxation in heart failure. Clinical implications.Circulation 1992;85:2326-2328.

22. Udelson JE, Bonow RO. Radionuclide angiographic evaluation of left ventricular diastolicfunction. In: Gaasch WH, LeWinter MM ,ed. Left ventricular diastolic dysfunction and heartfailure. Philadelphia: Lea & Febiger, 1994;167-191.

23. Assennato P, Candela B, Hoffmann E, Spano C, Traina M, Rotolo A, Raineri AA. Left ventricularfilling rate at rest and during exercise in patients with previous myocardial infarction. Int J Cardiol1993;41:219-223.

24. Bonow RO. Regional left ventricular nonuniformity. Effects on left ventricular diastolic function inischemic heart disease, hypertrophic cardiomyopathy, and the normal heart. Circulation1990;81(suppl III):III-54-III-65.

25. Caruana MP, Lahiri A, Cashman PMM, Altman DG, Raftery EB. Effects of chronic congestiveheart failure secondary to coronary artery disease on the circadian rhythm of blood pressure andheart rate. Am J Cardiol 1988;62:755-759.

26. Udelson JE, Bacharach SL, Cannon RO 3d, Bonow RO. Minimum left ventricular pressureduring ß-adrenergic stimulation in human subjects: Evidence for elastic recoil and diatolic"suction" in the normal heart. Circulation 1990;82:1174-1182.

27. Perrone-Filardi P, Bacharach SL, Dilsizian V, Bonow RO. Effects of regional systolicasynchrony on left ventricular global diastolic function in patients with coronary artery disease. JAm Coll Cardiol 1992;19:739-744

28. Hausmann D, Nikutta P, Trappe HJ, Daniel WG, Wenzlaff P, Lichtlen PR. Circadian distributionof the characteristics of ischemic episodes in patients with stable coronary artery disease. Am JCardiol 1990;66:668-672.

29. Mahmarian JJ, Pratt CM. Silent myocardial ischemia in patients with coronary artery disease:possible link with diastolic left ventricular dysfunction. Circulation 1990;81(suppl III):III-33-III-40.

30. Bonow RO, Vitale DF, Bacharach SL. Asynchronous left ventricular regional function andimpaired global diastolic filling in patients with coronary artery disease: reversal after coronaryangioplasty. Circulation 1985;71:297-307.

31. Braun S, Van Der Wall EE, Emanuelsson H, Kobrin I, on behalf of the mibefradil internationalstudy group. Effects of a new calcium antagonist, mibefradil (Ro 40-5967), on silent ischemia inpatients with stable chronic angina pectoris: a multicenter placebo-controlled study. J Am CollCardiol 1996;27:317-322.

Chapter 6Effect of Mibefradil on

left ventricular diastolic function inpatients with congestive heart failure

H.J. Muntinga1, J.A.F.M. van der Vring1, M.G. Niemeyer1,2, F. van den Berg1, H.R.Knol3, P.J.L.M. Bernink1, E.E. van der Wall4, P.K. Blanksma5, K.I. Lie5.

1 Department of Cardiology, Martini Hospital, Groningen, 2 Department of Diagnostic Radiology andNuclear Medicine, University Hospital, Leiden, 3 Northern Centre for Healthcare Research, University ofGroningen, 4 Department of Cardiology, University Hospital, Leiden, 5 Department of Cardiology,University Hospital, Groningen, The Netherlands.

Summary

Background: Calcium antagonists are known to have anti-hypertensive and anti-anginalproperties. In heart failure however, their use can be hazardous, as systolic function candeteriorate. This may not be true for the new calcium antagonist Mibefradil, which has a newchemical structure. Calcium antagonists may also be beneficial for diastolic left ventricularfunction in coronary artery disease. In order to investigate the possible effects of Mibefradil ondiastolic left ventricular function we performed the present study in the setting of amulticenter, double blind, placebo controlled, multiple dose, safety trial.Methods: Fifteen patients with NYHA class II or III for dyspnoea and depressed ejection fraction(<40%) due to a previous myocardial infarction were investigated. The measured nuclearangiographic parameters included Ejection Fraction, Peak Ejection Rate and Peak FillingRate. Systolic blood pressure, diastolic blood pressure and heart rate were also obtained.Group I (5 patients) received placebo medication , group IIA (6 patients) Mibefradil 6.25, 12.5 or25 mg daily and group IIB (4 patients) 50 or 100 mg daily. Measurements were made beforeand after the first dose and after one week of treatment before and after the final dose.Results: Mibefradil clearly depressed Heart Rate (repeated measures analysis of variance:p<0.05). No statistically significant effects of Mibefradil were measured on blood pressures,systolic and diastolic left ventricular function.Conclusions: In our study conditions Mibefradil caused no worsening of systolic function andpreserved diastolic function in short-term treatment of patients with decreased EF and heartfailure.

Journal of Cardiovascular Pharmacology 1996:27;652-656

Effect of Mibefradil on diastolic LV function in CHF 61

s a result of myocardial infarction congestive heart failure can develop. Thismay be primarily due to diastolic dysfunction, or to a combination of diastolicand systolic dysfunction of the left ventricle.1,2,3 Diastolic dysfunction, in thiscontext, implies that the left ventricle is not able to accept blood at low

pressures. 1,4,5 Pulmonary congestion can thus develop.1,4,5 The use of calciumantagonists for the treatment of stable angina in such patients can be hazardousbecause of a potential deterioration of systolic function.6,7,8 This may not necessarily betrue in the case of second- and third-generation calcium antagonists, which might in factbenefit diastolic function in coronary artery disease.2,8,9

Mibefradil, a calcium antagonist from a new chemical structural class ofbenzimidazolyl-substituted tetraline derivatives, has been found to possess strong anti-anginal properties.10 In addition, no deterioration of systolic function has been reportedin patients with chronic stable angina pectoris and also in rats with chronic myocardialinfarction.10,11 To date, no reports have occurred concerning the safety of Mibefradil onsystolic and diastolic left ventricular function in patients with chronic heart failure. Theeffects of Mibefradil on diastolic left ventricular function in patients with congestive heartfailure due to earlier myocardial infarction were investigated in the present study.

METHODSPatients. Fifteen patients with chronic congestive heart failure, NYHA class II or III wereincluded in this study. All patients had suffered myocardial infarction more than threemonths prior to the investigation, which resulted in a left ventricular ejection fraction of40 percent or less. Exclusion criteria included: unstable angina pectoris; clinicallysignificant valvar disease, including mitral regurgitation; hypertrophic obstructive ordilated cardiomyopathy; moderate or uncontrolled hypertension; clinically significantarrhythmias or bradycardia and 1st or higher degree AV-block.

Study design. On day one of investigation patients were randomised in a double blinddesign for their study medication, which was either placebo (Group I, five patients) orMibefradil (Group II, ten patients). Those patients receiving Mibefradil were randomlyassigned in pairs to different dose groups. Six patients received Mibefradil in therelatively low doses of 6.25, 12.5 and 25 mg daily (Group IIA). The remaining fourpatients received the higher doses of 50 and 100 mg Mibefradil daily (Group IIB). Theythen underwent multigated nuclear angiography at approximately 10 AM. Three hourslater (1 PM) a second nuclear angiography was performed. For the following sevendays patients received their study medication once daily. On day eight a multigatednuclear angiogram was made before and three hours after ingestion of the last studytablet (10 AM and 1 PM). During nuclear angiography resting systolic blood pressure(SBP), diastolic blood pressure (DBP) and heart rate (HR) were also obtained. Meanarterial pressure (MAP) was then calculated.

A

62 Chapter 6

Nuclear angiography. Some of each patient's red blood cells were labelled with 99mtechnetium-pertechnetate, after intravenous administration of stannous pyrophosphate.The total dosage range was 550-740 MBq. Left ventricular function was evaluated byradionuclide angiography using a gamma camera (Siemens Orbiter) with an all-purpose parallel-hole collimator and interfaced with a Pinnacle computer (Medasys Inc,Ann Arbor). Multigated images in supine rest position were obtained in the left anterioroblique view with a caudal tilt, such that both left and right ventricles appearedseparated. Only a 5% cycle-length-window with forward gating was accepted.12

Acquisition was completed after 150,000 counts per frame of 20 ms. Temporalsmoothing was achieved using five Fourier harmonics.13,14

Measurements. From the systolic part of the left ventricular time-activity curve (TAC)we measured the Ejection Fraction (EF) and the Peak Ejection Rate (PER). Thediastolic parameters included the Peak Filling Rate (PFR) and the time to Peak FillingRate (t-PFR). PFR was normalised to end-diastolic volume as well as stroke volume.Atrial Contribution (AC) was measured as a percentage of the diastolic filling volume.

Statistical analysis. Measured data are expressed as mean ± 1 SD unless otherwisestated. For statistical analysis a model of repeated measures analysis of variance wasused since the same variable was obtained on several occasions. In this design thevariability due to differences between subjects can be eliminated from the experimentalerror. A p-value <0.05 was considered statistically significant. We tested the differencesin the means with the various Mibefradil dosages. In the case of an immediate effect ofMibefradil on the tested parameters we would expect a third order polynomial forgroup II. This would be more pronounced in the higher dosage group IIB. In the case of adelayed medication effect we would expect a linear change in the parameters of groupII.RESULTSPatients. The patients' baseline characteristics are presented in Table 6.1. All calcium

TABLE 6.1. Baseline characteristics of the 15 study patients.Age (years) 68 ± 12 (mean ± SD)Sex male/female 12/3 (80/20%)Functional class

NYHA II 14NYHA III 1

LV ejection fraction 25 ± 7%Location of myocardial infarction*

Inferoposterior 4 (27%)Anterior 14 (93%)Unknown 2 (14%)

Previous interventionsCABG 2 (14%)

* more than one infarction location per patient possible.CABG = coronary artery bypass graft; NYHA = New York Heart Association; LV = left ventricular.

TABLE 6.1. Baseline characteristics of the 15 study patients.Age (years) 68 ± 12 (mean ± SD)Sex male/female 12/3 (80/20%)Functional class

NYHA II 14NYHA III 1

LV ejection fraction 25 ± 7%Location of myocardial infarction*

Inferoposterior 4 (27%)Anterior 14 (93%)Unknown 2 (14%)

Previous interventionsCABG 2 (14%)

* more than one infarction location per patient possible.CABG = coronary artery bypass graft; NYHA = New York Heart Association; LV = left ventricular.


antagonists and ß-blocking agents were withdrawn at least five half-lifes before trial. Allother oral medication was continued. At baseline there were no statistically significantdifferences between the three groups.

Effects of Mibefradil on mean arterial pressure, heart rate and ejection fraction.Table 6.2 shows the findings on heart rate, mean arterial pressure and ejection fraction.Heart rate was decreased by Mibefradil (p<0.05). This bradycardic effect was evidentonly after one week of treatment in the high dose group (Figure 6.1). No effect ofMibefradil was seen on SBP, DBP, MAP, EF and PER.

Effects of Mibefradil on the diastolic nuclear angiographic parameters. Thefindings on the diastolic nuclear angiographic parameters are summarised in Table 6.3.Mibefradil had no statistically significant effect on PFR, both when normalised to end-diastolic counts or to stroke counts (Figure 6.2A and 6.2B). Mibefradil had no effect onAC (Figure 6.3) or t-PFR.

DISCUSSIONThe aim of the present study was to investigate the effects of Mibefradil on diastolic leftventricular function in patients with heart failure and decreased ejection fraction due toprevious myocardial infarction. Diastolic function was assessed by measuring variousindices of diastole using nuclear angiography (PFR, t-PFR and AC). Several indices ofsystolic function were also measured using this technique (EF and PER).

The interpretation of diastolic parameters in patients with systolic dysfunction is

TABLE 6.2. Heart rate, mean arterial pressure and ejection fraction on both study days before and threehours after the medication gift.

HR(beats/min)

MAP(mmHg)

EF(%EDV)

Mibefradil Day 1 Day 8 Day 1 Day 8 Day 1 Day 8Patient (mg/day) BM AM BM AM BM AM BM AM BM AM BM AM

1 0 83 79 79 79 96 95 96 88 26 24 26 262 0 63 70 58 63 86 77 81 78 43 51 58 473 0 71 72 72 71 108 94 101 95 13 15 17 174 0 75 81 73 81 99 93 102 97 22 24 23 235 0 66 77 67 74 89 79 85 91 18 24 19 186 6.25 61 60 57 74 115 106 103 103 30 31 30 297 6.25 64 63 68 82 85 87 91 91 26 26 28 278 12.5 79 79 74 71 101 95 98 91 34 33 33 379 12.5 69 67 69 72 115 97 91 98 41 38 40 43

10 25 95 95 94 98 101 96 106 93 19 23 17 2411 25 58 67 71 65 93 79 79 84 38 33 37 3812 50 73 81 62 63 83 63 87 81 24 28 34 3713 50 99 93 88 81 107 101 109 101 25 29 23 2514 100 58 59 58 60 99 99 90 97 32 34 36 3915 100 56 55 53 52 94 97 90 81 24 23 27 28

HR = heart rate; MAP = mean arterial pressure; EF = ejection fraction; EDV = end-diastolic volume; BM= before medication; AM = three hours after medication.

TABLE 6.3. Diastolic nuclear angiographic parameters on both study days before and three hours afterthe medication gift.

PFR(EDV/s)

Normalised PFR(SV/s)

AC(%SV)

Mibefradil Day 1 Day 8 Day 1 Day 8 Day 1 Day 8Patient (mg/day) BM AM BM AM BM AM BM AM BM AM BM AM

1 0 0,98 0,92 0,92 1,01 3,8 3,8 3,5 3,9 38 65 43 372 0 1,51 1,40 1,10 1,61 3,5 2,7 1,9 3,4 29 25 28 203 0 0,51 0,70 0,59 0,79 4,0 4,6 3,5 4,6 20 18 19 194 0 0,61 0,64 0,72 0,85 2,8 2,7 3,1 3,7 71 44 38 365 0 0,39 1,53 0,63 0,94 2,1 6,4 3,3 5,2 65 56 52 766 6.25 0,63 0,52 0,67 0,83 2,1 1,7 2,2 2,9 32 37 35 527 6.25 0,87 1,12 1,15 1,13 3,4 4,3 4,1 4,2 22 27 31 528 12.5 1,02 1,17 0,91 1,15 3,0 3,5 2,8 3,1 63 57 53 579 12.5 1,02 1,26 1,52 1,35 2,5 3,3 3,8 3,1 42 41 44 34

10 25 0,70 0,95 0,84 1,14 3,7 4,1 5,0 4,8 36 26 30 2811 25 0,78 0,92 0,92 0,80 2,0 2,8 2,5 2,1 36 44 52 4212 50 0,78 1,48 0,77 0,66 3,2 5,3 2,3 1,8 58 87 44 4713 50 1,57 1,28 1,16 1,35 6,3 4,4 5,0 5,4 74 66 56 2914 100 0,93 0,85 1,10 1,10 2,9 2,5 3,1 2,8 37 34 37 3415 100 0,50 0,56 0,48 0,65 2,1 2,4 1,8 2,3 27 35 29 23

PFR = peak filling rate; AC = atrial contribution; EDV = end-diastolic volume; SV = Stroke volume; BM= before medication; AM = three hours after medication.

64 Chapter 6

complex. Not only physiological variations such as heart rate and blood pressure caninfluence the outcome of measurements on diastolic function,13,15,16,17 but alsomethodological aspects such as gating mode and normalisation parameters inradionuclide angiography, must be taken into account when studying diastolicparameters.12,13,14,17,18 The interpretation of these parameters must therefore be madewith great caution. In standardised situations, however, diastolic parameters in patientswith systolic dysfunction are likely to be reproducible.

The data show that Mibefradil had a statistically significant effect on HR,especially in the dosages of 50 and 100 mg daily. No such effect was found at dosagesof 6.25 to 100 mg daily on either systolic or diastolic parameters or blood pressure.

In previous studies performed on patients with chronic heart failure adversereactions of short-term treatment with calcium antagonists on systolic function has beenemphasized.6,7,8 The observed systolic deterioration may not however, be present intreatment with second-generation calcium antagonists.7,8 These results are inagreement with those of the present study, where no statistically significant effect ofMibefradil on EF and PER was found. The effect of calcium antagonists in heart failure on diastolic function has notbeen extensively reported,1,2,3,9 but a stabilising and beneficial effect has beenrecorded in several studies. This effect is probably due either to the negativechronotropic or to


60

65

70

75

80

HR

(b

eats

/min

)

Before med. After med. Before med. After med.

Day 1 Day 8

Placebo Low Mibefradil High Mibefradil

Mibefradil effect: p<0.05

0,7

0,8

0,9

1

1,1

1,2

PFR

(ED

V/s

)


Day 1 Day 8


No Mibefradil effect: p>0.05

2,0

3,0

4,0

5,0

PFR

' (S

V/s

)


Day 1 Day 8



30

40

50

60

70

AC

(%S

V)


Day 1 Day 8



FIGURE 6.1. Line graph of the heartrate (HR) on day 1 before and 3 hoursafter medication intake and after 1week of treatment before and 3 hoursafter the last medication intake. Thereis a statistically significant decrease inheart rate as a result of Mibefradil(p<0.05).

FIGURE 6.2B. Line graph of PeakFilling Rate (PFR’), normalised tostroke volume (SV) on day 1 beforeand 3 hours after medication and after1 week of treatment before and 3 hoursafter the last medication intake.Mibefradil had no statisticallysignificant effect on this parameter.

FIGURE 6.3. Line graph of the AtrialContribution to diastolic filling (AC) onday 1 before and 3 hours aftermedication and after 1 week oftreatment before and 3 hours after thelast medication intake. Mibefradil hadno statistical significant effect on thisparameter.

FIGURE 6.2A. Line graph of PeakFilling Rate (PFR) normalised to end-diastolic volume (EDV) on day 1 beforeand 3 hours after medication and after1 week of treatment before and 3 hoursafter the last medication intake.Mibefradil had no statisticallysignificant effect on this parameter.

66 Chapter 6

the coronary vasodilator properties of calcium antagonists.1,9 In the former explanationmore time is available to produce increased left ventricular filling whereas byvasodilatation a better balance between myocardial oxygen demand and supply isachieved. This is likely to be an important mechanism in the improvement of diastolicfunction in patients with coronary artery disease after coronary angioplasty or bypasssurgery.19,20 It could also be an important mechanism in the improvement of diastolicfunction after oral calcium antagonists in patients with recent myocardial infarction anddepressed left ventricular systolic function.9 Coronary vasodilatation was also observedin anaesthetised dogs during ischaemia after Mibefradil i.v.21 Moreover, in stableangina in humans, Mibefradil has proven to be effective in doses of 50, 100 and 200mg daily.10 Whether or not vasodilatation plays a role in chronic heart failure due to oldmyocardial infarction remains uncertain.

In the present data no evidence was found for improvement or worsening ofdiastolic function after short-term Mibefradil in patients with severely depressed systolicfunction. This may be due to the relatively low doses in which it was administered.However, no direct effect on PFR was found in group IIB either. The combined datasuggest that the short-term administration of the new calcium antagonist Mibefradil isrelatively safe in patients with coronary artery disease and depressed left ventricularfunction. The long-term treatment of patients with heart failure with Mibefradil remainshowever to be investigated. The mechanism for the unfavourable effects of calciumantagonists in the long-term treatment of patients with heart failure is probablymultifactorial.8 The negative inotropic effect of most calcium antagonists is probablyless with Mibefradil.11 However, the possible role of unfavourable neurohormonalactivation in patients with heart failure when treated with Mibefradil is not yet clear. Moreresearch is therefore needed on a larger study population at adequate doses for thetreatment of angina pectoris to be able to draw more definite conclusions on the safetyof Mibefradil in patients with heart failure.

Conclusion. Compared to a control group of patients who received placebomedication a clear statistically significant effect of Mibefradil was measured on heartrate, while no effect was measured on blood pressure, systolic and diastolic leftventricular function. We conclude therefore that short-term administration of Mibefradil inpatients with heart failure due to previous myocardial infarction is probably safe.

ACKNOWLEDGMENTThe authors wish to thank H. Louwes and H. Schuurman from the Department of NuclearMedicine of the Martini Hospital in Groningen for their technical support.

References

1. Gaasch WH, Blaustein AS, Le Winter MM. Heart failure and clinical disorders of left ventricular


diastolic function. In: Gaasch WH, LeWinter MM, ed. Left ventricular diastolic dysfunction andheart failure. Philadelphia: Lea & Febiger, 1994;245-258.

2. Lahiri A, Rodrigues EA, Carboni GP, Raftery EB. Effects of long-term treatment with calciumantagonists on left ventricular diastolic function in stable angina and heart failure. Circulation1990;81(suppl III):III-130-III-138.

3. Packer M. Abnormalities of diastolic function as a potential cause of exercise intolerance inchronic heart failure. Circulation 1990;81(suppl III):III-78-III-86.



6. Elkayam U, Amin J, Anilkumar M, Vasquez J, Weber L, Rahimtoola SH. A prospective,randomized, double-blind, crossover study to compare the efficacy and safety of chronicnifedipine therapy with that of isosorbide dinitrate and their combination in the treatment ofchronic congestive heart failure. Circulation 1990;82:1954-1961.

7. Packer M. Pathophysiological mechanisms underlying the adverse effects of calcium channel-blocking drugs in patients with chronic heart failure. Circulation 1989;80(supplIV):IV-59-IV-67.

8. Elkayam U, Shotan A, Mehra A, Ostrzega E. Calcium channel blockers in heart failure. J AmColl Cardiol 1993;22(suppl A):139A-144A.

9. Dienstl F,Motro M, Poole Wilson PA, Schartl M, Bassand JP, Van Dalen FJ, Emmott SN,Horowitz M, Hugenholtz PG. Improved diastolic function with the calcium antagonist nisoldipine(coat-core) in patients post myocardial infarction: results of the DEFIANT study. Eur Heart J1992;13:1496-1505.

10. Portegies MCM, Schmitt R, Kraay CJ, Braat SHJG, Gassner A, Hagemeyer F, Pozenel H,Prager G, Viersma JW, Wall EE van der, Kleinbloesem CH, Lie KI. Lack of negative inotropiceffects of the new calcium antagonist Ro 40-5967 in patients with stable angina pectoris. JCardiovasc Pharmacol 1991;18:746-751.

11. Véniant M, Clozel JP, Hess P, Wolfgang R. Ro 40-5967, in contrast to diltiazem, does notreduce left ventricular contractility in rats with chronic myocardial infarction. J CardiovascPharmacol 1991;17:277-284.


13. Udelson JE, Bonow RO. Radionuclide angiographic evaluation of left ventricular diastolicfunction. In: Gaasch WH, LeWinter MM ,ed. Left ventricular diastolic dysfunction and heartfailure. Philadelphia: Lea & Febiger, 1994;167-191.

14. Bacharach SL, Green MV, Vitale D, White G, Douglas MA, Bonow RO, Larson SM. Optimumfourier filtering of cardiac data: a minimum-error method: concise communication. J Nucl Med1983;24:1176-1184.

15. Levine HJ, Gaasch WH. Clinical recognition and treatment of diastolic dysfunction and heartfailure. In: Gaasch WH, LeWinter MM ,ed. Left ventricular diastolic dysfunction and heart failure.Philadelphia: Lea & Febiger, 1994;439-454.

16. Bonow RO. Radionuclide Angiographic Evaluation of Left ventricular diastolic function.Circulation 1991;84(suppl I):I-208-I-215.

17. Lee KJ, Southee AE, Bautovich GJ, Freedman B, McLaughlin AF, Rossleigh MA, Hutton BF,Morris JG. Normalised radionuclide measures of left ventricular diastolic function. Eur J NuclMed 1989;15:123-127.

18. Bonow RO, Bacharach SL, Crawford-Green C, Green MV. Influence of temporal smoothing onquantitation of left ventricular function by gated blood pool scintigraphy. Am J Cardiol1989;64:921-925.

68 Chapter 6

19. Bonow RO, Kent KM, Rosing DR, Lipson LC, Bacharach SL, Green MV, Epstein SE. Improvedleft ventricular diastolic filling in patients with coronary artery disease after percutaneoustransluminal coronary angioplasty. Circulation 1982;66:1159-1167.

20. Lawson WE, Seifert F, Anagnostopoulos C, Hills DJ, Swinford RD, Cohn PF. Effect of coronaryartery bypass grafting on left ventricular diastolic function. Am J Cardiol 1988;61:283-287.

21. Clozel JP, Banken L, Osterrieder W. Effects of Ro 40-5967, a novel calcium antagonist, onmyocardial function during ischemia induced by lowering coronary perfusion pressure in dogs:comparison with verapamil. J Cardiovasc Pharmacol 1989;14:713-721.

Chapter 7Left ventricular beat-to-beat performance

in atrial fibrillation: dependence oncontractility, preload and afterload

H.J. Muntinga1, A.T.M. Gosselink1, P.K. Blanksma1, P.J. de Kam1, E.E. van derWall, and H.J.G.M. Crijns1.

1Thoraxcentre, Department of Cardiology, University Hospital Groningen, Groningen, 2Department ofCardiology, University Hospital Leiden, Leiden, The Netherlands.

Summary

Objective:To assess independent determinants of beat-to-beat variation in left ventricularperformance during atrial fibrillation.Design: Prospective study.Setting: University hospital.Patients: Seven patients with chronic non-valvar atrial fibrillation.Interventions: Invasive and non-invasive haemodynamic variables were obtained using a non-imaging computerised nuclear probe, a balloon-tipped flow directed catheter, and a non-invasive fingertip blood pressure measurement system linked to a personal computer.Main outcome measures: Left ventricular ejection fraction, left ventricular volume, ventricularcycle length, pulmonary capillary wedge pressure, and measures for left ventricular afterload(end-systolic pressure / stroke volume) and contractility (end-systolic pressure / end-systolicvolume) were calculated on a beat-to-beat basis during 500 consecutive RR intervals. Withmultiple regression analysis a statistical model of the beat-to-beat variation of ejection fractionwas constructed, containing these variables.Results: Positive independent relations with ejection fraction were found for preceding RRinterval, contractility, and end-diastolic volume, whereas inverse relations were found forafterload, preceding end-systolic volume and preceding contractility (all variables p<0.0001). Arelatively strong interaction was found between end-diastolic volume and afterload, indicatingthat ejection fraction was relatively more enhanced by preload in the presence of low afterload.Conclusions: The varying left ventricular systolic performance during atrial fibrillation isindependently influenced by beat-to-beat variations of cycle length, preload, afterload andcontractility. Beat-to beat variations in preload exhibit their effect on ventricular performancemainly in the presence of a low afterload.

Heart 1999;82:575-580

70 Chapter 7

he randomly irregular ventricular response to atrial fibrillation1 not onlycauses an irregular but also an unequal pulse.2,3 The beat-to-beatvariations in ventricular performance have been ascribed to variations

in the length of the preceding heart period,2 beat-to-beat variations in preload bymeans of the Frank-Starling mechanism,4 beat-to-beat variations in afterload,5,6 andbeat-to-beat variations in contractility, acting either by the interval-contractilityrelationship,7 or by means of postextrasystolic potentiation.8 or to a combination ofthose factors.9 In a previous study of left ventricular beat-to-beat performance inpatients with nonvalvular atrial fibrillation, we have demonstrated that the interval-force relation rather than the Frank-Starling mechanism explained the varying leftventricular systolic performance during atrial fibrillation over the entire range of RRintervals.9 The contribution of beat-to-beat variation in preload, i.e. the Frank-Starlingmechanism, to the varying left ventricular function during atrial fibrillation was limitedto short preceding intervals. The contribution of aortic impedance, i.e. afterload, aswell as other beat-to-beat regulatory mechanisms, including the positive andnegative effects of ejection10,11 and the preceding beat contraction history12 couldnot be evaluated due to the absence of simultaneous measurement of leftventricular volume and (aortic) pressure measurement on a beat-to-beat basis.

Therefore, the aim of the present study was to assess in what proportion thehaemodynamic regulatory mechanisms determine the beat-to-beat variations in leftventricular performance during atrial fibrillation. For this purpose we used a non-imaging computerised nuclear probe13,14 allowing beat-to-beat left ventricular volumemeasurement and invasive and non-invasive haemodynamic monitoring enablingextensive haemodynamic data acquisition on a beat-to-beat basis in a large numberof consecutive beats.

METHODSPatients and study protocol. Seven patients with chronic atrial fibrillation wereincluded (Table 7.1). To avoid blunting of cycle length dependent haemodynamicchanges by valvular heart disease, in particular mitral stenosis,3,4 only patients withnon-valvular atrial fibrillation were studied. Prior to the study, all patients underwentM-mode and Doppler echocardiography. All antiarrhythmic drugs, including digitalisand calcium antagonists were stopped at least 5 drug half-lives before the study.The study was approved by the Institutional Review Board and written informedconsent was given by all 7 patients.

Nuclear probe. To measure relative left ventricular volumes on a beat-to-beat basis,a commercially available non-imaging computerised nuclear probe (NuclearStethoscope, Bios, Valhalla, NY) was used.14,15 Methods are previously described byour group.9,16 In short, equilibrium blood pool labelling was obtained by the in vivo

T

Haemodynamics in atrial fibrillation 71

labelling of red blood cells with 20 mCi 99mTc. To search for the optimal position ofthe detector, the technique recommended by the manufacturer was used, bymonitoring the continuously displayed values of stroke counts and ejection fraction.14

At the optimal left ventricular position, the values of stroke counts and ejectionfraction were maximal and they were minimal for the background position. Theanalogue output from the probe, as well as the electrocardiogram, were fed into apersonal computer with custom-developed software. This system allowed forcontinuous real-time display and permanent recording of a simultaneously acquiredhigh temporal resolution radionuclide left ventricular time-activity curve (orbackground activity level), and an electrocardiographic signal. After final probepositioning, beat-to-beat data were acquired during 500 consecutive beats. Beat-to-beat analysis of the time-activity curve allowed instantaneous assessment of relativeleft ventricular volume.

Haemodynamic measurements. A balloon-tipped flow directed catheter ("Swan-Ganz" catheter) was used to measure pulmonary capillary wedge pressure on abeat-to-beat basis. Directly before the start of the recording of 500 consecutivebeats per patient, cardiac output was measured using the thermodilution method.

Table 7.1. Baseline characteristics of the 7 study patients and mean values ± standard deviations of measured and calculated independent variables determining left ventricular ejection fraction over 500consecutive cycles arranged by ejection fraction (ranges between brackets)

Patient 1 Patient 2 Patient 3 Patient 4 Patient 5 Patient 6 Patient 7Gender M F M M M M MAge (years) 64 67 36 61 61 42 45Underlying heart disease IHD RHD lone lone IHD DCM loneArrhythmia duration (months) 2 168 4 6 56 1.5 10LA diameter, long axis view (mm) - 47 41 45 50 52 -LV end diastolic diameter (mm) 55 45 50 51 55 70 40LV end systolic diameter (mm) 42 29 38 38 37 58 30Mean RR interval (ms) Range

427 ± 138180-930

610 ± 131410-1150

589 ± 134180-1210

776 ± 259170-1670

632 ± 147330-1160

885 ± 302330-1880

729 ± 137510-1250

Mean LV ejection fraction (%) Range

25 ± 151-71

30 ± 97-54

31 ± 123-61

38 ± 122-62

41 ± 88-60

42 ± 102-67

46 ± 731-63

Mean LV end diastolic volume (ml) Range

136 ± 3652-230

215 ± 33152-335

142 ± 2476-190

211 ± 40112-314

136 ± 1767-164

180 ± 2782-231

184 ± 21125-228

Mean LV end systolic volume (ml) Range

99 ± 2541-178

148 ± 20111-221

97 ± 1857-137

129 ± 2373-226

80 ± 753-98

105 ± 2452-162

98 ± 1476-144

Mean PCWP (mmHg) Range

21 ± 510-35

22 ± 317-27

18 ± 213-28

9 ± 24-33

24 ± 63-38

9 ± 34-30

9 ± 24-17

Mean LV contractility (mmHg/ml) Range

1.2 ± 0.40.2-3.1

1.0 ± 0.20.7-1.5

1.1 ± 0.20.7-1.9

1.3 ± 0.30.7-2.4

1.2± 0.011.17-1.21

1.2 ± 0.40.2-2.8

1.5 ± 0.21.0-3.2

Mean LV aterload (mmHg/ml) Range

7.3 ± 12.70.5-114.9

2.9 ± 2.61.3-25.0

3.6 ± 4.31.1-41.6

3.2 ± 5.11.0-52.1

1.9 ± 1.00.8-13.8

1.8 ± 1.70.5-26.6

1.8 ± 0.41.0-2.0

DCM = dilating cardiomyopathy, IHD = ischaemic heart disease, LA = left atrial, lone = ‘lone’arrhythmia, LV = left ventricular, PCWP = pulmonary wedge pressure, RHD = rheumatic heart disease.

72 Chapter 7

The mean value of three measurements was used to calculate the individual factorwhich allows the conversion of measured counts to ml. This factor equals the ratio ofcardiac output and the product of stroke counts and heart rate. Peripheral bloodpressure was assumed to change similar to aortic pressure. To measure peripheralblood pressure a non-invasive fingertip blood pressure measurement system(Finapres, Ohmeda TM 2300, Inglewood, CO) was used which enables accuratemeasurement of systolic and diastolic blood pressure on a beat-to-beat basis ascompared to intra-arterial measurement.17 In essence, the method is based on acontrol loop, consisting of an inflatable finger cuff equipped with an infraredphotoplethysmographic device to measure the finger artery blood volume under thecuff.18 The system is set to maintain a null transmural pressure. Changes in arterialblood volume due to pressure changes, detected by plethysmography, arecounteracted by means of a fast electropneumatic servo system, which modulatesthe cuff pressure.

Data processing. The obtained data were simultaneously fed into a personalcomputer, enabling accurate, beat-to-beat calculation (and storage) of ejectionfraction, left ventricular volume, left ventricular cycle length, pulmonary capillary

0 500 1000 1500 2000 2500 3000

Time (ms)

QRS QRS QRS QRS

ppRR pRR Index interval

ppEDVppESVppESP

pEDVpESVpESP

EDVESVESP

FIGURE 7.1 Schematic drawing showing the relation between the QRS complex on theelectrocardiogram, the definition of the index interval, preceding and pre-preceding interval (pRR andppRR, respectively), and the moment of volume and pressure measurements. During the index intervalEDV (end-diastolic volume), ESP (end-systolic pressure), and ESV(end-systolic volume) weremeasured. During the preceding interval pEDV, pESP, and pESV (indicating preceding EDV, ESP,and ESV respectively), and during the pre-preceding interval ppEDV, ppESP, and ppESV (indicatingpre-preceding EDV, ESP, and ESV respectively) were measured.


wedge pressure and peripheral blood pressure in a large number of consecutivebeats (Figure 7.1). Using these pressure and volume data, left ventricular afterloadof the index interval was defined as the ratio of end-systolic pressure and strokevolume (ESP/SV),19,20 and left ventricular contractility of the index interval as theratio of end-systolic pressure and end-systolic volume (ESP/ESV).21 In this way, leftventricular afterload and left ventricular contractility values of the index interval,preceding and pre-preceding interval were calculated on a beat to beat basis.Pulmonary capillary wedge pressure (PCWP) and end-diastolic volume (EDV) wereused as parameters for left ventricular preload of the index interval.

Statistical analysis. The validity of a multivariate model of left ventricular beat-to-beat performance in atrial fibrillation described previously by our group,9 was testedusing the same non-invasive haemodynamic variables measured in the presentpatients, i.e. preceding RR interval, pre-preceding RR interval, end-diastolic volume,and preceding end-systolic volume. A measure of the fit of a model to the data, inthis study left ventricular ejection fraction, is the model correlation (R2).

To estimate univariate associations with the dependent variable leftventricular ejection fraction the independent factors from our previous model wereanalysed together with the following newly measured and calculated factors:pulmonary capillary wedge pressure, afterload, contractility of the index cycle,contractility of the preceding cycle, and contractility of the pre-preceding cycle. Thesignificant variables describing ejection fraction in our previous model were filled inwith the obtained haemodynamic factors of the present study showing significantunivariate associations with ejection fraction (P<0.05), to assess independentdeterminants of ejection fraction in a mixed effects model. In this model the variationbetween the patients was added as an additive variation term. Using a backwardselection method variables with a t-test parameter <10 were deleted from the model.Clinically relevant potential one-way interaction terms were evaluated in addition.The independent variables were introduced as centred terms, by subtracting themean.

RESULTSThe same results as before were found when the non-invasive parametersdescribing a previous multiple regression model of left ventricular ejection fraction9

were obtained from the present patients and were fitted into the previous model.The model correlation (R2) was now 0.70, comparable to the value found previously(0.73). After adding the newly studied parameters, pulmonary capillary wedgepressure, left ventricular afterload, left ventricular contractility of the index beat,contractility of the preceding beat, and contractility of the pre-preceding beat, thefinal model used the following equation:

74 Chapter 7

7

EF = Constant + � Pi + β1(pRR-pRR) + β2(pEDV-pEDV) + β3(pESV-pESV) + i=1

β4(ESP/SV-ESP/SV) + β5(ESP/ESV-ESP/ESV) + β6(pESP/pESV-pESP/pESV) +

β7(EDV-EDV)(ESP/SV-ESP/SV)

where EF = left ventricular ejection fraction; P1 to P7 = effects for the 7 patients; β1 to

β7 = regression coefficients for the independent variables and their interaction terms,which determine left ventricular ejection fraction; pRR = preceding RR-interval; EDV= end-diastolic volume; pESV = preceding end-systolic volume; ESP/SV = ratio ofend-systolic pressure and stroke volume; ESP/ESV and pESP/pESV = ratio of end-systolic pressure and end-systolic volume of the index beat, and of the precedingcycle, respectively. In addition, the model shows one interaction term. The modelcorrelation (R2) in the new model was 0.87.

Baseline characteristics of the 7 study patients, individual echocardiographicdimensions, ejection fraction and haemodynamic measurements as well as theirranges during 500 consecutive cardiac cycles are shown in table 7.1. Table 7.2summarises the results of the univariate correlation analysis. Pulmonary capillarywedge pressure was the only parameter which had only weak correlation withejection fraction in the univariate analysis (p>0.01). All other tested parameters hadsignificant univariate correlations with ejection fraction (p<0.01). The strongestunivariate correlations were present between ejection fraction and preceding RRinterval, end-diastolic volume, contractility, and afterload (all t-test parameters > 10).

TABLE 7.2 Independent variables and their univariate associations (t-values) with left ventricularejection fractionVariable TpRR 41.12ppRR 4.67EDV 21.74PCWP -2.37pESV -5.53Afterload -31.53Contractility 28.18p-Contractility 8.48pp-Contractility 9.75pRR = preceding RR interval; ppRR = pre-preceding RR interval; EDV = end-diastolic volume; PCWP= pulmonary capillary wedge pressure; pESV = preceding end-systolic volume; Afterload = end-systolicpressure/stroke volume; Contractility = end-systolic pressure/end-systolic volume; p-Contractility =preceding end-systolic pressure/preceding end-systolic volume; pp-Contractility = pre-preceding end-systolic pressure/pre-preceding end-systolic volume.


FIGURE 7.2 Relation betweenpreceding RR interval and leftventricular (LV) ejection fraction in oneof the patients, illustrating the positiverelation between preceding RRinterval and LV ejection fraction.There is a curvilinear relationship withejection fraction remaining ratherconstant at long RR intervals.

FIGURE 7.3 Relation between leftventricular (LV) end-diastolic volumeand LV ejection fraction in the samepatient, illustrating the positive relationbetween LV end-diastolic volume andLV ejection fraction.

FIGURE 7.4 Example of the relationbetween left ventricular (LV)contractility of the index cycle(ESP/ESV indicates the ratio of end-systolic pressure and end-systolicvolume) and LV ejection fraction in thesame patient, illustrating the positiverelation between LV contractility andLV ejection fraction.

FIGURE 7.5 Example of the relationbetween left ventricular (LV) afterloadof the index cycle (ESP/SV indicatesthe ratio of end-systolic pressure andstroke volume) and LV ejectionfraction, illustrating the negativerelation between LV afterload and LVejection fraction.

76 Chapter 7

The relations between these independent variables and the dependent variable leftventricular ejection fraction are illustrated in figures 7.2 to 7.5. Figure 7.2 illustratesthe positive relation between preceding RR interval and ejection fraction in one ofthe patients. Figure 7.3 illustrates the positive relation between end-diastolic volumeand ejection fraction in the same patient. Figure 7.4 shows an example of thepositive relation between contractility of the index cycle (ESP/ESV) and leftventricular ejection fraction, and figure 7.5 shows an example of the inverse relationbetween afterload (ESP/SV) of the index cycle and left ventricular ejection fraction.

Multiple regression analysis. Table 7.3 summarises the factors to which ejectionfraction was significantly related in the multiple regression analysis. Preceding RRinterval, left ventricular end-diastolic volume, and left ventricular contractility of theindex beat showed an independent positive relationship with ejection fraction,whereas preceding end-systolic volume, left ventricular afterload of the index beatand contractility of the preceding beat showed an independent inverse relation withejection fraction.

Effects of interactions on left ventricular ejection fraction. There was oneseparate statistically significant one-way interaction term which met the criteria tostay in the final model (Table 7.3). Figure 7.6 illustrates the influence of theinteraction between left ventricular end-diastolic volume and afterload of the indexbeat. For a given end-diastolic volume, the ejection fraction was relatively lessenhanced if the afterload was high, whereas with low afterload, end-diastolic volumeaffected the ejection fraction more.

TABLE 7.3 Independent variables and their association with left ventricular ejection fraction withmultiple regression analysis *

VariableRegressionCoefficient

SE ofRegression Coefficient T

pRR 0.0092 0.00057 16.13EDV 0.1771 0.00508 34.80pESV -0.2664 0.00724 -36.82afterload -1.0475 0.05173 -20.25contractility 21.9391 0.46395 47.29p-contractility -14.9105 0.52436 -28.44EDV * afterload -0.0081 0.00071 -11.52Intercept 37.1621 0.22925 162.10*p � 0.0001 for all comparisons.pRR indicates preceding RR interval; EDV, end-diastolic volume; pESV, preceding end-systolicvolume; afterload, end-systolic pressure / stroke volume; contractility, end-systolic pressure / end-systolic volume; p-contractility, preceding end-systolic pressure / preceding end-systolic volume.


DISCUSSIONThe present study describes a statistical model of the beat-to-beat variation ofejection fraction in atrial fibrillation. In this multivariate model of the beat-to-beatchanges of ejection fraction, beat-to-beat variation of contractility, precedingcontractility, as well as preload and afterload, determined ejection fraction of theindex beat. Considering the t-test parameters of the multiple regression analysis, themost important parameter appeared to be contractility of the index cycle. Thepresent model filled in the previously described model,9 and the addition ofmeasures of preload, afterload and contractility strongly improved the strength of themodel.

Effect of contractility on ejection fraction. The finding that left ventricular beat-to-beat variation of contractility during atrial fibrillation is strongly related to beat-to-beatvariation of left ventricular ejection fraction is in agreement with results of previousstudies.3,7,8,22,23 In addition, ejection fraction was related to contractility of thepreceding cycle. This observation confirms earlier reports which demonstrate thepossibility of cardiac performance to be dependent on the mechanical events ofmore than one preceding beat.8,12 In the univariate analysis left ventricular ejectionfraction was positively influenced by preceding contractility. However, themultivariate analysis showed an inverse relation between left ventricular ejectionfraction and preceding contractility. The positive univariate relation between ejectionfraction and preceding contractility is apparently described by other parameters inthe multivariate model, so that in the multivariate model the true negative relation isexposed. The positive univariate relation between preceding contractility andejection fraction may be explained by the model described by Hardman et al, inwhich the effect of postextrasystolic potentiation decays over a number of beats.[8]

0

10

20

30

40

50

60

70

1 3 5 7 9

ESP/SV (mmHg/ml)

LV

eje

ctio

n f

ract

ion

(%

)

EDV = 300

EDV = 250

EDV = 200

EDV = 150

EDV = 100

EDV = 50

FIGURE 7.6 Effect of theinteraction between left ventricular(LV) afterload (ESP/SV indicatesthe ratio of end-systolic pressureand stroke volume) and LV end-diastolic volume (EDV) on LVejection fraction. The curves arebased on the multiple regressionmodel. All other variables werefixed on their mean value. Withrising afterload ejection fractiondecreases. The interaction showsthat with low afterload, the effect ofend-diastolic volume on ejectionfraction is relatively large, whereaswith high afterload ejection fractionwas relatively less influenced byend-diastolic volume.

78 Chapter 7

The positive relation between contractility of the preceding beat and ejection fractionwas however not demonstrated in the present multivariate model. Possibly, this iscaused by the presence of measures of contractility of the index beat, which show arelatively strong effect on left ventricular ejection fraction compared to contractility ofthe preceding cycle.

The origin of the remaining negative multivariate relation between contractilityof the preceding cycle and left ventricular ejection fraction is more difficult toestablish, although it may be found in the same model by Hardman et al.8 A shortpre-preceding cycle will be poorly mechanically restituted, and is followed by a weakcontraction. The following ejection will be strengthened due to postextrasystolicpotentiation. However, the scope of the present model does not allow suchconclusions. In order to elucidate the origin of the complicated relation between leftventricular ejection fraction and preceding contractility a study needs to beperformed into the relative dependency of left ventricular contractility, precedingcontractility, and cycle length fluctuations.

Effect of pre- and afterload parameters on ejection fraction. The positive relationwhich existed between left ventricular end-diastolic volume and ejection fractionconfirms findings of previous studies in which the effect of the Frank-Starlingmechanism on ejection fraction in patients with non-valvular atrial fibrillation wasdemonstrated.4,5,9 In a previous study by our group this contribution of the Frank-Starling mechanism to the varying ejection fraction was however limited to situationsof short preceding cycle lengths and long pre-preceding cycle lengths.9 Theinteractions between end-diastolic volume and preceding and pre-preceding RR-interval were however not demonstrable in the present model and were described bythe remaining interaction between end-diastolic volume and afterload.

The interaction between left ventricular end-diastolic volume and afterloadindicates that ejection fraction was relatively more influenced by variations in preloadin the presence of a low afterload, compared to the influence of preload in thepresence of a high afterload. This suggests that the role of the Frank-Starlingmechanism in the determination of ejection fraction in patients with atrial fibrillation isrestricted to situations in which afterload is low, which may be the situation in thepresence of a short pre-preceding RR interval. After a short pre-preceding intervalonly a small amount of blood volume will be ejected and the rise in aortic pressurewill be small. As a consequence, the runoff in the aorta will be considerable andaortic impedance (i.e. afterload) during the next beat will be relatively low, resultingin an increased ejection fraction.5,6

Effect of cycle length fluctuations on left ventricular ejection fraction. After theaddition of the abovementioned determinants of contractility and afterload to our


previous model,9 the influence of the preceding RR interval on left ventricularejection fraction, measured by its t-test parameter, tended to be less in the presentmultivariate model. In addition, the relation between pre-preceding RR interval andejection fraction did not even reach a t-test parameter high enough to remain in thefinal model. Part of the relation between preceding RR-interval and ejection fractionand the entire relation between pre-preceding RR-interval and ejection fraction wereapparently described by other parameters, which had not been included in ourprevious model. In our opinion, this related to the predominating effect of theinterval-force relation determining the variable left ventricular performance in atrialfibrillation. In the presence of a long preceding RR interval, mechanical restitutionwill be complete and the following ejection will be strengthened, whereas in thepresence of a short pre-preceding RR interval contractility of the index beat will behigh ("postextrasystolic potentiation").8,24,25 These cycle length fluctuations influenceejection fraction indirectly by their effect on contractility, but also by their effect onafterload. Taking contractility and afterload into account when assessing the origin ofthe fluctuations of ejection fraction in atrial fibrillation, this strongly replaces theinfluence of random cycle length fluctuations.

Clinical implications. In earlier reports the effects of beat-to-beat variations of cyclelength, contractility, preload an afterload on the variation of the pulse during atrialfibrillation have been described.8,9,10 These mechanisms are probably responsiblefor the adverse haemodynamics produced by the irregularity of the pulse in atrialfibrillation.26,27 The mutual proportions in which these mechanisms contribute to thebeat-to-beat left ventricular systolic performance in patients with nonvalvar atrialfibrillation as measured by left ventricular ejection fraction are demonstrated in thepresent study. In order to optimise haemodynamics in patients with atrial fibrillationthe origin of the beat-to-beat variations of contractility, preload and afterload, i.e.random cycle length fluctuations, may be a starting-point for therapeutic options, aswas demonstrated for transcatheter ablation of the atrioventricular junction andpacemaker implant resulting in a regular ventricular rhythm.28 Another starting-pointfor therapeutic options in these patients may be optimisation of left ventricularafterload without reduction of preload and contractility. The latter option howeverremains to be investigated.

Limitations. First, although patients with valvar heart disease were excluded fromthis study, patients were still relatively heterogeneous with respect to underlyingheart disease. The latter, in combination with the small number of patients,precluded subgroup analysis. Moreover, this was not the primary target of thepresent study. Secondly, the description of contractility and afterload by ESP/ESVand ESP/SV depends on the accuracy of the approximation of end-systolic aortic

80 Chapter 7

pressure by non-invasive measurement of peripheral blood pressure. Although thevalues of ESP equal that of intra-arterial blood pressure measurement,18 they differsignificantly from aortic systolic pressure on physiologic grounds. However, the beat-to-beat variability of ESP will equal that of aortic systolic pressure. Therefore, theapproximation of aortic systolic pressure by ESP would change the values of theregression coefficients of the multiple regression analysis for afterload, contractilityof the index cycle, and contractility of the preceding cycle, but not the value of the t-test parameter.

Conclusions. The varying left ventricular systolic performance measured by leftventricular ejection fraction in atrial fibrillation is dominated by variations ofcontractility, probably caused by the interval-force relation. Beat-to-beat variations inpreload and afterload play a more modest role. The presence of pre- and afterloadvariations may result from random cycle length fluctuations as well. Beat-to-beatvariations in preload, consistent with the Frank-Starling mechanism, exhibit theireffect on ventricular performance mainly in the presence of a reduced afterload.

References

1. Bootsma BK, Hoelen AJ, Strackee J, et al. Analysis of R-R intervals in patients with atrialfibrillation at rest and during exercise. Circulation 1970;41:783-94.

2. Einthoven W, Korteweg AJ. On the variability of the size of the pulse in cases of auricularfibrillation. Heart 1915;6:107-20.

3. Van Dam IM. Left ventricular dimensions during atrial fibrillation. Thesis, Utrecht, TheNetherlands, 1988.

4. Braunwald E, Frye RL, Aygen MM, et al. Studies on Starling's Law of the heart. III.Observations in patients with mitral stenosis and atrial fibrillation on the relationships betweenleft ventricular end-diastolic segment length, filling pressure, and the characteristics ofventricular contraction. J Clin Invest 1960;39:1874-1884.

5. Buchbinder WC, Sugerman H. Arterial blood pressure in case of auricular fibrillation,measured directly. Arch Intern Med 1940;66:625-642.

6. Karliner JS, Gault JH, Bovchard RJ, et al. Factors influencing the ejection fraction and themean rate of circumferential fibre shortening during atrial fibrillation. Cardiovasc Res1974;8:18-25.

7. Meijler FL, Strackee J, Van Capelle FJL, et al. Computer analysis of the RR interval-contractility relationship during random stimulation of the isolated heart. Circ Res 1968;22:695-702.

8. Hardman SMC, Noble MIM, Seed WA. Postextrasystolic potentiation and its contribution to thebeat-to-beat variation of the pulse during atrial fibrillation. Circulation 1992;86:1223-1232.

9. Gosselink ATM, Blanksma PK, Crijns HJGM, et al. Left ventricular beat-to-beat performancein atrial fibrillation: contribution of frank-starling mechanism after short rather than long RRintervals. J Am Coll Cardiol 1995;26:1516-21.


10. Hunter WC. Endsystolic pressure as a balance between opposing effects of ejection. Circ Res1989;64:265-275.

11. Slinker BK, Shroff SG, Kirkpatrick RD, et al. Left ventricular function depends on previous beatejection but not previous beat pressure load. Circ Res 1991;69:1051-1057.

12. Slinker BK. Cardiac cycle length modulates cardiovascular regulation that is dependent onprevious beat contraction history. Circ Res 1991;69:2-11.

13. Schneider J, Berger HJ, Sands MJ, et al. Beat-to-beat left ventricular performance in atrialfibrillation: radionuclide assessment with the computerized probe. Am J Card 1983;51:1189-1195.

14. Berger HJ, Davies RA, Batsford WP, et al. Beat-to-beat left ventricular performance assessedfrom the equilibrium cardiac blood pool using a computerized nuclear probe. Circulation1981;63:133-141.

15. Wagner HN, Wake R, Nickoloff E, et al. The Nuclear Stethoscope: a simple device forgeneration of left ventricular volume curves. Am J Cardiol 1976;38:747-750.

16. Van Gelder IC, Crijns HJGM, Blanksma PK, et al. Time course of hemodynamic changes andimprovement of exercise tolerance after cardioversion of chronic atrial fibrillation unassociatedwith cardiac valve disease. Am J Cardiol 1993;72:560-566.

17. Imholz BP, Wieling W, Langewouters GJ, et al. Continuous finger arterial pressure: utility inthe cardiovascular laboratory. Clin Auton Res 1991;1:43-53.

18. Omboni S, Parati G, Frattola A, et al. Spectral and sequence analysis of finger blood pressurevariability. Hypertension 1993;22:26-33.

19. Suga H, Sagawa K, Shoukas AA. Load independence of the instantaneous pressure-volumeratio of the canine left ventricle and effects of epinephrine and heart rate on the ratio. Circ Res1973;32:314-322.

20. Suga H, Sagawa K. Instantaneous pressure-volume relationships and their ratio in theexcised, supported canine left ventricle. Circ Res 1974;35:117-126.

21. Sunagawa K, Maughan WL, Burkhoff D, et al. Left ventricular interaction with arterial loadstudied in isolated canine ventricle. Am J Physiol 1983;245(Heart Circ Physiol 14):H773-H780.

22. Edmands RE, Greenspan K, Fisch C. The role of inotropic variation in ventricular functionduring atrial fibrillation. J Clin Invest 1970;49:738-746.

23. Gibson DG, Broder G, Sowton E. Effect of varying pulse interval in atrial fibrillation on leftventricular function in man. Br Heart J 1971;33:388-393.

24. Seed WA, Walker JM: Review: Relations between beat interval and force of the heartbeat andits clinical implications. Cardiovasc Res 1988;22:303-314.

25. Schneider F, Martin DT, Schick EC, et al. Interval dependent changes in left ventricularcontractile state and mechanical restitution during atrial fibrillation in human beings. Circulation1996;94(suppl I):I-387. Abstract.

26. Clark DM, Plumb VJ, Epstein AE, et al. Hemodynamic effects of an irregular sequence ofventricular cycle lengths during atrial fibrillation. J Am Coll Cardiol 1997;30:1039-45.

27. Daoud EG, Weiss R, Bahu M, et al. Effect of an irregular ventricular rhythm on cardiac output.Am J Cardiol 1996;78:1433-36.

28. Natale A, Zimerman L, Tomassoni G, et al. Impact on ventricular function an quality of life oftranscatheter ablation of the atrioventricular junction in chronic atrial fibrillation with a normalventricular response. Am J Cardiol;78:1431-33.

Letter to the editor about chapter 7Beat to beat left ventricular performance

in spontaneous atrial fibrillationdoes not depend on afterload

Accepted for publication in HeartProfessor Mark I.M. NobleImperial College School of Medicine, London

The recent article in Heart on this subject by Muntinga et al.1 confirms, using non-invasive techniques, the findings obtained more directly by Brookes et al 2 that beattot beat left ventricular performance in spontaneous atrial fibrillation depends on beatto beat variation in cycle length, left ventricular end- diastolic volume(EDV), andcontractility. Their contention that afterload is another independent determinant restson their Figure 7.5 (page 75) in which ejection fraction is plotted against ESP/SVwhere ESP is end-systolic pressure and SV is stroke volume. However, ejectionfraction is SV/EDV, so they have SV on both axes, which is invalid. A plot of avariable, in this case SV, against its reciprocal 1/SV will lead inevitably to an inversehyperbolic relationship as a mathematical necessity. In this case it is accompaniedby a small amount of scatter caused by the other variables, but the main relationshipshown is a mathematical artefact and not an actual dependence of ejection fractionon afterload.

Figure 7.3 is also incorrect because SV/EDV is plotted against EDV (EDVappearing on both axes), but in this case a plot of SV against EDV would haveresulted in a positive relationship and the same conclusion.

In the study of Brookes et al, the left ventricular systolic pressure (directlymeasured with catheter-tip manometer) varied rather little during atrial fibrillation,which the authors attributed to clamping of the arterial systolic pressure by peripheralmechanisms. This finding and the invalidity of using ESP/SV (above) raises thequestion , “What is the afterload in the intact mammal?” I prefer not to use the termbecause there are so many different indices that purport to be “afterload”; I prefer touse the measured variable. Many workers in the field would accept using leftventricular systolic pressure. However, the term was invented to describe theconstant force or stress in an isolated strip of muscle when shortening duringcontraction in a particular experimental set-up. The only corresponding variablevariable in the intact human is systolic wall stress, but this declines during ejectionthat there is no single value. However inspection of directly measured left ventricularpressure-volume loops during spontaneous atrial fibrillation (2) shows that, most ofthe time, contraction takes place over similar mid-systolic values implying the samewall stress.

83 Letter to the editor about chapter 7

Of course, left ventricular performance will depend on “afterload” (If that canbe defined) if “afterload” changes, but my conclusion is that in spontaneous atrialfibrillation there are no important changes.

References

1. Muntinga HJ, Gosselink ATM, Blanksma PK, De Kam PJ, Van Der Wall EE, Crijns HJGM. Leftventricular beat to beat performance in atrial fibrillation: dependence on contractility, preloadand afterload. Heart 1999;82:575-580.

2. Brookes CIO, White PA, Staples M, Oldershaw PJ, Redington AN, Collins PD, Noble MIM.Myocardial contractility is not constant during spontaneous atrial fibrillation in patients.Circulation 1998;98:1762-1768.

Response

H.J. Muntinga, P.K. BlanksmaUniversity Hospital, Groningen

In his letter, Professor Noble concludes, based on a recent study by Brookes et al,1

that in chronic atrial fibrillation there are no important beat to beat changes of leftventricular afterload measured by left ventricular systolic pressure. Based on figure7.5 of our article he concludes that the relation between afterload as calculated bythe ratio of end systolic pressure (ESP) and stroke volume (SV), and ejection fractionis a mathematical artefact, instead of a physiological relationship.2 He rejects leftventricular afterload as an important measure of left ventricular performance. Theconclusion that afterload, preload and contractility are independent determinants ofleft ventricular ejection fraction however rests on a multivariate model of leftventricular ejection fraction.

In our study we non-invasively measured peripheral blood pressure, which weassumed to change, parallel with aortic pressure.2 We calculated left ventricularvolumes in millilitres from results of measurements with a “Nuclear Stethoscope” anda “Swan Ganz” catheter. Left ventricular afterload was then calculated as the fractionof ESP and SV. The validity of the ESP to SV relationship as a measure of effectivearterial elastance was tested by Sunagawa et al.3 The arterial system properties thenconsist of three elements: peripheral resistance, arterial compliance andcharacteristic impedance of the arterial system. In table 7.1 of our article wedemonstrate that this measure of afterload varies significantly from beat to beat ineach investigated patient. The argumentation that a limited variation of end systolicpressure in patients with atrial fibrillation (in our patient group the average ESP was126 and the average standard deviation of ESP was 13 mmHg) is measuredbecause arterial pressure is damped by baroreceptor and other reflex mechanisms

Letter to the editor about chapter 7 84

confirms in our opinion the observation that afterload, which unites the abovementioned properties of the arterial system changes from beat to beat.

In experimental models of the intact cardiovascular system a uniquerelationship exists in the left ventricle between SV and ESP when preload andcontractility remain constant.4,5 In this situation, afterload determines the exact valuefor SV and ESP. SV is therefore inevitably both a measure of left ventricularperformance, and a measure of left ventricular afterload. Indeed, this results in amathematical relationship between ejection fraction and afterload. The othermeasured factors, end diastolic volume and ESP, also influence the measure ofinterdependence between ejection fraction and afterload, and can determine thevalue of this relationship. The conclusion in our article that ejection fraction inpatients with atrial fibrillation is dependent on afterload (apart from the dependencyon preload and contractility) is based on the multivariate analysis. Figure 7.5 isadded to illustrate the univariate relation between ejection fraction and afterload. Itfurther illustrates that SV importantly determines the direction and nature of therelation between ejection fraction and afterload. The multivariate analysis is also thebasis for the conclusion that ejection fraction is dependent on preload. The univariaterelation between ejection fraction and preload is illustrated with figure 7.3. If thisrelation had only a mathematical nature, a negative hyperbolic relation would havebeen found (1/EDV versus EDV) instead of a linear positive relation. SV apparentlydetermines the direction of this relation. This finding can physiologically be explainedby the Frank-Starling mechanism.6 The differences between figures 7.3 and 7.5contribute in our opinion largely to their importance to the present article.

We agree with Professor Noble that a mathematical relationship existsbetween ejection fraction and the described measures of afterload and preload, butwe do not share his opinion that we may not use these measures in a model todescribe left ventricular beat to beat performance.

References

1. Brookes CIO, White PA, Staples M, Oldershaw PJ, Redington AN, Collins PD, Noble MIM.Myocardial contractility is not constant during spontaneous atrial fibrillation in patients.Circulation 1998;98:1762-1768.

2. Muntinga HJ, Gosselink ATM, Blanksma PK, De Kam PJ, Van der Wall EE, Crijns HJGM. Leftventricular beat to beat performance in atrial fibrillation: dependence on contractility, preload,and afterload. Heart 1999;82:575-580.

3. Sunagawa K, Maughan WL, Burkhoff D, Sagawa K. Left ventricular interaction with arterialload studied in isolated canine ventricle. Am J Physiol 1983;245(Heart Circ Physiol 14):H773-H780.

4. Suga H, Sagawa K, Shoukas AA. Load independence of the instantaneous pressure-volumeratio of the canine left ventricle and effects of epinephrine and heart rate on the ratio. Circ Res1973;32:314-322.

85 Letter to the editor about chapter 7

5. Suga H, Sagawa K. Instantaneous pressure-volume relationships and their ratio in the excised,supported canine left ventricle. Circ Res 1974;35:117-126.

6. Gosselink ATM, Blanksma PK, Crijns HJGM, Van Gelder IC, De Kam PJ, Hillege HL,Niemeyer MG, Lie KI, Meijler FL. Left ventricular beat-to-beat performance in atrial fibrillation:contribution of frank-starling mechanism after short rather than long RR intervals.

Chapter 8Evaluation of

left ventricular diastolic function afterelectrical cardioversion of atrial fibrillation

H.J. Muntinga1,2, F. van den Berg1, M.G. Niemeyer1, P.K. Blanksma2, E.E. van derWall3, H.J.G.M. Crijns2.

1 Department of Cardiology, Martini Hospital, Groningen, 2 Department of Cardiology, University Hospital,Groningen, 3 Department of Cardiology, University Hospital, Leiden, The Netherlands.

Summary

Objectives: This study evaluated left ventricular (LV) diastolic function after direct currentelectrical cardioversion of atrial fibrillation (AF).Background: Tachycardia associated left ventricular dysfunction in AF is characterized bydecreased LV systolic performance, but the data concerning diastolic function are discordant.Methods: Serial radionuclide angiography was performed in 23 patients (mean age 65 ± 13years) with chronic (mean duration 198 ± 242 days) lone AF after cardioversion. Twenty patientswith normal coronary angiograms and no myocardial or valvular disease served as controlsubjects.Results: Early after cardioversion patients with AF had lower LV ejection fraction (EF) comparedto controls (54 ± 10 % vs. 60 ± 8 %, P<0.05), higher peak filling rate (PFR, 3.91 ± 0.68 fillingvolume (FV)/s vs. 3.29 ± 0.64 FV/s, P<0.01), shorter time to PFR (TPFR, 162 ± 46 ms vs. 196 ± 42ms, P<0.05), and lower additional filling fraction (AFF, 24 ± 7 % FV vs. 32 ± 7 % FV, P < 0.05). In5 patients remaining in sinus rhythm for 3 months, PFR decreased from 4.03 ±± 0.63 FV/s to 3.13± 0.54 FV/s (P<0.01).Conclusions: These data do not confirm the existence of diastolic LV dysfunction early afterelectrical cardioversion in patients with lone AF, but are consistent with decreased atrialcompliance and contractility due to atrial remodeling. Persistent sinus rhythm in patients 3months after electrical cardioversion for chronic lone AF is probably accompanied by reversalof atrial remodeling. Alternatively, the findings in patients with persistent sinus rhythm may bedue to development of post-tachycardia LV diastolic dysfunction.

Submitted for publication

Diastolic function after cardioversion 87

n chronic atrial fibrillation (AF) left ventricular function may gradually diminish as aresult of an intrinsic tachycardiomyopathy.1,2,3 This tachycardia-inducedcardiomyopathy is characterized by decreased systolic function, ventricular dilation,

and elevated ventricular filling pressures.2,4,5,6 Not only an inadequate high heart rate inrest but also tachycardia with exercise may contribute to this cardiomyopathy.7 In suchpatients left ventricular systolic function may gradually improve after electricalcardioversion or adequate rate control.2,3,5 Little is known about left ventricular diastolicfunction in AF.8,9 In animal models with chronic rapid pacing induced cardiomyopathy,systolic remodeling is accompanied by an increased diastolic wall stress and impairedmyocardial relaxation.10,11,12,13 In addition, recovery from supraventricular tachycardiainduced cardiomyopathy in animal models is associated with hypertrophy, reducedmyocardial blood flow and diastolic dysfunction.11,14,15 Whether the same changes indiastolic left ventricular function hold true for humans with AF, remains to beinvestigated. The present study was performed to investigate diastolic function inpatients with AF after successful cardioversion compared to a control group. To thispurpose, we performed radionuclide angiography, which may be used to serially studydiastolic filling.16 The findings are compared to findings in other studies of diastolic leftventricular filling in patients with chronic AF early after cardioversion published inEnglish language literature.

METHODSStudy patients. Twenty-three patients who had successfully undergone direct currentelectrical cardioversion for persistent AF or atrial flutter were studied. Onset of AF oratrial flutter was defined as the moment when complaints indicating AF first occurred, orin absence of any complaints the first 12-lead electrocardiogram showing AF or atrial flutter. All patients were examined physically, electrocardiographically andradiographically. Lone AF was diagnosed when based on medical history, routinephysical, laboratory and radiographic examination, 12-lead electrocardiography,echocardiography and when indicated coronary arteriography and ventriculography, theabsence of hypertension, significant valvular disease, coronary artery disease,hyperthyroidism or other cardiovascular disease could be established. Patientcharacteristics are presented in Table 8.1. Informed consent was given by all patients.Oral anticoagulation was instituted for at least 4 weeks before as well as aftercardioversion. Rate-controlling drug therapy and antiarrhythmic drug therapy wasinstituted singly or in combination, in the setting of a serial electrical cardioversionprotocol, and was kept constant during follow-up.17,18 Direct-current electricalcardioversion was performed with a shock sequence of 100, 200, and 360 Joules ifnecessary. Six hours after successful cardioversion left ventricular radionuclideangiography was performed. Results were compared with a control group of 20 patients(mean age 57 ± 12 years, 8 males) with normal findings in coronary arteriography andleft ventriculography. This control group has been described before.16 Patients were

I

88 Chapter 8

followed in the outpatient department. Recurrence of AF was established by physicalexamination and a 12-lead electrocardiogram. When sinus rhythm persisted,radionuclide angiography was repeated after 3 months.

Radionuclide angiography. Left ventricular function was evaluated by equilibriumblood pool scintigraphy using a gamma camera (Siemens Orbiter) with an all-purposeparallel-hole collimator interfaced with a Pinnacle computer (Medasys Inc, Ann Arbor).Technetium 99m with a total dosage of 550-740 MBq was used for labeling the patients'red blood cells. Rest supine multigated images were obtained in left anterior obliqueview with a caudal tilt, so that the left and right ventricle were adequately separated. Onlya 5% cycle-length-window with forward gating was accepted.19 Acquisition wascompleted after 150,000 counts per frame of 20 ms. Temporal smoothing wasperformed by 5 Fourier harmonics.20,21

Radionuclide image measurements. From the left ventricular time-activity curve wedirectly measured left ventricular ejection fraction (EF). In order to make interindividualcomparison of volume fractions possible, measured filling counts were converted topercentages of the filling volume (FV, measured filling counts = 100% FV). All patientsshowed separation of the diastolic portion of the curve into early filling phase andadditional filling phase. In 8 patients a long pronounced diastasis period was observed.This allowed volume measurement of early diastolic filling fraction (EDF), diastasis fillingfraction (DF), and additional filling fraction (AFF, Figure 8.1A) relative to the leftventricular filling volume. Atrial contraction only partially contributes to AFF, sincepassive left ventricular filling may continue after the diastasis period.22,23 The firstderivative of the time versus relative filling volume curve expresses instantaneous fillingrates measured in filling volume/sec. With this normalization the measurement does notvary with the residual volume and the ejection fraction.24 This curve was used tocalculate peak filling rate (PFR) and time to peak filling rate (TPFR, Figure 8.1B). PFRis the rapid instantaneous filling rate during early diastolic filling. Timing was performedin relation to the beginning of filling.

Statistical analysis. The data are expressed as mean ± 1 SD. For comparison of thedata between the control group and the lone AF group statistical analysis wasperformed with Student's t-test for independent samples assuming equal variances. Forcomparison of the follow-up data a paired Student's t-test was used. All P-values weretwo-tailed. A P-value <0.05 was considered statistically significant.


-5

-4

-3

-2

-1

0

1

2

3

4

5

0 200 400 600 800 1000 1200

Time (ms)

LV

fil

lin

g r

ate

(FV

/s)

PFR

TPFR

0

20

40

60

80

100

0 200 400 600 800 1000 1200Time (ms)

LV

vo

lum

e (%

FV

)

EDF

AFF

DF

FIGURE 8.1A. Radionuclide timeactivity curve of one of the patientsshowing left ventricular (LV) volumerelative to the filling volume (FV) inpercentages. In this example earlydiastolic filling fraction (EDF) andadditional filling fraction (AFF) areseparated by a relative long andpronounced diastasis filling (DF).B. First derivative curve showing leftventricular filling rate expressed asfilling volume/s (FV/s). Peak filling rate(PFR) is the rapid instantaneous fillingrate during early diastolic filling. Timeto peak filling rate (TPFR) is measuredfrom the beginning of left ventricularfilling.

TABLE 8.1. Characteristics of patients with lone atrial fibrillation (AF) compared to normal controls(mean ± standard deviation)

Control (n=20) Lone AF (n=23)Age (years) 57 ± 12 65 ± 13*Gender (male/female) 8 / 12 16 / 7Atrial fibrillation duration (range, days) - 198 ± 242 (60 - 1215)Echocardiography (mm): LVEDD 48 ± 5 49 ± 7 LVESD 29 ± 4 33 ± 7 FS (% of LVEDD) 39 ± 6 34 ± 12 LA (LA) 35 ± 6 45 ± 7* LA (AP) 51 ± 2 65 ± 10†Heart rate (min-1)‡ 68 ± 12 68 ± 11Electrocardiography: PR interval (ms)‡ 167 ± 20 211 ± 30§ QRS width (ms)‡ 90 ± 9 90 ± 20Systolic blood pressure (mmHg)‡ 136 ± 14 125 ± 24Diastolic blood pressure (mmHg)‡ 84 ± 7 68 ± 11§Radionuclide angiography LVEF (% of EDV)‡ 60 ± 8 54 ± 10**P < 0.05 vs. control. †P < 0.01 vs. control. ‡ for AF patients: after cardioversion. §P < 0.001 vs. control.

EDV = end-diastolic volume; FS = fractional shortening; LA (AP) = left atrium apical view; LA(LA) = left atrium long axis; LVEDD = left ventricular end-diastolic diameter; LVESD = left ventricular end-systolic diameter; LVEF = left ventricular ejection fraction.

90 Chapter 8

RESULTSPatients’ Characteristics. Twenty-three patients with lone AF were included (Table8.1). AF patients were older (P<0.05) and had higher left atrial dimensions onechocardiography (P<0.05 in the long axis view, P<0.01 in the apical view) compared tocontrol subjects. All patients had clear atrial depolarization waves on theelectrocardiogram after cardioversion with a relatively long PR relationship. Aftercardioversion resting ventricular heart rate had decreased in all patients (mean heartrate before cardioversion: 91 ± 22 min-1, after cardioversion 68 ± 11 min-1, P<0.00001).After cardioversion AF patients had lower diastolic blood pressure (P<0.01). None ofthe patients had clinical signs of cerebral or peripheral embolism.

Systolic and diastolic left ventricular function after cardioversion. In Table 8.2individual and mean values of systolic and diastolic left ventricular function parametersof patients with lone AF are compared to values of normal controls. The control grouphad a higher mean left ventricular ejection fraction (P<0.05). After cardioversion meanpeak filling rate was higher (P<0.01) and mean time to peak filling rate was shorter(P<0.05) in AF patients. Additional left ventricular filling fraction was less in AF patientswithin 6 hours after cardioversion compared to normal controls (P<0.001). In 1 AFpatient no additional filling was detectable early after cardioversion.

Radionuclide angiography at follow-up. After 3 months, 9 patients (39%) were still insinus rhythm. Of these, 5 patients could be restudied. The other 4 refused follow-up.Individual initial and follow-up data on systolic and diastolic left ventricular function arelisted in Table 8.3. In patients with AF remaining in sinus rhythm for 3 months mean PFRdecreased from 4.03 ± 0.63 FV/s 6 hours after cardioversion to 3.13 ± 0.54 FV/s 3months after cardioversion (P<0.01) and mean EDF decreased from 72 ± 6% to 65 ±7% (p<0.05). There were no statistically significant differences between the values ofEF, TPFR and AFF 6 hours and 3 months after cardioversion.

Potential factors determining recurrence of AF. There were no differences in patientcharacteristics or medical treatment between the patients with failure of electricalcardioversion after 3 months and those with persistent sinus rhythm. No statisticallysignificant differences in systolic (134 ± 31 mmHg and 117 ± 16 mmHg respectively)and diastolic blood pressure (69 ± 12 mmHg and 68 ± 11 mmHg), atrial fibrillationduration (158 ± 112 days and 221 ± 294 days respectively), and cardioversion energy(238 ± 122 and 307 ± 90 Joules respectively) were present between groups of patientswith success and failure of cardioversion after 3 months. Also, no differences werepresent between these patient groups with regard to systolic left ventricular function anddiastolic filling characteristics measured with radionuclide angiography.


TABLE 8.2. Radionuclide angiography derived left ventricular filling parameters of 23 patients with loneatrial fibrillation (AF) direct post cardioversion compared to 20 normal controlsPatient HR (min-1) EF (%) PFR (FV/s) TPFR (ms) EDF (%FV) DF (%FV) AFF (%FV)

Contr AF Contr AF Contr AF Contr AF Contr AF contr AF contr AF1 62 82 61 29 2.51 4.70 175 183 67 91 - - 33 02 55 69 65 58 2.94 3.38 115 109 47 76 22 - 31 243 79 69 60 67 4.36 4.02 198 129 71 62 - 15 29 234 60 54 68 54 3.21 3.42 186 165 72 64 - 23 28 135 67 54 50 67 2.44 3.41 312 122 61 58 - 21 39 216 72 64 42 43 3.37 3.95 228 220 69 74 - - 31 267 71 73 68 50 2.74 5.14 129 180 69 78 - - 31 228 81 97 61 63 3.55 5.17 207 126 63 64 - - 37 369 59 53 57 57 233 3.56 209 142 69 63 - 14 31 2310 69 68 64 53 3.51 3.49 202 193 62 70 9 - 29 3011 64 73 58 56 4.42 4.45 216 96 75 74 10 - 16 2612 61 72 61 58 3.49 3.55 211 117 64 71 - - 36 2913 52 71 48 56 3.45 3.94 184 132 63 70 14 - 24 3014 64 83 68 39 2.50 4.30 223 186 69 63 - - 31 3715 105 60 74 36 4.27 4.82 130 86 48 78 - - 52 2216 64 75 55 60 3.02 3.87 181 256 65 71 - - 35 2917 62 60 65 57 3.35 3.05 180 219 74 73 - - 26 2818 58 65 54 52 2.88 3.29 222 202 71 74 - - 29 2619 81 66 65 66 4.15 4.68 204 121 67 66 - 22 33 1220 69 59 61 60 3.33 2.95 213 167 68 59 - 22 32 1921 - 57 - 46 - 2.92 - 222 - 58 - 25 - 1622 - 56 - 60 - 3.48 - 159 - 72 - 9 - 1923 - 76 - 48 - 4.34 - 191 - 76 - - - 24

mean 68 68 60 54* 3.29 3.91† 196 162* 66 70 14 19 32 23‡SD 12 11 8 10 0.64 0.68 42 46 7 8 6 6 7 8

*P < 0.05 vs. control. †P < 0.01 vs. control. ‡P < 0.001 vs. control.- = Data not available or applicable; AFF = additional filling fraction; DF = diastasis filling

fraction; EDF = early diastolic filling fraction; EF = ejection fraction of the left ventricle; FV = fillingvolume; HR = heart rate; PFR = peak filling rate; SD = standard deviation; TPFR = time to peak fillingrate.

TABLE 8.3. Radionuclide angiography follow-up of 5 restudied patients in sinus rhythm 3 months afterdirect current cardioversionPatien

tHR (min-1) EF (%) PFR (FV/s) TPFR (ms) EDF (%FV) AFF (%FV)

6hours

3months

6hours

3months

6hours

3months

6hours

3months

6hours

3months

6hours

3months

3 69 58 67 58 4.02 2.70 129 251 62 59 23 4111 73 73 56 57 4.45 3.46 96 180 74 69 26 3112 72 59 58 56 3.55 3.31 117 127 71 58 29 2815 60 63 36 59 4.82 3.74 86 179 78 74 22 2618 65 54 52 57 3.29 2.46 202 200 74 63 26 22

Mean 68 61 54 57 4.03 3.13† 126 187 72 65* 25 30SD 5 7 11 1 0.63 0.54 46 44 6 7 3 7

*P < 0.05 vs. 6 hours. †P < 0.01 vs. 6 hours.AFF = additional filling fraction; EDF = early diastolic filling; EF = ejection fraction of the left

ventricle; FV = Filling Volume; HR = heart rate; PFR = peak filling rate; SD = standard deviation; TPFR =time to peak filling rate.

92 Chapter 8

DISCUSSIONLV filling early after cardioversion. The present study does not provide evidence forleft ventricular diastolic dysfunction early after direct current electrical cardioversion inpatients with chronic lone AF.

Previous studies of left ventricular filling characteristics early after cardioversionfor chronic atrial fibrillation with Doppler echocardiography either found increased peakearly filling velocity25,26,27,28 or no increase.29 These findings have only scarcely beendiscussed. Iuchi et al suggested that augmented early left ventricular filling may be acompensation for decreased inflow during atrial systole after electrical cardioversion forchronic atrial fibrillation.27 Xiong et al suggested their findings on the transmitral flowpattern in early diastole with Doppler echocardiography in patients early aftercardioversion for chronic atrial fibrillation to result from a diastolic relaxation abnormalitywith increased left atrial pressure changing to a relaxation abnormality after follow-up of1 month.30 In animal models with rapid pacing-induced heart failure reduced cardiacoutput and decreased left ventricular ejection fraction early after deactivation of pacingwere paralleled by diastolic dysfunction consisting of decreased relaxation, increasedmyocardial stiffness, and end-diastolic wall stress.10,12,14 Diastolic dysfunction could bereversed by normalizing increased left ventricular loading.10 In addition, diastolicdysfunction remained measurable 4 weeks after discontinuation of pacing.11,31

Increased ventricular relaxation velocity which could explain increased PFR anddecreased TPFR32 early after cardioversion in chronic AF patients in the present studyis therefore rather unlikely. We also reject the difference in age, ejection fraction andblood pressure between the patients with AF after cardioversion and the controlsubjects as important underlying factors for a different relaxation velocity. Thedifferences in age and left ventricular ejection fraction between the control subjects andthe patients with lone AF in the present study would produce higher PFR and lowerTPFR instead of the found lower PFR and higher TPFR in the control subjects.16 Therelative higher blood pressure in control subjects compared to patients with lone AFafter cardioversion could produce increased afterload to the left ventricle withdecreased relaxation velocity. Although previous investigators reported a significantdecrease of left ventricular relaxation due to increased afterload,33,34 early left ventricularfilling responds variably to blood pressure elevations in subjects with normal leftventricular systolic function without mitral regurgitation.35 It is therefore not likely thatincreased relaxation velocity of the left ventricle explains the observed filling pattern inthe present study with increased PFR and decreased TPFR in patients with chronic AFearly after direct current electrical cardioversion.

The atrioventricular pressure difference is not only influenced by left ventricularrelaxation, but also by left atrial pressure.36 Increased left atrial pressure may produceincreased atrioventricular pressure gradient and concomitant increased early diastolicfilling.22 Increased atrial pressure has recently been proposed to be an explanation forthe observed changes of left ventricular filling in Doppler echocardiography early after


cardioversion in patients with AF.30 This may either be produced by increased atrialloading and volume or decreased atrial compliance.36 In chronic AF atrial enlargementis commonly described.37,38 Also long-standing AF may produce interstitial fibrosis anddegeneration of atrial myocytes.39,40 We therefore consider decreased atrialcompliance (producing an increased v-wave) a probable explanation for the increasedearly diastolic filling rate measured in the present study.

The finding of decreased additional filling fraction compared to control subjects inthe present study is probably explained by decreased atrial contraction. In Table 8.4 thefindings on left ventricular early and additional filling in previous studies aftercardioversion for chronic atrial fibrillation are presented. In our patient group additionalfilling fraction was higher early after cardioversion compared to many studies ofadditional filling in patients with chronic atrial fibrillation aftercardioversion.2,29,41,42,43,44,45 In the studies reporting additional filling fraction of > 20 %of the filling volume after cardioversion,25,26,29,42,43,46,47 AF duration was generallyshorter, and the interval between cardioversion and measurement of filling wasgenerally longer than in studies which report additional filling fractions of < 20 %. Factorswith potential influence on atrial mechanical include duration of atrial fibrillation41 and theproportion of patients with underlying disease, i.e. coronary artery disease,47

hypertrophic cardiomyopathy,49 hypertension50 and congestive heart failure.51 Additionalfilling of the left ventricle may be influenced by the properties of the left ventricle, i.e. leftventricular relaxation and compliance.22 We therefore do not claim the entire additionalfilling fraction in patients after cardioversion for chronic lone atrial fibrillation in thepresent study to be due to atrial contraction. Probably, additional filling partly originatesfrom passive diastasis filling.23

LV filling late after cardioversion. As decreased atrial compliance may beresponsible for increased early diastolic filling rate directly after cardioversion of chronicAF, normalized atrial compliance may be responsible for normalized early diastolicfilling rate after 3 months of sinus rhythm.

94 Chapter 8

TABLE 8.4. Values of left ventricular diastolic filling parameters early after cardioversion in patients withchronic atrial fibrillation in the present study compared with that of previous studies in English languageliterature.Study(referencenumber)

TimeafterCV

Atrial fibrillationduration

n %loneAF

ModeOfCV

Tech-nique

EDF(%FV)

AFF%FV

nAFF= 0

Interval several hours Manning (40) <1h 5 mo (3 wk - 24 mo) 21 14 E D - 16 ± 12 1 Manning (40) 3-4h 5 mo (3 wk - 24 mo) 9 - E D - 12 ± 6 0 Manning (41) <2 h >6 wk 21 - E+C D 83* 17 ± 11 - Vgelder (2) < 4 h 1-24 mo 8 63 E NS - 3 ± 5 6 Harjai (45) < 6h 77% > 28 d 40 15 E D 72* 28† - Harjai (45) < 6h 58% > 28 d 12 17 C D 60* 40† - O’Neill (42) 5 min 31 ± 31 d (2 - 100 d) 14 0 E D 85* 15 ± 13 4 O’Neill (42) 30 min 31 ± 31 d (2 - 100 d) 14 0 E D 87* 13 ± 11 3 Manning (41) <2 h 2 - 6 wk 22 - E+C D 78* 22 ± 15 - Manning (41) <2 h 0.3 - 2 wk 17 - E+C D 70* 30 ± 11 - Present study 6h 198 ± 242 23 100 E RNA 70 24 ± 7 1Interval 24 hours Ito (46) 24 h 15 ± 20 mo (1 - 71 mo) 20 45 E+C D - 31 ± 14 - Shapiro (29) 24 h 429 ± 245 d (> 1 wk) 14 - E+C D 85* 15‡ 5 Manning (40) 24 h 5 mo (3 wk - 24 mo) 14 - E D - 16 ± 8 0 Pollak (43) 24 h 133 ± 108 d (2 wk - 1 yr) 20 - E D 80* 20 ± 8 - Pollak (43) 24 h 100 ± 54 (2 wk - 1 yr) 17 - E§ D 90* 10 ± 4 - Jovic (25) < 24 h < 6 mo� 21 57 C D 74* 26 ± 7 0 Omran (44) 1 d 78 ± 89 d (>3 d) 20 25 IAD D 80* 20¶ 1 Manning (41) <24 h >6 wk 21 - E+C D 83* 17 ± 12 - Navazio (26) 24 h > 5 wk 12 - C D 71* 29 ± 3 - O’Neill (42) 24 h 31 ± 31 d (2 - 100 d) 14 0 E D 78* 22 ± 14 1 Manning (41) <24 h 2 - 6 wk 22 - E+C D 78* 22 ± 13 - Manning (41) <24 h 0.3 - 2 wk 17 - E+C D 68* 32 ± 11 - Shapiro (29) 24 h < 1 wk 18 - E D 67* 33‡ 0Studies are categorized according to the interval between cardioversion and measurement of filling.Within the categories patients are arranged according to atrial fibrillation duration.*Indirectly calculated by the formula: AFF + EDF = 100% filling volume when only AFF is given in thearticle; †Calculated by the formula AFF = Doppler echocardiographic A wave time-velocity integral / sumof E and A wave time velocity integrals; ‡Calculated by the formula: AFF = (AFF / EDF) / (1 + (AFF /EDF)) when the AFF / EDF ratio is given in the article; §Group with additional medication (Sotalol);�According to the text: patients had chronic atrial fibrillation, so we assume atrial fibrillation duration ofat least 24 hours according to the definition by Van Gelder (51); ¶Measured in a graph;- = data not applicable or not mentioned in the article; AFF = additional filling fraction; C = chemical; CV= cardioversion; D = Doppler echocardiography of transmitral flow; d = days; E = electrical; EDF = earlydiastolic filling fraction; FV = filling volume; IAD = internal atrial defibrillation; mo = months; n = number ofpatients; NS = nuclear stehoscope; RNA = radionuclide angiography; wk = weeks.

In Doppler echocardiographic diastolic transmitral left ventricular inflow studiesafter cardioversion in humans with AF, a decreasing E/A ratio was observed, indicatingincreasing atrial contractile function and/or progression of left ventricular diastolicfunction abnormalities.29,30 Evidence favoring progression of diastolic dysfunction andleft ventricular hypertrophy after termination of supraventricular tachycardia wasproduced in animal studies.11,14,15 A decreased left ventricular relaxation velocity incombination with persistent decreased atrial compliance may also explain thenormalized PFR 3 months after cardioversion. Probably, both factors play an importantrole. The current data may form therefore the first step in the recognition of earlydiastolic filling abnormalities to be an important feature in tachycardia induced leftventricular dysfunction.


Limitations. The limited number of measurements in the present study represents oneof the limitations of this study. More measurements after cardioversion could haverevealed a time course of changed diastolic left ventricular properties. Also, the limitednumber of patients available for restudying after 3 months influences the statisticalpower of the follow-up data late after cardioversion. Further, in radionuclide angiographytechnical limitations of various kind, like e.g. gating mode and temporal smoothing arepresent. The application of forward gating and temporal smoothing with 5 Fourierharmonics are however relatively reliable techniques in the radionuclide angiographicevaluation of early diastolic filling, and result in reproducible results.19,21 In addition, theuse of only one normalization parameter (FV) may make interpretation to physiologicalevents uncertain. Since the patient groups in the present study had different meanejection fractions of the left ventricle, measurements normalized to the end-diastolicvolume would have been dependent on this variable.

Conclusion. The present study reports increased early diastolic left ventricular fillingrate shortly after cardioversion of AF, followed by a long term decreasing early leftventricular filling rate. Thus, decreased atrial compliance probably plays an importantrole early after cardioversion, whereas LV diastolic dysfunction may develop late afterelectrical cardioversion.

References

1. Van den Berg MP, Van Veldhuisen DJ, Crijns HJGM, Lie KI. Reversion of tachycardiomyopathyafter ß-blocker. Lancet 1993;341:1667.

2. Van Gelder IC, Crijns HJGM, Blanksma PK, et al. Time course of hemodynamic changes andimprovement of exercise tolerance after cardioversion of chronic atrial fibrillation unassociatedwith cardiac valve disease. Am J Cardiol 1993;72:560-566.

3. Rodriguez LM, Smeets JLRM Xie B, et al. Improvement in left ventricular function by ablation ofatrioventrcular nodal conduction in selected patients with lone atrial fibrillation. Am J Cardiol1993;72:1137-1141.

4. Packer DL, Bardy GH, Worley SJ, et al. Tachycardia-induced cardiomyopathy: a reversible formof left ventricular dysfunction. Am J Cardiol 1986;57:563-570.

5. Grogan M, Smith HC, Gersh BJ, Wood DL. Left ventricular dysfunction due to atrial fibrillation inpatients initially believed to have idiopathic dilated cardiomyopathy. Am J Cardiol 1992;69:1570-1573.

6. Ohno M, Cheng CP, Little WC. Mechanism of altered patterns of left ventricular filling during thedevelopment of congestive heart failure. Circulation 1994;89:2241-2250.

7. Van den Berg MP, Crijns HJGM, Gosselink ATM, et al. Chronotropic response to exercise inpatients with atrial fibrillation: relation to functional state. Br Heart J 1993;70:150-153.

8. Zile MR. The development of diastolic dysfunction in chronic tachycardia. In: Spinale FG, editor.Pathophysiology of tachycardia-induced heart failure. Armonk, NY: Futura Publishing Company,Inc:25-44.

96 Chapter 8

9. Shinbane JS, Wood MA, Jensen DN, Ellenbogen KA, Fitzpatrick AP, Scheinman MM.Tachycardia-induced cardiomyopathy: a review of animal models and clinical studies. J Am CollCardiol 1997;29:709-715.

10. Komamura K, Shannon RP, Pasipoularides A, et al. Alterations in left ventricular diastolicfunction in conscious dogs with pacing-induced heart failure. J Clin Invest 1992;89:1825-1838.

11. Tomita M, Spinale FG, Crawford FA, Zile MR. Changes in left ventricular volume, mass, andfunction during the development and regression of supraventricular tachycardia-inducedcardiomyopathy: disparity between recovery of systolic versus diastolic function. Circulation1991;83:635-644.

12. Sasayama S, Asanoi H, Ishizaka S. Mechanics of contraction and relaxation of the ventricle inexperimental heart failure produced by rapid ventricular pacing in the conscious dog. Eur Heart J1991;12(Supplement C):35-41.

13. Zile MR, Mukherjee R, Clayton C, Kato S, Spinale FG. Effects of chronic supraventricular pacingtachycardia on relaxation rate in isolated cardiac muscle cells. Am J Physiol 1995;268 (HeartCirc Physiol 37):H2104-H2113.

14. Spinale FG, Zellner JL, Tomita M, Crawford FA, Zile MR. Relation between ventricular andmyocyte remodeling with the development and regression of supraventricular tachycardia-induced cardiomyopathy. Circ Res 1991;69:1058-1067.

15. Spinale FG, Tanaka R, Crawforde FA, Zile MR. Changes in myocardial blood flow duringdevelopment of and recovery from tachycardia-induced cardiomyopathy. Circulation 1992;85:717-729.

16. Muntinga HJ, Van den Berg F, Knol HR, et al. Normal values and reproducibility of left ventricularfilling parameters by radionuclide angiography. Int J Cardiac Imaging 1997;13:165-171.

17. Crijns HJGM, Van Gelder IC, Van Gilst WH, Hillege H, Gosselink ATM, Lie KI. Serialantiarrhythmic drug treatment to maintain sinus rhythm after electrical cardioversion for chronicatrial fibrillation or atrial flutter. Am J Cardiol 1991;68:335-341.

18. Crijns HJGM, Van Gelder IC, Tieleman RG, et al. Long-term outcome of electrical cardioversionin patients with chronic atrial flutter. Heart 1997;77:56-61.


20. Udelson JE, Bonow RO. Radionuclide angiographic evaluation of left ventricular diastolic function.In: Gaasch WH, LeWinter MM ,editors. Left ventricular diastolic dysfunction and heart failure.Philadelphia: Lea & Febiger, 1994:167-191.

21. Bacharach SL, Green MV, Vitale D, et al. Optimum Fourier filtering of cardiac data: a minimum-error method: concise communication. J Nucl Med 1983;24:1176-1184.

22. Yellin EL, Meisner JS, Nikolic SD, Keren G. The scientific basis for the relations betweenpulsed-Doppler transmitral velocity patterns and left heart chamber properties. Echocardiography1992;9:313-338.

23. Muntinga HJ, Van den Berg F, Knol HR, Schuurman JJ, Van der Wall EE. Quantification of theatrial contribution to diastolic filling during radionuclide angiography. Nuclear MedicineCommunications 1997;18:642-647.

24. Stewart RAH, McKenna WJ. Assessment of diastolic filling indexes obtained by radionuclideventriculography. Am J Cardiol 1990;65:226-230.

25. Jovic A, Troskot R. Recovery of atrial systolic function after pharmacological conversion ofchronic atrial fibrillation to sinus rhythm: a Doppler achocardiographic study. Heart 1997;77:46-49.

26. Navazio A, Chesi G, Montanari P, Marani L. Recovery of left atrial mechanical function afterchemical cardioversion of chronic atrial fibrillation. J Am Coll Cardiol 1998;31: 296A (abstr).


27. Iuchi A, Oki T, Fukuda N, et al. Changes of transmitral and pulmonary venous flow velocitypatterns after cardioversion of atrial fibrillation. Am heart J 1996;131:270-275.

28. Tabata T, Oki T, Iuchi A, et al. Evaluation of left atrial appendage function by measurement ofchanges in flow velocity patterns after electrical cardioversion in patients with isolated atrialfibrillation. Am J Cardiol 1997;79:615-620.

29. Shapiro EP, Effron MB, Lima S, Ouyang P, Siu CO, Bush D. Transient atrial dysfunction afterconversion of chronic atrial fibrillation to sinus rhythm. Am J Cardiol 1988;62:1202-1207.

30. Xiong C, Sonnhag C, Nylander E, Wranne B. Atrial and ventricular function after cardioversion ofatrial fibrillation. Br Heart J 1995;74:254-260.

31. Moe GW, Grima EA, Howard RJ, Seth R, Armstrong PW. Left ventricular remodeling anddisparate changes in contractility and relaxation during development of and recovery fromexperimental heart failure. Cardiovasc Res 1994;28:66-71.

32. Magorien DJ, Shaffer P, Bush C, Magorien RD, Kolibash AJ, Unverferth DV, Bashore TM.Hemodynamic correlates for timing intervals, ejection rate and filling rate derived from theradionuclide angiographic volume curve. Am J Cardiol 1984;53:567-571.



35. Takenaka K, Shiota T, Sakamoto T, et al. Effect of acute systemic blood pressure elevation onleft ventricular filling with and without mitral regurgitation. Am J Cardiol 1989;63:623-625.

36. Ishida Y, Meisner JS, Tsujioka K, et al. Left ventricular filling dynamics: influence of leftventricular relaxation and left atrial pressure. Circulation 1986;74:187-196.

37. Van Gelder IC, Crijns HJGM, Van Gilst WH, Hamer HPM, Lie KI. Decrease of right and left atrialsizes after direct-current electrical cardioversion in chronic atrial fibrillation. Am J Cardiol1991;67:93-95.

38. Gosselink ATM, Crijns HJGM, Hamer HPM, Hillege HJ, Lie KI. Changes in left and right atrialsize after cardioversion of atrial fibrillation: role of mitral valve disease. J Am Coll Cardiol1993;22:1666-1672.

39. Davies MJ, Pomerance A. Pathophysiology of atrial fibrillation in man. Br Heart J 1972;34:520-525.

40. White CW, Kerber RE, Weiss HR, Marcus ML. The effects of atrial fibrillation on atrial pressure-volume and flow relationships. Circ Res 1982;51:205-215.

41. Manning WJ, Leeman DE, Gotch PJ, Come PC. Pulsed Doppler evaluation of atrial mechanicalfunction after electrical cardioversion of atrial fibrillation. J Am Coll Cardiol 1989;13:617-623.

42. Manning WJ, Silverman DI, Katz SE, et al. Impaired left atrial mechanical function aftercardioversion: relation to the duration of atrial fibrillation. J Am Coll Cardiol 1994;23:1535-1540.

43. O’Neill, Puleo PR, Bolli R, Rokey R. Return of mechanical function following electrical conversionof atrial fibrillation. Am Heart J 1990;120:353-359.

44. Pollak A, Falk RH. Aggravation of postcardioversion atrial dysfunction by sotalol. J Am CollCardiol 1995;25:665-671.

45. Omran H, Jung W, Rabahieh R, et al. Left atrial chamber and appendage function after internalatrial defibbrillation: a prospective and serial transesophageal echocardiographic study. J AmColl Cardiol 1997;29:131-138.

46. Harjai KJ, Mobarek SK, Cheirif J, Boulos LM, Murgo JP, Abi-samra F. Clinical variables affectingrecovery of left atrial mechanical function after cardioversion from atrial fibrillation. J Am CollCardiol1997;30:481-486.

98 Chapter 8

47. Ito T, Suwa M, Otake Y, et al. Assessment of left atrial appendage function after cardioversion ofatrial fibrillation: relation to left atrial mechanical function. Am Heart J 1998;135:1020-1026.

48. Fragasso G, Chierchia SL, Pizetti G, et al. Impaired left ventricular filling dynamics in patientswith angina and angiographically normal coronary arteries: effect of beta adrenergic blockade.Heart 1997;77:32-39.

49. Bonow RO, Frederick TM, Bacharach SL, et al. Atrial systole and left ventricular filling inhypertrophic cardiomyopathy: effect of verapamil. Am J Cardiol 1983;51:1386-1391.

50. Hoit BD, Walsh RA. Diastolic function in hypertensive heart disease. In: Gaasch WH, LeWinterMM ,editors. Left ventricular diastolic dysfunction and heart failure. Philadelphia: Lea & Febiger,1994:354-372.

51. Kono T, Sabbah HN, Rosman H, Alam M, Stein PD, Goldstein S. Left atrial contribution toventricular filling during the course of evolving heart failure. Circulation 1992;86:1317-1322.

52. Van Gelder IC. Characteristics of patients with chronic atrial fibrillation and the prediction ofsuccessful DC electrical cardioversion. In: Kingma JH et al, editors. Atrial fibrillation, a treatabledisease? Dordrecht, The Netherlands: Kluwer academic publishers, 1992:67-86.

Chapter 9Summary and conclusions

eart failure due to diastolic left ventricular dysfunction is a public health problem ofincreasing importance. In 30 to 40% of patients presenting with heart failureprimary diastolic left ventricular dysfunction is found. Patients presenting withheart failure and normal left ventricular systolic function are assumed to have

diastolic dysfunction when no valvular dysfunction is present. When abnormal systolic leftventricular function or valvular heart disease are diagnosed diastolic dysfunction of theleft ventricle may be present as well. Heart failure due to diastolic left ventriculardysfunction may lead to typical symptoms and signs of congestive heart failure includingintolerance of exercise, and pulmonary oedema. Diagnosis is related to the underlyingcardiac abnormality, i.e. diastolic left ventricular dysfunction in hypertrophy due tohypertension or hypertrophic cardiomyopathy. Therapy is also depending on theunderlying cardiac abnormality. To date, no ‘gold standard’ is available to diagnosediastolic heart failure. The probability of diastolic dysfunction as the cause of heartfailure in individual patients can be increased by direct and indirect measures of leftventricular function. In clinical situations this requires a thorough knowledge of thephysiology and pathophysiology of diastolic left ventricular function, as well asknowledge of advantages and limitations of the used technique (chapter 2).

Radionuclide angiography of left ventricular function is a well known technique forstudying left ventricular function and is commonly used to estimate left ventricularejection fraction. With modifications this technique is also used to measure leftventricular diastolic function. In chapter 3 the normal values and reproducibility of leftventricular filling parameters by radionuclide angiography are described. Since age,heart rate and left ventricular ejection fraction may influence left ventricular filling rate, wedetermined normal values in relation to these variables. For determining reproducibilitywe performed radionuclide angiography before and after exercise after a period of rest(measurement 2) in 20 patients with normal findings at coronary angiography and leftventriculography. We also compared our findings with values of diastolic fillingparameters measured with radionuclide angiography in normal individuals by otherauthors. Normal values for peak filling rate (mean and standard deviation of the 20patients) were 2.2 ± 0.6 end-diastolic volume / second (measurement 2: 2.4 ± 0.7 end-diastolic volume / second , correlation coefficient r = 0.82), for time to peak filling rate

H

100 Chapter 9

198 ± 22 msec (measurement 2: 203 ± 24 msec, r = 0.45), and for atrial contribution todiastolic filling 31 ± 11% (measurement 2: 31 ± 10%, r = 0.72). These findings arecomparable with values of diastolic left ventricular filling parameters reported by otherauthors using the same technique. Both peak filling rate and time to peak filling ratewere correlated to age (respectively r = -0.68 and r = 0.48, P<0.05). Peak filling ratewas also correlated with heart rate and ejection fraction (respectively r = 0.51 and r =0.50). Time to peak filling rate was however not correlated with heart rate and ejectionfraction. Atrial contribution to diastolic filling was correlated with heart rate (r = 0.79, P <0.01). It is concluded that parameters of diastolic left ventricular filling can reliably andreproducibly be assessed in individual patients using radionuclide angiography. Thistechnique may be used to serially follow diastolic function. Interindividual comparison ofdiastolic filling must be performed with caution, and be placed in perspective of thedependency on age, heart rate and ejection fraction.

In chapter 4 two techniques used for the quantification of atrial contribution todiastolic filling during radionuclide angiography are compared. The moment of onset ofatrial contraction is usually determined by the shape of the time activity curve. Ameasure outside the curve has not been applied to verify the exact duration of the atrialcontribution phase. In 34 patients with a variety of diseases, the atrial contribution wasquantified using a flow-volume loop which was constructed from the radionuclide timeactivity curve of the left ventricle and its first derivative. The duration of the atrialcontribution phase derived from the flow-volume loop correlated well with the PQ-intervalon the electrocardiogram. Also the relative filling volumes within these time periods wereclosely correlated. In a subset of patients these time intervals were however not exactlythe same. In these patients the duration of the atrial contribution phase was alwayslonger than the PQ-interval on the electrocardiogram, indicating the existence of passivediastasis flow before the onset of atrial contraction. This was more apparent in patientswith low heart rates than in those with high heart rates. Despite the close correlationbetween the PQ-interval on the electrocardiogram and the duration of the atrialcontribution phase with radionuclide angiography, they may represent different entities.The PQ-interval on the electrocardiogram is better for determining the moment of onsetof atrial mechanical activity during radionuclide angiography than the interval betweenthe onset of the atrial contribution phase and the end of diastole in the flow-volume loop.

The diagnostic and therapeutic implications of a circadian rhythm in leftventricular relaxation, observed in 15 patients with congestive heart failure, aredescribed in chapter 5. Since afterload is an important determinant of left ventricularrelaxation, the physiological circadian variation of arterial blood pressure may inducediurnal variability of left ventricular filling parameters in patients with congestive heartfailure. At 10 AM compared to 1 PM a higher systolic blood pressure (P < 0.01) anddiastolic pressure (P < 0.01), a lower peak filling rate (P < 0.01), and a longer time topeak filling rate (P<0.05) were found. Using repeated measures analysis of variance,the changes of peak filling rate were related to the daily changes of systolic and diastolic

Summary and conclusions 101

blood pressure (each P < 0.05). These data suggest that left ventricular relaxation asmeasured by peak filling rate and time to peak filling rate is inversely correlated with acircadian fall in blood pressure in patients with congestive heart failure. This could havethe important diagnostic implication that timing should be taken into account whenassessing diastolic left ventricular function in patients with congestive heart failure, aswell as the therapeutic implication that treatment regimens should account for anincreased risk of pulmonary oedema in these patients early in the morning.

Patients with chronic congestive heart failure and decreased systolic leftventricular performance often have diastolic dysfunction as well. Decreased diastolicfunction in such patients may lead to worsening of exercise tolerance and prognosis.Increased diastolic function may prevent this. In patients with coronary artery disease,calcium antagonists may be beneficial for diastolic dysfunction. The potentialdeterioration of systolic function on administration of calcium antagonists in patients withcongestive heart failure is reported not to be present in mibefradil. We thereforeinvestigated the effect of this calcium antagonist on left ventricular diastolic function inpatients with congestive heart failure and depressed ejection fraction due to previousmyocardial infarction (chapter 6). In 15 patients with New York Heart Association classII or III for dyspnoea radionuclide angiography was performed to obtain measures ofsystolic and diastolic left ventricular function. Systolic and diastolic blood pressure andheart rate were also obtained. Patients were randomly divided into three groups. Group I(5 patients) received placebo medication; group IIA (6 patients) received mibefradil6.25, 12.5, or 25 mg/day; and group IIB (4 patients) received mibefradil 50 or 100mg/day. Measurements were made before and after the first dose and after one week oftreatment before and after the final dose. Mibefradil clearly decreased heart rate(repeated measures analysis of variance p < 0.05). No statistically significant effects ofmibefradil were noted on blood pressure or left ventricular function. In our studyconditions, mibefradil caused no worsening of systolic function and preserved diastolicfunction in short-term treatment of patients with decreased ejection fraction and heartfailure. In 1998 mibefradil was withdrawn from the market because of seriousinteractions with other drugs.

In the next chapters the effect of atrial fibrillation on left ventricular function isdiscussed. Chapter 7 is focussed on systolic left ventricular function. Since atrialfibrillation is characterised by a randomly irregular ventricular response, resulting incontinuous variation in left ventricular beat-to-beat mechanical behaviour andhaemodynamic variables, we used a non-imaging computerised probe, together with aballoon-tipped flow directed catheter and a non-invasive fingertip blood pressuremeasurement system linked to a personal computer in order to perform beat-to-beat leftventricular volume measurement and invasive and non-invasive haemodynamicmonitoring. To assess independent determinants of beat-to-beat variations in leftventricular performance during atrial fibrillation a statistical model was constructedcontaining haemodynamic variables of 500 consecutive RR intervals in seven patients

102 Chapter 9

with chronic non-valvular atrial fibrillation. Positive independent relations with leftventricular ejection fraction were found for the preceding RR interval, left ventricularcontractility (end-systolic pressure / end-systolic volume), and end-diastolic volume,whereas inverse relations were found for afterload (end-systolic pressure / strokevolume), preceding end-systolic volume and preceding contractility. A relative stronginteraction was found between end-diastolic volume and afterload, indicating thatejection fraction was relatively more enhanced by preload in the presence of lowafterload. The varying left ventricular systolic performance during atrial fibrillation istherefore independently influenced by beat-to-beat variations of cycle length, preload,afterload and contractility. Beat-to-beat variations in preload exhibit their effect onventricular performance mainly in the presence of a low afterload.

Left ventricular diastolic function after direct current electrical cardioversion ofatrial fibrillation is evaluated in chapter 8. Tachycardia associated left ventriculardysfunction in atrial fibrillation is characterised by decreased left ventricular (LV) systolicperformance, but the data in the literature concerning diastolic function are discordant.Serial radionuclide angiography was performed in 23 patients (mean age 65 ± 13years) with chronic (mean duration 198 ± 242 days) lone atrial fibrillation aftercardioversion. Twenty patients with normal coronary angiograms and no myocardial orvalvular disease served as control subjects. Early after cardioversion patients with atrialfibrillation had lower left ventricular ejection fraction (EF) compared to controls (54 ± 10% end-diastolic volume vs. 60 ± 8 % end-diastolic volume, P<0.05), higher peak fillingrate (PFR, 3.91 ± 0.68 filling volume/s vs. 3.29 ± 0.64 filling volume/s, P<0.01), shortertime to PFR (TPFR, 162 ± 46 ms vs. 196 ± 42 ms, P<0.05), and lower additional fillingfraction (AFF, 24 ± 7 % filling volume vs. 32 ± 7 % filling volume, P < 0.05). In 5 patientsremaining in sinus rhythm for 3 months, PFR decreased from 4.03 ± 0.63 filling volume/sto 3.13 ± 0.54 filling volume/s (P<0.01). The increase of EF from 54 ± 11 to 57 ± 1 %,TPFR from 126 ± 46 ms to 187 ± 44 ms and of AFF from 25 ± 3 % filling volume to 30 ±7 % filling volume were not statistically significant. None of the factors recorded earlyafter cardioversion predicted recurrence of atrial fibrillation. These data do not confirmthe existence of diastolic left ventricular dysfunction early after electrical cardioversion inpatients with lone atrial fibrillation, but are consistent with decreased atrial complianceand contractility due to atrial remodelling. Persistent sinus rhythm in patients 3 monthsafter electrical cardioversion for chronic lone atrial fibrillation may be accompanied withaltered diastolic LV function as an important feature of post-tachycardia left ventriculardysfunction. The measurements in patients with persistent sinus rhythm may alternativelybe due to (partial) reversibility of atrial remodelling.

CONCLUSIONSIn summary, the following conclusions can be drawn from the above findings of thepresent thesis:


• Radionuclide angiography is a reliable and reproducible technique to assessdiastolic left ventricular function. It may be used to serially follow diastolic function inindividual patients. Interindividual comparison of diastolic filling can be performedprovided that the results are placed in perspective of the dependency on age, heartrate and ejection fraction.

• Additional filling of the left ventricle is partly caused by atrial contraction and partlycaused by passive filling.

• The time of day on which diastolic left ventricular function is assessed in patients withdecreased systolic function has influence on the result.

• The calcium antagonist mibefradil has no effect on diastolic left ventricular functionafter short-term treatment of patients with heart failure and decreased systolicfunction.

• The beat-to-beat variation of systolic left ventricular function in patients with atrialfibrillation is not only caused by variation of preload, but also by variation ofcontractility, afterload and RR-interval.

• The changed inflow profile in the left ventricle early after cardioversion in patientswith chronic “lone” atrial fibrillation is due to atrial remodelling. Diastolic leftventricular dysfunction and partial reversibility of atrial remodelling are probably bothresponsible for a changed inflow in the left ventricle late after cardioversion.

FUTURE DIRECTIONSFrom the above conclusions some future directions can be formulated:• Reliable limits of the normal range of parameters of diastolic left ventricular function

must be formulated in order to increase clinical applicability.• The contribution of passive filling to left ventricular filling must be quantified. Does a

relation exist with left ventricular compliance, left ventricular relaxation or clinicalparameters?

• Collect details of the variation of diastolic left ventricular filling parameters inpatients with heart failure and decreased systolic left ventricular function. Is thisvariation present in other patient groups? Is there a relation with the moment of drugintake?

• What effect have calcium antagonists on de diastolic left ventricular function inpatients with diseases other than heart failure, e.g. after an episode with atrialfibrillation?

• The research of mechanisms behind the beat-to-beat variation of left ventricularcontractility. Can the variability be reduced in order to improve left ventricularfunction?

• What is the role of reduced atrial compliance in chronic “lone” atrial fibrillation onatrial fibrillation recurrence after cardioversion?

104 Chapter 9

Many unknown factors are still present In the knowledge of left ventricular function. Aselection is discussed in the present thesis. Some findings will need to be confirmed infuture work. One of the important questions of this moment is how to trace patients withdiastolic dysfunction as a cause of heart failure. Whether radionuclide angiography canplay a role here depends on the clinical availability of this technique, as well as thesuccess of future work to clearly indicate diagnostic borders of existing parameters.Also new reliable diagnostic parameters of diastolic heart failure need be found. Sincepatients with diastolic left ventricular dysfunction often present with exercise intolerance,these measurements must preferably also be possible with exercise. Finally, relationswith clinical parameters and clinical end-points need to be strengthened.

Hoofdstuk 9Samenvatting en conclusies

artfalen door diastolische linker hartkamerdisfunctie is eengezondheidsprobleem van toenemende omvang. Bij 30 tot 40% van depatiënten met hartfalen wordt primair een diastolische linker

hartkamerdisfunctie gevonden. Aangenomen wordt dat patiënten met hartfalen eendiastolische disfunctie van de linker hartkamer hebben wanneer de systolische functietussen de uiterste normaalwaarden ligt en geen klepvitia aanwezig zijn. Systolischelinker hartkamerdisfunctie sluit diastolische linker hartkamerdisfunctie echter niet uit.Hartfalen door diastolische linker hartkamerdisfunctie gaat gepaard met de typischesymptomen van hartfalen door systolische linker hartkamerdisfunctie zoalsinspanningsintolerantie en longoedeem. De diagnose heeft betrekking op deonderliggende hartaandoening, bijvoorbeeld diastolische linker hartkamerdisfunctie bijhypertrofie ten gevolge van hypertensie of hypertrofische cardiomyopathie. Ook detherapie is afhankelijk van de onderliggende hartaandoening. Tot op heden is nog geen‘gouden standaard’ beschikbaar om diastolisch hartfalen te diagnosticeren. Bijindividuele patiënten kan door directe en indirecte meting van de linker hartkamerfunctiede waarschijnlijkheid dat diastolische disfunctie bijdraagt aan het ontstaan van hartfalenworden geschat. In klinische situaties is hiervoor een gedegen kennis van de fysiologieen pathofysiologie van diastolische linker kamer functie noodzakelijk, maar ook kennisvan de voor- en nadelen van de gebruikte diagnostische techniek (hoofdstuk 2).

Radionuclide angiografie is een bekende techniek voor de studie van de linkerhartkamerfunctie en wordt met name gebruikt voor de meting van de ejectiefractie vande linker hartkamer. Met bepaalde aanpassingen is deze techniek ook bruikbaar voorstudie van de diastolische linker hartkamerfunctie. In hoofdstuk 3 worden denormaalwaarden en de reproduceerbaarheid van de vullingparameters van de linkerhartkamer beschreven zoals gemeten met radionuclide angiografie. Omdat leeftijd,hartfrequentie en ejectiefractie van de linker hartkamer het vullingtempo van de linkerhartkamer beïnvloeden bepaalden wij normaalwaarden in relatie tot deze variabelen.Om de reproduceerbaarheid te bepalen werd radionuclide angiografie van de linkerhartkamer verricht voor inspanning en na een rustperiode na inspanning (meting 2) bij20 patiënten met normale bevindingen bij coronair angiografie en ventriculografie. Debevindingen werden vergeleken met de waarden van diastolische linker

H

106 Hoofdstuk 9

hartkamervulling van gezonde personen door andere onderzoekers gemeten metradionuclide angiografie. Gemiddelde en standaard deviatie van de 20 patiënten voorhet hoogste vullingtempo van de linker hartkamer (PFR = peak filling rate) was 2,2 ± 0,6einddiastolisch volume / seconde (meting 2: 2,4 ± 0,7 einddiastolisch volume / seconde,r = 0,82), voor tijdsduur tot het hoogste vullingtempo (TPFR = time to peak filling rate)198 ± 22 ms (meting 2: 203 ± 24 ms, r = 0,45), en voor de boezembijdrage aan dediastolische vulling 31 ± 11% (meting 2: 31 ± 10%, r = 0,72). Deze bevindingen zijnvergelijkbaar met waarden van diastolische vullingparameters van de linker hartkamerdie door andere onderzoekers werden gerapporteerd bij gebruik van dezelfde techniek.Zowel PFR als TPFR waren gecorreleerd met de leeftijd (respectievelijk r = -0,68 en r =0,48, P<0,05). PFR was ook gecorreleerd met hartfrequentie en ejectiefractie(respectievelijk r = 0,51 and r = 0,50). TPFR was echter niet gecorreleerd methartfrequentie en ejectiefractie, terwijl de boezembijdrage aan de diastolische linkerhartkamervulling wel was gecorreleerd met de hartfrequentie (r = 0,79, P < 0,01). In dithoofdstuk wordt geconcludeerd dat diastolische vullingparameters van de linkerhartkamer in individuele patiënten betrouwbaar en reproduceerbaar verkregen kunnenworden met radionuclide angiografie. Deze techniek kan daarom gebruikt worden omdiastolische functieparameters serieel te volgen. Interindividuele vergelijking vandiastolische vulling moet echter met grote voorzichtigheid worden uitgevoerd, engeplaatst worden tegen de achtergrond van de afhankelijkheid van de parameters vanleeftijd, hartfrequentie en ejectiefractie.

In hoofdstuk 4 worden twee radionuclide angiografische technieken vergelekendie de boezembijdrage aan de linker hartkamervulling kwantificeren. Het momentwaarop de boezemcontractie begint wordt meestal bepaald door de vorm van de curvewaarin de gemeten stralingsactiviteit van het met 99mtechnetium gelabelde bloed in delinker hartkamer uitgezet is tegen de tijdsduur van een hartcyclus. Een referentiewaardeom de duur van de fase waarin de linker boezem de linker hartkamer vult te verifiërenwordt daarbij niet toegepast. Van 34 patiënten met verschillende onderliggendehartziekten werd de boezembijdrage gekwantificeerd met gebruikmaking van eenbloedstroom-volumelus welke geconstrueerd was uit de activiteitcurve van de linkerhartkamer in de tijd en de eerste afgeleide curve hiervan. De duur van deboezembijdragefase gemeten in de bloedstroom-volumelus correleerde goed met hetPQ interval op het elektrocardiogram. Ook het relatieve vullingvolume van deze fasecorreleerde goed met het relatieve vullingvolume na het begin van de P-top op hetelektrocardiogram. In een subgroep van patiënten waren de duur van deboezembijdragefase en het PQ interval op het elektrocardiogram niet exact hetzelfde.Bij deze patiënten was de duur van de boezembijdragefase altijd langer dan het PQinterval op het elektrocardiogram. Deze waarneming werd vaker gedaan bij patiëntenmet een lage hartfrequentie. Hiermee wordt bevestigd dat het PQ interval op hetelektrocardiogram en de duur van de boezembijdragefase met radionuclide angiografieverschillende entiteiten vertegenwoordigen ondanks hun onderling goede correlatie.

Samenvatting en conclusies 107

Mogelijk is een passieve diastasisstroom in de linker hartkamer voor het begin van deboezemcontractie mede verantwoordelijk voor het ontstaan van de boezembijdragefaseop het radionuclide angiogram. Het PQ interval op het elektrocardiogram is daarombeter voor het bepalen van het begin van boezemactiviteit tijdens radionuclideangiografie dan de duur van de boezembijdrage bij de diastolische vulling van de linkerhartkamer in de bloedstroom-volumelus.

De diagnostische en therapeutische implicaties van een circadiaans ritme vande linker hartkamerrelaxatie welke werd geobserveerd bij 15 patiënten met hartfalen,worden beschreven in hoofdstuk 5. Omdat "afterload" aanzienlijke invloed heeft op delinker hartkamerrelaxatie, zou de fysiologische circadiaanse variatie van de arteriëlebloeddruk bij patiënten met hartfalen variabiliteit van de vullingparameters van de linkerhartkamer kunnen veroorzaken. In vergelijking met 1 uur ‘s middags werd om 10 uur ‘sochtends een hogere systolische (P < 0,01) en diastolische bloeddruk (P < 0,01), eenlager hoogste vullingtempo van de linker hartkamer (PFR = peak filling rate, P < 0,05)en een langere tijd tot het hoogste vullingtempo (TPFR = time to peak filling rate, P <0,05) gemeten. Met gebruikmaking van variantieanalyse bij herhaalde metingen, warende veranderingen van PFR gerelateerd aan de dagelijkse veranderingen van desystolische en diastolische bloeddruk (iedere P < 0,05). Deze gegevens suggereren datde linker hartkamerrelaxatie gemeten met de PFR en de TPFR omgekeerdgecorreleerd is met een circadiaans patroon in het dagelijks bloeddrukverloop bijpatiënten met hartfalen. Dit kan de belangrijke diagnostische implicatie hebben datrekening moet worden gehouden met het tijdstip van meten bij het bepalen vandiastolische linker hartkamerfunctie van patiënten met hartfalen, en de therapeutischeconsequentie dat behandelstrategieën rekening zouden moeten houden met eentoegenomen risico op longoedeem bij deze patiënten vroeg in de ochtend.

Patiënten met chronisch hartfalen en afgenomen systolische linkerhartkamerfunctie hebben vaak daarnaast ook diastolische disfunctie. Dit kan bij dezepatiënten tot verslechtering van de inspanningstolerantie en de prognose leiden.Verbetering van de diastolische linker hartkamerfunctie zou dit kunnen voorkomen. Bijpatiënten met coronaire hartziekte kunnen calcium antagonisten een gunstig effecthebben op de diastolische linker hartkamerfunctie. De potentiële verslechtering van desystolische linker hartkamerfunctie na toediening van calcium antagonisten bij patiëntenmet hartfalen wordt niet bij mibefradil gerapporteerd. Daarom werd het effect van dezecalciuantagonist op de diastolische linker hartkamerfunctie onderzocht bij patiënten methartfalen en verminderde ejectiefractie na een eerder myocardinfarct (hoofdstuk 6).Radionuclide angiografie werd verricht om maten voor de systolische en diastolischelinker hartkamerfunctie te meten bij 15 patiënten met kortademigheidklasse II of IIIvolgens de indeling van de New York Heart Association. Ook werden de systolische ende diastolische bloeddruk en de hartfrequentie gemeten. De patiënten werdenwillekeurig verdeeld in 3 groepen. Groep I (5 patiënten) kreeg placebo medicatie; groepIIA (6 patiënten) kreeg mibefradil 6,25, 12,5 of 25 mg / dag; en groep IIB (4 patiënten)

108 Hoofdstuk 9

kreeg mibefradil 50 of 100 mg / dag. De metingen werden voor en na de eerste gift enna 1 week behandeling voor en na de laatste gift verricht. Mibefradil verminderdeduidelijk de hartfrequentie (variantieanalyse van herhaalde metingen P < 0,05). Geenstatistisch significant effect van mibefradil werd opgemerkt op de bloeddruk of op delinker hartkamerfunctie. Onder deze studievoorwaarden van korte termijnbehandelingvan patiënten met verminderde ejectiefractie en hartfalen veroorzaakte mibefradil geenverslechtering van de systolische hartfunctie en werd de diastolische functie behouden.In verband met ernstige interacties met andere geneesmiddelen werd mibefradil doorde fabrikant medio 1998 van de markt gehaald.

In de volgende hoofdstukken wordt het effect van boezemfibrilleren op de linkerhartkamerfunctie behandeld. In hoofdstuk 7 ligt het accent op de systolische linkerhartkamerfunctie. Omdat boezemfibrilleren gekarakteriseerd wordt door volstrektonregelmatig kamervolgen, resulterend in een continu van slag tot slag variëren van hetmechanisch gedrag van de linker hartkamer en hemodynamische variabelen, werdeneen zogenaamde nucleaire stethoscoop, een Swan-Ganz katheter en een niet-invasiefbloeddrukmeetsysteem gekoppeld aan een personal computer, ten einde van slag totslag volumemetingen, en invasieve en niet-invasieve hemodynamische metingen tekunnen verrichten. Om onafhankelijke variabelen te verkrijgen die het van slag tot slagvariëren van de linker hartkamerfunctie tijdens boezemfibrilleren bepalen, werd eenstatistisch model geconstrueerd van de hemodynamische variabelen van 500opeenvolgende RR intervallen bij zeven patiënten met chronisch niet-valvulairboezemfibrilleren. Positief onafhankelijke relaties met de ejectiefractie van de linkerhartkamer werden gevonden voor het voorafgaande RR interval, de contractiliteit van delinker hartkamer (einddiastolische druk / eindsystolisch volume), en het einddiastolischevolume, terwijl omgekeerde relaties gevonden werden voor de “afterload”(eindsystolische druk / slagvolume), het voorafgaande eindsystolische volume en devoorafgaande contractiliteit. Een relatief sterke interactie werd gevonden tusseneinddiastolisch volume en “afterload”, hetgeen een aanwijzing vormt dat deejectiefractie relatief meer toeneemt door de “preload” wanneer de “afterload” laag is.De variatie van de systolische linker hartkamerfunctie tijdens boezemfibrilleren wordtdaarom onafhankelijk beïnvloed door het van slag tot slag variëren van het RR interval,de “preload”, de “afterload” en de contractiliteit. Het van slag tot slag variëren van de“preload” heeft met name effect op de linker hartkamerfunctie wanneer de “afterload”laag is.

De diastolische linker hartkamerfunctie na elektrische cardioversie van patiëntenmet chronisch boezemfibrilleren wordt besproken in hoofdstuk 8. De met tachycardiegeassocieerde linker hartkamerdisfunctie wordt gekarakteriseerd door eenverminderde systolische functie, maar de literatuurgegevens over diastolische disfunctiezijn tegenstrijdig. Seriële radionuclide angiografie werd verricht bij 23 patiënten(gemiddelde leeftijd 65 ± 13 jaar) met chronisch (gemiddelde duur 198 ± 242 dagen)boezemfibrilleren zonder bekende oorzaak (“lone”) na cardioversie. Twintig patiënten


met normale bevindingen tijdens coronair angiografie en zonder myocardziekte ofklepvitium werden als controlepatiënten gebruikt. Vroeg na cardioversie haddenpatiënten met boezemfibrilleren een lagere ejectiefractie (EF) van de linker hartkamerdan de controlepatiënten (54 ± 10 % vs. 60 ± 8 % einddiastolisch volume, P < 0,05),een hoger hoogste vullingtempo (peak filling rate = PFR, 3,91 ± 0,68 vs. 3,29 ± 0,64vullingvolume / s, P < 0,01), een kortere tijd tot het hoogste vullingtempo (time to PFR =TPFR, 162 ± 46 vs. 196 ± 42 ms, P < 0,05), en een lager additionele vullingfractie (AFF= additional filling fraction, 24 ± 7 % vs. 32 ± 7 % vullingvolume, P < 0,05). Bij 5patiënten die 3 maanden sinus ritme behielden daalde PFR van 4,03 ± 0,63vullingvolume / s naar 3,13 ± 0,54 vullingvolume / s (P < 0,01). De toename van EF van54 ± 11 tot 57 ± 11 %, TPFR van 126 ± 46 ms tot 187 ± 44 ms en van AFF van 25 ± 3% tot 30 ± 7 % vullingvolume was niet statistisch significant. Geen van de factorengemeten vroeg na cardioversie voorspelde de terugkeer van boezemfibrilleren. Dezegegevens bevestigen het bestaan van diastolische linker hartkamerdisfunctie vroeg naelektrische cardioversie bij patiënten met “lone” boezemfibrilleren niet, maar komen welovereen met verminderde boezemcompliantie en contractiliteit ten gevolge vanboezemremodellering. De metingen bij patiënten met persisterend sinus ritme 3maanden na elektrische cardioversie voor chronisch “lone” boezemfibrilleren passen bijveranderde diastolische linker hartkamerfunctie als belangrijke uitingsvorm van linkerhartkamerdisfunctie na tachycardie. Deze metingen bij patiënten met persisterend sinusritme zouden ook een uiting kunnen zijn van partiele reversibiliteit vanboezemremodellering.

CONCLUSIESSamenvattend kan men uit bovenstaande bevindingen in deze thesis het volgendeconcluderen:• Radionuclide angiografie is een betrouwbare en reproduceerbare techniek om

diastolische linker hartkamerfunctie in kaart te brengen. Hierdoor kan dediastolische linker hartkamerfunctie serieel gevolgd worden in individuele patiënten.Tevens kan deze techniek gebruikt worden voor het interindividueel bestuderen vande diastolische linker hartkamerfunctie, mits rekening wordt gehouden met deeffecten die hartfrequentie, leeftijd en ejectiefractie hierop hebben.

• De additionele vullingfase van de linker hartkamer wordt deels doorboezemcontractie maar ook deels door passieve vulling veroorzaakt.

• Het tijdstip van de dag waarop bij patiënten met verminderde systolische linkerhartkamerfunctie de diastolische functie gemeten wordt is van invloed op deverkregen meetwaarde.

• De calciumantagonist mibefradil heeft bij patiënten met hartfalen en verminderdelinker hartkamerfunctie na korte behandeling geen effect op de diastolischehartkamerfunctie.

110 Hoofdstuk 9

• Het van slag tot slag variëren van de systolische linker hartkamerfunctie bij patiëntenmet boezemfibrilleren wordt niet alleen door variatie van de “preload” veroorzaakt,maar ook door variatie van de contractiliteit, “afterload” en het RR-interval.

• Vroeg na cardioversie bij patiënten met chronisch “lone” boezemfibrilleren wordt hetveranderde instroomprofiel van de linker hartkamer bepaald doorboezemremodellering. Diastolische linker hartkamerdisfunctie en gedeeltelijkeomkeerbaarheid van boezemremodellering zijn beide waarschijnlijk verantwoordelijkvoor het veranderde instroomprofiel van de linker hartkamer laat na cardioversie.

TOEKOMSTIGE ONDERZOEKSRICHTINGUitgaande van bovenstaande conclusies kan een aantal toekomstigeonderzoeksrichtingen en onderzoeksvragen vragen concreet worden geformuleerd:• Het formuleren van betrouwbare grenswaarden van het normale bereik van

parameters van de diastolische linker hartkamerfunctie met radionuclide angiografieten einde de klinische toepasbaarheid te vergroten.

• Het kwantificeren van de passieve vullingfase. Is er een relatie te leggen met dehartkamercompliantie, hartkamerrelaxatie of klinische parameters?

• Details verzamelen betreffende de variatie van diastolische vullingparameters bijpatiënten met hartfalen en systolisch verminderde linker hartkamerfunctie. Is dezevariatie ook aanwezig bij andere patiëntengroepen? Is er een relatie met de hettijdstip van inname van medicatie?

• Wat is het effect van calciumantagonisten op de diastolische hartfunctie bij patiëntenmet andere ziektebeelden dan hartfalen, bijvoorbeeld na een episode metboezemfibrilleren?

• Het onderzoeken van de mechanismen achter het van slag tot slag variëren van decontractiliteit van de linker hartkamer. Kan deze variabiliteit worden verminderdteneinde de linker hartkamerfunctie te verbeteren?

• Welke rol speelt de veranderde boezemcompliantie bij chronisch “lone”boezemfibrilleren met betrekking tot de recidiefkans op boezemfibrilleren nacardioversie?

In de kennis van de linker hartkamerfunctie zijn nog vele onbekende factoren. Een aantalfactoren wordt in deze thesis besproken. Sommige bevindingen zullen in toekomstigonderzoek moeten worden bevestigd. Centraal zal de vraag moeten komen te staanhoe men patiënten met diastolische disfunctie als oorzaak van hartfalen kan opsporen.Of radionuclide angiografie hierbij een rol kan spelen hangt niet alleen af van debeschikbaarheid van deze techniek in de kliniek, ook de diagnostische grenzen vanbestaande parameters zullen duidelijk aangegeven moeten kunnen worden. Daarnaastzal men op zoek moeten gaan naar nieuwe betrouwbare diagnostische parameters vandiastolisch hartfalen. Omdat patiënten met diastolische linker hartkamerdisfunctie zich


vaak met inspanningsintolerantie presenteren, moeten metingen ook bij inspanningmogelijk zijn. Een relatie zal moeten worden gelegd met klinische parameters eneindpunten.

Chapter 9Summary and conclusions

eart failure due to diastolic left ventricular dysfunction is a public health problem ofincreasing importance. In 30 to 40% of patients presenting with heart failureprimary diastolic left ventricular dysfunction is found. Patients presenting withheart failure and normal left ventricular systolic function are assumed to have

diastolic dysfunction when no valvular dysfunction is present. When abnormal systolic leftventricular function or valvular heart disease are diagnosed diastolic dysfunction of theleft ventricle may be present as well. Heart failure due to diastolic left ventriculardysfunction may lead to typical symptoms and signs of congestive heart failure includingintolerance of exercise, and pulmonary oedema. Diagnosis is related to the underlyingcardiac abnormality, i.e. diastolic left ventricular dysfunction in hypertrophy due tohypertension or hypertrophic cardiomyopathy. Therapy is also depending on theunderlying cardiac abnormality. To date, no ‘gold standard’ is available to diagnosediastolic heart failure. The probability of diastolic dysfunction as the cause of heartfailure in individual patients can be increased by direct and indirect measures of leftventricular function. In clinical situations this requires a thorough knowledge of thephysiology and pathophysiology of diastolic left ventricular function, as well asknowledge of advantages and limitations of the used technique (chapter 2).

Radionuclide angiography of left ventricular function is a well known technique forstudying left ventricular function and is commonly used to estimate left ventricularejection fraction. With modifications this technique is also used to measure leftventricular diastolic function. In chapter 3 the normal values and reproducibility of leftventricular filling parameters by radionuclide angiography are described. Since age,heart rate and left ventricular ejection fraction may influence left ventricular filling rate, wedetermined normal values in relation to these variables. For determining reproducibilitywe performed radionuclide angiography before and after exercise after a period of rest(measurement 2) in 20 patients with normal findings at coronary angiography and leftventriculography. We also compared our findings with values of diastolic fillingparameters measured with radionuclide angiography in normal individuals by otherauthors. Normal values for peak filling rate (mean and standard deviation of the 20patients) were 2.2 ± 0.6 end-diastolic volume / second (measurement 2: 2.4 ± 0.7 end-diastolic volume / second , correlation coefficient r = 0.82), for time to peak filling rate

H

100 Chapter 9

198 ± 22 msec (measurement 2: 203 ± 24 msec, r = 0.45), and for atrial contribution todiastolic filling 31 ± 11% (measurement 2: 31 ± 10%, r = 0.72). These findings arecomparable with values of diastolic left ventricular filling parameters reported by otherauthors using the same technique. Both peak filling rate and time to peak filling ratewere correlated to age (respectively r = -0.68 and r = 0.48, P<0.05). Peak filling ratewas also correlated with heart rate and ejection fraction (respectively r = 0.51 and r =0.50). Time to peak filling rate was however not correlated with heart rate and ejectionfraction. Atrial contribution to diastolic filling was correlated with heart rate (r = 0.79, P <0.01). It is concluded that parameters of diastolic left ventricular filling can reliably andreproducibly be assessed in individual patients using radionuclide angiography. Thistechnique may be used to serially follow diastolic function. Interindividual comparison ofdiastolic filling must be performed with caution, and be placed in perspective of thedependency on age, heart rate and ejection fraction.

In chapter 4 two techniques used for the quantification of atrial contribution todiastolic filling during radionuclide angiography are compared. The moment of onset ofatrial contraction is usually determined by the shape of the time activity curve. Ameasure outside the curve has not been applied to verify the exact duration of the atrialcontribution phase. In 34 patients with a variety of diseases, the atrial contribution wasquantified using a flow-volume loop which was constructed from the radionuclide timeactivity curve of the left ventricle and its first derivative. The duration of the atrialcontribution phase derived from the flow-volume loop correlated well with the PQ-intervalon the electrocardiogram. Also the relative filling volumes within these time periods wereclosely correlated. In a subset of patients these time intervals were however not exactlythe same. In these patients the duration of the atrial contribution phase was alwayslonger than the PQ-interval on the electrocardiogram, indicating the existence of passivediastasis flow before the onset of atrial contraction. This was more apparent in patientswith low heart rates than in those with high heart rates. Despite the close correlationbetween the PQ-interval on the electrocardiogram and the duration of the atrialcontribution phase with radionuclide angiography, they may represent different entities.The PQ-interval on the electrocardiogram is better for determining the moment of onsetof atrial mechanical activity during radionuclide angiography than the interval betweenthe onset of the atrial contribution phase and the end of diastole in the flow-volume loop.

The diagnostic and therapeutic implications of a circadian rhythm in leftventricular relaxation, observed in 15 patients with congestive heart failure, aredescribed in chapter 5. Since afterload is an important determinant of left ventricularrelaxation, the physiological circadian variation of arterial blood pressure may inducediurnal variability of left ventricular filling parameters in patients with congestive heartfailure. At 10 AM compared to 1 PM a higher systolic blood pressure (P < 0.01) anddiastolic pressure (P < 0.01), a lower peak filling rate (P < 0.01), and a longer time topeak filling rate (P<0.05) were found. Using repeated measures analysis of variance,the changes of peak filling rate were related to the daily changes of systolic and diastolic


blood pressure (each P < 0.05). These data suggest that left ventricular relaxation asmeasured by peak filling rate and time to peak filling rate is inversely correlated with acircadian fall in blood pressure in patients with congestive heart failure. This could havethe important diagnostic implication that timing should be taken into account whenassessing diastolic left ventricular function in patients with congestive heart failure, aswell as the therapeutic implication that treatment regimens should account for anincreased risk of pulmonary oedema in these patients early in the morning.

Patients with chronic congestive heart failure and decreased systolic leftventricular performance often have diastolic dysfunction as well. Decreased diastolicfunction in such patients may lead to worsening of exercise tolerance and prognosis.Increased diastolic function may prevent this. In patients with coronary artery disease,calcium antagonists may be beneficial for diastolic dysfunction. The potentialdeterioration of systolic function on administration of calcium antagonists in patients withcongestive heart failure is reported not to be present in mibefradil. We thereforeinvestigated the effect of this calcium antagonist on left ventricular diastolic function inpatients with congestive heart failure and depressed ejection fraction due to previousmyocardial infarction (chapter 6). In 15 patients with New York Heart Association classII or III for dyspnoea radionuclide angiography was performed to obtain measures ofsystolic and diastolic left ventricular function. Systolic and diastolic blood pressure andheart rate were also obtained. Patients were randomly divided into three groups. Group I(5 patients) received placebo medication; group IIA (6 patients) received mibefradil6.25, 12.5, or 25 mg/day; and group IIB (4 patients) received mibefradil 50 or 100mg/day. Measurements were made before and after the first dose and after one week oftreatment before and after the final dose. Mibefradil clearly decreased heart rate(repeated measures analysis of variance p < 0.05). No statistically significant effects ofmibefradil were noted on blood pressure or left ventricular function. In our studyconditions, mibefradil caused no worsening of systolic function and preserved diastolicfunction in short-term treatment of patients with decreased ejection fraction and heartfailure. In 1998 mibefradil was withdrawn from the market because of seriousinteractions with other drugs.

In the next chapters the effect of atrial fibrillation on left ventricular function isdiscussed. Chapter 7 is focussed on systolic left ventricular function. Since atrialfibrillation is characterised by a randomly irregular ventricular response, resulting incontinuous variation in left ventricular beat-to-beat mechanical behaviour andhaemodynamic variables, we used a non-imaging computerised probe, together with aballoon-tipped flow directed catheter and a non-invasive fingertip blood pressuremeasurement system linked to a personal computer in order to perform beat-to-beat leftventricular volume measurement and invasive and non-invasive haemodynamicmonitoring. To assess independent determinants of beat-to-beat variations in leftventricular performance during atrial fibrillation a statistical model was constructedcontaining haemodynamic variables of 500 consecutive RR intervals in seven patients

102 Chapter 9

with chronic non-valvular atrial fibrillation. Positive independent relations with leftventricular ejection fraction were found for the preceding RR interval, left ventricularcontractility (end-systolic pressure / end-systolic volume), and end-diastolic volume,whereas inverse relations were found for afterload (end-systolic pressure / strokevolume), preceding end-systolic volume and preceding contractility. A relative stronginteraction was found between end-diastolic volume and afterload, indicating thatejection fraction was relatively more enhanced by preload in the presence of lowafterload. The varying left ventricular systolic performance during atrial fibrillation istherefore independently influenced by beat-to-beat variations of cycle length, preload,afterload and contractility. Beat-to-beat variations in preload exhibit their effect onventricular performance mainly in the presence of a low afterload.

Left ventricular diastolic function after direct current electrical cardioversion ofatrial fibrillation is evaluated in chapter 8. Tachycardia associated left ventriculardysfunction in atrial fibrillation is characterised by decreased left ventricular (LV) systolicperformance, but the data in the literature concerning diastolic function are discordant.Serial radionuclide angiography was performed in 23 patients (mean age 65 ± 13years) with chronic (mean duration 198 ± 242 days) lone atrial fibrillation aftercardioversion. Twenty patients with normal coronary angiograms and no myocardial orvalvular disease served as control subjects. Early after cardioversion patients with atrialfibrillation had lower left ventricular ejection fraction (EF) compared to controls (54 ± 10% end-diastolic volume vs. 60 ± 8 % end-diastolic volume, P<0.05), higher peak fillingrate (PFR, 3.91 ± 0.68 filling volume/s vs. 3.29 ± 0.64 filling volume/s, P<0.01), shortertime to PFR (TPFR, 162 ± 46 ms vs. 196 ± 42 ms, P<0.05), and lower additional fillingfraction (AFF, 24 ± 7 % filling volume vs. 32 ± 7 % filling volume, P < 0.05). In 5 patientsremaining in sinus rhythm for 3 months, PFR decreased from 4.03 ± 0.63 filling volume/sto 3.13 ± 0.54 filling volume/s (P<0.01). The increase of EF from 54 ± 11 to 57 ± 1 %,TPFR from 126 ± 46 ms to 187 ± 44 ms and of AFF from 25 ± 3 % filling volume to 30 ±7 % filling volume were not statistically significant. None of the factors recorded earlyafter cardioversion predicted recurrence of atrial fibrillation. These data do not confirmthe existence of diastolic left ventricular dysfunction early after electrical cardioversion inpatients with lone atrial fibrillation, but are consistent with decreased atrial complianceand contractility due to atrial remodelling. Persistent sinus rhythm in patients 3 monthsafter electrical cardioversion for chronic lone atrial fibrillation may be accompanied withaltered diastolic LV function as an important feature of post-tachycardia left ventriculardysfunction. The measurements in patients with persistent sinus rhythm may alternativelybe due to (partial) reversibility of atrial remodelling.

CONCLUSIONSIn summary, the following conclusions can be drawn from the above findings of thepresent thesis:


• Radionuclide angiography is a reliable and reproducible technique to assessdiastolic left ventricular function. It may be used to serially follow diastolic function inindividual patients. Interindividual comparison of diastolic filling can be performedprovided that the results are placed in perspective of the dependency on age, heartrate and ejection fraction.

• Additional filling of the left ventricle is partly caused by atrial contraction and partlycaused by passive filling.

• The time of day on which diastolic left ventricular function is assessed in patients withdecreased systolic function has influence on the result.

• The calcium antagonist mibefradil has no effect on diastolic left ventricular functionafter short-term treatment of patients with heart failure and decreased systolicfunction.

• The beat-to-beat variation of systolic left ventricular function in patients with atrialfibrillation is not only caused by variation of preload, but also by variation ofcontractility, afterload and RR-interval.

• The changed inflow profile in the left ventricle early after cardioversion in patientswith chronic “lone” atrial fibrillation is due to atrial remodelling. Diastolic leftventricular dysfunction and partial reversibility of atrial remodelling are probably bothresponsible for a changed inflow in the left ventricle late after cardioversion.

FUTURE DIRECTIONSFrom the above conclusions some future directions can be formulated:• Reliable limits of the normal range of parameters of diastolic left ventricular function

must be formulated in order to increase clinical applicability.• The contribution of passive filling to left ventricular filling must be quantified. Does a

relation exist with left ventricular compliance, left ventricular relaxation or clinicalparameters?

• Collect details of the variation of diastolic left ventricular filling parameters inpatients with heart failure and decreased systolic left ventricular function. Is thisvariation present in other patient groups? Is there a relation with the moment of drugintake?

• What effect have calcium antagonists on de diastolic left ventricular function inpatients with diseases other than heart failure, e.g. after an episode with atrialfibrillation?

• The research of mechanisms behind the beat-to-beat variation of left ventricularcontractility. Can the variability be reduced in order to improve left ventricularfunction?

• What is the role of reduced atrial compliance in chronic “lone” atrial fibrillation onatrial fibrillation recurrence after cardioversion?

104 Chapter 9

Many unknown factors are still present In the knowledge of left ventricular function. Aselection is discussed in the present thesis. Some findings will need to be confirmed infuture work. One of the important questions of this moment is how to trace patients withdiastolic dysfunction as a cause of heart failure. Whether radionuclide angiography canplay a role here depends on the clinical availability of this technique, as well as thesuccess of future work to clearly indicate diagnostic borders of existing parameters.Also new reliable diagnostic parameters of diastolic heart failure need be found. Sincepatients with diastolic left ventricular dysfunction often present with exercise intolerance,these measurements must preferably also be possible with exercise. Finally, relationswith clinical parameters and clinical end-points need to be strengthened.

Hoofdstuk 9Samenvatting en conclusies

artfalen door diastolische linker hartkamerdisfunctie is eengezondheidsprobleem van toenemende omvang. Bij 30 tot 40% van depatiënten met hartfalen wordt primair een diastolische linker

hartkamerdisfunctie gevonden. Aangenomen wordt dat patiënten met hartfalen eendiastolische disfunctie van de linker hartkamer hebben wanneer de systolische functietussen de uiterste normaalwaarden ligt en geen klepvitia aanwezig zijn. Systolischelinker hartkamerdisfunctie sluit diastolische linker hartkamerdisfunctie echter niet uit.Hartfalen door diastolische linker hartkamerdisfunctie gaat gepaard met de typischesymptomen van hartfalen door systolische linker hartkamerdisfunctie zoalsinspanningsintolerantie en longoedeem. De diagnose heeft betrekking op deonderliggende hartaandoening, bijvoorbeeld diastolische linker hartkamerdisfunctie bijhypertrofie ten gevolge van hypertensie of hypertrofische cardiomyopathie. Ook detherapie is afhankelijk van de onderliggende hartaandoening. Tot op heden is nog geen‘gouden standaard’ beschikbaar om diastolisch hartfalen te diagnosticeren. Bijindividuele patiënten kan door directe en indirecte meting van de linker hartkamerfunctiede waarschijnlijkheid dat diastolische disfunctie bijdraagt aan het ontstaan van hartfalenworden geschat. In klinische situaties is hiervoor een gedegen kennis van de fysiologieen pathofysiologie van diastolische linker kamer functie noodzakelijk, maar ook kennisvan de voor- en nadelen van de gebruikte diagnostische techniek (hoofdstuk 2).

Radionuclide angiografie is een bekende techniek voor de studie van de linkerhartkamerfunctie en wordt met name gebruikt voor de meting van de ejectiefractie vande linker hartkamer. Met bepaalde aanpassingen is deze techniek ook bruikbaar voorstudie van de diastolische linker hartkamerfunctie. In hoofdstuk 3 worden denormaalwaarden en de reproduceerbaarheid van de vullingparameters van de linkerhartkamer beschreven zoals gemeten met radionuclide angiografie. Omdat leeftijd,hartfrequentie en ejectiefractie van de linker hartkamer het vullingtempo van de linkerhartkamer beïnvloeden bepaalden wij normaalwaarden in relatie tot deze variabelen.Om de reproduceerbaarheid te bepalen werd radionuclide angiografie van de linkerhartkamer verricht voor inspanning en na een rustperiode na inspanning (meting 2) bij20 patiënten met normale bevindingen bij coronair angiografie en ventriculografie. Debevindingen werden vergeleken met de waarden van diastolische linker

H

106 Hoofdstuk 9

hartkamervulling van gezonde personen door andere onderzoekers gemeten metradionuclide angiografie. Gemiddelde en standaard deviatie van de 20 patiënten voorhet hoogste vullingtempo van de linker hartkamer (PFR = peak filling rate) was 2,2 ± 0,6einddiastolisch volume / seconde (meting 2: 2,4 ± 0,7 einddiastolisch volume / seconde,r = 0,82), voor tijdsduur tot het hoogste vullingtempo (TPFR = time to peak filling rate)198 ± 22 ms (meting 2: 203 ± 24 ms, r = 0,45), en voor de boezembijdrage aan dediastolische vulling 31 ± 11% (meting 2: 31 ± 10%, r = 0,72). Deze bevindingen zijnvergelijkbaar met waarden van diastolische vullingparameters van de linker hartkamerdie door andere onderzoekers werden gerapporteerd bij gebruik van dezelfde techniek.Zowel PFR als TPFR waren gecorreleerd met de leeftijd (respectievelijk r = -0,68 en r =0,48, P<0,05). PFR was ook gecorreleerd met hartfrequentie en ejectiefractie(respectievelijk r = 0,51 and r = 0,50). TPFR was echter niet gecorreleerd methartfrequentie en ejectiefractie, terwijl de boezembijdrage aan de diastolische linkerhartkamervulling wel was gecorreleerd met de hartfrequentie (r = 0,79, P < 0,01). In dithoofdstuk wordt geconcludeerd dat diastolische vullingparameters van de linkerhartkamer in individuele patiënten betrouwbaar en reproduceerbaar verkregen kunnenworden met radionuclide angiografie. Deze techniek kan daarom gebruikt worden omdiastolische functieparameters serieel te volgen. Interindividuele vergelijking vandiastolische vulling moet echter met grote voorzichtigheid worden uitgevoerd, engeplaatst worden tegen de achtergrond van de afhankelijkheid van de parameters vanleeftijd, hartfrequentie en ejectiefractie.

In hoofdstuk 4 worden twee radionuclide angiografische technieken vergelekendie de boezembijdrage aan de linker hartkamervulling kwantificeren. Het momentwaarop de boezemcontractie begint wordt meestal bepaald door de vorm van de curvewaarin de gemeten stralingsactiviteit van het met 99mtechnetium gelabelde bloed in delinker hartkamer uitgezet is tegen de tijdsduur van een hartcyclus. Een referentiewaardeom de duur van de fase waarin de linker boezem de linker hartkamer vult te verifiërenwordt daarbij niet toegepast. Van 34 patiënten met verschillende onderliggendehartziekten werd de boezembijdrage gekwantificeerd met gebruikmaking van eenbloedstroom-volumelus welke geconstrueerd was uit de activiteitcurve van de linkerhartkamer in de tijd en de eerste afgeleide curve hiervan. De duur van deboezembijdragefase gemeten in de bloedstroom-volumelus correleerde goed met hetPQ interval op het elektrocardiogram. Ook het relatieve vullingvolume van deze fasecorreleerde goed met het relatieve vullingvolume na het begin van de P-top op hetelektrocardiogram. In een subgroep van patiënten waren de duur van deboezembijdragefase en het PQ interval op het elektrocardiogram niet exact hetzelfde.Bij deze patiënten was de duur van de boezembijdragefase altijd langer dan het PQinterval op het elektrocardiogram. Deze waarneming werd vaker gedaan bij patiëntenmet een lage hartfrequentie. Hiermee wordt bevestigd dat het PQ interval op hetelektrocardiogram en de duur van de boezembijdragefase met radionuclide angiografieverschillende entiteiten vertegenwoordigen ondanks hun onderling goede correlatie.


Mogelijk is een passieve diastasisstroom in de linker hartkamer voor het begin van deboezemcontractie mede verantwoordelijk voor het ontstaan van de boezembijdragefaseop het radionuclide angiogram. Het PQ interval op het elektrocardiogram is daarombeter voor het bepalen van het begin van boezemactiviteit tijdens radionuclideangiografie dan de duur van de boezembijdrage bij de diastolische vulling van de linkerhartkamer in de bloedstroom-volumelus.

De diagnostische en therapeutische implicaties van een circadiaans ritme vande linker hartkamerrelaxatie welke werd geobserveerd bij 15 patiënten met hartfalen,worden beschreven in hoofdstuk 5. Omdat "afterload" aanzienlijke invloed heeft op delinker hartkamerrelaxatie, zou de fysiologische circadiaanse variatie van de arteriëlebloeddruk bij patiënten met hartfalen variabiliteit van de vullingparameters van de linkerhartkamer kunnen veroorzaken. In vergelijking met 1 uur ‘s middags werd om 10 uur ‘sochtends een hogere systolische (P < 0,01) en diastolische bloeddruk (P < 0,01), eenlager hoogste vullingtempo van de linker hartkamer (PFR = peak filling rate, P < 0,05)en een langere tijd tot het hoogste vullingtempo (TPFR = time to peak filling rate, P <0,05) gemeten. Met gebruikmaking van variantieanalyse bij herhaalde metingen, warende veranderingen van PFR gerelateerd aan de dagelijkse veranderingen van desystolische en diastolische bloeddruk (iedere P < 0,05). Deze gegevens suggereren datde linker hartkamerrelaxatie gemeten met de PFR en de TPFR omgekeerdgecorreleerd is met een circadiaans patroon in het dagelijks bloeddrukverloop bijpatiënten met hartfalen. Dit kan de belangrijke diagnostische implicatie hebben datrekening moet worden gehouden met het tijdstip van meten bij het bepalen vandiastolische linker hartkamerfunctie van patiënten met hartfalen, en de therapeutischeconsequentie dat behandelstrategieën rekening zouden moeten houden met eentoegenomen risico op longoedeem bij deze patiënten vroeg in de ochtend.

Patiënten met chronisch hartfalen en afgenomen systolische linkerhartkamerfunctie hebben vaak daarnaast ook diastolische disfunctie. Dit kan bij dezepatiënten tot verslechtering van de inspanningstolerantie en de prognose leiden.Verbetering van de diastolische linker hartkamerfunctie zou dit kunnen voorkomen. Bijpatiënten met coronaire hartziekte kunnen calcium antagonisten een gunstig effecthebben op de diastolische linker hartkamerfunctie. De potentiële verslechtering van desystolische linker hartkamerfunctie na toediening van calcium antagonisten bij patiëntenmet hartfalen wordt niet bij mibefradil gerapporteerd. Daarom werd het effect van dezecalciuantagonist op de diastolische linker hartkamerfunctie onderzocht bij patiënten methartfalen en verminderde ejectiefractie na een eerder myocardinfarct (hoofdstuk 6).Radionuclide angiografie werd verricht om maten voor de systolische en diastolischelinker hartkamerfunctie te meten bij 15 patiënten met kortademigheidklasse II of IIIvolgens de indeling van de New York Heart Association. Ook werden de systolische ende diastolische bloeddruk en de hartfrequentie gemeten. De patiënten werdenwillekeurig verdeeld in 3 groepen. Groep I (5 patiënten) kreeg placebo medicatie; groepIIA (6 patiënten) kreeg mibefradil 6,25, 12,5 of 25 mg / dag; en groep IIB (4 patiënten)

108 Hoofdstuk 9

kreeg mibefradil 50 of 100 mg / dag. De metingen werden voor en na de eerste gift enna 1 week behandeling voor en na de laatste gift verricht. Mibefradil verminderdeduidelijk de hartfrequentie (variantieanalyse van herhaalde metingen P < 0,05). Geenstatistisch significant effect van mibefradil werd opgemerkt op de bloeddruk of op delinker hartkamerfunctie. Onder deze studievoorwaarden van korte termijnbehandelingvan patiënten met verminderde ejectiefractie en hartfalen veroorzaakte mibefradil geenverslechtering van de systolische hartfunctie en werd de diastolische functie behouden.In verband met ernstige interacties met andere geneesmiddelen werd mibefradil doorde fabrikant medio 1998 van de markt gehaald.

In de volgende hoofdstukken wordt het effect van boezemfibrilleren op de linkerhartkamerfunctie behandeld. In hoofdstuk 7 ligt het accent op de systolische linkerhartkamerfunctie. Omdat boezemfibrilleren gekarakteriseerd wordt door volstrektonregelmatig kamervolgen, resulterend in een continu van slag tot slag variëren van hetmechanisch gedrag van de linker hartkamer en hemodynamische variabelen, werdeneen zogenaamde nucleaire stethoscoop, een Swan-Ganz katheter en een niet-invasiefbloeddrukmeetsysteem gekoppeld aan een personal computer, ten einde van slag totslag volumemetingen, en invasieve en niet-invasieve hemodynamische metingen tekunnen verrichten. Om onafhankelijke variabelen te verkrijgen die het van slag tot slagvariëren van de linker hartkamerfunctie tijdens boezemfibrilleren bepalen, werd eenstatistisch model geconstrueerd van de hemodynamische variabelen van 500opeenvolgende RR intervallen bij zeven patiënten met chronisch niet-valvulairboezemfibrilleren. Positief onafhankelijke relaties met de ejectiefractie van de linkerhartkamer werden gevonden voor het voorafgaande RR interval, de contractiliteit van delinker hartkamer (einddiastolische druk / eindsystolisch volume), en het einddiastolischevolume, terwijl omgekeerde relaties gevonden werden voor de “afterload”(eindsystolische druk / slagvolume), het voorafgaande eindsystolische volume en devoorafgaande contractiliteit. Een relatief sterke interactie werd gevonden tusseneinddiastolisch volume en “afterload”, hetgeen een aanwijzing vormt dat deejectiefractie relatief meer toeneemt door de “preload” wanneer de “afterload” laag is.De variatie van de systolische linker hartkamerfunctie tijdens boezemfibrilleren wordtdaarom onafhankelijk beïnvloed door het van slag tot slag variëren van het RR interval,de “preload”, de “afterload” en de contractiliteit. Het van slag tot slag variëren van de“preload” heeft met name effect op de linker hartkamerfunctie wanneer de “afterload”laag is.

De diastolische linker hartkamerfunctie na elektrische cardioversie van patiëntenmet chronisch boezemfibrilleren wordt besproken in hoofdstuk 8. De met tachycardiegeassocieerde linker hartkamerdisfunctie wordt gekarakteriseerd door eenverminderde systolische functie, maar de literatuurgegevens over diastolische disfunctiezijn tegenstrijdig. Seriële radionuclide angiografie werd verricht bij 23 patiënten(gemiddelde leeftijd 65 ± 13 jaar) met chronisch (gemiddelde duur 198 ± 242 dagen)boezemfibrilleren zonder bekende oorzaak (“lone”) na cardioversie. Twintig patiënten


met normale bevindingen tijdens coronair angiografie en zonder myocardziekte ofklepvitium werden als controlepatiënten gebruikt. Vroeg na cardioversie haddenpatiënten met boezemfibrilleren een lagere ejectiefractie (EF) van de linker hartkamerdan de controlepatiënten (54 ± 10 % vs. 60 ± 8 % einddiastolisch volume, P < 0,05),een hoger hoogste vullingtempo (peak filling rate = PFR, 3,91 ± 0,68 vs. 3,29 ± 0,64vullingvolume / s, P < 0,01), een kortere tijd tot het hoogste vullingtempo (time to PFR =TPFR, 162 ± 46 vs. 196 ± 42 ms, P < 0,05), en een lager additionele vullingfractie (AFF= additional filling fraction, 24 ± 7 % vs. 32 ± 7 % vullingvolume, P < 0,05). Bij 5patiënten die 3 maanden sinus ritme behielden daalde PFR van 4,03 ± 0,63vullingvolume / s naar 3,13 ± 0,54 vullingvolume / s (P < 0,01). De toename van EF van54 ± 11 tot 57 ± 11 %, TPFR van 126 ± 46 ms tot 187 ± 44 ms en van AFF van 25 ± 3% tot 30 ± 7 % vullingvolume was niet statistisch significant. Geen van de factorengemeten vroeg na cardioversie voorspelde de terugkeer van boezemfibrilleren. Dezegegevens bevestigen het bestaan van diastolische linker hartkamerdisfunctie vroeg naelektrische cardioversie bij patiënten met “lone” boezemfibrilleren niet, maar komen welovereen met verminderde boezemcompliantie en contractiliteit ten gevolge vanboezemremodellering. De metingen bij patiënten met persisterend sinus ritme 3maanden na elektrische cardioversie voor chronisch “lone” boezemfibrilleren passen bijveranderde diastolische linker hartkamerfunctie als belangrijke uitingsvorm van linkerhartkamerdisfunctie na tachycardie. Deze metingen bij patiënten met persisterend sinusritme zouden ook een uiting kunnen zijn van partiele reversibiliteit vanboezemremodellering.

CONCLUSIESSamenvattend kan men uit bovenstaande bevindingen in deze thesis het volgendeconcluderen:• Radionuclide angiografie is een betrouwbare en reproduceerbare techniek om

diastolische linker hartkamerfunctie in kaart te brengen. Hierdoor kan dediastolische linker hartkamerfunctie serieel gevolgd worden in individuele patiënten.Tevens kan deze techniek gebruikt worden voor het interindividueel bestuderen vande diastolische linker hartkamerfunctie, mits rekening wordt gehouden met deeffecten die hartfrequentie, leeftijd en ejectiefractie hierop hebben.

• De additionele vullingfase van de linker hartkamer wordt deels doorboezemcontractie maar ook deels door passieve vulling veroorzaakt.

• Het tijdstip van de dag waarop bij patiënten met verminderde systolische linkerhartkamerfunctie de diastolische functie gemeten wordt is van invloed op deverkregen meetwaarde.

• De calciumantagonist mibefradil heeft bij patiënten met hartfalen en verminderdelinker hartkamerfunctie na korte behandeling geen effect op de diastolischehartkamerfunctie.

110 Hoofdstuk 9

• Het van slag tot slag variëren van de systolische linker hartkamerfunctie bij patiëntenmet boezemfibrilleren wordt niet alleen door variatie van de “preload” veroorzaakt,maar ook door variatie van de contractiliteit, “afterload” en het RR-interval.

• Vroeg na cardioversie bij patiënten met chronisch “lone” boezemfibrilleren wordt hetveranderde instroomprofiel van de linker hartkamer bepaald doorboezemremodellering. Diastolische linker hartkamerdisfunctie en gedeeltelijkeomkeerbaarheid van boezemremodellering zijn beide waarschijnlijk verantwoordelijkvoor het veranderde instroomprofiel van de linker hartkamer laat na cardioversie.

TOEKOMSTIGE ONDERZOEKSRICHTINGUitgaande van bovenstaande conclusies kan een aantal toekomstigeonderzoeksrichtingen en onderzoeksvragen vragen concreet worden geformuleerd:• Het formuleren van betrouwbare grenswaarden van het normale bereik van

parameters van de diastolische linker hartkamerfunctie met radionuclide angiografieten einde de klinische toepasbaarheid te vergroten.

• Het kwantificeren van de passieve vullingfase. Is er een relatie te leggen met dehartkamercompliantie, hartkamerrelaxatie of klinische parameters?

• Details verzamelen betreffende de variatie van diastolische vullingparameters bijpatiënten met hartfalen en systolisch verminderde linker hartkamerfunctie. Is dezevariatie ook aanwezig bij andere patiëntengroepen? Is er een relatie met de hettijdstip van inname van medicatie?

• Wat is het effect van calciumantagonisten op de diastolische hartfunctie bij patiëntenmet andere ziektebeelden dan hartfalen, bijvoorbeeld na een episode metboezemfibrilleren?

• Het onderzoeken van de mechanismen achter het van slag tot slag variëren van decontractiliteit van de linker hartkamer. Kan deze variabiliteit worden verminderdteneinde de linker hartkamerfunctie te verbeteren?

• Welke rol speelt de veranderde boezemcompliantie bij chronisch “lone”boezemfibrilleren met betrekking tot de recidiefkans op boezemfibrilleren nacardioversie?

In de kennis van de linker hartkamerfunctie zijn nog vele onbekende factoren. Een aantalfactoren wordt in deze thesis besproken. Sommige bevindingen zullen in toekomstigonderzoek moeten worden bevestigd. Centraal zal de vraag moeten komen te staanhoe men patiënten met diastolische disfunctie als oorzaak van hartfalen kan opsporen.Of radionuclide angiografie hierbij een rol kan spelen hangt niet alleen af van debeschikbaarheid van deze techniek in de kliniek, ook de diagnostische grenzen vanbestaande parameters zullen duidelijk aangegeven moeten kunnen worden. Daarnaastzal men op zoek moeten gaan naar nieuwe betrouwbare diagnostische parameters vandiastolisch hartfalen. Omdat patiënten met diastolische linker hartkamerdisfunctie zich


vaak met inspanningsintolerantie presenteren, moeten metingen ook bij inspanningmogelijk zijn. Een relatie zal moeten worden gelegd met klinische parameters eneindpunten.

DANKWOORDBij de totstandkoming van dit proefschrift zijn veel mensen betrokken geweest. Al dezemensen wil ik hartelijk bedanken. Een aantal mensen wil ik daarnaast in het bijzonderbedanken.

Mijn promotor Prof. H.J.G.M. Crijns. Beste Harry, het is een groot genoegen omde cardioloog die mij ooit begeleidde in het onderzoek als co-assistent, nu als promotoren opleider te mogen hebben. Je inhoudelijke kritiek is vaak pittig, maar is altijdopbouwend. Je niet aflatende energie om het onderzoek te structureren weerhoudt je erniet van ook voor het menselijke aspect aandacht te hebben.

Mijn tweede promotor Prof. E.E. van der Wall. Beste Ernst, je vermogen in kortetijd een probleem geheel te overzien, dit te plaatsen in een context, en het vermogenvervolgens een oplossing te bedenken, hebben ertoe bijgedragen dat ik niet langdurigben verzand in details. De motiverende werking die uitgaat van je werklustige enpositieve houding is op momenten dat het tempo uit het onderzoek wegviel van cruciaalbelang geweest.

Mijn allereerste kennismaking met diastolische functie en disfunctie werdverzorgd door Fred van de Berg. Fred, jij hebt het diastolisch functieonderzoek metisotopen “in huis” gehaald in het Martini Ziekenhuis. Je hebt de hiervoor noodzakelijkkennis en techniek persoonlijk uit Michigan opgehaald en je door Frank Starling en JackJuni laten begeleiden. Oorspronkelijk wilde je zelf op dit onderzoek promoveren, maar jehebt hiervan afgezien. Ik hoop dat ik een waardig opvolger van je onderzoek bengebleken en dat het huidige resultaat je bevalt.

Van groot belang is ook de inzet van mijn referent Dr. M.G. Niemeyer voor hettotstandkomen van dit proefschrift geweest. Beste Menco, je bent in staat om dezwakke punten van een zelfstandig opgestart onderzoek op te sporen en aan te vullenmet kennis van buitenaf. Je vermogen positieve aspecten te zien en te benutten maaktehet mogelijk dit onderzoek voort te zetten.

Mijn co-promotor Dr. P.K. Blanksma. Paul, mede dankzij jouw fysiologischekennis en kennis van nucleaire technieken was het mogelijk de gevonden resultaten inzinvolle medische begrippen te vertalen. Ook heb jij een doorslaggevende rol gespeeldbij de uitvoering en analyse van studies met de “nuclear stethoscope”.

Verder wil ik de leden van de promotiecommissie Prof. Dr. T. van der Werf, Prof.Dr. W.H. van Gilst en Prof. Dr. F.L. Meijler bedanken voor het doorlezen en vancommentaar voorzien van het manuscript. In het bijzonder wil ik Professor Meijlerbedanken voor zijn persoonlijke aanpak hiervan.

Van wezenlijk belang is natuurlijk de statistische ondersteuning van Hans Knolgeweest. Hans, mede dankzij jouw doorzettingsvermogen en vindingrijkheid is hetgelukt de gegevens over mibefradil te analyseren. Dankzij jouw analysetechniekontdekten we een diastolisch dagritme bij patiënten met een slechte systolischehartkamerfunctie. Mede hierdoor is het gelukt een tweede conclusie uit ditpatiëntenmateriaal te formuleren, en dit resulteerde uiteindelijk in een tweede artikel.Ook voor statistische ondersteuning wil ik Pieter Jan de Kam van harte bedanken.

Pieter Jan, dankzij jouw is de ingewikkelde analyse van de slag-op-slag variatie van delinker hartkamerfunctie tot zinvolle proporties teruggebracht.

Op deze plek wil ik Prof. Dr. K.I. Lie bedanken voor zijn positieve houding tenaanzien van het opzetten van een promotieonderzoek buiten het academische centrum.Ik wil graag Dr. P.J.L.M. Bernink vanaf hier bedanken voor zijn inzet voor mijnproefschrift. Beste Peter, dankzij jouw was het mogelijk onderzoeksgegevens uit WCN(Working Group on Cardiovascular Research The Netherlands) studies die in hetMartini Ziekenhuis werden uitgevoerd te gebruiken voor mijn proefschrift. Verder wil ikop deze plek ook Dr. A.J.M. Cleophas bedanken voor zijn enthousiasme waarmee hijbijgedragen heeft aan het schrijven van het proefschrift. Verder wil ik Dr. A.T.M.Gosselink bedanken voor zijn betrokkenheid bij de uitvoering van de studies naar hetslag op slag mechanisme achter de wisselende systolische linker hartkamerfunctie bijboezemfibrilleren.

Hartelijke dank ben ik zeker schuldig aan Oebele Dijkstra van de afdelingnucleaire geneeskunde in het Martini Ziekenhuis. Beste Oebele, jij bent betrokken bijhet opzetten van diastolisch functieonderzoek in het Martini Ziekenhuis. Jij hebt er voorgezorgd dat de nodige technische kennis aanwezig was en bleef. Dit is een belangrijkvoorwaardescheppend feit geweest. Ook wil ik je bedanken voor je betrokkenheid bijde uitvoering van de verschillende studies, en je medewerking bij het beantwoorden vansoms lastige post hoc vragen. Van de afdeling nucleaire geneeskunde wil ik J.J.Schuurman en H. Louwes bedanken voor hun inzet, en ook Nanette Werbata-Jansemaen haar collega’s medisch nucleair werkers in het Martini Ziekenhuis die de studieshebben uitgevoerd.

Mijn paranimfen Anton Tuinenburg en Sieneke Bult dank ik alvast hartelijk voorhun bereidheid mij overeind te houden tijdens de verdediging van mijn proefschrift.

Mijn verdere dank gaat uit naar G.P. Molhoek die mijn eerste opleidingsjaarCardiologie aan het Medisch Spectrum Twente in Enschede verzorgde en naar Dr. M.van Marwijk Kooy die mijn Interne vooropleiding aan de Isala Klinieken in Zwolleverzorgt. Mijn verdere dank gaat uit naar de toenmalige collega’s arts-assistenten aanhet Martini Ziekenhuis, en aan het Medisch Spectrum, en mijn huidige collega’s aan deIsala Klinieken, die soms een paar stapjes harder moesten lopen, zodat ik aan mijnproefschrift kon werken.

Monique, jouw wil ik graag bedanken voor al het geduld dat je had. Gelukkig wistje mij ervan te doordringen dat naast het werk ook andere zaken van belang zijn.

UITNODIGINGvoor het bijwonenvan de openbare

verdedigingvan het proefschrift

Left ventriculardiastolic function

and cardiac diseasea radionuclide angiography study

door H.J. Muntinga

woensdag

10 mei 200014.15 uur

AcademiegebouwBroerstraat 5

Groningen

Wij verzoeken u om14.00 uur

aanwezig te zijn i.v.m.een korte inleiding

door de promovendusvoorafgaande aan de

verdediging

Receptie na afloopin het Academiegebouw

de paranimfenG.T. Bult-Muntinga

(0591-676775)A.E. Tuinenburg(050-5270264)

Stellingen behorende bij het proefschrift

Left ventricular diastolic function and cardiac disease

A radionuclide angiography studyHarm Jans MuntingaGroningen, 10 mei 2000

1. Een normale sytolische linker hartkamerfunctie is onvoldoende om de diagnosehartfalen te verwerpen.

2. Diastolische linker hartkamerdisfunctie behoort geen bevinding per exclusionumte zijn.

3. De grote variatie van de opgegeven prevalentie van diastolisch hartfalen is eenuiting van gebrek aan uniformiteit in de diagnostische criteria.

4. Hartfalen veroorzaakt door systolische en diastolische disfunctie hebben duidelijkverschillende pathofysiologie, therapie en prognose. (W.H. Gaasch en M.M. LeWinter1994)

5. Interpretatie van resultaten van radionuclide functieonderzoek betreffende dediastolische linker hartkamerfunctie van individuele patiënten is een zaak volvalkuilen.

6. Laatdiastolische linker hartkamervulling ontstaat niet alleen doorboezemcontractie.

7. De wisselende linker hartkamerfunctie tijdens boezemfibrilleren wordt door eensamenspel van factoren bepaald.

8. Door tachycardie geïnduceerde veranderingen in het hart bij boezemfibrillerenhebben voor een deel betrekking op de diastolische hartfunctie.

9. Voor het hart geldt dat ontspanning inspannender is dan inspanning ontspannendis. Voor de mens als geheel geldt het omgekeerde.

10. Wetenschappelijke vorderingen in de geneeskunde vergroten de kans opvorderingen in de geneeskunst.

11. De medicus dient patiënten ook als klanten te zien.

12. Bodemdaling door grondstofwinning in Groningen is geen eigentijds probleem.

13. Wel golden raand nait eert is pronkjewail nait weerd.

Left ventricular diastolic function and cardiac diasease · diastolic left ventricular dysfunction...

Documents

Transcript of Left ventricular diastolic function and cardiac diasease · diastolic left ventricular dysfunction...