Articles in PresS. J Appl Physiol (September 29, 2005). doi: /japplphysiol

Transcription

1 Articles in PresS. J Appl Physiol (September 29, 2005). doi: /japplphysiol Assessment of Left Ventricular Diastolic Function by Early Diastolic Mitral Annulus Peak Acceleration Rate: Experimental Studies and Clinical Application Qinyun Ruan, MD, Liyun Rao, PhD, Katherine J Middleton, RCT, Dirar S Khoury, PhD, Sherif F. Nagueh, MD From the First Affiliated Hospital of Fujian Medical University, Fuzhou, China, from the Methodist Debakey Heart Center and from the Section of Cardiology, Department of Medicine, Baylor College of Medicine, Houston, Texas Correspondence and reprint requests Sherif F. Nagueh, MD Methodist DeBakey Heart Center 6550 Fannin Street, Suite-677 Houston, Texas Phone: snagueh@tmh.tmc.edu Copyright 2005 by the American Physiological Society.

2 2 Abstract We sought to examine the hemodynamic determinants and clinical application of the peak acceleration rate of early (Ea) diastolic velocity of the mitral annulus by tissue Doppler. Simultaneous LA and left ventricular (LV) catheterization and Doppler echocardiography (DE) were performed in 10 dogs. Preload was altered using volume infusion and caval occlusion, whereas myocardial lusitropic state was altered with dobutamine and esmolol. The clinical application was examined in 190 consecutive patients (control: 55, impaired relaxation: 41, pseudonormal: 46 and restrictive LV filling: 48). In addition, in 60 consecutive patients, we examined the relation between it and mean wedge pressure (PCWP) with simultaneous DE and right heart catheterization. In canine studies, a significant positive relation was present between peak Acc rate of Ea and transmitral pressure gradient only in the stages with normal or enhanced LV relaxation, but with no relation in the stages where was 50 ms. Its hemodynamic determinants were:, LV minimal pressure and transmitral pressure gradient. In clinical studies, peak Acc rate of Ea was significantly lower in patients with impaired LV relaxation irrespective of filling pressures (p<0.001), and with similar accuracy to peak Ea velocity (AUC for septal and lateral peak Acc rates: both 0.78) in identifying these patients. No significant relation was observed between peak Acc rate and PCWP. Peak acceleration rate of Ea appears to be a useful index of LV relaxation but not of filling pressures and can be applied to identify patients with impaired LV relaxation irrespective of their filling pressures.

3 3 Left ventricular (LV) diastolic function plays an important role in determining LV filling pressures and pulmonary congestion symptoms (14) in patients with and without depressed LV ejection fraction (EF). Novel non-invasive indices utilizing tissue Doppler imaging (TDI) have been recently applied for the evaluation of LV diastolic function and the prediction of filling pressures (3, 6, 9-10). Two of these indices include combined measurements of mitral inflow and TD signals of annular early diastolic excursion (3, 6, 9-10). The above methods have the disadvantage of multiple measurements from different cardiac cycles. A single measurement, if accurate, could be more applicable clinically. In this regard, a single measurement of the peak acceleration rate (P Acc) of early diastolic annular velocity (Ea) was shown in an animal model to provide a good assessment of left atrial pressure (4). Its clinical utility however, was not examined. We therefore sought to examine the hemodynamic determinants of the peak acceleration rate of Ea in experimental studies and to assess its clinical application in consecutive patients, including those referred for right heart catheterization. Methods Animal Preparation After the Baylor College of Medicine Animal Research Committee approved the study, 10 healthy adult mongrel dogs (weight: 19 to 28 kg) were anesthetized with sodium pentobarbital (30mg/kg), intubated, and mechanically ventilated with an external respirator. After a midline sternotomy, the heart was exposed and after calibration, a high fidelity 5-F pressure catheter (Millar Instruments, Houston, TX) was introduced into the left atrium (LA) through its appendage. Likewise, to measure LV pressures, a calibrated 5-F 12-electrode pressure catheter (Millar) was advanced into the LV by crossing the aortic valve (positioned along LV long axis). Surface electrocardiogram (lead II), atrial, and ventricular pressure signals were simultaneously acquired on

4 4 computer based data acquisition system and LA and LV pressures were digitized with recordings acquired at end expiration. The inferior vena cava (IVC) was dissected and a ring was placed around it to allow for gradual occlusion of the vein. Hemodynamic Measurements LV minimal pressure and LVEDP, maximal instantaneous diastolic transmitral pressure and LA mean pressure were noted. Also ascertained were the first derivatives of LV pressure in diastole (-dp/dt), and the time constant of LV relaxation (tau: ) (13). Echocardiography The animals were imaged from the epicardium using an ultrasound system equipped with TD imaging capability. In the apical 4-chamber view, PW Doppler was applied to record mitral inflow at the valve tips (intraobserver variability for measuring mitral peak E velocity: 3±1%). TD program was applied in PW Doppler to record mitral annular velocities at the septal and lateral areas (6). The peak velocity of Ea (intraobserver variability 5±2%) at above sites of the annulus was measured under the different loading conditions. The time interval between the peak of R wave and onset of mitral E velocity and between peak of R wave and onset of early diastolic velocity (Ea) at the 4 areas of the mitral annulus were measured. Subsequently the difference between these time intervals (T Ea E ) was calculated (10) for each of the 4 areas and an average value was derived (intraobserver variability 4±2%, interobserver variability 6±2%). The peak acceleration rate of Ea velocity was derived by the computer after the Ea velocity was traced. The computer algorithm used divided the time from the onset of Ea to its peak into 20 equal intervals and measured the change in velocity per unit of time in each interval and reported the highest value (interobserver variability 8±4%). In an attempt to simplify the measurement of acceleration rate of Ea, we also calculated

5 5 the mean acceleration rate as the Ea velocity divided by acceleration time. Interobserver variability was 6±3%. Experimental Protocols Initially LA pressure was increased with intravenous infusion of isotonic saline and decreased with inferior vena caval (IVC) external compression. Both the infusions and compressions were performed in a sequential manner with data acquired at predetermined increments and decrements of mean LA pressure. After achieving a stable hemodynamic state at each LA pressure level, the LA and the LV pressures and Doppler data, were acquired. After a stable hemodynamic state was achieved, to evaluate the influence of LV relaxation on Ea and its peak acceleration rate, dobutamine was administered at a dose of 5 ug/kg/min with Doppler and pressure data acquired. Dobutamine infusion was then terminated, and after the animals returned to their baseline state, esmolol with its negative lusitropic properties was administered (0.5 mg/kg intravenously) with subsequent reacquisition of data. To assess the possible interaction between atrial pressure and ventricular relaxation on the peak acceleration rate of Ea, fluid administration and IVC compression were repeated during dobutamine and esmolol infusion. Human Studies Study Group The institutional review board of Baylor College of Medicine approved the protocol and all patients provided written informed consent. The study sample was comprised of 190 consecutive patients referred for echocardiographic evaluation at our laboratory. Patients were divided according to the mitral inflow pattern, clinical data and echocardiographic examination into 4 groups (control, impaired relaxation or IR, pseudonormal or PN, and restrictive or Res.). Patients in the control group (n=55) had no history or evidence of cardiovascular disease with normal LV dimensions, wall

6 6 thickness, mass, normal PA systolic pressure and no evidence of significant valvular pathology on echocardiography. The IR group consisted of 41 patients with hypertension, coronary artery disease and/or LV hypertrophy, and a mitral inflow pattern with an early to late transmitral flow velocity (E/A) ratio <1.0. The PN group consisted of 46 patients with symptoms of pulmonary congestion and an E/A ratio 1.0. The Res. LV filling group consisted of 48 patients with symptoms of pulmonary congestion and an E/A ratio 2.0. An additional group of 60 consecutive patients who were undergoing right heart catheterization in the cardiac catheterization laboratory (n=25) or the intensive care unit (n=35) was included. All were in sinus rhythm ( beats/min) and with satisfactory Doppler and pressure recordings and no evidence of mitral stenosis, prosthetic mitral valve or severe mitral regurgitation. Patients had simultaneous echocardiographic and hemodynamic measurements. Echocardiographic studies: Patients were imaged in a supine position with an ultrasound system equipped with harmonic imaging, a multifrequency transducer and TD imaging capability. After acquiring parasternal and apical views, pulse-doppler was utilized to record transmitral and pulmonary venous flow in the apical 4-chamber view as previously described (8). TD imaging was applied in the PW mode to record the mitral annular velocities at the septal and lateral areas. Studies were recorded for later playback and analysis. Echocardiographic Analysis The analysis was performed offline without knowledge of hemodynamic data. LV EF and LA maximum volume were performed per the recommendations of the American Society of Echocardiography (12). All Doppler values represent the average of 3 beats. Pulmonary artery (PA) systolic pressure was estimated using the tricuspid regurgitation jet. Mitral inflow was analyzed for peak E, peak A velocities, E/A ratio and deceleration

7 7 (DT) time of E velocity. From the pulmonary vein flow velocities, the systolic filling fraction (SFF) and Ar A duration were computed (5, 11). Ea and Aa velocities at the 2 areas of the mitral annulus were measured (septal, and lateral). Peak and mean acceleration rates of Ea at septal and lateral sides of mitral annulus were also measured. In addition, the time interval between peak of R wave and onset of mitral E velocity as well as the time interval between peak of R wave and onset of Ea at the 4 areas of the mitral annulus were measured (10). Hemodynamic Measurements Hemodynamic data were collected by an investigator unaware of the echocardiographic measurements, at end expiration and represent the average of 5 cycles. PCWP (wedge verified by fluoroscopy, phasic changes in pressure waveforms and oxygen saturation) was determined using balanced transducers (0 level at mid-axillary line). Statistical Analysis Ea and its peak and mean acceleration rates were correlated with hemodynamic parameters using regression analysis. Stepwise regression was then used to determine the hemodynamic parameters that correlated best with the individual Doppler variables. The study was powered to detect a correlation coefficient of at least 0.4 between the transmitral pressure gradient and peak acceleration rate of Ea (power =80%, p =0.05). Repeated measures ANOVA with Bonferroni correction were used to compare Doppler and hemodynamic parameters at the different lusitropic states (baseline, dobutamine and esmolol) and loading conditions. ANOVA with Bonferroni correction was used to compare the control group with each of the other 3 groups (IR, PN and Res.). ROC curves were applied to examine the accuracy of TDI signals in identifying patients with increased LV filling pressures. Linear regression analysis was used to relate mean wedge pressure to Doppler measurements. Significance was set at a p value <0.05.

8 8 Results Hemodynamics and Doppler measurements: Animal studies Table 1 summarizes the hemodynamic and TD data (average of the 2 annular sites shown in the table since similar observations were noted at each separate site) at the different experimental stages. Volume expansion resulted in an increase in LV filling pressures, annular Ea and its peak and mean acceleration rate. IVC occlusion resulted in significantly opposite changes in the above measurements. Dobutamine infusion resulted in shorter along with an increase in annular Ea and its peak and mean acceleration rate. Esmolol infusion resulted in significantly opposite changes in the above measurements. In all of the above interventions, only esmolol infusion resulted in a significant change in T E-Ea, prolonging it. Relation of TD signals to LV Hemodynamics: Animal studies In individual dogs, the correlation coefficient of Ea velocity, peak and mean acceleration rate of Ea with LA mean pressure ranged from 0.4 to 0.75 (p value range: 0.1 to 0.03). As for LV relaxation, peak Ea velocity (r=-0.76) and its mean (r=-0.73) and peak acceleration rates (r=-0.75) exhibited strong relations to (all p <0.001) and -dp/dt (r ranging from 0.65 to 0.79, all p <0.001) (Figure 1). Peak and mean acceleration rates of Ea were also significantly related to LV minimal pressure (r = and r=-0.65, respectively, both p<0.01). The relation of peak acceleration rate of Ea to the transmitral pressure gradient was evaluated in all the experimental stages where was 50 ms and in those where was <50 ms. There was no significant relation between peak acceleration rate of Ea and the pressure gradient in the stages where 50 ms, whereas a significant positive relation emerged in the stages where was <50 ms (Figure 2).

9 9 On multiple regression analysis (model R 2 =0.79, p <0.01), the predictors of peak acceleration rate of Ea velocity were:, LV minimal pressure and transmitral pressure gradient. Human studies Clinical characteristics and Doppler echocardiographic variables for the 190 patients, divided into 4 groups are listed in table 2. No significant differences were observed between the 4 groups in age, heart rate or blood pressure. The control, pseudonormal (PN) and restrictive (Res.) groups displayed a higher E velocity, E/A ratio, and a shorter deceleration time (DT) compared with the impaired relaxation (IR) group. Mitral inflow however, did not differentiate between normal and PN groups. However, as previously reported, LA volume, PA systolic pressure and pulmonary venous SFF, and Ar-A duration were useful in that regard (see table 2). As expected, septal and lateral Ea velocities were significantly lower in patients with PN and restrictive LV filling in comparison with controls, whereas T E-Ea was significantly longer in the patient groups (table 2). Similar to peak Ea velocity, peak and mean acceleration rate of Ea were significantly lower in patients with PN and Res. LV filling when compared to controls (ANOVA, p<0.001; Bonferroni t-test p<0.05 for control group versus each of the other 3 groups). Figure 3 illustrates examples of TDI signals in 4 representative cases. Table 3 summarizes the accuracy of peak Ea velocity and its acceleration rate in identifying patients with impaired LV relaxation despite elevated filling pressures (ROC curves shown in figure 4). Relation to mean PCWP The mean age of this group was 60±15 years. The mean arterial pressure was 81±15 mmhg, whereas the mean pulmonary artery pressure was 33±10 mmhg. Mitral inflow pattern was that of IR in 20 patients, PN filling in 25 patients and restrictive filling

10 10 in 15 patients. Mean PCWP was 20±10 (range: 5-45) mmhg. A non significant trend for an inverse relation was noted between mean PCWP and septal (and lateral) peak Acc rate (peak acceleration rate: r=-0.23, p=0.08, figure 5). However significant correlations were noted with mitral E/A ratio (r=0.5, p<0.05), E/Ea ratio (r=0.69, p<0.01), isovolumic relaxation time (r=-0.55, p<0.01), and SFF (r=-0.53, p<0.05). Discussion The animal experiments show that the hemodynamic determinants of peak acceleration rate of Ea are transmitral pressure gradient, LV minimal pressure and LV relaxation. The relation between peak acceleration rate of Ea and transmitral pressure gradient is a direct linear relation only in the presence of normal or enhanced LV relaxation, but no direct correlation was noted in the presence of impaired LV relaxation. The clinical studies confirm and extend these animal observations. The peak acceleration rate of Ea, at either side of the mitral annulus, in patients with impaired LV relaxation (IR, PN and Res. groups) was significantly lower than in the age matched control group. Overall its accuracy in identifying patients with impaired relaxation and elevated filling pressures was similar to peak Ea velocity. Furthermore, in the subgroup with simultaneous invasive measurements, no relation was noted between mean PCWP and peak acceleration rate of Ea. Animal studies We used the PW signal at each side of the mitral annulus to calculate the peak Acc. rate of Ea. PW -unlike color Doppler- has the major advantage of superior temporal resolution which suits well the need to calculate the peak acceleration rate given the very short time (frequently <50 ms) when acceleration of Ea is occurring. An additional advantage to our method lies in the higher sampling rate afforded by using 20 intervals

11 11 for deriving the peak Acc rate, making it much less likely to miss the highest acceleration rate when limiting the analysis to longer time intervals. In the previous study (4), the peak acceleration rate of the Ea velocity, particularly at the septal side of the mitral annulus, increased significantly with blood infusion but was not altered by dobutamine or metoprolol administration. Unlike these findings, we noted that changes in the myocardial lusitropic state with either dobutamine or esmolol were effective in producing significant changes in peak acceleration rate of Ea. The magnitude of change in Ea peak acceleration rate was comparable to the change in peak Ea velocity induced by these drugs. Furthermore, the peak acceleration rate of Ea had a significant inverse correlation with the invasive measurements of LV relaxation that was almost identical to those of Ea peak velocity. The relation between the peak acceleration rate of Ea and filling pressures is a complicated one, and overall very similar to that of the peak Ea velocity itself, as previously reported in animal (3, 6) and human studies (1, 2). The relation between LA pressure and peak acceleration rate of Ea in the presence of cardiac dysfunction is very important for the clinical application of Ea Acc rate in the clinical setting, since the assessment of LV filling pressures is frequently needed in patients with cardiac disease. The previous animal work was limited in that regard due to the occurrence of only small changes in LA pressure with beta-blockers. However, the peak acceleration rate of Ea can be used to predict LV filling pressures in normal subjects, given the direct relation between this Doppler measurement and preload in normal states. On the other hand, our animal study provides the pathophysiologic reasons for not applying the peak acceleration rate of Ea to the estimation of LV filling pressures, in the presence of impaired LV relaxation (figure 2). Similar to the acceleration rate of Ea, T E-Ea duration was altered by dobutamine and esmolol but unlike the acceleration rate, IVC occlusion and esmolol infusion had no significant effect on this time interval. Overall

12 12 these results are similar with the observations of Hasegawa et al (3) who noted that Ea was progressively delayed after LA to LV pressure crossover and was significantly related to tau in an animal model of pacing induced heart failure. Human Studies In the human studies, the mean and peak acceleration rate of Ea at either side of the mitral annulus were significantly lower in patients with impaired LV relaxation, irrespective of LV filling pressures. It is interesting that the relation between mean wedge pressure and peak acceleration rate of Ea, shown in figure 5, in patients with cardiac disease was very similar to that observed in the canine experiments between transmitral pressure gradient and peak acceleration rate of Ea when LV relaxation was impaired as shown in figure 2 (interrupted line and open circles). Our study also shows that the accuracy of peak acceleration rate of Ea at either side of the mitral annulus for identifying patients with impaired LV relaxation despite elevated filling pressures (PN and restrictive LV filling groups), is similar to that of the peak Ea velocity. There was no incremental information gained over peak Ea velocity by measuring its acceleration rate. Given the above findings and the complexity of its measurement, peak Ea velocity rather than its acceleration rate is more suitable for the daily application in the laboratory.

13 13 References 1-Firstenberg MS, Levine BD, Garcia MJ, Greenberg NL, Cardon L, Morehead AJ, Zuckerman J, Thomas JD. Relationship of echocardiographic indices to pulmonary capillary wedge pressures in healthy volunteers. J Am Coll Cardiol 36: , Graham RJ, Gelman JS, Donelan L, Mottram PM, Peverill RE. Effect of preload reduction by haemodialysis on new indices of diastolic function. Clin Sci (Lond). 105: , Hasegawa H, Little WC, Ohno M, Brucks S, Morimoto A, Cheng HJ, Cheng CP. Diastolic mitral annular velocity during the development of heart failure. J Am Coll Cardiol 41: , Hashimoto I, Bhat AH, Li X, Jones M, Davies CH, Swanson JC, Schindera ST, Sahn DJ. Tissue Doppler-derived myocardial acceleration for evaluation of left ventricular diastolic function. J Am Coll Cardiol 44: , Kuecherer HF, Muhiudeen IA, Kusumoto FM, Lee E, Moulinier LE, Cahalan MK, Schiller NB. Estimation of mean left atrial pressure from transesophageal pulsed Doppler echocardiography of pulmonary venous flow. Circulation 82: , Nagueh SF, Middleton KJ, Kopelen HA, Zoghbi WA, Quinones MA. Doppler tissue imaging :A noninvasive technique for evaluation of left ventricular relaxation and estimation of filling pressures. J Am Coll Cardiol 30: , Nagueh SF, Sun H, Kopelen HA, Middleton KJ, Khoury DS. Hemodynamic determinants of the mitral annulus diastolic velocities by tissue Doppler. J Am Coll Cardiol 37: , Nishimura RA, Tajik AJ. Evaluation of diastolic filling of left ventricle in health and disease: Doppler echocardiography is the clinician's Rosetta Stone. J Am Coll Cardiol

14 14 30:8-18, Ommen SR, Nishimura RA, Appleton CP, Miller FA, Oh JK, Redfield MM, Tajik AJ. Clinical utility of Doppler echocardiography and tissue Doppler imaging in the estimation of left ventricular filling pressures: A comparative simultaneous Doppler-catheterization study. Circulation 102: , Rivas-Gotz C, Khoury DS, Manolios M, Rao L, Kopelen HA, Nagueh SF. Time interval between onset of mitral inflow and onset of early diastolic velocity by tissue Doppler: a novel index of left ventricular relaxation: experimental studies and clinical application. J Am Coll Cardiol 42: , Rossvoll O, Hatle LK. Pulmonary venous flow velocities recorded by transthoracic Doppler ultrasound: relation to left ventricular diastolic pressures. J Am Coll Cardiol 21: , Schiller NB, Shah PM, Crawford MH, DeMaria A, Devereux R, Feigenbaum H, Gutgesell H, Reichek N, Sahn D, Schnittger I, et al. Recommendations for quantitation of the left ventricle by two-dimensional echocardiography: American Society of Echocardiography Committee on Standards, Subcommittee on quantitation of twodimensional echocardiograms. J Am Soc Echocardiog 2: , Weiss JL, Frederiksen JW, Weisfeldt ML. Hemodynamic determinants of the time-course of fall in canine left ventricular pressure. J Clin Invest 58: , Zile MR, Brutsaert DL. New concepts in diastolic dysfunction and diastolic heart failure: Part I: diagnosis, prognosis, and measurements of diastolic function. Circulation 105: , 2002.

15 15 Figure Legends Figure 1 Relation of versus septal peak Ea velocity (left), and its peak acceleration rate (right). Figure 2 Relation of peak acceleration rate of Ea versus transmitral pressure gradient. Solid line and circles correspond to the stages where was < 50ms. The interrupted line and open circles correspond to the stages where was 50 ms. In the presence of normal or enhanced relaxation, a direct significant relation was present (r=0.71, p<0.01), whereas with impaired relaxation, no relation was observed between peak acceleration rate of Ea and transmitral pressure gradient (p>0.3). Figure 3 Examples of mitral inflow and TDI at septal annulus from patients in each of the 4 groups. Note the reduced Ea velocity in patients with PN and Res. LV filling, despite increased LV filling pressures. Peak acceleration rate of Ea in the normal subject was 330 cm/s 2, whereas it was reduced to 140 cm/s 2 in the patient with IR pattern, to 138 cm/s 2 in the patient with PN pattern and to 160 cm/s 2 in the patient with Res. filling. Figure 4 ROC curves showing the accuracy of septal peak acceleration rate of Ea (left), and lateral peak acceleration rate of Ea (right) in identifying patients with impaired LV relaxation despite increased LV filling pressures (PN and Res. filling groups). Figure 5 Plot of mean PCWP versus septal Ea peak acceleration rate.

16 16 Table 1. Hemodynamics and TD Velocities During Volume Loading, IVC Occlusion, Dobutamine and Esmolol Infusions Hemodynamic Parameter Baseline Volume Loading Data: Mean±SD; p <0.05 vs volume loading, IVC occlusion, dobutamine and esmolol stages; p <0.05 vs IVC occlusion, dobutamine and esmolol stages; p <0.05 vs IVC occlusion and esmolol stages; p<0.05 vs dobutamine and esmolol stages, p< 0.05 vs baseline, volume loading and IVC occlusion stages, p <0.05 vs esmolol stage; p <0.05 vs baseline, volume infusion, IVC occlusion and esmolol, p <0.05 vs baseline, volume loading, IVC occlusion and dobutamine stages. IVC Occlusion Dobutamine Esmolol LV Stroke Volume (ml) 20±5 28±4 14±4 23±4.5 16±5.4 Heart rate (beats/min) 115±6 115±8 118±12 136±3 103±8 LVEDP (mmhg) 5±3 11±5 2.1± ± ±3 LA mean pressure (mmhg) 7±4 11±6.4 3±2 3.5±2.2 15±4 Tau (ms) 42±10 44±9 39±8 26±7 87±8 LV -dp/dt (mmhg/s) 1840± ± ± ± ±367 Ea (cm/s) 5.2±1 5.9± ±1 8.8±0.8 3±0.7 Ea mean acceleration rate (cm/s 2 ) 74±15 90±19 63±20 100±21 53±20 Ea peak acceleration rate (cm/s 2 ) 105±25 129±21 91±19 143±33 69±18 T E-Ea (ms) 1±5 0±5 1±3 0±3 23±4

17 17 Table 2. Clinical and Doppler findings in the 4 patient groups Control (n=55) IR (n=41) PN (n=46) Res. (n=48) Age (years) 61±15 62±15 61±15 67±12 Heart rate (/min) 73±12 75±13 75±13 70±13 Mean arterial pressure (mmhg) 90±14 97±19 94±16 87±19 LV EF (%) 65±3* 53±17 55±17 50±15 Left atrial volume (ml) 43±11* 73±28 81±26 89±22 PA systolic pressure (mmhg) 30±3 35±13 44± ±10 Mitral E/A ratio 1.1± ± ±0.29 3±0.9 Deceleration time (ms) 220±86 252±84 188±69 169±43 Pulmonary veins (SFF) 0.6± ± ± ±0.1 Ar A duration (ms) 0±8* 18±10 45±11 58±10 Septal Ea (cm/s) 9.3±3.2* 5.2± ± ±1.5 Septal mean acceleration rate (cm/s 2 ) 131±53* 77±41 84±43 74±32 Septal peak acceleration rate (cm/s 2 ) 259±99* 159±88 178±87 149±75 Lateral Ea (cm/s) 11.3±4.7* 7.5± ± ±3.2 Lateral mean acceleration rate (cm/s 2 ) 161±68* 98±38 134±49 100±48 Lateral peak acceleration rate (cm/s 2 ) 311±116* 185±79 210± ±112 T E-Ea (ms) 3±5* 29±5 33±3 40±3 *: p <0.05 versus IR, PN and Res. groups, : p <0.05 vs control, IR and PN groups, : p <0.05 vs control and PN groups, : p <0.05 vs PN and Res groups, : p<0.05 vs Res. group, : p <0.05 vs PN

18 18 Table 3 Accuracy of peak Ea velocity and acceleration rate of Ea in identifying patients with impaired LV relaxation despite elevated filling pressures Sensitivity (%) Specificity (%) Area under ROC curve Septal Ea (<5.6 cm/s) ( ) * Septal peak acceleration rate (<195.5 cm/s 2 ) ( ) * Lateral Ea (<8.6 cm/s) ( ) * Lateral peak acceleration rate (<252 cm/s 2 ) ( ) * *: p<0.0001, numbers between parenthesis refer to 5-95% confidence intervals.

19 19 Peak Ea velocity (cm/s) r = p< Peak Acceleration Rate of Ea (cm/s 2 ) r = p< (ms) 1 Figure 1

20 20 Peak Acceleration Rate of Ea (cm/s 2 ) Maximum Transmitral Pressure Gradient (mmhg) 2 Figure 2

21 21 NL NL NL IR PN Res. E Ea Aa 3 Figure 3

22 22 ROC Curve for Septal Ea Pk Acc Rate ROC Curve for Lateral Ea Peak Acc Rate Figure 4

23 23 Peak Acc. Rate of Septal Ea (cm/s 2 ) Mean PCWP (mmhg) r = p = 0.08 Y = X 5 Figure 5