Tissue Doppler echocardiographic evidence of atrial mechanical dysfunction in coronary artery disease
Cheuk-Man Yu,Jeffrey Wing-Hon Fung,Qing Zhang,Leo C.C. Kum,Hong Lin,Gabriel Wai-Kwok Yip,Maggie Wang,John E. Sanderson
DOI: https://doi.org/10.1016/j.ijcard.2004.12.077
IF: 4.039
2006-01-01
International Journal of Cardiology
Abstract:Results The V A in the LA (5.0 ± 2.6 Vs 7.7 ± 2.6 cm/s), IAS (4.8 ± 1.7 Vs 5.7 ± 1.5 cm/s) and RA (6.8 ± 3.1 Vs 9.2 ± 2.9 cm/s) were significantly decreased in patients with CAD when compared with controls (all p < 0.001). Patients with impaired systolic function (ejection fraction ≤ 50%) had significantly lower V A in the LA and IAS than those with ejection fraction > 50% (both p < 0.001); and were lower in those with restrictive filling pattern (RFP) than non-RFP of diastolic dysfunction (both p < 0.05). The V A in all the subgroups was lower than controls. In contrast, transmitral atrial velocity was unable to reveal any abnormality except in the subgroup with a RFP. The LA dimension, area and volume were increased in the disease groups, but were largely unchanged in the RA despite abnormal V A . The physiological inter-atrial delay for the onset and peak atrial contraction between the RA and LA were unaffected by CAD. Conclusions The atrial contractile function in both atria was impaired in the presence of CAD, especially in the LA. This was detected even in patients with preserved systolic function or mild diastolic dysfunction such as non-RFP. Direct assessment of atrial velocity by TDI may better reflect atrial mechanical function than transmitral atrial velocity. Keywords Atrium Coronary artery disease Tissue Doppler echocardiography Diastole 1 Introduction Assessment of atrial function is an integral part in the examination of cardiac function. In the presence of cardiac diseases, atrial function could be primarily or secondarily affected. Hemodynamic assessment of atrial function using high fidelity catheter is limited by its invasive nature and may not be able to reflect the true function of the atrium, such as the contractility of the chamber. Doppler echocardiography is one of the commonest modality to assess atrial function and estimate hemodynamic difference between the atrium and the ventricle, which indirectly reflects atrial function [1] . However, atrial mechanical function is not assessed directly by this method. Recently, tissue Doppler imaging (TDI) has been employed as a sensitive and reproducible tool for the assessment of cardiac function [2] . In experimental, animal and clinical settings, TDI-derived indices have been proven valuable to assess both regional and global functions of the heart [2–6] . Therefore, TDI may be a useful non-invasive tool for the assessment of atrial function, and investigate the relation between right (RA) and left (LA) atrial mechanical function. Coronary artery disease (CAD) is the commonest cardiovascular disease resulting in myocardial dysfunction and heart failure [7] . However, the possible presence of atrial dysfunction in CAD has only been explored by one invasive study previously [8] . The aim of our study was to investigate whether atrial mechanical function was impaired in patients with CAD; and compared the degree of inter-atrial electromechanical delay with normal controls. 2 Methods 2.1 Study population and design One hundred and eighteen patients (mean age 64.7 ± 10.4 years, 72% male) with known CAD were recruited prospectively from the cardiology clinics in two teaching hospitals. Among them, 75% had a history of myocardial infarction (MI) (anterior in 43, inferior in 31, anterior and inferior in 13 and posterior in 1), 53% had hypertension and 41% had diabetes mellitus. Coronary angiography was performed in 77% of patients. Among those who had angiogram performed, single vessel disease was found in 28%, 2-vessel disease in 26% and triple vessel disease in 46%. Medications included antiplatelet drugs in 96%, diuretics in 39%, nitrate in 80%, β-blockers in 57%, calcium channel blocker in 26%, angiotensin converting enzyme inhibitors in 65% and statins in 72%. Patients were excluded if they had atrial fibrillation, significant aortic and mitral valvular lesions especially atrioventricular valve stenosis or regurgitation, presence of non-ischemic cardiomyopathies, and poor imaging quality that precluded satisfactory echocardiographic assessment. The echocardiographic results were compared with 100 normal subjects who were recruited from the community. These subjects did not have any history of cardiovascular or systemic illness and were not on any regular medications. They had normal physical examination including blood pressure as well as normal hemoglucostix, ECG and echocardiographic examinations. The study was approved by the human research committee of the institution and informed consent was obtained from each patient. The control group had similar age (mean: 64.1 ± 9.6 years, p = NS) and gender (65% male, p = NS) distribution to the disease group. The heart rate was also similar between patients and controls (64.8 ± 10.6 Vs 63.7 ± 12.7 beats/min, p = NS). 2.2 Echocardiography Standard echocardiography with Doppler studies were performed (Vivid 5, Vingmed-General Electric, Horten, Norway). The left ventricular (LV) dimensions and ejection fraction were measured by 2-dimension-guided M-mode method according to the guidelines of American Society of Echocardiography [9] . LV diastolic dysfunction was classified as restrictive filling pattern (RFP) (defined as transmitral early to atrial filling [MV-E/A] ratio ≥ 2 or MV-E/A = 1 to 2 and deceleration time of early filling [MV-DT] < 140 ms), abnormal relaxation pattern (defined as MV-E/A ratio < 1, MV-E/A ratio = 1 to 2 and MV-DT > 240 ms), or pseudonormal pattern (normal transmitral pattern but abnormal pulmonary venous flow profile [reverse in systolic to diastolic forward ratio] or atrial reversal time exceeds that of transmitral atrial wave time) [10–14] . The latter 2 patterns were collectively labeled as non-restrictive filling pattern (non-RFP) as previously described [10,12,15,16] . At least 3 consecutive beats of sinus rhythm were measured and the average values were taken. Atrial function was assessed at the apical 4-chamber view [17] . In both left and right atria, the following parameters were measured: medial–lateral (LAD-ML and RAD-ML) and superior–inferior atrial (LAD-SI and RAD-SI) dimensions, the atrial area just before the onset of active atrial contraction (LAA-PRE and RAA-PRE), the minimal atrial area after atrial contraction is completed (LAA-POST and RAA-POST), and the atrial emptying fraction based on the change in areas (LAA-EF and RAA-EF) of both LA and RA. Using the modified Simpson rule, the atrial volume just before (LAV-PRE and RAV-PRE) and after active atrial contraction (LAV-POST and RAV-POST) and atrial emptying fraction based on the change in volumes (LAV-EF and RAV-EF) were also calculated [17] . TDI was performed at the apical 4-chamber view for the long-axis motion of the heart as previously described [4,6] . Two-dimension echocardiography with TDI-color imaging was performed with a 2.5 or 3.5 MHz phase-array transducer. The imaging angle was adjusted to ensure a parallel alignment of the sampling window with the myocardial segment of interest. Gain settings, filters, pulse repetitive frequency, sector size and depth were adjusted to optimize color saturation. The mean frame rate was 106 ± 12 s − 1 (range: 80–140 frames/s). At least 3 consecutive beats were stored and the images were digitized and analyzed off-line by a computer (EchoPac 6.3.6, Vingmed-General Electric, Horten, Norway). Atrial Doppler velocity profile signals were reconstituted off-line by placing a 1.2 mm sampling window at the mid levels of the LA, inter-atrial septum (IAS) and the RA, respectively. The peak regional atrial contraction velocities ( V A ) after the onset of p wave of electrocardiogram were measured. The time to the onset ( T O ), peak ( T P ) and the end ( T E ) of atrial contraction were measured using the onset of p wave in ECG as the reference point. For the V A , the inter-observer and intra-observer variability has been compared in 60 consecutive measurements which were 1.4% and 3.9% for LA, 3.8% and 8.2% for RA, and 1.4% and 5.5%, respectively for IAS. The inter-observer and intra-observer variability of the timing of atrial events was similar for all the atrial sites, which were within 4.8% and 5.7%, respectively. 2.3 Statistics The difference in mean between 2 groups were examined by paired or unpaired t -test where appropriate (SPSS for Windows, ver. 11.0, SPSS Inc., Chicago, Illinois, USA). The differences in mean among various groups were analyzed by Scheffe's multiple comparisons in one-way ANOVA. Linear regression analysis was used to compare the correlation between atrial function and other echocardiographic parameters. Stepwise multiple regression analysis was performed to investigate the independent effect of various covariates on atrial function. All data were expressed as mean ± SD. A p value < 0.05 was considered statistically significant. 3 Results 3.1 Atrial function in normal subjects ( Tables 1 and 2 ) In the normal subjects, the peak V A was significantly higher in the RA than the LA ( p < 0.001), and both sites were significantly higher than the IAS (both p < 0.001). Regarding the mechanical sequence of atrial contraction, the onset ( T O ) was the earliest at the RA, which was followed by the IAS, and was the latest at the LA (all p < 0.001 when compared between sites). The atrial contractions ( T P ) was peaked nearly simultaneously at the RA and IAS, which were significantly earlier than the LA (both p < 0.001). However, atrial systole was completed ( T E ) first at the IAS (both p < 0.001 Vs LA or RA), which was followed by the LA, and was the latest in the RA. The total duration of atrial systole was the shortest at the LA, and was the longest at the RA ( p < 0.001 Vs LA). There was a positive correlation between V A in LA and age ( r = 0.30, p = 0.004), though such relationship was found for IAS and RA. 3.2 Atrial function in patients with coronary artery disease ( Tables 1 and 2 ) The transmitral and transtricuspid peak atrial velocities were not different between patients with CAD and normal controls (both p = NS). On the other hand, the V A was significantly lower in patients than controls in all the three atrial sites ( p < 0.001). The V A in the LA was lower than the RA ( p < 0.001). For the timing of atrial mechanical function, as similar to normal controls, the T O was the earliest at the RA ( p < 0.001 Vs IAS or LA), followed by the IAS, and was the latest at the LA ( p < 0.001 Vs IAS). The T P was comparable in the RA and IAS, and was followed by the peak LA contraction ( p < 0.01 Vs RA, p < 0.001 Vs IAS). The atrial contraction was accomplished first in the IAS ( p < 0.001 Vs LA or RA) and was, however, significantly delayed at the LA when compared with controls ( p < 0.01). The total duration of atrial contraction was therefore prolonged in the LA ( p = 0.02 Vs controls), though was the longest in the RA ( p < 0.001 Vs LA or IAS). There was no correlation between age and V A in any atrial sites. In subgroup analysis, patients were stratified to those with ( n = 89) and without ( n = 29) previous MI. It was observed that the V A in the atrial sites remained abnormal in either subgroup and to a similar extent ( Tables 1 and 2 ). Atrial function was assessed by other echocardiographic methods. The LA dimension, area and volume were consistently enlarged in patients with CAD ( Table 1 ). The LA emptying fraction for both area and volume measurements were significantly reduced. In patients without previous myocardial infarction, there was enlargement of LA cavity, though the emptying fraction by area or volume failed to reveal any difference ( Table 1 ). On the other hand, all the LA parameters were significantly abnormal in those with previous myocardial infarction ( Table 1 ). In the RA, increase in atrial dimension and area were observed in the CAD group, though the volumetric measurements as well as the RA emptying fraction were unable to reveal any change ( Table 1 ). These abnormalities were attributed to those with previous myocardial infarction. 3.3 Relationship between LV systolic function and atrial function ( Table 3 ) Patients with CAD had significantly lower ejection fraction (78 ± 6 Vs 59 ± 14%, p < 0.001) as well as larger LV end-diastolic (4.6 ± 0.5 Vs 5.3 ± 1.0 cm, p < 0.001) and end-systolic (2.8 ± 0.4 Vs 3.9 ± 1.1 cm, p < 0.001) diameters. There was a positive correlation between ejection fraction ( r = 0.50, p < 0.001), LV end-systolic diameter ( r = − 0.48, p < 0.001) and the V A of the LA. When patients with CAD were stratified by the LV ejection fraction, 79 (67%) had preserved ejection fraction (> 50%) and was impaired (≤ 50%) in 39 (33%) patients. For those with impaired ejection fraction, the V A was significantly lower in all the three atrial sites (all p < 0.001). In contrast, the V A was only decreased in the LA and RA (both p < 0.001) for those with preserved ejection fraction, and the V A was significantly higher in these patients in the LA and IAS than those with impaired systolic function ( p < 0.001). For other echocardiographic parameters of LA function, patients with either preserved or impaired LV ejection fraction had LA enlargement as shown by the increase in dimension and area. However, more severe and consistent changes were observed in those with impaired ejection fraction ( Table 3 ). In the RA, increased RA dimension, area, volume and RV emptying fraction were nearly only seen in those with impaired LV ejection fraction, and were further impaired in these patients than those with preserved ejection fraction. 3.4 Relationship between LV diastolic function and atrial function ( Table 4 ) With respect to LV diastolic function, 81 patients had an abnormal relaxation pattern (ARP), 10 had a pseudonormal pattern, 11 had a normal pattern and 16 had a restrictive filling pattern (RFP). The first three groups were collectively labeled as non-restrictive filling pattern (non-RFP) as it has been shown that RFP is the worst form of diastolic dysfunction [18] . Among patients with a RFP, all but one had a history of previous MI. When the transmitral peak atrial filling velocity was compared, as expected, it was reduced in patients with a RFP than those with a non-RFP (47.0 ± 16.1 Vs 73.9 ± 16.7 cm/s, p < 0.001). On the contrary, the V A was significantly lower in the IAS and LA in patients with a RFP than the non-RFP ones, and the V A was decreased in all the three atrial sites in both groups when compared with controls (all p < 0.001) ( Table 4 ). When individual patterns of diastolic function were compared, the V A in the LA was consistently abnormal in patients with a RFP (3.4 ± 2.8 cm/s, p < 0.001), abnormal relaxation (5.6 ± 2.5 cm, p < 0.001) or pseudonormal patterns (3.6 ± 1.7 cm, p < 0.001) when compared with controls (7.7 ± 2.6 cm/s) ( Fig. 1 ). In both RFP and non-RFP subgroups, the LA dimension, area and volume were significantly increased, while the LA emptying fraction by area or volume were significantly reduced ( Table 4 ). There was no difference in these parameters between patients with RFP and non-RFP. In the RA, only one dimension was found to be abnormal. Other parameters of RA function could not detect any abnormality in either patient subgroup. By stepwise multiple regression analysis, the parameters that independently predicted the reduction of V A in the LA was the ejection fraction ( p = 0.001) and the presence of diastolic dysfunction, in particular the RFP ( p = 0.002) ( Table 5 ). 3.5 Relationship between TDI and other echocardiographic parameters of atrial function In the LA, the V A correlated modestly and negatively with LA dimensions (LAD-SI: r = − 0.20, p = 0.008; LAD-ML: r = − 0.18, p = 0.02), and was better with LA area (LAA-PRE: r = − 0.33, p < 0.001; LAA-POST: r = − 0.35, p < 0.001) and LA volume (LAA-PRE: r = − 0.30, p < 0.001; LAA-POST: r = − 0.32, p < 0.001). The correlation was only modest for LA emptying fraction by area ( r = 0.24, p = 0.001) or by volume ( r = 0.22, p = 0.003). In the RA, the V A only correlated weakly with RA dimension (RAA-ML: r = 0.20, p = 0.01) and RA area (RAV-POST: r = 0.16, p = 0.04), but not RA emptying fraction (RAA-EF: r = 0.01, p = NS; RAV-EF: r = 0.04, p = NS). All the normal controls had a V A in LA > 3.5 cm/s. Therefore, prevalence of abnormal V A in LA below this cutoff value in patients with CAD was 31% (37 / 118 patients). Similar, based on a V A in the RA > 4.0 cm/s as the cutoff value, the prevalence of abnormal RA function was 20% (24 / 118 patients). 4 Discussion This study assessed atrial function non-invasively and illustrated the presence of atrial dysfunction in patients with CAD. The regional atrial velocities in both atria decreased significantly. This was consistently seen in patients with and without previous MI, and was worse in the presence of LV systolic dysfunction or RFP of diastolic dysfunction. The latter two were also independent predictors of LA contractile dysfunction. However, atrial electromechanical delay was unaffected in CAD. 4.1 Assessment of atrial pump function Despite the availability of various tools for the assessment of ventricular function, quantitative assessment of atrial function remains limited. Echocardiography has been employed to quantify LA function, such as transmitral atrial filling velocity and atrial filling fraction by Doppler study [19] , the measurement of LA area by planimetry [17] or automatically by acoustic quantification [20,21] . However, the methods are not without drawbacks: the transmitral atrial velocity is more dependent on the atrioventricular pressure gradient rather than atrial contractility per se [1] , while far-field artifact may affect the accuracy of atrial area calculation. On the other hand, TDI has been proven to be useful to assess global and regional cardiac function [2,5,22] as well as the timing of events in the cardiac cycle, including that of atrial contraction [23–25] . In normal subjects, the RA has a higher V A than LA, and it was the lowest at the IAS. Therefore, the free walls of both atria have higher mobility and hence higher velocities than the relatively attached IAS. In the presence of CAD, the V A in the LA and IAS were drastically reduced. The RA was affected by the disease, albeit to a lesser extent. Therefore, atrial function was jeopardized in the presence of CAD, even in those without previous MI. A previous invasive hemodynamic study observed that LA function was abnormal in the presence of myocardial infarction or heart failure, though no data was available in patients without infarction [8] . In our patients, a low LV ejection fraction and the presence of RFP in diastole were important determinants of LA contractile dysfunction. From the regression coefficient, each percentage of increase in ejection fraction was associated with an increase in V A in LA by 0.06 cm/s while the occurrence of RFP was associated with a reduction of V A in LA by 3.17 cm/s ( Fig. 2 ). The present study was not intended to study the mechanism of producing atrial dysfunction in CAD. However, it is possibly a direct result of atrial myocyte ischemia or infarction. In acute ischemic model, LA mechanical dysfunction is evident in patients with left circumflex territory ischemia or transient occlusion due to the jeopardized blood supply, but not in those with left anterior descending coronary artery disease [8] . The impairment of RA function may be partially explained by atherosclerotic disease of the right coronary artery which was present in over half of the patients underwent coronary angiography. Also, dysfunction of either atria may also caused by structural atrial abnormalities, such as the accumulation of interstitial fibrosis, or secondary to atrial pressure overload in the presence of ventricular systolic or diastolic dysfunction [11] . RFP of diastolic dysfunction has been regarded as the worst form of diastolic dysfunction which is associated with a poor prognosis [10,16] . Our patients with RFP had a dramatic reduction of V A in all the atrial sites, especially the LA, implicating the presence of atrial dysfunction. LA failure has been suggested in these patients. Such hypothesis is further supported by the increased LA dimension and size, and reduced atrial emptying fraction by area. On the other hand, the redistribution of LV diastolic filling in patients with delayed early diastolic relaxation will increase atrial preload in late diastole, while atrial afterload (LV pressure) could be normal. As a result, the volumetric increase in the LA may render it to act partially as a conduit for filling during late diastolic phase, apart from active atrial contraction. Therefore, transmitral atrial velocity could appear to be normal or even paradoxically increased ( Fig. 1 ). In fact, left atrial velocity were confirmed to be abnormal in these patients by TDI. Therefore, one needs to analyze atrial contractile function with caution when based on transmitral profile alone. Recently abnormal LA function in diastolic dysfunction was also confirmed by acoustic quantification [21] . The use of transmitral atrial velocity to assess atrial mechanical function might be more appropriate for serial assessment of atrial function in the same study subject where the change in hemodynamic condition is minimal, such as after cardioversion for atrial fibrillation [19,26] . 4.2 Interatrial delay in atrial contraction The time domain of atrial mechanical function between RA and LA has not been investigated systemically in the presence of cardiac diseases. This study revealed the time sequence of atrial mechanical function and how it was affected by the presence of CAD. In normal subjects, the onset of atrial systole began in the RA, followed by the IAS, and was the latest in the LA. The interatrial delay was 24 ms. This represents the propagation of the atrial depolarization from the sinoatrial node to the LA [27] . Previous studies based on M-mode echocardiography and invasive catheterization reported an inter-atrial delay of 25 ms and 20 ms, respectively, in normal subjects [28,29] . Since these studies were performed in a small number of normal controls (20 or less) [27,29] , this study adds further confirmatory information over a large control population. In the presence of CAD, the onset (25 ms) and peak (10 ms) inter-atrial electromechanical delay was unaltered when compared with the controls. On the other hand, the total duration of atrial contraction in the LA was prolonged, indicating that a longer period of contraction is needed to overcome the reduced contractile force and possibly elevated LV late-diastolic pressure in these patients. Therefore, although atrial mechanical function is jeopardized in the presence of CAD, inter-atrial electrical conduction is preserved. 4.3 Clinical implications Atrial contractile dysfunction is consistently observed in patients with CAD, disregarding the status of previous myocardial infarction, coexisting systolic dysfunction or severity of diastolic dysfunction. Therefore, it appears to be an early echocardiographic feature in CAD, even before the occurrence of obvious LV systolic dysfunction. In addition, this study provides evidence that atrial contractile function could be abnormal, even in condition with a normal or large transmitral atrial wave. It appears that the measurement of regional atrial velocities could provide supplementary information on mechanical function of both atria while conventional assessment such as planimetry could be tedious and time consuming. Therefore, for a comprehensive assessment of cardiac function, a quick measurement of atrial function by TDI is helpful. Furthermore, study of the atrial contractile function will help clinicians to better understand the impact of CAD on cardiac function. References [1] J.D. Thomas A.E. Weyman Echocardiographic Doppler evaluation of left ventricular diastolic function. Physics and physiology Circulation 84 3 1991 977 990 [2] P. Trambaiolo G. Tonti A. Salustri F. Fedele G. Sutherland New insights into regional systolic and diastolic left ventricular function with tissue Doppler echocardiography: from qualitative analysis to a quantitative approach J Am Soc Echocardiogr 14 2 2001 85 96 [3] L. Rodriguez M. Garcia M. Ares B.P. Griffin S. Nakatani J.D. Thomas Assessment of mitral annular dynamics during diastole by Doppler tissue imaging: comparison with mitral Doppler inflow in subjects without heart disease and in patients with left ventricular hypertrophy Am Heart J 131 5 1996 982 987 [4] R.G. Pai K.S. Gill Amplitudes, durations, and timings of apically directed left ventricular myocardial velocities: II. Systolic and diastolic asynchrony in patients with left ventricular hypertrophy J Am Soc Echocardiogr 11 2 1998 112 118 [5] J. Gorcsan V.K. Gulati W.A. Mandarino W.E. Katz Color-coded measures of myocardial velocity throughout the cardiac cycle by tissue Doppler imaging to quantify regional left ventricular function Am Heart J 131 6 1996 1203 1213 [6] C.M. Yu Q. Wang C.P. Lau Reversible impairment of left and right ventricular systolic and diastolic function during short-lasting atrial fibrillation in patients with an implantable atrial defibrillator: a tissue Doppler imaging study Pacing Clin Electrophysiol 24 6 2001 979 988 [7] D.L. Mann Mechanisms and models in heart failure: a combinatorial approach Circulation 100 9 1999 999 1008 [8] C. Stefanadis J. Dernellis E. Tsiamis P. Toutouzas Effects of pacing-induced and balloon coronary occlusion ischemia on left atrial function in patients with coronary artery disease J Am Coll Cardiol 33 3 1999 687 696 [9] D.J. Sahn A. DeMaria J. Kisslo A. Weyman Recommendations regarding quantitation in M-mode echocardiography: results of a survey of echocardiographic measurements Circulation 58 6 1978 1072 1083 [10] G.Y. Xie M.R. Berk M.D. Smith J.C. Gurley A.N. DeMaria Prognostic value of Doppler transmitral flow patterns in patients with congestive heart failure J Am Coll Cardiol 24 1 1994 132 139 [11] C.M. Yu J.E. Sanderson S. Chan L. Yeung Y.T. Hung K.S. Woo Right ventricular diastolic dysfunction in heart failure Circulation 93 8 1996 1509 1514 [12] F. Nijland O. Kamp A.J. Karreman M.J. van Eenige C.A. Visser Prognostic implications of restrictive left ventricular filling in acute myocardial infarction: a serial Doppler echocardiographic study J Am Coll Cardiol 30 7 1997 1618 1624 [13] O. Rossvoll L.K. Hatle Pulmonary venous flow velocities recorded by transthoracic Doppler ultrasound: relation to left ventricular diastolic pressures J Am Coll Cardiol 21 7 1993 1687 1696 [14] P. Poirier P. Bogaty C. Garneau L. Marois J.G. Dumesnil Diastolic dysfunction in normotensive men with well-controlled type 2 diabetes: importance of maneuvers in echocardiographic screening for preclinical diabetic cardiomyopathy Diabetes Care 24 1 2001 5 10 [15] C.M. Yu B.M. Cheung R. Leung Q. Wang W.H. Lai C.P. Lau Increase in plasma adrenomedullin in patients with heart failure characterised by diastolic dysfunction Heart 86 2 2001 155 160 [16] H.C. Yu J.E. Sanderson Different prognostic significance of right and left ventricular diastolic dysfunction in heart failure Clin Cardiol 22 8 1999 504 512 [17] S.J. Lester E.W. Ryan N.B. Schiller E. Foster Best method in clinical practice and in research studies to determine left atrial size Am J Cardiol 84 7 1999 829 832 [18] R.A. Nishimura M.D. Abel L.K. Hatle A.J. Tajik Assessment of diastolic function of the heart: background and current applications of Doppler echocardiography: Part II. Clinical studies Mayo Clin Proc 64 2 1989 181 204 [19] H.F. Tse Q. Wang C.M. Yu G.M. Ayers C.P. Lau Time course of recovery of left atrial mechanical dysfunction after cardioversion of spontaneous atrial fibrillation with the implantable atrial defibrillator Am J Cardiol 86 9 2000 1023 1025 [A10] [20] A.D. Waggoner B. Barzilai J.G. Miller J.E. Perez On-line assessment of left atrial area and function by echocardiographic automatic boundary detection Circulation 88 3 1993 1142 1149 [21] C.M. Yu H. Lin L.C. Kum W.F. Lam W.H. Fung J.E. Sanderson Evidence of atrial mechanical dysfunction by acoustic quantification in abnormal relaxation and restrictive filling patterns of diastolic dysfunction in patients with coronary artery disease Eur J Echocardiog 4 4 2003 272 278 [22] J. Gorcsan D.P. Strum W.A. Mandarino V.K. Gulati M.R. Pinsky Quantitative assessment of alterations in regional left ventricular contractility with color-coded tissue Doppler echocardiography. Comparison with sonomicrometry and pressure–volume relations Circulation 95 10 1997 2423 2433 [23] J. Zamorano D.R. Wallbridge J. Ge J. Drozd J. Nesser R. Erbel Non-invasive assessment of cardiac physiology by tissue Doppler echocardiography. A comparison with invasive haemodynamics Eur Heart J 18 2 1997 330 339 [24] K. Miyatake M. Yamagishi N. Tanaka New method for evaluating left ventricular wall motion by color-coded tissue Doppler imaging: in vitro and in vivo studies J Am Coll Cardiol 25 3 1995 717 724 [25] M.A. Garcia-Fernandez J. Azevedo M. Moreno Regional diastolic function in ischaemic heart disease using pulsed wave Doppler tissue imaging Eur Heart J 20 7 1999 496 505 [26] W.J. Manning D.E. Leeman P.J. Gotch P.C. Come Pulsed Doppler evaluation of atrial mechanical function after electrical cardioversion of atrial fibrillation J Am Coll Cardiol 13 3 1989 617 623 [27] M.E. Josephson D.L. Scharf J.A. Kastor J.G. Kitchen Atrial endocardial activation in man. Electrode catheter technique of endocardial mapping Am J Cardiol 39 7 1977 972 981 [28] K. Wang H.B. Xiao S. Fujimoto D.G. Gibson Atrial electromechanical sequence in normal subjects and patients with DDD pacemakers Br Heart J 74 4 1995 403 407 [29] E. Braunwald A.P. Fishman A. Cournand Time relationship of dynamic events in the cardiac chambers, pulmonary artery and aorta in man Circ Res 4 1956 100 107