Abnormally high mean left atrial pressure (MLAP) due

Transcription

1 J Vet Intern Med 2004;18: Echocardiographic Estimation of Mean Left Atrial Pressure in a Canine Model of Acute Mitral Valve Insufficiency Mark A. Oyama, D. David Sisson, Barret J. Bulmer, and Peter D. Constable High mean left atrial pressure (MLAP) due to canine degenerative mitral valve disease is associated with clinically relevant morbidity and mortality. The ability to noninvasively measure MLAP would assist in the diagnosis and treatment of disease. Doppler echocardiography allows measurement of early transmitral blood flow (E) and the velocity of the mitral valve annulus (E a ). The ratio of early mitral inflow velocity to early mitral annular velocity (E : E a ) correlates well with MLAP in human subjects. We sought to determine the ability of E : E a to predict MLAP in dogs with experimentally induced mitral regurgitation. Nine anesthetized purpose-bred dogs underwent placement of a Swan-Ganz catheter into the left atrium and recording of MLAP. Simultaneous transthoracic echocardiographic and hemodynamic studies were performed after acute chordae tendineae rupture and during IV infusion with nitroprusside ( g kg 1 min 1 ) or hydralazine (1 1.5 mg/kg). Mitral regurgitant fraction, measured by single-plane angiography and thermodilution, ranged from 17% to 81%. MLAP increased from mm Hg to mm Hg after creation of mitral valve regurgitation (MR; P.018). Forty sets of echocardiographic measurements were obtained from 7 dogs, and E, as well as E : E a, were linearly related to MLAP. The R 2 value for the linear regression equation containing E : E a as the dependent variable (0.83) was greater than that for E (0.73). The 95% confidence intervals were calculated for predicting MLAP 20 mm Hg from E : E a, and E : E a 9.1 or 6.0 indicated a 95% probability that MLAP was 20 mm Hg or 20 mm Hg, respectively. Echocardiography can be used to predict MLAP in isoflurane-anesthetized dogs with experimentally induced acute mitral valve insufficiency. Key words: Dog; Filling pressure; Mitral regurgitation; Tissue Doppler. Abnormally high mean left atrial pressure (MLAP) due to heart disease is associated with pulmonary edema, respiratory distress, and poor prognosis. 1 3 Reduction of MLAP and resolution of clinical signs is achieved by diuresis, angiotensin converting enzyme (ACE) inhibition, and vasodilator therapy. In dogs with valvular heart disease, measurement of MLAP is of considerable value in assessing the severity of cardiac dysfunction and monitoring response to therapy. Direct measurement of MLAP (or its estimation from pulmonary capillary wedge pressure) involves cardiac catheterization, an invasive procedure requiring specialized equipment and technical ability. Echocardiography is the predominant noninvasive modality used to gather information about cardiac function. Previous studies in humans have correlated various echocardiographic measures with invasive measures of MLAP. 4 8 Commonly, the velocity of early diastolic flow across the mitral valve (E) is used, but this method is hindered by weak correlation due to the complex interactions among velocity, pressure, and diastolic function. Ventricular diastolic function is affected by many factors, including the atrioventricular pressure gradient, active myocardial relaxation, passive diastolic recoil, and the compliance of the From the Department of Veterinary Clinical Medicine, College of Veterinary Medicine, University of Illinois, Urbana, IL. Dr Bulmer is currently affiliated with the Veterinary Teaching Hospital, Kansas State University, Manhattan, KS. Preliminary data were presented at the Annual Scientific Session of the American College of Cardiology, Chicago, IL, March 30 April 2, Reprint requests: Mark A. Oyama, DVM, DACVIM-Cardiology, Department of Veterinary Clinical Medicine, 1008 West Hazelwood Drive, College of Veterinary Medicine, University of Illinois, Urbana, IL 61801; oyama@uiuc.edu. Submitted January 9, 2004; Revised March 4, 2004; Accepted April 6, Copyright 2004 by the American College of Veterinary Internal Medicine /04/ /$3.00/0 myocardial tissue. As the mitral valve opens, MLAP helps establish an initial atrioventricular pressure gradient that causes early ventricular filling, but the velocity of the resultant mitral inflow also is affected by the diastolic properties of the receiving ventricle. In patients with reduced diastolic function, the ventricular influence on mitral flow becomes increasingly important, and mitral inflow velocity becomes a less reliable predictor of MLAP. 9,10 Various echocardiographic techniques, such as analysis of pulmonary venous flow, have been devised to help circumvent this problem. 11,12 The drawback of this approach is that pulmonary vein inspection often is technically difficult in veterinary patients, and the inclusion of additional variables leads to the tedious task of solving complex multivariate equations in a busy clinical setting. Doppler tissue imaging is an echocardiographic technique that quantifies the velocity of cardiac wall motion. The longitudinal velocity of the mitral valve annulus during early diastole (E a ) correlates well with the time constant of ventricular isovolumic pressure decay ( ) and is a relatively load-independent index of early ventricular diastolic function. 9,13 This measure can be used to remove the influence of diastolic function from the mitral inflow measurement (ie, measurements of mitral inflow velocity can be corrected for changes in ventricular diastolic function). When E a is used to form a ratio with the velocity of mitral blood inflow (E : E a ), the result is a strong correlation with MLAP. 14 In prior studies of human patients with hypertrophic 15 and dilated 16 cardiomyopathy, E : E a demonstrated a moderate to good correlation (r 0.76 and 0.60) with left heart filling pressure. 15,16 Despite these findings, E : E a studies have routinely excluded patients with clinically relevant valvular disease, and the correlation between E : E a and MLAP during moderate or severe mitral valve regurgitation (MR) is unknown. The current study was designed to evaluate the ability of Doppler tissue imaging to predict MLAP in dogs with experimentally induced acute MR.

2 668 Oyama et al the lateral mitral valve annulus and included the peak early (E a ) and late annular (A a ) velocity (Fig 1B). The peak rate of increase (centimeters per second) of the left ventricular chamber dimension (dd/dt) was calculated from the right short-axis M-mode study. M-mode echocardiograms were imported into a PC-based digitizing program, f and the endocardial borders of the left ventricular cavity were manually traced using an electronic cursor. Based on the position of the cursor, the computer created a digital coordinate system, and the instantaneous dimension of the left ventricular chamber could be calculated. A thirdorder polynomial equation was fitted to the plot of dimension versus time and its derivative (dd/dt) solved. The maximum dd/dt value was normalized for heart size by dividing by the instantaneous dimension [(dd/dt)/d]. 17 All echocardiographic measurements were performed in triplicate, and the resultant average was used for analysis. Creation of Mitral Regurgitation Fig 1. (A) Representative Doppler mitral inflow study demonstrating the measurement of early (E) and late (A) diastolic mitral inflow velocity and the deceleration time of early diastolic inflow (E DT ). (B) Representative tissue Doppler study demonstrating the measurement of early (E a ) and late (A a ) mitral annulus velocity. Materials and Methods Patient Preparation The study was approved by the University of Illinois, Division of Animal Resources, and all animals were treated in compliance with the NIH Guidelines for the Care and Use of Laboratory Animals. Nine adult purpose-bred dogs were anesthetized with IV sodium pentobarbital (15 mg/kg), intubated, and maintained on inhalant isoflurane anesthesia. The right jugular vein was surgically exposed and 7 Fr Swan- Ganz catheters a were advanced into the proximal pulmonary artery and left atrium (via transvenous atrial septostomy) under fluoroscopic guidance. Correct placement of the left atrial catheter was confirmed by manual injection of 3 5 ml of angiographic contrast medium, b after which the catheter balloon was inflated to preclude accidental withdrawal of the catheter from the atrium. Cardiac output was determined by the thermodilution method. c MLAP measurement and the lead II surface electrocardiogram were recorded using a digital physiologic monitoring system. d Echocardiographic Studies Echocardiograms were performed using a commercial ultrasound system. e Routine M-mode measurements were performed using the right parasternal short-axis views for measurement of left ventricular end-diastolic (LVDd) and end-systolic (LVDs) dimensions and the ratio of the left atrial and aortic diameter (LA/Ao) and for calculation of the left ventricular fractional shortening percentage (FS). Doppler studies were performed from the left apical 4-chamber view. For each dog, gain and filter settings were adjusted to reduce background noise and provide clear blood flow and tissue signals. Studies were recorded with a sweep speed of 100 mm/s and stored on an optical disk for analysis. Mitral inflow measurements (centimeters per second) were acquired at the tips of the mitral valve leaflets and included peak early (E) and late (A) velocity, the ratio of early to late velocity (E : A), and the deceleration time (seconds) of early mitral inflow (E DT ; Fig 1A). E DT was measured as the time from peak E to the zero baseline. Tissue Doppler measurements (centimeters per second) were performed on Echocardiograms and MLAP recordings were first performed during the baseline-anesthetized state. MLAP pressure was calculated as the mean pressure over the duration of the echocardiographic study. After baseline evaluation, the right carotid artery was surgically isolated and 7 Fr myocardial biopsy forceps g were advanced retrograde across the aortic valve and used to tear mitral valve chordae tendineae. 18 Creation of MR was confirmed by echocardiography, and the regurgitant fraction (RF) calculated using single plane angiography and thermodilution. 19 Sequential biopsy procedures were performed to obtain increasing degrees of MR. IV nitroprusside (2.5 5 g kg 1 min 1 )orhydralazine (1 3 mg/kg) was used to alter loading conditions and MLAP. The severity of MR was arbitrarily defined as moderate (RF 65%) or severe (RF 65%). Statistical Analysis Data are reported as mean standard deviation. Statistical significance was defined as P.05. Paired Student s t-tests were used to compare echocardiographic measurements obtained during baseline and after the first creation of mitral regurgitation, and before the administration of vasodilator therapy. Bonferroni-adjusted P-values were used for the paired Student s t-test comparisons within each family (eg, hemodynamic values, conventional echocardiography, Doppler tissue echocardiography) of comparisons. The relationship between E and MLAP and between E : E a and MLAP was examined using multivariable linear regression with dummy variable coding, which accounts for between-subjects variability, thereby increasing the precision with which slope and intercept coefficients for the regression line can be estimated. 20 This analysis of covariance (ANCOVA) approach enforces a uniform slope for each dog and is reasonable whenever the slopes are similar. 21 Because data were available from 7 dogs, 6 dummy variables (D 1 D 6 ) were defined in the following way: C i 1ifdogi 7, 1 if dog 7, and 0 otherwise. The following regression equation was used to analyze the linear relationship between the dependent variable (y) and MLAP: y b o b i D i b MLAP (MLAP), where b o is the intercept value, b i is the coefficient value for ith dog, and b MLAP is the coefficient value for MLAP. Dummy variables were entered into the model first to account for between-dog differences before analyzing the main factor of interest (MLAP). Coefficients associated with D i describe how much the intercept values for each dog vary from the average, but this information was of minimal interest in this study, and only the coefficients for b o and b MLAP are reported. The adequacy of the final regression models was evaluated by examining residual plots, and the normal probability plot of the standardized residuals and the presence of outliers or influential observations were identified by calculating Cook s distance. Interobserver variability (M.A.O. versus D.D.S.) was calculated from the baseline values for each dog as the difference between the measured values divided by the average value. Correlation between

3 Estimation of Left Atrial Pressure 669 Table 1. Hemodynamic and echocardiographic values at baseline and after the first episode of chordae tendineae rupture and creation of mitral regurgitation (n 7). Baseline Hemodynamic values HR (bpm) MLAP (mmhg) RF (%) NA Conventional echocardiography LVDd (cm) LVDs (cm) LA/Ao FS (%) E (cm/s) A (cm/s) E:A E DT (s) dd/dt (cm/s) dd/dt/d (s 1 ) Doppler tissue echocardiography E a (cm/s) E:E a Mitral Regurgitation NA a NA a P-value NS (0.0251) NS NS (0.0263) NS NS HR, heart rate; MLAP, mean left atrial pressure; RF, regurgitant fraction; LVDd, left ventricular end diastolic, LVDs, left ventricular end systolic; LA/Ao, left atrial and aortic diameter; FS, fractional shortening; E, early transmittal blood flow; A, late velocity; E DT, early transmittal blood flow deceleration time; dd/dt, left ventricular chamber dimension; E a, velocity of the mitral valve annulus; NA, not applicable; NS, not significant. Values are meam SD. a Following the first creation of mitral regurgitation, 6 of 7 dogs possessed fused E and A waves, and analysis of A velocity and E/A ratio was not performed. the 2 observers was examined by linear regression analysis. Intraobserver variability was calculated for a single observer (M.A.O.) as the maximal difference between 3 measured baseline values for each dog divided by the average of the observations. Results One dog developed complete atrioventricular block after atrial septostomy, and 1 dog yielded only 2 MLAP measurements; both dogs were excluded from further analysis. A total of 40 separate echocardiographic studies were performed on the 7 remaining dogs during the presence of acute MR. Hemodynamic and Echocardiographic Data Creation of acute MR increased LVDd (Table 1) and tended to increase heart rate (HR) (P.0251) and left atrial and aortic diameter (LA/Ao) (P.0263). MLAP increased from mm Hg at baseline to mm Hg during MR (P.018). Twenty studies were performed during moderate MR, and 20 studies during severe regurgitation. Satisfactory echocardiographic studies were obtained in all 40 instances. In 31 of 40 instances (78%), sinus tachycardia caused complete merging of the E and A waves, and the resulting peak velocity was used as the E value. In 21 of 40 instances (53%), the mitral annular E a Fig 2. Scatterplot of the relationship between E : E a and mean left atrial pressure (MLAP) for 7 dogs under isoflurane anesthesia. Values for each dog were obtained after varying degrees of acute experimentally induced mitral regurgitation and after administration of nitroprusside or hydralazine to change loading conditions. E : E a is the ratio of E (early diastolic mitral inflow velocity) to E a (the early mitral annulus velocity) and is dimensionless. The linear regression line (solid line) and 95% confidence interval for prediction (dashed line) of MLAP from E : E a also are depicted. and A a waves were fused, and the peak velocity was used as the E a value. The deceleration time of early mitral flow did not markedly change after creation of acute MR, but both the peak rate of ventricular dimension change and early mitral valve annular velocity increased from baseline [(dd/dt)/d, to cm/s, P.0003; E a, to cm/s, P.0006). Relationship to MLAP Examination of residual plots and the normal probability plot of the standardized residuals, as well as calculation of Cook s distance, indicated the presence of 2 outliers for E : E a, and these were removed before statistical analysis was completed. Regression analysis therefore used 40 data points from 7 dogs for E, and 38 data points from 7 dogs for E : E a. Both E and E : E a (Fig 2) were linearly related with MLAP, but the R 2 value for the linear regression equation containing E : E a as the dependent variable (0.83) was greater than that for E (0.73). The following regression lines were developed, relating E (meters per second) or E : E a (unitless) to the independent variable MLAP (in millimeters of mercury [mm Hg]), with the standard error of the estimate in parentheses: E (0.050) (0.0020) MLAP; E : E a 5.14 (0.29) (0.011) MLAP. To permit clinical prediction of MLAP (in mm Hg) from the measured value for E : E a, the regression equation was reformulated, producing the following equation: MLAP predicted 6.38 (E : E a ) 28.3, and guidelines for predicting MLAP (and the 95% confidence interval for the predicted value) from an echocardiographic measurement of E : E a were developed (Table 2).

4 670 Oyama et al Table 2. Guidelines for predicting mean left atrial pressure (MLAP) from an echocardiographic measurement of E:E a. E:E a % Confidence Interval for Predicted MLAP (mm Hg) Prediction (mm Hg) Interobserver and Intraobserver Variability Interobserver variability of baseline E : E a was % (r 0.87, P.005) and of baseline E was % (r.98, P.001). Intraobserver variability of baseline E : E a was %, and of E was %. Discussion Previous studies have shown that traditional Doppler mitral inflow velocities can be used to predict filling pressures but are frequently confounded by the effects of altered relaxation. Doppler tissue imaging is a new technique that applies the principles of Doppler measurement to myocardial tissue. The velocity of the mitral valve annulus in early diastole (E a ) represents the relaxation of muscle fibers in the left ventricular wall and is relatively independent of the atrioventricular pressure gradient. By creating a ratio of mitral inflow to annular velocity, the effects of ventricular relaxation can be removed, allowing a more accurate index of MLAP. The results of this study indicated that both E : E a and E were linearly related with MLAP in a model of acute canine MR, but the R 2 value for E : E a (0.83) was greater than that for E (0.73). This finding suggests that echocardiographic determination of E : E a would more accurately predict MLAP than echocardiographic determination of E. To our knowledge, this is the first study to evaluate E : E a in dogs with substantial mitral valvular insufficiency. These results, obtained from a small number of dogs, compare favorably with previous studies of larger numbers of patients with hypertrophic and dilated cardiomyopathy. Nagueh et al 15 reported that use of E : E a increased the correlation between E and filling pressure from r 0.40 to r 0.76 in 35 patients with hypertrophic cardiomyopathy. Similarly, Ommen et al 16 reported an increase in correlation from r 0.46 (E : A versus filling pressure) to r 0.60 (E : E a versus filling pressure) in 100 patients with dilated cardiomyopathy. Previous results suggest that when ventricular diastolic function is altered, E a can be used to index early diastolic mitral inflow velocity and achieve improved correlation with MLAP. 9 In this manner, E a corrects a mitral inflow measure that is being increasingly confounded by altered diastolic function. The effect of acute MR on ventricular relaxation is controversial. Creation of MR in a porcine model did not markedly change ventricular diastolic elastance, 22 whereas Zile et al 10 indicated that relaxation is impaired during acute MR, as was increased over baseline (29 3 versus 40 8 ms, P.05). The same study documented a rightward shift along the diastolic pressuredimension curve, which indicates an increase in the operating stiffness of the ventricle. Similarly, Katayama et al 23 reported a decrease in ventricular diastolic elastance after creation of acute MR. These alterations would impact diastolic performance negatively, but in cases of severe MR these relaxation abnormalities can be overcome both by large increases in MLAP and enhanced passive diastolic recoil. High MLAP during severe MR augments the atrioventricular pressure gradient, but decreased afterload reduces the end-systolic ventricular dimension, increasing the subsequent diastolic elastic recoil. 10 The end result is a net improvement in overall diastolic function despite impaired relaxation. In this study, this improvement is indicated by increased (dd/dt)/d and is consistent with previously reported results involving acute MR. 10,22,24 Although overall diastolic function appears to have been increased, we cannot state with certainty if left ventricular relaxation was enhanced, impaired, or unchanged, because was not measured. The marked increase in E a suggests that the relaxation properties of the ventricle were improved after creation of MR. If so, correction of E occurred in the opposite direction from what has been observed in patients with impaired relaxation abnormalities such as hypertrophic or dilated cardiomyopathy. Compared with the aforementioned studies in which E : E a substantially improved the correlation with filling pressures, the effect in this study was more modest. Inasmuch as the degree of correction gained by E:E a depends on the extent of relaxation abnormalities, the degree of impairment in acute MR appears to be less than that observed in hypertrophic or dilated cardiomyopathy. The value of E : E a also is reliant on the independence of E a from loading conditions and its specificity as a surrogate marker of relaxation. Interestingly, Ohte et al 25 reported that in human patients with moderate to severe MR, the relationship between relaxation and E a is weakened. In patients with high filling pressures (pulmonary capillary wedge pressure [PCWP] 16 mm Hg) and reduced ventricular relaxation ( 46 ms), a paradoxically increased E a was found compared with values in patients with reduced relaxation but normal pressures. Use of E : E a to estimate MLAP during acute MR appears valid over a wide range of conditions. E : E a appeared to be linearly related to MLAP over varying degrees of MR severity, as well as during IV vasodilator therapy (Fig 2). E : E a retained a strong correlation with MLAP despite the presence of sinus tachycardia and fused E and A signals. This preliminary finding extends those of a previous study by Nagueh et al 26 who reported a correlation between E : E a and filling pressures (r 0.86) in 35 human patients with sinus tachycardia but without clinically relevant MR. The 95% confidence intervals for predicting MLAP from the measured E : E a value for dogs with acute experimental MR were wide and approximated 20 mm Hg (Table 2; Fig 2), indicating that only a very large increase in MLAP could be accurately detected by measuring E : E a during echocardiography. The large 95% confidence intervals therefore decrease the clinical utility of measuring E : E a,

5 Estimation of Left Atrial Pressure 671 but E : E a values 9.1 or 6.0 predicted MLAP 20 mm Hg or 20 mm Hg, respectively. Validation of these findings in a clinical setting has not yet been established. Limitations of this study were the small number of dogs examined and the use of an experimental model of acute MR, which may not accurately reflect the clinical condition of chronic valvular regurgitation. Chronic MR causes ventricular hypertrophy, decreased contractility, and alterations of myocardial stiffness and relaxation, all of which may influence the relationship between E : E a and MLAP. 27,28 Nonetheless, evaluation of this technique in dogs with naturally occurring disease appears warranted. In the diseased heart, abnormally high MLAP and loss of ventricular compliance cause E DT to decrease. 16 Our results did not find a marked relationship between E DT and MLAP, which is in contrast to previous studies. Pozzoli et al 29 and Temporelli et al, 30 studying patients with MR and chronic dilated cardiomyopathy, reported correlations between E DT and filling pressures of 0.67 and 0.92, respectively. It is possible that differences in the chronicity of MR, concurrent medical therapy, heart rate, systolic function, or some combination of these factors influenced these results. The majority of these measurements (78%) were performed during sinus tachycardia and fusion of mitral velocity signals. Interestingly, in human patients with sinus tachycardia, Nagueh et al 26 found a much weaker correlation between E DT and pulmonary capillary wedge pressure (r 0.36, P.05) than was found by Pozzoli et al 29 and Temporelli et al. 30 We also note that systolic function in our subjects was preserved, whereas patients reported by Pozzoli et al 29 and Temporelli et al 30 had ejection fractions 35%. Preserved systolic function acts to maintain early diastolic recoil and can contribute to changes in E DT, independent of filling pressures. Ommen et al 16 reported that the correlation between E DT and left ventricular end-diastolic pressure fell from 0.60 to 0.17 when comparing subsets of patients with ejection fractions 50% versus 50%. In a model of acute canine MR, E : E a was of greater explanatory ability than E, and E : E a 9.1 indicated MLAP 20 mm Hg. These results suggest that E : E a can be of use in estimating MLAP in dogs with acute MR, and additional study is warranted to define the clinical utility of this measurement in dogs with naturally occurring disease. Footnotes a Baxter Healthcare Corporation, Irvine, CA b Hypaque-76, Nycomed Inc., Princeton, NJ c E for M V2213A, Honeywell Inc., Morristown, NJ d MP150, Biopac Systems, Santa Barbara, CA e System V, GE Medical Systems, Waukesha, WI f DigitizeIt 1.4, Bormisoft, Heidelburg, Germany g Cordis Corporation, Miami Lakes, FL Acknowledgments The authors would like to thank Robyn Ostapkowicz and Judie Walker for their technical assistance. References 1. The IMPROVE Study Group. Acute and short-term hemodynamic, echocardiographic, and clinical effects of enalapril maleate in dogs with naturally acquired heart failure: Results of the Invasive Multicenter Prospective Veterinary Evaluation of Enalapril study. J Vet Intern Med 1995;9: Vanoverschelde JL, Raphael DA, Robert AR, Cosyns JR. Left ventricular filling in dilated cardiomyopathy: Relation to functional class and hemodynamics. J Am Coll Cardiol 1990;15: Calvert CA, Pickus CW, Jacobs GJ, Brown J. Signalment, survival, and prognostic factors in Doberman pinschers with end-stage cardiomyopathy. J Vet Intern Med 1997;11: Gorcsan J, Snow FR, Paulsen W, Nixon JV. Noninvasive estimation of left atrial pressure in patients with congestive heart failure and mitral regurgitation by Doppler echocardiography. Am Heart J 1991;121: Appleton CP, Galloway JM, Gonzalez MS, et al. Estimation of left ventricular filling pressures using two-dimensional and Doppler echocardiography in adult patients with cardiac disease. Additional value of analyzing left atrial size, left atrial ejection fraction and the difference in duration of pulmonary venous and mitral flow velocity at atrial contraction. J Am Coll Cardiol 1993;22: Mulvagh S, Quinones MA, Kleiman NS, et al. Estimation of left ventricular end-diastolic pressure from Doppler transmitral flow velocity in cardiac patients independent of systolic performance. J Am Coll Cardiol 1992;20: Nagueh SF, Kopelen HA, Zoghbi WA. Feasibility and accuracy of Doppler echocardiographic estimation of pulmonary artery occlusive pressure in the intensive care unit. Am J Cardiol 1995;75: Vanoverschelde JL, Robert AR, Gerbaux A, et al. Noninvasive estimation of pulmonary arterial wedge pressure with Doppler transmitral flow velocity pattern in patients with known heart disease. Am J Cardiol 1995;75: Nagueh SF, Sun H, Kopelen HA, et al. Hemodynamic determinants of the mitral annulus diastolic velocities by tissue Doppler. J Am Coll Cardiol 2001;37: Zile MR, Tomita M, Nakano K, et al. Effects of left ventricular volume overload produced by mitral regurgitation on diastolic function. Am J Physiol 1991;261:H1471 H Pu M, Griffin BP, Vandervoort PM, et al. The value of assessing pulmonary venous flow velocity for predicting severity of mitral regurgitation: A quantitative assessment integrating left ventricular function. J Am Soc Echocardiogr 1999;12: Schwerzmann M, Wustmann K, Zimmerli M, Seiler C. Accurate determination of mitral regurgitation by assessing its influence on the combined diastolic mitral and pulmonary venous flow: just looking twice. Eur J Echocardiogr 2001;2: Sohn DW, Chai IH, Lee DJ, et al. Assessment of mitral annulus velocity by Doppler tissue imaging in the evaluation of left ventricular diastolic function. J Am Coll Cardiol 1997;30: Nagueh SF, Middleton KJ, Kopelen HA, et al. Doppler tissue imaging: A noninvasive technique for evaluation of left ventricular relaxation and estimation of filling pressures. J Am Coll Cardiol 1997; 30: Nagueh SF, Lakkis NM, Middleton KJ, et al. Doppler estimation of left ventricular filling pressures in patients with hypertrophic cardiomyopathy. Circulation 1999;99: Ommen SR, Nishimura RA, Appleton CP, et al. Clinical utility of Doppler echocardiography and tissue Doppler imaging in the estimation of left ventricular filling pressures: A comparative simultaneous Doppler-catheterization study. Circulation 2000;102: Gibson DG, Brown D. Measurement of instantaneous left ventricular dimension and filling rate in man, using echocardiography. Br Heart J 1973;35: Kleaveland JP, Kussmaul WG, Vinciguerra T, et al. Volume

6 672 Oyama et al overload hypertrophy in a closed-chest model of mitral regurgitation. Am J Physiol 1988;254:H1034 H Lord PF, Carmichael JA, Tashjian RJ. Left ventricular volume measurements by single-plane cineangiocardiography: Hemodynamic study of normal canine left ventricle. Am J Vet Res 1970;31: Glantz SA, Slinker BK. Primer of Applied Regression and Analysis of Variance. New York: McGraw-Hill; 1990: Feldman HA. Families of lines: Random effects in linear regression analysis. J Appl Physiol 1988;64: Hennein HA, Jones M, Stone CD, Clark RE. Left ventricular function in experimental mitral regurgitation with intact chordae tendineae. J Thorac Cardiovasc Surg 1993;105: Katayama K, Tajimi T, Guth BD, et al. Early diastolic filling dynamics during experimental mitral regurgitation in the conscious dog. Circulation 1988;78: Jeresaty RM. Left ventricular function in acute non-ischaemic mitral regurgitation. Eur Heart J 1991;12(Suppl B): Ohte N, Narita H, Akita S, et al. Striking effect of left ventricular high filling pressure with mitral regurgitation on mitral annular velocity during early diastole. A study using colour M-mode tissue Doppler imaging. Eur J Echocardiogr 2002;3: Nagueh SF, Mikati I, Kopelen HA, et al. Doppler estimation of left ventricular filling pressure in sinus tachycardia. A new application of tissue doppler imaging. Circulation 1998;98: Carabello BA, Nakano K, Corin W, et al. Left ventricular function in experimental volume overload hypertrophy. Am J Physiol 1989;256:H974 H Tsutsui H, Urabe Y, Mann DL, et al. Effects of chronic mitral regurgitation on diastolic function in isolated cardiocytes. Circ Res 1993;72: Pozzoli M, Capomolla S, Pinna G, et al. Doppler echocardiography reliably predicts pulmonary artery wedge pressure in patients with chronic heart failure with and without mitral regurgitation. J Am Coll Cardiol 1996;27: Temporelli PL, Scapellato F, Corra U, et al. Chronic mitral regurgitation and Doppler estimation of left ventricular filling pressures in patients with heart failure. J Am Soc Echocardiogr 2001;14: