Significance of Left Atrial Pressure and Left Ventricular Relaxation as Determinants of Left Ventricular Early Diastolic Filling Flow in Man

Size: px

Start display at page:

Download "Significance of Left Atrial Pressure and Left Ventricular Relaxation as Determinants of Left Ventricular Early Diastolic Filling Flow in Man"

Morgan Riley
5 years ago
Views:

1 Significance of Left Atrial Pressure and Left Ventricular Relaxation as Determinants of Left Ventricular Early Diastolic Filling Flow in Man Toshiyuki TAKAHASHI, M.D., Masahiko IIZUKA, M.D., Takashi SERIZAWA, M.D., Tetsuo OHYA, M.D., Hiroshi SATO, M.D., Osami KOHMOTO, M.D., Takatoshi MOCHIZUKI, M.D., Tsuguya SAKAMOTO, M.D., and Tsuneaki SUGIMOTO, M.D. SUMMARY We analyzed the relationships among parameters of left ventricular (LV) early diastolic filling flow (EDF) obtained with pulsed Doppler echocardiography, mean pulmonary wedge pressure (PCWP) and the time constant of LV pressure fall calculated by either Weiss' (Tw) or Thompson's (Tb) method. PCWP correlated with the peak velocity (R) (r=0.537, p<0.05), acceleration (Ac) (r=0.545, p<0.05) and deceleration (Dc) (r=0.606, p<0.01) of LVEDF. In contrast, Tb correlated only with the time to the peak of LVEDF (TPF) (r=0.487, p< 0.05), and Tw did not correlate with the Doppler-derived indices significantly. After correcting for the effect of PCWP, significant partial correlations between R and Tw (r=-0.535, p<0.05), and between Ac and both Tw (r=-0.606, p<0.01) and Tb (r=-0.569, p<0.05) were found. Dc did not correlate with Tw or Tb. These results suggest that the level of left atrial pressure may mask the relationship between parameters of LVEDF and LV relaxation, and that the relations among these variables vary with individual indices of LVEDF. Additional Indexing Words: Doppler echocardiography Multivariate analysis ARAMETERS of left ventricular early diastolic filling have been reported to be closely related to the rate of left ventricular relaxation in chronic stable patients.1)-3) On the other hand, many clinical4)-6) and experimental7)-9) studies have suggested that changes in left atrial pressure should also be taken into consideration when left atrial pressure is altered acutely From the Second Department of Internal Medicine, Faculty of Medicine, University of Tokyo, Tokyo, Japan. Address for correspondence: M. Iizuka, M.D., Second Department of Internal Medicine, Faculty of Medicine, University of Tokyo, Hongo, Tokyo 113, Japan. Received for publication September 22, Accepted February 13,

2 320 TAKAHASHI, ET AL. Jpn. Heart J. May 1990 or subacutely. Experimental studies7)-9) have shown that parameters of left ventricular early diastolic filling are determined by both changes in left atrial pressure and the rate of left ventricular relaxation, and that changes in left atrial pressure might mask the relationships between parameters of left ventricular early diastolic filling and relaxation. However, these interactions among left atrial pressure, the rate of left ventricular relaxation and indices of left ventricular early diastolic filling remain uncertain in the clinical settings of chronic patients. Accordingly, in this study we analyzed the relationships among parameters of left ventricular early diastolic filling flow obtained with pulsed Doppler echocardiography, pulmonary capillary wedge pressure and the time constant of left ventricular isovolumic pressure fall in chronic stable patients with cardiac diseases. METHODS Subjects: We analyzed 19 patients who had undergone diagnostic cardiac catheterization and angiography (Table I). Informed consent was obtained from all the subjects and the protocol was approved by the Ethical Committee of the Hospital of the University of Tokyo. Patients with severe heart failure (New York Heart Association, grade 3 or 4) or with significant valvular diseases were excluded from this study. All patients showed normal sinus rhythm at the time of study. Cardiac catheterization: Left and right heart catheterization was performed by the femoral approach. A 7 French catheter-tip micromanometer (PC370, Millar Inc., Houston, Texas) was inserted into the left ventricle and a 7 French balloontipped thermodilution catheter (Gould Inc., Cleveland, Ohio) was advanced into the peripheral pulmonary artery. The balloon-tipped catheter was connected to a Statham P23 D transducer (Gould Inc.), which was calibrated against a mercury reference. Zero level was determined as 5cm below the anterior axillar line and frequently checked. The micromanometer was also calibrated against a mercury reference. Both the left ventricular and pulmonary capillary wedge pressures were recorded simultaneously on a thermal array recorder at a paper speed of 100mm/sec. Signals were also recorded on a FM magnetic tape recorder (Sony Magnescale Inc., Tokyo). Pressure signals on the magnetic tapes were digitized at 5 msec intervals, and left ventricular peak systolic and end-diastolic pressures and the time

3 Vol.31 No.3 DETERMINANTS OF LV EARLY DIASTOLIC FILLING FLOW 321 Fig. 1. Measurements of Doppler-derived parameters of left ventricular early diastolic filling flow obtained with pulsed Doppler echocardiography. The peak velocity (R), acceleration (Ac) and deceleration (Dc) of early diastolic filling flow, and the time to peak early diastolic filling flow (TPF) were determined as shown in the figure. See text for details. MV=mitral valve; SV=sample volume. constant of left ventricular isovolumic pressure fall were calculated with a personal computer system (Packet IIe, Anritsu Ltd., Tokyo). We calculated the time constant by both Weiss'10) and Thompson's11) methods. Fitting was done from the time of left ventricular peak negative dp/dt to the time when the left ventricular pressure reached a level 5mmHg above the succeeding left ventricular end-diastolic pressure. Correlation coefficients of fit were more than 0.99 by both methods in all subjects. The average values of at least 5 successive beats were used for the following analyses. As an index of left atrial pressure, we measured mean pulmonary capillary wedge pressure. Doppler echocardiography: Doppler echocardiography was performed within 24 hours before or after the cardiac catheterization. No patients showed significant changes in clinical conditions during this period. A color flow mapping echocardiograph (SSD 880, Aloka Ltd., Tokyo) was used. The ultrasonic frequency was 2.5MHz and the pulse repetition rate was 2KHz. A long-axis view of the left ventricle was imaged by the apical approach, and the sample volume was placed in the center of the mitral annulus during diastole. The detected Doppler signals were analyzed by a spectral analyzer based on the fast Fourier transform, and were recorded on a strip chart re-

322 TAKAHASHI, ET AL. Jpn. Heart J. May 1990 corder (SSZ 95, Aloka Ltd.) at a paper speed of 50mm/sec. The recordings were done while the breath was held at end-expiration.

4 322 TAKAHASHI, ET AL. Jpn. Heart J. May 1990 corder (SSZ 95, Aloka Ltd.) at a paper speed of 50mm/sec. The recordings were done while the breath was held at end-expiration. The following indices were measured as shown in Fig. 1: peak velocity (R), acceleration (Ac) and deceleration (Dc) of left ventricular early diastolic filling flow, the time interval from the aortic component of the second heart sound to the peak of early diastolic filling flow (TPF). The peak velocity was measured at the mid-point of the Doppler spectrum. The acceleration or deceleration was determined as the slope of the straight line drawn from the peak of the early filling flow to the half point of the peak on the upward or downward side of the Doppler signal, respectively.2) The velocity was calculated from the frequency shift by the Doppler equation.2) Average values of at least 5 successive beats were used also in analyzing Doppler-derived indices. Doppler data were analyzed by an observer who was blind to the hemodynamic data. Statistical analysis: First, we calculated simple correlation coefficients among the hemodynamic variables and Doppler-derived indices. Partial correlation coefficients between the indices of left ventricular early filling flow and the time constant of left ventricular isovolumic pressure fall or pulmonary capillary wedge pressure were then calculated. Statistical analyses were performed with a statistical program for a personal computer system (Multivariate analysis pack, Maruzen Ltd., Tokyo). Significance was accepted at the 0.05 level. RESULTS Clinical data and hemodynamic and Doppler-derived parameters of the subjects are summarized in Table I. Simple correlation coefficients among hemodynamic indices (Table II): There were significant correlations between the time constant of left ventricular pressure fall calculated by Weiss' method (Tw) and both left ventricular end-diastolic pressure (r=0.712, p<0.01) and mean pulmonary capillary wedge pressure (r=0.820, p<0.01) in the subjects of the present study. However, no significant correlations were demonstrated between the time constant calculated by Thompson's method (Tb) and left ventricular end-diastolic or pulmonary capillary wedge pressure. Simple correlation coefficients between hemodynamic and Doppler-derived indices (Table III): There were significant correlations between mean pulmonary capillary wedge pressure and the peak velocity (R) (r=0.537, p<0.05), acceleration

5 Vol.31 No.3 DETERMINANTS OF LV EARLY DIASTOLIC FILLING FLOW 323

6 324 TAKAHASHI, ET AL. Jpn. Heart J. May 1990 Table II. Simple Correlation Coefficients among Hemodynamic Indices p<0.01. Abbreviations are shown in Table I. Table III. Simple Correlation Coefficients between Doppler-derived and Hemodynamic Indices p<0.05, p<0.01. Abbreviations are shown in Table I. Table IV. Partial Correlation Coefficients between Doppler-derived Parameters and mpcwp or the Time Constant (a: Tw, b: Tb) p<0.05, p<0.01. Abbreviations are shown in Table I. (Ac) (r=0.545, p<0.05) and deceleration (Dc) (r=0.606, p<0.01) of left ventricular early diastolic filling flow. Similar correlations were found between left ventricular end-diastolic pressure and the Doppler-derived parameters. No significant correlations were demonstrated between Tw and the Doppler-derived indices. Tb correlated significantly only with time to peak

7 Vol.31 No.3 DETERMINANTS OF LV EARLY DIASTOLIC FILLING FLOW 325 early diastolic filling flow (TPF) (r=0.487, p<0.05). Partial correlation coefficients between hemodynamic and Doppler-derived indices (Table IV): Significant partial correlations were found between Tw and R (r= , p<0.05), Ac (r=-0.606, p<0.01) and TPF (r=0.514, p<0.05) when pulmonary capillary wedge pressure was assumed to be constant (Table IV-a). There were also significant partial correlations between Tb and both Ac (r=-0.569, p<0.05) and TPF (r=0.529, p<0.05) under the assumption of constant pulmonary wedge pressure (Table IV-b). DISCUSSION We demonstrated that the level of mean pulmonary capillary wedge pressure, which was used as a substitute for left atrial pressure, influenced the relationship between the time constant of left ventricular pressure fall and Doppler-derived indices of left ventricular early diastolic filling flow in chronic stable patients. We also found that the relations between hemodynamic and Doppler-derived indices varied with each variable concerned. Comparison of the previous and present studies: The previous clinical studies1)-3) failed to demonstrate significant correlations between left ventricular filling pressure and parameters of left ventricular early diastolic filling in chronic patients. Since left ventricular diastolic filling is a multifactorial process, results of studies may vary with the methods used or the subjects selected. Fioretti et al1) used contrast ventriculography and Magorien et al3) used radionuclide ventriculography. These two studies in which they calculated only univariate correlation coefficients may not be appropriate in analyzing a multifactorial process. A study of Tanouchi et al2) had a similar design to ours. They used Doppler echocardiography and calculated partial correlation coefficients. Thus, differences between the results of their study and ours would be attributable mainly to differences in the patient selection. They studied the subjects with a more severe and wider spectrum of relaxation abnormalities. Comparison of effects of hemodynamic indices on various Doppler-derived indices: In this study, partial correlation analysis disclosed that the peak velocity and acceleration of left ventricular early diastolic filling flow were influenced by both mean pulmonary capillary wedge pressure and the time constant of left ventricular pressure fall (Table IV). However, the deceleration of early diastolic filling flow was mainly determined by pulmonary capillary wedge pressure.

8 326 TAKAHASHI, ET AL. Jpn. Heart J. May 1990 From the fluid dynamic point of view, the driving force of left ventricular diastolic filling flow is the atrioventricular pressure difference.7),12) Left atrial pressure represents the upstream pressure, and the rate of relaxation is an important determinant of left ventricular, that is, the downstream pressure, since the process of left ventricular relaxation can be assumed to continue after mitral valve opening.13) Since the influence of the relaxation process may diminish as time passes in diastole, we were able to find significant relations only between the time constant of pressure fall and the indices of the early rapid filling phase, such as the acceleration and peak velocity. Deceleration of early diastolic filling flow has been suggested to be caused by atrioventricular pressure gradient reversal during early diastole.14) This reversal was shown to be related to left atrial pressure, left ventricular chamber stiffness and mitral valve resistance in an experimental and computer simulation study,14) indicating that deceleration may increase with increases in these hemodynamic variables. Mean pulmonary capillary wedge pressure reflects not only left atrial pressure, but left ventricular passive properties, thereby explaining the significant correlation between deceleration of early filling flow and pulmonary wedge pressure. Time to peak filling (TPF) is the only index that showed a significant simple correlation with Tb, and did not correlate with pulmonary capillary wedge pressure. This unique nature of TPF was stressed also by Ishida et al.7) Recently, a computer simulation study15) showed that slow relaxation delayed the peak of early filling, since the peak of atrioventricular pressure difference occurred later. Other investigators have used different indices for assessing the timing of early filling, such as the time from peak negative dp/dt to the end of the rapid filling phase16) or the time from the beginning to the peak of early diastolic filling flow.9) We felt, however, that it was difficult to determine the beginning or end of early diastolic filling flow precisely on the Doppler spectrum. Comparison of the methods for calculation of the time constant of left ventricular pressure fall: There are several different methods for calculating the "monoexponential" time constant of left ventricular pressure decline.10),11),17)-19) However, there has been no study comparing time constants calculated with different methods in the light of the effects on early diastolic filling dynamics. In this study, the simple correlation coefficient between Tw and either R or Ac was positive although not significant, contrary to our expectation. In contrast, Tb showed a negative low correlation with R or Ac as expected. After correcting for the effect of pulmonary capillary wedge pressure, partial

Vol.31 DETERMINANTS OF LV EARLY DIASTOLIC FILLING FLOW 327 No.3 correlation between Tw and either R or Ac were not overly different from those between Tb and these indices.

9 Vol.31 DETERMINANTS OF LV EARLY DIASTOLIC FILLING FLOW 327 No.3 correlation between Tw and either R or Ac were not overly different from those between Tb and these indices. These results suggest that as far as the effects on early diastolic filling dynamics are concerned, Tb is equivalent to Tw after being corrected for the level of left ventricular filling pressure. Limitations of the study: Some technical problems should be considered in this study. First, we used mean pulmonary capillary wedge pressure as a substitute for left atrial pressure. Although ideally we should have measured a phasic left atrial pressure, we thought that a transseptal catheter technique was too invasive for this study. Second, we did not record hemodynamic and Doppler-derived indices simultaneously. To investigate the instantaneous filling dynamics, we should have measured these variables at the same time. However, a considerable number of clinical studies have been conducted to evaluate the chronic status of diastolic properties of patients rather than to determine the instantaneous hemodynamics. For this purpose, our study will provide useful information. Clinical implication: From the results of this study, we recommend that one should take the level of left atrial pressure into consideration when using some indices of left ventricular early diastolic filling as parameters of left ventricular relaxation in the chronic setting. Since correlation coefficients between the indices of filling and the time constant of relaxation were still low even with the assumption of constant filling pressure, one should pay attention to factors other than these two in chronic patients. REFERENCES 1. Fioretti P, Brower RW, Meester GT, Serruys PW: Interaction of left ventricular relaxation and filling during diastole in human subjects. Am J Cardiol 46: 197, Tanouchi J, Kitabatake A, Asao M, Morita T, Masuyama T, Hori M, Inoue M, Abe H: Role of left ventricular relaxation on transmitral flow dynamics during early diastole: a study with pulsed Doppler flowmetry. J Cardiography 13: 301, Magorien DJ, Shaffer P, Bush C, Magorien RD, Kolibash AJ, Unverferth DV, Bashore TM: Hemodynamic correlates for time intervals, ejection rate and filling rate derived from the radionuclide angiographic volume curve. Am J Cardiol 53: 567, Carrol JD, Hess OM, Hirzel HO, Krayenbuehl HP: Dynamics of left ventricular filling at rest and during exercise. Circulation 68: 59, Takahashi T, Iizuka M, Sato H, Serizawa T, Momomura S, Ohya T, Mochizuki T, Kohmoto O, Aoyagi T, Matsui H, Ikenouchi H, Sakamoto T, Sugimoto T: Doppler echocardiographic determined changes in left ventricular diastolic filling velocity during lower body positive and negative pressure method. Am J Cardiol 65: 237, 1990

328 TAKAHASHI, ET AL. Jpn. Heart J. May 1990 6. Choong CY, Herrmann HC, Weyman AE, Fifer MA: Preload dependence of Doppler-derived indexes of left ventricular diastolic function in humans.

10 328 TAKAHASHI, ET AL. Jpn. Heart J. May Choong CY, Herrmann HC, Weyman AE, Fifer MA: Preload dependence of Doppler-derived indexes of left ventricular diastolic function in humans. J Am Coll Cardiol 10: 800, Ishida Y, Meisner JS, Tsujioka K, Gallo JI, Yoran C, Frater RWM, Yellin EL: Left ventricular filling dynamics: influence of left ventricular relaxation and left atrial pressure. Circulation 74: 187, Mochizuki T, Iizuka M, Takahashi T, Sato H, Ohya T, Aoyagi T, Ma Y-X, Serizawa T, Sugimoto T: Analysis of the role of left atrial pressure in the regulation of left ventricular rapid filling in the dog (abstr). Jpn Circ J 51: 933, Choong CY, Abascal VM, Thomas JD, Guerrero JL, McGlew S, Weyman AE: Combined influence of ventricular loading and relaxation on the transmitral flow velocity profile in dogs measured by Doppler echocardiography. Circulation 78: 672, Weiss JL, Frederiksen JW, Weisfeldt ML: Hemodynamic determinants of the time-course of fall in canine left ventricular pressure. J Clin Invest 58: 751, Thompson DS, Waldron CB, Juul SM, Naqvi N, Swanton RH, Coltart DJ, Jenkins BS, Web- Poploe MM: Analysis of left ventricular pressure during isovolumic relaxation in coronary artery disease. Circulation 65: 690, Yellin EL, Sonnenblick EH, Frater RWM: Dynamic determinants of left ventricular filling: an overview. in Cardiac Dynamics, ed by Baan J, Arzenius AC, Yellin EL, Nijhoff, The Hague, p , Pasipoularides A, Mirsky I, Hess OM, Grimm J, Krayenbuehl HP: Myocardial relaxation and passive diastolic properties in man. Circulation 74: 991, Meisner JS, Pajaro OE, Frater RWM, Yellin EL: Determinants of atrioventricular pressure gradient reversal and S3 in early diastole (abstr). Circulation 76 (suppl IV): IV-512, Bayer R, Sideman S: Atrioventricular interactions: a theoretical simulation study. Am J Physiol 252: H653, Bahler RC, Martin P: Effects of loading conditions and inotropic state on rapid filling phase of left ventricle. Am J Physiol 248: H523, Yellin EL, Hori M, Yoran C, Sonnenblick EH, Gabby S, Frater RWM: Left ventricular relaxation in filling and nonfilling intact canine heart. Am J Physiol 250: H620, Raff GL, Glantz SA: Volume loading slows left ventricular isovolumic relaxation rate: evidence of load-dependent relaxation in the intact dog heart. Circ Res 48: 813, Serizawa T, Momomura S, Kohmoto O, Ohya T, Sato H, Takahashi T, Mochizuki T, Iizuka M, Sugimoto T: Mechanisms of abnormal myocardial relaxation induced by ischemia: comparison of low flow ischemia and hypoxia in isolated rabbit heart. Jpn Circ J 51: 90, 1987

Left ventricular diastolic dysfunction is an increasingly

Left ventricular diastolic dysfunction is an increasingly Review J Vet Intern Med 1999;13:5 13 Clinical Assessment of Left Ventricular Relaxation Peter Constable, William Muir III, and David Sisson Ventricular relaxation is altered in a number of cardiac disorders