Article Text
PDF
Abstract
Background Left atrial (LA) maximal volume is of prognostic value in patients after acute myocardial infarction (AMI). Recently, LA mechanical function and LA strain have been introduced as alternative methods to assess LA performance more accurately.
Objective To evaluate the relation between LA volume, mechanical function and strain, and adverse events in patients after AMI.
Methods Patients with AMI underwent two-dimensional echocardiography within 48 h of admission. LA volume and LA performance (mechanical function and systolic strain) were quantified. The endpoint was a composite of all-cause mortality, reinfarction and hospitalisation for heart failure.
Results 320 patients (mean age 60±12 years, 78% men) were followed up for 27±14 months. During follow-up, 48 patients (15%) reached the composite endpoint. After adjustment for clinical and echocardiographic parameters, LA maximal volume (HR 1.05, CI 1.00 to 1.10, p=0.04) and LA strain (HR 0.94, CI 0.89 to 0.99, p=0.02) were independently associated with adverse outcome. In addition, LA strain provided incremental value to LA maximal volume (p=0.03) for the prediction of adverse outcome.
Conclusions After AMI treated with primary percutaneous coronary intervention, LA strain provides additional prognostic value beyond LA maximal volume.
- Echocardiography
- left atrium
- myocardial infarction
- prognosis
- cardiac remodelling
- diastolic dysfunction
- risk stratification
- stemi
Statistics from Altmetric.com
- Echocardiography
- left atrium
- myocardial infarction
- prognosis
- cardiac remodelling
- diastolic dysfunction
- risk stratification
- stemi
Introduction
Left atrial (LA) maximal volume has been recognised as a powerful predictor of mortality and hospitalisation for heart failure in patients after acute myocardial infarction (AMI).1–3 Normal LA volumes are associated with good outcome, even in patients with depressed left ventricular (LV) function.1 4 On the other hand, larger LA volumes are associated with chronic increased LV filling pressures and adverse outcome post-AMI.2 Besides LA volume, recent studies have shown the value of several LA functional parameters.5 For example, LA ejection force is a measure of LA mechanical function that is strongly related to LV diastolic function. In the Strong Heart Study, LA ejection force was an independent predictor of cardiovascular events.5 6 Therefore, besides quantification of LA size, assessment of LA mechanical function may have additional prognostic value in post-AMI patients. However, the assessment of these parameters involves numerous geometrical assumptions and often results in underestimation of the atrial size.7
Direct evaluation of atrial myocardial function is currently feasible with speckle-tracking imaging. This novel technique permits assessment of active myocardial deformation which may provide additive value concerning LA function when compared to conventional echocardiographic measurements.8 9 Accordingly, the purpose of the current evaluation was to investigate the association between LA performance expressed in LA volumes, mechanical function and strain, and adverse events in post-AMI patients.
Methods
Patient selection and data collection
Consecutive patients admitted with ST-segment elevation AMI treated with primary percutaneous coronary intervention were evaluated. Diagnosis of ST-segment elevation AMI was based on typical electrocardiographic changes with clinical symptoms associated with elevation of cardiac biomarkers.10 Clinical and echocardiographic data were prospectively entered into the departmental cardiology information system (EPD-Vision, Leiden University Medical Center) and the echocardiography database, respectively, and retrospectively analysed.11 12 All patients were treated according to the institutional AMI protocol (MISSION!).11 This protocol, designed to improve care around AMI, includes structurised medical therapy, two-dimensional echocardiography performed <48 h of admission and standardised follow-up, as described previously.11 The baseline echocardiogram was used to assess LA and LV function. Specifically, LA function was assessed with phasic volumes by conventional echocardiography and with LA strain and strain rate by speckle-tracking imaging. Of note, patients with atrial fibrillation were excluded.
In addition, 35 normal controls selected from an echocardiographic database were included to provide the normal reference values of LA phasic volumes, strain and strain rate.13 The group of controls comprised individuals matched for age and gender who were referred for echocardiography with atypical chest pain, palpitations or syncope without murmur, and did not show structural heart disease. Those individuals who showed LV dilatation, had known hypertension, or were referred for echocardiographic evaluation of known valvular disease, murmur or heart failure were excluded.
Echocardiography
All patients were imaged in the left lateral decubitus position using a commercially available system (Vivid 7, General Electric-Medical Systems, Horton, Norway). Images were obtained with a simultaneous ECG signal, using a 3.5-MHz transducer at a depth of 16 cm in the parasternal and apical views. Standard M-mode and two-dimensional images were acquired during breath hold and saved in cine-loop format. Analysis of echocardiographic images was performed offline by two independent observers using dedicated software (EchoPac version 108.1.5, General Electric-Vingmed).
LV end-systolic volume, end-diastolic volume and ejection fraction were assessed using the biplane Simpson method in the apical 4- and 2-chamber views.14
In addition, the LV was divided into 16 segments; each segment was analysed individually and scored based on its motion and systolic thickening (1=normokinesis, 2=hypokinesis, 3=akinesis, 4=dyskinesis). Wall motion score index was calculated as the sum of the segment scores divided by the number of segments scored.14
Severity of mitral regurgitation was graded semi-quantitatively from the jet area of colour-flow Doppler data and by measuring the width of the vena contracta. Mitral regurgitation was characterised as: mild=jet area/LA area <20% and vena contracta width <0.30 cm; moderate=jet area/LA area 20–40% and vena contracta width 0.30–0.69 cm; and severe=jet area/LA area >40% and vena contracta width ≥0.70 cm.15
To assess diastolic function, pulsed-wave Doppler of the mitral valve inflow was obtained by placing the Doppler sample volume between the tips of the mitral leaflets. The early (E) and late (A) peak diastolic velocities and E-wave deceleration time were measured. E/E′ ratio was obtained by dividing E by E′, which was measured using colour-coded tissue Doppler imaging at the septal side of the mitral annulus in the apical 4-chamber view.16
Analysis of LA function
LA function consists of the reservoir period (inflow during ventricular systole), conduit period (passive emptying during ventricular relaxation and diastasis) and contractile period (active emptying). To analyse all components of LA function, LA volumes were calculated according to the biplane Simpson method at three time points: (1) maximal volume (LAmax) at end-systole, just before mitral valve opening; (2) minimal volume (LAmin) at end-diastole, just before mitral valve closure; and (3) volume before atrial active contraction (LApreA) obtained from the last frame before mitral valve reopening or at time of the P wave on the surface electrocardiogram. All LA volumes were indexed to the body surface area.14
LA mechanical function was derived from the LA volumes and expressed with the following formulae: (1) total atrial emptying fraction: LA total ejection fraction=((LAmax−LAmin)/LAmax)×100; (2) active atrial emptying fraction: LA active ejection fraction=((LApreA−LAmin)/LApreA)×100, which is considered an index of LA active contraction; (3) passive atrial emptying fraction: LA passive ejection fraction=((LAmax−LApreA)/LAmax)×100, which is considered an index of LA conduit function; and (4) atrial expansion index: LA expansion index=((LAmax−LAmin)/LAmin)×100, which is considered an index of LA reservoir function.17
Longitudinal LA wall deformation was assessed in the apical views using speckle-tracking analysis.18 This novel software analyses motion by tracking frame-to-frame movement of natural acoustic markers in two dimensions. All images were recorded with a frame rate of >40 fps (range 40–100 fps) for reliable analysis. The LA endocardial border was manually traced and the automatically created region of interest was adjusted to the thickness of the myocardium. The extent of LA wall stretching during the reservoir period may be important for maintaining adequate LV filling.19 Therefore, LA peak systolic longitudinal strain and strain rate were assessed at each mid-LA segment (septal, lateral, anterior, inferior and posterior) in the apical views and averaged as a measure of LA compliance.8 20 Segments were discarded if tracking was of poor quality. Strain and strain rate analysis was feasible in 79% of segments.
Follow-up and endpoint definitions
All patients were followed up according to the protocol and the occurrence of adverse events was noted. Patients in whom more than 6 months follow-up data were lacking, were considered as lost to follow-up, and excluded from further analysis. The endpoint was defined as a composite of all-cause mortality, non-fatal reinfarction and hospitalisation for heart failure. Non-fatal reinfarction was defined based on criteria of typical chest pain, elevated cardiac enzyme levels, and typical changes on the electrocardiogram.10 Hospitalisation for heart failure was defined as hospitalisation for new-onset or worsening heart failure.
Statistical analysis
Continuous data are presented as mean±SD and categorical data are presented as frequencies and percentages. Differences in characteristics between patient groups were evaluated using the unpaired Student t test and χ2 test.
The primary aim was to assess the association between LA performance and adverse events after adjusting for clinical and echocardiographic covariates. Separate multivariable models were constructed for LA volumes, mechanical function, strain and strain rate using Cox proportional hazards analysis to evaluate the individual prognostic importance of the different LA measurements. Because of the relative low number of events, the number of covariables had to be limited. Accordingly, based on both clinical judgement and univariable statistical significance, age, Killip class, multivessel disease, peak cardiac troponin T level, LV ejection fraction, E/E′ ratio and mitral regurgitation were introduced in the model.
In addition, the potential relationship between renal function and LA volumes, phasic and mechanical function, strain and strain rate was assessed with ANOVA tests. For this purpose, estimated glomerular filtration rate (eGFR) was calculated using the standard formula by Cockcroft and Gault and expressed in ml/min/1.73 m2.21 Patients were divided into three subgroups according to the cut-off values proposed by the National Kidney Foundation practice guidelines: eGFR ≥90 ml/min/1.73 m2 for normal kidney function, eGFR 60–90 ml/min/1.73 m2 for mildly decreased, and eGFR<60 ml/min/1.73 m2 for moderately to severely decreased kidney function.22
To further investigate the clinical relevance of LA performance, the population was stratified into two groups according to LA dysfunction. The cut-off value for LA maximal volume was chosen at 32 ml/m2 which corresponds to 2 SDs from the normal LA size and has been previously validated in relation to clinically relevant endpoints.1 2 4 The normal value of LA strain in the group of 35 normal controls was 39±10%. Patients were therefore divided according to the mean value minus 2 SDs, which corresponds to the lower limit of normal LA strain (19%).
Event rates were plotted in Kaplan–Meier curves for the composite endpoint and the study population divided by the previously mentioned cut-off values, and groups were compared using the log-rank test.
The incremental value of LA performance to known risk factors for adverse outcome (age, Killip class, multivessel disease, peak cardiac troponin T level, LV ejection fraction, E/E′ ratio and mitral regurgitation) was established. For this purpose, those characteristics were entered in the Cox proportional hazard model in a stepwise fashion. Subsequently, LA maximal volume and LA strain were entered individually, to test further incremental value. Global χ2 values including significance levels were calculated.
Finally, 15 patients were randomly selected to test the intra- and inter-observer reproducibility of LA measurements. Bland–Altman analyses were performed.
All statistical tests were two-sided, and a p value <0.05 was considered statistically significant.
Results
Patient characteristics
A total of 368 consecutive AMI patients treated with primary percutaneous coronary intervention were evaluated. Three (0.8%) patients died before echocardiographic examination could be performed, and in 8 (2.2%) patients echocardiographic assessment was not available <48 h of admission due to logistic reasons. Another 19 (5.7%) patients were excluded from further analysis because image quality was not sufficient for analysis and 18 (4.9%) patients were lost to follow-up. The study population therefore comprised 320 patients. Tables 1 and 2 summarise the clinical and echocardiographic characteristics. Mean age of the patients was 60±12 years and 78% were male. Baseline echocardiography revealed an LV ejection fraction of 46±8%; maximal LA volume was 25±7 ml/m2. Mean LA strain and strain rate were 33±11% and 2.3±0.7 per second, respectively.
Baseline clinical characteristics
Baseline echocardiographic characteristics
Echocardiographic data obtained in AMI patients were compared with the group of 35 normal controls. Post-AMI patients had significantly larger LA maximal volume (25±7 vs 22±6 ml/m2, p=0.04) and lower LA total ejection fraction (56±11 vs 61±5%, p=0.002) compared with the group of normal controls. Interestingly, passive emptying fraction was significantly reduced in the post-AMI patients (28±10 vs 39±13%, p<0.001), which was compensated by an increased active emptying fraction (38±11 vs 34±14%, p=0.04). LA reservoir function was significantly reduced (140±65 vs 164±41%, p=0.03), also reflected by the lower LA strain (39±10 vs 33±11%, p=0.002) in comparison with the normal controls (table 2).
In addition, the relationship between renal function LA volumes and LA mechanical function was evaluated. There were no significant differences in LA volumes and phasic function among the three categories of eGFR. However, decreasing eGFR was associated with significantly lower LA strain and strain rate (from 35±11 to 32±11 and 25±11%, p<0.001 for LA strain; and from 2.4±0.7 to 2.2±0.7 and 1.9±0.7 per second, p=0.005 for LA strain rate).
Fifteen patients were randomly identified for inter- and intra-observer agreement. According to the Bland–Altman analysis, intra-observer variability was good, with mean differences of 1.6±2.4 ml/m2 for LA maximal volume, −1.2±1.6 ml/m2 for LA minimal volume, 1.0±1.4 ml/m2 for LApreA volume, 0.5±3.0% for LA strain and 0.06±0.22 per second for LA strain rate. Inter-observer reproducibility was also good, with mean differences for LA maximal volume, LA minimal volume, LApreA volume, LA strain and LA strain rate of 2.1±4.2 ml/m2, −1.8±2.1 ml/m2, 1.6±2.6 ml/m2, 1.0±4.4% and −0.12±0.24 per second, respectively.
Follow-up
During a mean follow-up of 27±14 months, 48 patients (15%) reached the composite endpoint: 29 patients died (9%), 11 patients (3%) had a non-fatal reinfarction, and 14 patients (4%) were hospitalised for heart failure. Differences in clinical and echocardiographic characteristics between patients who reached the composite endpoint and patients who remained event-free are shown in tables 1 and 2.
Relation between LA performance and outcome
Table 3 shows the significant univariable predictors of the composite endpoint. In addition to clinical characteristics and LV function measurements, LA maximal volume, ejection fraction, strain and strain rate were univariable predictors of the composite endpoint. After adjusting LA maximal volume, ejection fraction, strain and strain rate for other variables that predicted adverse outcome, LA maximal volume and LA strain independently predicted the occurrence of the composite endpoint (HR 1.05, 95% CI 1.00 to 1.10, p=0.04 and HR 0.94, 95% CI 0.89 to 0.99, p=0.02, respectively). However, LA ejection fraction and LA strain rate did not remain significant in the multiple variable analysis (HR 0.99, 95% CI 0.96 to 1.03, p=0.63 and HR 0.55, 95% CI 0.28 to 1.06, p=0.07, respectively). To further investigate the prognostic value of LA function, LA maximal volume and LA strain were dichotomised according to normal and abnormal LA function with the above described cut-off values. Kaplan–Meier curves for LA maximal volume divided into >32 ml/m2 and ≤32 ml/m2 and LA strain divided into <19% and ≥19% are shown in figure 1. The 3-year event rate in patients with LA maximal volume >32 ml/m2 (n=50) was 35% compared to 14% in patients with LA maximal volume ≤32 ml/m2 (n=270, p=0.008). The incidence of adverse events at 3 years was 26% in patients with LA strain <19% (n=23) and 12% in patients with LA strain ≥19% (n=228, p=0.001).
Cox univariable predictors for the composite endpoint
Cumulative incidence of adverse events in patients stratified by left atrial (LA) maximal volume (panel A) and LA strain (panel B).
Incremental value of LA strain to traditional risk factors and LA volume
Global χ2 values were calculated to assess the incremental value of LA function. LA maximal volume provided incremental value to traditional risk factors (age, Killip class ≥2, multivessel disease, peak cardiac troponin T level, LV ejection fraction, E/E′ ratio and moderate or severe mitral regurgitation) by increasing the global χ2 value from 35.5 to 40.2 (p=0.046). In addition, when LA strain was added to the previous model with LA maximal volume, the predictive power of the model increased even further, reflected by the increase in the global χ2 from 40.2 to 43.1 (p=0.03).
Discussion
The main findings of the present retrospective evaluation can be summarised as follows: (1) LA reservoir function assessed with LA strain provides useful information in patients with AMI treated with primary percutaneous coronary intervention; and (2) LA strain is a promising novel technique to quantify LA function; it provides additional value to baseline risk factors and LA maximal volume for the prediction of adverse events after AMI.
Assessment of LA function and outcome
In the current evaluation, LA function was assessed using LA volumes, mechanical function and strain. Currently, guidelines recommend measuring LA volume with the ellipsoid model or Simpson method.14 Indeed, LA volume has been found to be strongly related to cardiovascular disease.23 Several studies have demonstrated that LA volume, measured early after AMI, provides prognostic value incremental to known risk factors.1 2 4 Recently, Meris et al demonstrated the strong relationship between LA volume and outcome in patients with LV dysfunction or heart failure after AMI.1 LA indexed volume ≥32ml/m2 was independently associated with death or heart failure (HR 2.35, 95% CI 1.28 to 4.31, p=0.006).1 These results were extended in the present evaluation, including patients with AMI treated with primary percutaneous coronary intervention and relatively preserved LV function.
Beyond LA size, LA mechanical function may improve the risk stratification. LA function consists of the reservoir period (inflow during ventricular systole), conduit period (passive emptying during ventricular relaxation and diastasis) and contractile period (active emptying). In post-AMI patients, LV remodelling occurs with concomitant effects on the LA. In the present evaluation, comparisons of LA function with matched normal controls demonstrated that besides LA dilatation, LA total ejection fraction deteriorated post-AMI. Interestingly, assessment of phasic changes of LA volumes demonstrated that LA passive ejection fraction is significantly diminished, which is compensated by an increase in active contractile function of the LA. As a result, LV stroke volume can be maintained despite LV dysfunction.24 25 This phenomenon has been reported previously by Bozkurt et al, who performed serial echocardiography in 73 patients with an anterior AMI at four time points (at admission, and after 1 week, 1 month and 3 months).26 The authors demonstrated that remodelling of the LA starts during the first week post-AMI and continues gradually up to 3 months.
Besides active contraction, LA relaxation reflected by the reservoir function is particularly important during acute ischaemia.24 Due to increased LV chamber stiffness and LV filling pressures, LA pressure may be increased.27 To maintain adequate LV filling, a preserved LA reservoir function, which can withstand the impact of the increased LA pressure, is crucial. In contrast, in patients with non-compliant LA and reduced reservoir function, LV filling may be significantly impaired, increasing the risk of heart failure and death. However, evaluation of LA reservoir function relies on LA volume measurements and is therefore challenging. Measurements may be inaccurate as they depend on geometrical assumptions and are load dependent.28 In contrast, speckle-tracking is a comprehensive imaging tool that permits LA reservoir function assessment by direct evaluation of the atrial myocardium and may better reflect intrinsic LA function properties.8 19 29
LA strain
Speckle-tracking derived strain has been used extensively to detect subtle LV dysfunction and associated outcome in different patient populations.30–32 Recently, several studies have demonstrated that strain measurements are feasible and useful for the detection of changes in LA performance.20 33 34 In particular, the assessment of LA reservoir function post-AMI by direct evaluation of LA myocardium deformation may provide clinically relevant information. Peak positive longitudinal strain of the LA reflects the stretch of the wall during the reservoir period. During acute LV ischaemia, atrial contraction is initially increased and compensates LV dysfunction. However, with further progression of LV dysfunction and increased LV filling pressures, the LA distensibility becomes more important. Previous studies have demonstrated that the LA reservoir function is determined by the preceding LA contraction, LV contraction through the descent of the base during systole and influenced by the LA chamber stiffness.27 LA strain reflects all those components by directly evaluating the amount of deformation of the myocardium as reported by previous studies.9 33 34
Recently, in 36 patients with systolic heart failure, Cameli et al reported that LA strain correlated better with pulmonary capillary wedge pressure than the traditional E/E′ ratio.35 In addition, excellent sensitivity and specificity of 100% and 93%, respectively, were observed for LA strain <15.1% to predict elevated filling pressures. The strong correlation between LA strain and LV diastolic dysfunction may explain the strong relation observed with adverse outcome in the current study. Although this is the first study to evaluate the prognostic value of LA strain in patients after AMI, several studies have related LA strain to outcome in other patient populations. For example, in patients with atrial fibrillation, studies have demonstrated the predictive value of LA strain for maintenance of sinus rhythm after catheter ablation.20
Limitations
The cut-off value for LA strain was chosen at 2 SDs from the normal LA strain in a group of 35 normal controls, corresponding to 19%. These results may not apply to larger populations. The addition of LA strain to the model including LA volume yielded a significant but modest increase in the global χ2 value. Therefore, the clinical relevance of these measurements needs to be further investigated. In addition, measurement of LA strain may be challenging as demonstrated by the reported feasibility. However, the semi-automated assessment of LA strain is promising and provides a comprehensive assessment of LA function. In addition, improvements of the software may improve the feasibility of the application in clinical practice. Finally, mitral annular velocities were assessed with colour-coded tissue Doppler imaging.
Conclusion
The current retrospective evaluation demonstrates that LA strain provides additional prognostic value beyond LA maximal volume in patients with AMI treated with primary percutaneous intervention.
References
Footnotes
Funding JJB received grants from GE Healthcare, Lantheus Medical Imaging, St Jude Medical, Medtronic, Boston Scientific, Biotronik and Edwards Lifesciences. MJS received grants from Boston Scientific, Medtronic and Biotronik.
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.