Prognostic Significance of the Combination of Left Atrial Reservoir Strain and Global Longitudinal Strain Immediately After Onset of ST-Elevation Acute Myocardial Infarction

Abstract

Background: The role of left atrial (LA) function in the long-term prognosis of ST-elevation acute myocardial infarction (STEMI) is still unclear.

Methods and Results: Percutaneous coronary intervention (PCI) was performed in 433 patients with the first episode of STEMI within 12 h of onset. The patients underwent echocardiography 24 h after admission. LA reservoir strain and other echocardiographic parameters were analyzed. Follow up was performed for up to 10 years (mean duration, 91 months). The primary endpoint was major adverse cardiovascular events (MACE): cardiac death or hospitalization due to heart failure (HF). MACE occurred in 90 patients (20%) during the follow-up period. Multivariate Cox hazard analyses showed LA reservoir strain, global longitudinal strain (GLS), age and maximum B-type natriuretic peptide (BNP) were the significant predictors of MACE. Kaplan-Meier curves demonstrated that LA reservoir strain <25.8% was a strong predictor (Log rank, χ²=76.7, P<0.0001). Net reclassification improvement (NRI) demonstrated that adding LA reservoir strain had significant incremental effect on the conventional parameters (NRI and 95% CI: 0.24 [0.11–0.44]) . When combined with GLS >−11.5%, the patients with LA reservoir strain <25.8% were found to be at extremely high risk for MACE (Log rank, χ²=126.3, P<0.0001).

Conclusions: LA reservoir strain immediately after STEMI onset was a significant predictor of poor prognosis in patients, especially when combined with GLS.

Echocardiography enables healthcare professionals to evaluate cardiac function and hemodynamic status, even for ST-segment elevation myocardial infarction (STEMI). The ratio of early diastolic flow velocity of transmitral flow (TMF) (E) to early diastolic mitral annulus velocity (e’) (E/e’) and left atrial (LA) dilatation are well recognized as strong predictors of the adverse outcomes after STEMI.^1,2 Recently, atrial function using tissue Doppler imaging (TDI) immediately after STEMI onset was reported as a predictor of STEMI.³ Recent studies have revealed that 2-dimensional speckle tracking strain echocardiography (2D-STE) imaging quantifies even subtle LV dysfunction in patients with various cardiac diseases, including STEMI.⁴ Furthermore, LA reservoir strain measured by 2D-STE has been proposed as a supplementary marker of LV filling pressure.⁵ LA reservoir strain was reported to provide compensatory information for global longitudinal strain (GLS); however, it did not add further information regarding adverse outcomes in patients with acute myocardial infarction (AMI).⁶ In contrast, a recent report has shown the clinical usefulness of peak LA reservoir strain in patients with acute heart failure;^7,8 however, there are no data on the role of the LA strain parameter that is exclusively involved in patients with STEMI. In this study, we assessed the clinical role of LA strain parameters in predicting long-term outcomes in patients with a first STEMI who underwent reperfusion therapy with percutaneous coronary intervention (PCI).

Methods

Patients and Protocols

The study protocol is shown in Supplementary Figure 1. The retrospective observational study group comprised 615 consecutive patients with STEMI admitted to the Yokohama City University Medical Center, Yokohama, Japan, between April 2008 and October 2014. The exclusion criteria were as follows: (1) previous myocardial infarction (MI; n=31); (2) severe valvular heart disease (n=20); (3) chronic atrial fibrillation (AF; n=35); (4) death before examination (n=10); (5) incapable of undergoing strain analysis (n=50); (6) undergoing hemodialysis (n=5); (7) in-hospital death after echocardiography (n=12); and (8) incapable of undergoing LA strain analysis (n=36). Finally, 433 patients were included in this study. All patients successfully underwent reperfusion therapy with PCI within 12 h after symptom onset and were discharged from the hospital. STEMI was defined as: (1) typical chest pain lasting >30 min; (2) ST-segment elevation >0.1 mV within 2 contiguous leads on initial electrocardiogram (ECG); and (3) elevation of creatinine phosphokinase (CPK) levels to twice the upper limit of the normal range. Hemodynamic status was defined using Killip’s classification. Echocardiography was performed 24 h after patient admission (within 48 h). ^{99 m}Tc-sestamibi single-photon emission computed tomography (SPECT) was performed 7–14 days after PCI to estimate the infarct size in 264 patients. Patients were administered 16 mCi (600 MBq) ^{99 m}Tc-sestamibi intravenously, and SPECT was performed 1 h after injection of the radioactive agent. Infarct size was defined as <50% of uptake area, which was measured by an experienced radiological technologist (NT). A 12-lead ECG was recorded after the PCI (QRS score) at a paper speed of 25 mm/s and an amplification of 10 mm/mV. We calculated the 32-point QRS score, which has been validated in patients with AMI and strongly correlates with infarct size and prognosis.⁹ The patients were followed up for up to 10 years (mean follow up, 91 months) at regular visits to their attending physicians or by telephone interview. The primary endpoint was a major adverse cardiac event (MACE). MACE was defined as the incidence of cardiac death or hospitalization due to heart failure (HF). The study protocol was approved by our institute’s ethics committee (B200900018) and complied with the provisions of the Declaration of Helsinki. The ethics committee waived the requirement for individual informed consent. This was a retrospective study and patients could opt out of the database if they wished.

Echocardiography

All patients were examined by using standard precordial 2-dimensional and Doppler echocardiography while positioned in the left lateral or supine position due to their condition. Echocardiography was performed by experienced cardiologists (N.I. and M.G.) using the iE33 (Philips Medical System, Andover, MA, USA), Vivid q, and VividE9 (GE Vingmed Ultrasound, Horten, Norway) ultrasound systems. Left ventricular ejection fraction (LVEF) was calculated using the biplane modified Simpson’s method. LV mass and LA volume were calculated using the area-length method from apical 4- and 2-chamber views and were indexed for body surface area (BSA). These measurements are based on published guidelines.¹⁰ TMF was assessed in the apical 4-chamber view, using pulsed-wave Doppler echocardiography. E- and A-wave peak velocities, deceleration time (Dct-TMF), and duration of the A-wave were measured from the mitral inflow profile. The Dct of pulmonary venous flow was measured (Dct-PVF). The TDI of the mitral annulus was obtained from the apical 4-chamber view, using a 1- to 2-mm sample volumes placed at the septal side by careful use of the spectral pulse Doppler method averaged from 3cardiac cycles. We checked systolic velocity (s’) and early (e’) and late (a’) diastolic velocities of the mitral annulus at the septal site, then E/e’ was calculated. Mitral regurgitation (MR) was graded as absent, slight, mild, moderate, or severe. LA total, passive, and active emptying fractions were measured.¹¹ LA volumes were measured: (1) just before mitral valve opening, at end-systole (maximal LA volume, LAVmax); (2) at the onset of the P wave on electrocardiography (pre-atrial contraction volume, LAVpreA); and (3) at mitral valve closure, at end diastole (minimal LA volume, LAVmin).¹² From these, the following measurements were calculated: LA total emptying volume (LATEV) = LAVmax − LAVmin; and LA total emptying fraction (TEF) = LATEV / LAVmax × 100. Right ventricular indices included fractional area change and tricuspid annular plane systolic excursion (TAPSE). LA reservoir strain and GLS were measured by using 2D-STE. LA strain was calculated from the apical 4-chamber view, as recommended in the American Society of Echocardiography/European Association of Cardiovascular Imaging (ASE/EACVI) consensus report. Strain analysis was performed using 2D Cardiac performance (TomTec imaging systems) by the two experienced cardiologists (N.I. and J.I.) who were blinded to the clinical baseline data. Peak LA reservoir strain was calculated from end diastole in the ECG at the sharp downslope of the strain trace. The endocardial border was manually traced in apical views, delineating a region of interest (ROI), with the lowest width, composed of 6 segments for each view. Then, necessary manual adjustments of the ROI were performed, and the longitudinal strain curves for each segment were generated by the software. GLS was calculated as the average of 4-chamber, 2-chamber and 3-chamber longitudinal strain curves. LA pump strain was calculated after onset of the P wave in the ECG at the sharp downslope of the strain trace.

Supplementary Figure 2 shows a representative case of LA strain analysis. The left panel shows an apical 4-chamber image with a color-coded region of interest for LA strain, and the right panel shows LA strain curves. In patients in which an optimal visualization of LA could not be guaranteed, the breath-hold technique was applied.

Biochemical Markers

Blood samples were obtained at 24 h after onset. Creatinine and plasma B-type natriuretic peptide (BNP) were measured (admission BNP). Renal function was assessed based on the estimated glomerular filtration rate (eGFR). CPK and myocardial band (MB) were obtained on admission, at 3-h intervals during the first 24 h, at 6-h intervals during the next 2 days, and daily until discharge. BNP was measured at 1 week after onset (maximum BNP).

Reproducibility

Intra-observer and inter-observer (N.I. and J.K.) variability in 2D-STE measurements were assessed in 20 randomly selected participants in this study. They are reported as inter-class correlation coefficients (ICCs).

Statistical Analyses

For reasons of uniformity, summary statistics for all continuous variables are therefore presented as medians together with the 25th and 75th percentiles (interquartile range, IQR) or expressed as mean (standard deviation, SD) for continuous variables and frequency (%) for categorical variables. For comparisons of groups divided by the endpoint, either with or without MACE, the Student’s t-test or Mann-Whitney U-test and χ² test were used, as appropriate.

To identify a parameter associated with MACE, we used Cox proportional hazard models. First, we performed univariate analysis with all parameters. Then, we established the multivariable model. We selected the most significant variables from each category below: (1) clinical variables included age, sex, BSA, hypertension; (2) biochemical markers included BNP and eGFR; (3) infarct size, QRS score, anterior MI, multivessel disease (MVD), Killip >1; (4) systolic function markers included s’, LVEF, GLS; (5) LA-related variables included LAV max index, left atrial total empty fraction (LATEF), LA reservoir strain, Dct-PVF, a’; and (6) the other conventional echocardiography variables included E/e’, MR≥moderate, right ventricular fractional area change (RVFAC), TAPSE. Hazard ratio (HR) and their 95% confidence intervals (CI) were calculated. To avoid multicollinearity among parameters, we calculated variance inflation factors (VIFs). We defined the presence of multicollinearity as a VIF ≥10.¹³ In addition, the incremental effect of LA reservoir strain and GLS on the significant indices in the prediction of future MACE was evaluated using net reclassification improvement (NRI), as previously described.¹⁴

To determine the optimal threshold of LA reservoir and GLS for the prediction of the endpoints, receiver operating characteristic (ROC) analysis was applied, and the areas under the curve (AUCs) and 95% CI for LA reservoir and GLS were calculated. Patients were categorized into two groups based on the cut-off values for LA reservoir strain and GLS values determined by the ROC curve. The survival curves for each group were estimated using the Kaplan-Meier method. The log-rank test was used for group comparisons. The multivariate analyses were also conducted for patients with and without LA enlargement (determined by the median value of the LAVmax index), because an enlarged LA is thought to be a robust marker of increased LV filling pressure.

For all analyses, a 2-sided P value <0.05 was considered statistically significant. Statistical analyses were performed using JMP pro 15.0 software (SAS Institute, Cary, NC, USA) and the R packages ‘nricens’ and ‘timeROC’ (R Foundation for Statistical Computing, Vienna, Austria).

Results

Characteristics of Patients

Table 1 shows the characteristics of the 433 patients (90 with MACE and 343 without MACE). Significant differences were observed between the patients with and without MACE in terms of age, BSA, presence of hypertension and MVD, Killip classification, infarct size, BNP levels, and renal function. Significant differences were also found between the patients with and without MACEs for all echocardiographic characteristics except E/A and TMF deceleration time (Table 2).

Table 1. Patient Characteristics According to the Presence of MACE

Category	All patients	MACE (+)	MACE (−)	P value*
Number	433	90	343
Patient background factors
Age (years)	64.5 (56–73)	72 (63–78)	62 (54–71)	<0.0001
Sex male, n (%)	364 (84)	77 (86)	287 (83)	0.66
BSA (m2)	1.71 (1.58–1.83)	1.67 (1.55–1.80)	1.72 (1.60–1.84)	0.03
sBP (mmHg)	110 (100–121)	107 (96–120)	110 (100–120)	0.47
dBP (mmHg)	60 (56–68)	60 (51–64)	60 (56–69)	0.39
HR (beats/min)	72 (66–80)	75 (66–83)	72 (66–80)	0.15
Risk factors
Hypertension (yes)	256 (59)	67 (74)	189 (55)	0.001
Diabetes mellitus (yes)	121 (28)	29 (32)	92 (27)	0.31
Anterior MI (yes)	229 (52)	55 (61)	174 (51)	0.07
MVD, n (%)	144 (34)	44 (51)	100 (30)	0.0003
Killip >1	95 (21)	32 (36)	63 (19)	0.0005
Reperfusion time (min)	135 (95–216)	158 (107–255)	130 (93–210)	0.05
Peak CPK (IU/L)	3,058±2,933	4,130±4,019	2,775±2,486	0.002
Peak CPK-MB (IU/L)	283±248	382±339	258±211	0.001
Biochemical markers
Admission BNP (pg/mL)	26.8 (11.8–78.2)	58.0 (28.3–162.4)	26.5 (11.8–61.9)	<0.0001
Maximum BNP (pg/mL)	91.7 (42.5–204.3)	220.1 (91.1–408.4)	85.6 (42.8–172.6)	<0.0001
eGFR (mL/min/1.73 m2)	63.1 (52.1–77.7)	53.1 (45.1–68.9)	64.9 (54.8–79.5)	<0.0001
MIBI SPECT
Infarct size (IS)	6 (2–18)	11.6 (0.5–53.6)	5.8 (0–15)	0.008
Electrocardiography
QRS score at 24 h	2.9±3.3	4.3±4.5	2.5±2.7	0.0003
Mediation at discharge, n (%)
β-blocker	357 (82)	77 (86)	280 (82)	0.37
ACE inhibitor/ARB	409 (94)	84 (93)	325 (95)	0.6
Statins	423 (98)	88 (98)	335 (98)	0.95

*P values were for patients with vs. without MACE. ACE, angiotensin-converting enzyme; ARB, angiotensin II receptor blocker; BNP, B-type natriuretic peptide; BP, blood pressure; BSA, body surface area; CPK, creatine phosphokinase; eGFR, estimated glomerular filtration rate; HR, heart rate; MACE, adverse cardiac events; MB, myocardial band; MI, myocardial infarction; MIBI, ^{99 m}Tc methoxy-isobutyl-isonitrile; MVD, multivessel disease; SPECT, single photon emission computed tomography.

Table 2. Echocardiography Data Stratified by the Presence of MACE

Category	All patients	MACE (+)	MACE (−)	P value*
Number	433	90	343
Echocardiography
LVEDVI (mL/m2)	61.2 (49.1–73.0)	59.7 (49.9–75.7)	53.7 (43.6–66.3)	0.003
LVESVI (mL/m2)	27.1 (20.0–38.0)	32.4 (24.6–48.1)	26.6 (19.3–36.5)	0.007
LVEF (%)	52 (44–62)	45 (35–61)	53 (45–62)	0.03
LAVI (mL/m2)	39.0 (30.1–48.2)	39.3 (31.3–51.5)	33.1 (26.8–44.1)	0.02
E/A	0.85 (0.69–1.13)	0.89 (0.68–1.31)	0.84 (0.68–1.1)	0.23
Dct-TMF (ms)	204 (169–254)	187 (155–256)	204 (173–253)	0.14
Dct-PVF (ms)	193 (137–244)	145 (111–208)	204 (157–250)	0.0001
s’ (cm/s)	6.7 (5.8–7.3)	5.9 (4.5–7.2)	7.0 (5.6–8.0)	<0.0001
e’ (cm/s)	5.7 (4.4–6.9)	4.8 (4.0–6.0)	5.9 (4.6–7.0)	<0.0001
a’ (cm/s)	9.4 (7.9–10.8)	7.0 (6.0–8.9)	9.9 (7.1–12.6)	<0.0001
E/e’	12.5 (9.7–15.6)	16.1 (12.7–19.9)	11.8 (9.3–14.7)	<0.0001
MR moderate/severe	38 (8)	22 (24)	16 (5)	<0.0001
LA-related indices
LAV max index (mL/m2)	30.1 (22.3–35.6)	31.3 (23.8–39.9)	27.3 (21.9–34.9)	<0.0001
LAV preA index (mL/m2)	19.3 (15.3–24.6)	21.2 (16.8–30.6)	19.0 (15.1–23.9)	<0.0001
LAV min index (mL/m2)	12.8 (9.8–16.9)	14.5 (10.8–21.8)	12.5 (9.4–15.8)	<0.0001
LA total fraction (%)	54 (46–60)	48 (39–56)	55 (49–60)	<0.0001
RV-related indices
RV-FAC (%)	42 (38–45)	37 (35–41)	44 (40–47)	<0.0001
TAPSE (cm)	17 (15–18)	15 (13–18)	18 (15–18)	<0.0001
Speckle tracking
GLS (%)	−13.0 (−15.0 to −11.0)	−10.4 (−12.1 to −8.6)	−13.8 (−15.5 to −12.0)	<0.0001
LA reservoir strain (%)	28.6 (23.6–35.8)	20.3 (18.6–25.4)	30.3 (18.6–36.1)	<0.0001
LA pump strain (%)	13.8 (12.4–15.1)	11.4 (9.4–14.1)	14.2 (12.8–15.2)	<0.0001

*P values were for patients with vs. without MACE. A, late diastolic; a’, late diastolic velocities; Dct, deceleration time; E, early diastolic wave velocity; e’, early diastolic velocity of mitral annulus; EDVI, end-diastolic volume index; EF, ejection fraction; ESVI, end-systolic volume index; FAC, fractional area change; GLS, global longitudinal strain; LA, left atrium; LAV, left atrial volume; LAVI, left atrial volume index; LV, left ventricular; MACE, major adverse cardiac events; max, maximum volume; min, minimum volume; MR, mitral regurgitation; preA, pre-atrial contraction volume; PVF, pulmonary venous flow; RV, right ventricle; s’, systolic velocity; TAPSE, tricuspid annular plane systolic excursion; TMF, transmitral flow.

Predictors of Outcome

During follow up (median=91 months), 90 patients (21%) of the 433 patients experienced MACEs (cardiac death, n=23 patients; HF hospitalization, n=67 patients).

Univariate analysis revealed significant associations between MACE and several factors, including age, infarct size, anterior MI, LVEF, LAVmax index, E/e’, GLS and LA reservoir strain (Table 3; see Supplementary Table A for LAVmax index; median value=27.1 mL/m²). The multivariable models revealed significant associations between MACE and the variables of age, male gender, GLS, and LA reservoir strain (Table 4). Multivariate analysis of all the patients (n=433) and the multivariate analyses according to LAVmax index divided by the median value (27.1 mL/m²) revealed the strongest association with the LA reservoir strain, especially in patients with a small LA. GLS was the strongest in patients with a large LA. We then performed a multivariate Cox hazard proportional analysis of the LA-related indices (LAVmax index, LATEF, LA reservoir strain, Dct-PVF and a’) to predict MACE. LA reservoir strain was a stronger predictor compared with other indices (HR=0.86, 95% CI; 0.82–0.91, P<0.0001) (Supplementary Table B). Table 5 shows the incremental effect of adding LA reservoir strain or GLS to the base model, which included age, gender, anterior MI, infarct size, E/e’, and LAVmax index, which were the strongest indicators in each category. LA reservoir strain and GLS were both significant predictors of MACE (NRIs and 95% CI: 0.24 [0.11–0.40] and 0.32 [0.13–0.65], respectively).

Table 3. Univariate Cox Hazard Proportional Analysis to Predict MACE

	HR	95% CI	P value
Clinical
Age (years)	1.06	1.04–1.09	<0.0001
Gender, male (yes)	1.04	0.57–1.91	0.88
HT	2.05	1.28–3.29	0.001
Biochemical
Maximum BNP	1.002	1.001–1.003	<0.0001
eGFR	0.98	0.97–0.99	0.007
Infarction
Infarct size (%)	1.04	1.02–1.06	<0.0001
QRS score	1.13	1.07–1.19	<0.0001
MVD	2.37	1.54–3.61	<0.0001
Anterior MI (yes)	1.46	0.95–2.25	0.08
Systolic
s’	0.71	0.54–0.92	0.008
LVEF (%)	0.98	0.96–0.99	0.02
GLS (%)	1.43	1.32–1.54	<0.0001
LA related
LAV max index (mL/m2)	1.01	0.99–1.03	0.21
LATEF	0.94	0.91–0.97	0.002
LA reservoir strain (%)	0.84	0.80–0.89	<0.0001
Dct-PVF	0.99	0.98–0.998	0.01
a’	0.65	0.59–0.71	<0.0001
Conventional
E/e’	1.09	1.07–1.12	<0.0001
MR=moderate/severe	5.55	3.41–9.02	<0.0001
RV FAC	0.83	0.77–0.89	0.0001
TAPSE	0.87	0.82–0.94	0.001

LATEF, left atrial total empty fraction. Other abbreviations as in Tables 1,2.

Table 4. Multiple Cox Hazard Proportional Analysis to Predict MACE

	All patients (n=433)			Large LA (n=217)			Small LA (n=216)
	HR	95% CI	P value	HR	95% CI	P value	HR	95% CI	P value
Age (years)	1.03	1.01–1.06	0.006	1.04	1.01–1.08	0.004	1.01	0.97–1.05	0.57
Maximum BNP (pg/mL)	1.01	1.00–1.01	0.02	1.01	0.99–1.01	0.19	1.01	1.00–1.01	0.03
QRS score	1.01	0.95–1.08	0.68	1.01	0.91–1.09	0.99	1.02	0.91–1.13	0.68
GLS (%)	1.21	1.07–1.35	0.0001	1.23	1.06–1.45	0.001	1.18	1.01–1.39	0.02
LA reservoir strain (%)	0.93	0.86–0.98	0.01	0.94	0.88–0.99	0.04	0.89	0.82–0.97	0.001
E/e’	1.01	0.97–1.04	0.71	0.99	0.95–1.05	0.94	1.02	0.98–1.07	0.35

CI, confidence interval; HR, hazard ratio. Other abbreviations as in Tables 1,2.

Table 5. Incremental Effect of Adding LA Reservoir Strain and GLS to the Base Model

	NRI	95% CI
Indices
Base model + LA reservoir strain	0.24	0.11–0.40
Base model + GLS	0.32	0.18–0.65

Base model includes age, gender, anterior MI, infarct size, E/e’, LAVmax index. CI, confidence interval; NRI, net reclassification improvement. Other abbreviations as in Tables 1,2.

Figure 1 shows the ROC curves for predicting MACE. The optimal cut-off value for LA reservoir strain (A) was 25.8% (sensitivity, 0.73; 1-specificity, 0.48; and AUC, 0.80 [95% CI: 0.78–0.84]; P<0.0001) and the optimal cut-off value for GLS (B) was −11.5% (sensitivity, 0.78; 1-specificity, 0.58; and AUC, 0.82 [95% CI: 0.75–0.87]; P<0.0001).

Figure 1.

Receiver operating characteristic curves (ROC) for predicting MACE. (A) ROC curve of LA reservoir strain, and the optimal cut-off value was 25.8 (sensitivity, 0.73; 1-specificity, 0.48; and AUC, 0.80 [95% CI 0.78–0.84]; P<0.0001). (B) ROC curve of GLS, and the optimal cut-off value was −11.5 (sensitivity, 0.78; 1-specificity, 0.58; and AUC, 0.82 [95% CI 0.75–0.87]; P<0.0001). AUC, area under the curve; LA, left atrial; MACE, major adverse cardiovascular event.

Figure 2 shows the Kaplan-Meier curves for two groups categorized by the cut-off values for the LA reservoir and GLS, determined by the ROC curves shown in Figure 1. Figure 2A shows the Kaplan-Meier curves for the prediction of MACE using LA reservoir strain. A significant difference was found between the curves for the patients with LA reservoir strains ≤25.8% and >25.8% (Log rank, χ²=76.7; P<0.0001). Figure 2B shows the curves using the cut-off value of GLS=−11.5. A significant difference was observed between the curves for the patients with GLS ≤−11.5 and >11.5 (Log rank, χ²=108.8; P<0.0001).

Figure 2.

Kaplan-Meier curves for predicting MACE. These figures show the Kaplan-Meier curves for 2 groups categorized by the cut-off values determined by the ROC curves. (A) Kaplan-Meier curves for the prediction of MACE using LA reservoir strain. There was a significant difference between the curves for the patients with LA reservoir strains ≤25.8 and >25.8 (Log rank, χ²=76.7; P<0.0001). (B) Kaplan-Meier curves using the cut-off value of GLS=−11.5. There was a significant difference between the curves for the patients with GLS ≤−11.5 and >11.5 (Log rank, χ²=108.8; P<0.0001). (C) Kaplan-Meier curves according to the 4 groups: group A (LA reservoir strain ≥25.8, GLS ≤−11.5, n=237); group B (LA reservoir strain <25.8 and GLS ≤−11.5, n=57), group C (LA reservoir strain ≥25.8 and GLS >−11.5, n=38) and group D (LA reservoir strain <25.8 and GLS >−11.5, n=101). There were significant differences among 4 groups (Log rank χ²=126.3, P<0.0001). GLS, global longitudinal strain; LA, left atrial; MACE, major adverse cardiovascular event.

The patients were then divided into the following 4 groups using the cut-off values for both LA reservoir strain=25.8% and GLS=−11.5% (Figure 3): group A (LA reservoir strain ≥25.8%, GLS ≤−11.5%, n=237), group B (LA reservoir strain <25.8% and GLS ≤−11.5%, n=57), group C (LA reservoir strain ≥25.8% and GLS >−11.5%, n=38), and group D (LA reservoir strain <25.8% and GLS >−11.5%, n=101). Significant differences were detected in the frequency of MACE among the 4 groups (Pearson; χ² =107.2, P<0.0001). Figure 2C shows the Kaplan-Meier curves according to the 4 groups (group A: orange line, group B: blue line, group C: green line, group D: red line); significant differences were observed among the 4 groups (Log rank; χ²=126.3, P<0.0001). Pairwise model analysis revealed that group D had the worst outcome of all the groups (P<0.0001). Groups B and C had significantly poorer outcomes than were observed in group A (P=0.04 and P<0.0001, respectively). Significant differences were noted in the frequency of MACE among the 4 groups, except between groups B and C. These results suggested that patients with lower absolute values for both LA reservoir strain and GLS were at a high risk for MACE. We explored the patients’ characteristics and echocardiographic characteristics according to the 4 groups (Supplementary Table C,D). Then, we found that the patients in group B had a higher incidence of hypertension and diabetes and a relatively larger infarct size. We think these facts may affect the differences between groups A and B; however, we think further surveys are needed to confirm this issue.

Figure 3.

Plots for left atrial (LA) reservoir strain and global longitudinal strain (GLS). The red plot indicates patients with MACE and the blue plot indicates patients without MACE. The patients were divided into 4 groups using cut-off values: group A (LA reservoir strain ≥25.8, GLS ≤−11.5, n=237); group B (LA reservoir strain <25.8 and GLS ≤−11.5, n=57); group C (LA reservoir strain ≥25.8 and GLS >−11.5, n=38) and group D (LA reservoir strain <25.8 and GLS >−11.5, n=101). There were significant differences in the frequency of MACE among the 4 groups (Pearson, χ²=107.2, P<0.0001).

Reproducibility

The ICC values for intra-observer variability for GLS and LA reservoir strain were 0.91 and 0.90, respectively. The ICC values for overall inter-observer variability were 0.90, and 0.88, respectively.

Discussion

The findings presented from this study, complete with long-term follow-up data, identified the LA reservoir strain immediately after the onset of first STEMI as a strong predictor of MACE. To the best of our knowledge, this study is the first to examine long-term prognosis based on the LA reservoir strain immediately obtained after the onset of STEMI in patients with a first STEMI. The LA reservoir strain value had an incremental effect on the conventional predictor (age, gender, anterior MI, infarct size, E/e’, and LAVmax index). The LA reservoir strain was also a strong predictor when combined with GLS; therefore, we recommend examining both of these strain values in one examination at the bedside.

Currently, measuring the peak LA reservoir strain is considered a reliable method for the evaluation of LA function and is viewed as applicable to patients with AMI.^6,15 However, no long-term follow-up data have been presented exclusively on patients with STEMI. The current study demonstrated, with long-term follow-up data, that LA reservoir strain was a strong predictor of MACE.

In this study, we performed both conventional echocardiography and 2D-STE at 24 h after STEMI onset. Strain analysis may enable the precise evaluation of LV damage and function and allow prediction of the long-term prognosis in these patients. We also reduced the effects of confounding factors by studying only patients who had undergone their first STEMI with reperfusion within 12 h of symptom onset. Furthermore, we measured and analyzed the infarct size using ^{99 m}Tc-sestamibi SPECT to obtain a precise infarct size.¹⁶ Thus, we believe that our data have demonstrated the use of LA strain as a prognostic parameter for patients with STEMI.

We demonstrated that LA reservoir dysfunction immediately after the onset of STEMI was a strong predictor of MACE. This could indicate that strain imaging can measure cardiac function more accurately than conventional LA imaging.¹⁷ Furthermore, the LA reservoir function determined by 2D-STE has been reported as a prognostic marker in patients with various heart diseases; hence, it is reasonable to consider 2D-STE an ideal method.^18–20 Recently, Inoue et al reported the clinical usefulness of the LA reservoir strain in predicting the LV filling pressure by catheterization analysis.⁵ One study on LA reservoir strain did not include significant adjunct information regarding the patients with AMI;⁶ however, we have shown the clinical usefulness of the LA strain, which may represent an elevation of the filling pressure due to the sudden onset of a first-time STEMI. We think this finding reflects our inclusion of only first-time STEMI cases in our study. Acute MI causes a sudden increase in the LV filling pressure, leading to higher atrial pressure before the occurrence of LA enlargement. A difference in the hemodynamic status of patients with non-STEMI compared with STEMI is therefore plausible. Tan et al also reported LA reservoir strain as a sensitive parameter for estimating LV filling pressure in patients with HF and preserved EF.²⁰

We assume that the patients required further treatment for the deterioration of their LV function due to STEMI damage. As a consequence of decreased LV compliance, LA pressure rises, increasing LA wall tension and stretching the atrial myocardium. There is a close interaction between LA and LV function in each phase of LA function; therefore, an enlarged LA is a robust marker of increased LV filling pressure.^2,6 Adaptive LA functional changes become evident with worsening LV systolic and diastolic function, which may help to characterize the elevation of LV filling pressure.^21,22 Among the patients with preserved GLS, we could detect the high-risk patients by the value of LA reservoir strain, even having preserved GLS. We think the timing of the deterioration of LA function is different from that of LV function. Multivariate analyses show that LA strains are useful, especially in patients without LA enlargement; hence, we consider that a deterioration in the LA strain suggests that the current myocardial damage has damaged the LA (Table 4). Recently, peak LA reservoir strain has been reported as a predictor in patients with acute HF, regardless of age, LAVmax index, E/e’, or LVEF.⁷ Therefore, we believe that our study makes a significant contribution to the literature in the contemporary PCI era regarding the clinical implications of LA strain after the onset of STEMI.

Chu et al reported that a combination of LA strain and GLS could predict LV remodeling, but clinical outcomes may not be superior to those determined using a single indicator.²³ We think that this difference may mainly reflect differences between their study population and our population in terms of the patient number, reperfusion time, and length of follow-up periods. Previous guidelines have recommended GLS as a reproducible index that can be used to identify the improvement or worsening of patients’ conditions.²⁴ As we demonstrated previously, GLS has prognostic value, as it indicates the infarct size after MI.²⁵ However, we showed that LA reservoir strain can be a significant predictor even in the multivariate analysis with GLS. We think LA reservoir strain is a strong predictor even adjustment by GLS. Actually, LVEF and GLS are important predictors to evaluate cardiac function.²⁶ Both LVEF and strain are sensitive to myocardial contractility, although both are known to be load-dependent. However, GLS is less dependent on afterload in the failing heart; therefore, it can provide a more accurate reflection of contractility than LVEF in a situation associated with exposure to acute pressure changes, such as AMI.²⁷ Moreover, for this reason, we feel that the entire left-side cardiac function can be estimated using the combination of LA reservoir function and GLS. Moreover, LA reservoir strain enables us to estimate LA function precisely.

TDI imaging after onset of STEMI is a simple technique and determines strong predictors, however, we think our study has merit. Because strain analysis requires standardization and validation, and there is inter-vendor variability in strain imaging,²⁸ our results would be able to provide useful information about strain analysis. Multivariate analyses revealed the strongest association between LA reservoir strain and MACE, especially in patients with small LA. We believe this is due to the fact that LA function is more sensitive to detect cardiac dysfunction compared with LA volume.

Clinical Implication

Our study has shown the clinical usefulness of LA reservoir strain immediately after reperfusion therapy in patients with STEMI. Furthermore, LA reservoir strain had a statistically significant incremental effect on the conventional predictor. Moreover, we can estimate the prognosis of STEMI accurately by using a combination of LA reservoir strain and GLS. All examinations could be performed using 2D-STE in a one-image measurement. Thus, patients with STEMI can be effectively assessed by using bedside echocardiographic examinations. We can administer a range of cardioprotective drugs to prevent adverse events.²⁶

Study Limitations

This study had many limitations. First, this was a single-center retrospective observational study with a small sample size. Prospective observational studies about the usefulness of the LA reservoir strain with a larger sample are warranted. Second, chronic AF cases were excluded from this study; however, they were high-risk cases. Third, 2D measurements of LA size and function are not always as precise 3D analysis. We did not investigate 3D speckle tracking for LA evaluation; however, Tsujiuchi et al reported the clinical usefulness of 3D speckle tracking echocardiography (STE) to calculate the function of LA.²⁹ Fourth, in this study, the infarct size was estimated by using SPECT and CPK, not cardiac magnetic resonance imaging and troponin. Nonetheless, this study revealed the clinical usefulness of LA strain parameters in patients with STEMI immediately after PCI, which may lead to a better prognosis.

Conclusions

This 10-year follow-up study concluded that LA strain parameters obtained immediately after onset were strong predictor in patients with a first-time STEMI. LA reservoir strain was an especially strong predictor when combined with GLS. Furthermore, LA reservoir strain had an incremental effect on conventional strong parameters (age, gender, anterior MI, infarct size, E/e’). Further studies are required to confirm the role of strain imaging in STEMI.

Sources of Funding

This study did not receive any specific funding.

Disclosures

K.K., M.K. are editorial members of Circulation Journal.

Author Contributions

All authors read and approved the final manuscript.

IRB Information

This study was approved by the Yokohama City University, Center for Novel and Exploratory Clinical Trials (Reference number: B200900018, Clinical trial registration: UMIN000041995).

Data Availability

1. The individual deidentified participant data will be shared.

2. All analyzable data sets related to the study will be shared.

3. The study protocol and statistical analysis plan will be available.

4. The data will be available immediately following publication, and ending 10 years after the publication date.

5. Anyone who wishes to access the data should contact the corresponding author.

6. The data will be shared as Excel files via E-mail.

Supplementary Files

Please find supplementary file(s);

http://dx.doi.org/10.1253/circj.CJ-21-0907

References

• 1.   Iwahashi N, Kimura K, Kosuge M, Tsukahara K, Hibi K, Ebina T, et al. E/e’ two weeks after onset is a powerful predictor of cardiac death and heart failure in patients with a first-time ST elevation acute myocardial infarction. J Am Soc Echocardiogr 2012; 25: 1290–1298.
• 2.   Moller JE, Hillis GS, Oh JK, Seward JB, Reeder GS, Wright RS, et al. Left atrial volume: A powerful predictor of survival after acute myocardial infarction. Circulation 2003; 107: 2207–2212.
• 3.   Iwahashi N, Gohbara M, Abe T, Kirigaya J, Horii M, Hanajima Y, et al. Clinical significance of late diastolic tissue Doppler velocity at 24 hours or 14 days after onset of ST-elevation acute myocardial infarction. Circ Rep 2021; 3: 396–404.
• 4.   Cimino S, Canali E, Petronilli V, Cicogna F, De Luca L, Francone M, et al. Global and regional longitudinal strain assessed by two-dimensional speckle tracking echocardiography identifies early myocardial dysfunction and transmural extent of myocardial scar in patients with acute ST elevation myocardial infarction and relatively preserved LV function. Eur Heart J Cardiovasc Imaging 2013; 14: 805–811.
• 5.   Inoue K, Khan FH, Remme EW, Ohte N, García-Izquierdo E, Chetrit M, et al. Determinants of left atrial reservoir and pump strain and use of atrial strain for evaluation of left ventricular filling pressure. Eur Heart J Cardiovasc Imaging 2021; 23: 61–70.
• 6.   Ersbøll M, Andersen MJ, Valeur N, Mogensen UM, Waziri H, Møller JE, et al. The prognostic value of left atrial peak reservoir strain in acute myocardial infarction is dependent on left ventricular longitudinal function and left atrial size. Circ Cardiovasc Imaging 2013; 6: 26–33.
• 7.   Park JH, Hwang IC, Park JJ, Park JB, Cho GY. Prognostic power of left atrial strain in patients with acute heart failure. Eur Heart J Cardiovasc Imaging 2021; 22: 210–219.
• 8.   Yuda S. Current clinical applications of speckle tracking echocardiography for assessment of left atrial function. J Echocardiogr 2021; 19: 129–140.
• 9.   Iwahashi N, Gohbara M, Kirigaya J, Abe T, Horii M, Takahashi H, et al. Prognostic significance of a combination of QRS score and E/e’ obtained 2 weeks after the onset of ST-elevation myocardial infarction. Circ J 2020; 84: 1965–1973.
• 10.   Lang RM, Badano LP, Mor-Avi V, Afilalo J, Armstrong A, Ernande L, et al. Recommendations for cardiac chamber quantification by echocardiography in adults: An update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging. Eur Heart J Cardiovasc Imaging 2015; 16: 233–270.
• 11.   Leung DY, Boyd A, Ng AA, Chi C, Thomas L. Echocardiographic evaluation of left atrial size and function: Current understanding, pathophysiologic correlates, and prognostic implications. Am Heart J 2008; 156: 1056–1064.
• 12.   Blume GG, McLeod CJ, Barnes ME, Seward JB, Pellikka PA, Bastiansen PM, et al. Left atrial function: Physiology, assessment, and clinical implications. Eur J Echocardiogr 2011; 12: 421–430.
• 13.   Katz MH. Multivariable analysis: A practical guide for clinicians and public health researchers, Third Edition. Cambridge University Press: Milton Keynes, UK, 2011; 90–91.
• 14.   Iwahashi N, Kirigaya J, Gohbara M, Abe T, Horii M, Hanajima Y, et al. Global strain measured by three-dimensional speckle tracking echocardiography is a useful predictor for 10-year prognosis after a first ST-elevation acute myocardial infarction. Circ J 2021; 85: 1735–1743.
• 15.   Fernandes RM, Le Bihan D, Vilela AA, Barretto RBM, Santos ES, Assef JE, et al. Association between left atrial strain and left ventricular diastolic function in patients with acute coronary syndrome. J Echocardiogr 2019; 17: 138–146.
• 16.   Almeida AG, Carpenter JP, Cameli M, Donal E, Dweck MR, Flachskampf FA, et al. Multimodality imaging of myocardial viability: An expert consensus document from the European Association of Cardiovascular Imaging (EACVI). Eur Heart J Cardiovasc Imaging 2021; 22: e97–e125.
• 17.   Antoni ML, Ten Brinke EA, Marsan NA, Atary JZ, Holman ER, van der Wall EE, et al. Comprehensive assessment of changes in left atrial volumes and function after ST-segment elevation acute myocardial infarction: Role of two-dimensional speckle-tracking strain imaging. J Am Soc Echocardiogr 2011; 24: 1126–1133.
• 18.   Deferm S, Martens P, Verbrugge FH, Bertrand PB, Dauw J, Verhaert D, et al. LA mechanics in decompensated heart failure: Insights from strain echocardiography with invasive hemodynamics. JACC Cardiovasc Imaging 2020; 13: 1107–1115.
• 19.   Reddy YNV, Obokata M, Verbrugge FH, Lin G, Borlaug BA. Atrial dysfunction in patients with heart failure with preserved ejection fraction and atrial fibrillation. J Am Coll Cardiol 2020; 76: 1051–1064.
• 20.   Tan TS, Akbulut IM, Demirtola AI, Serifler NT, Ozyuncu N, Esenboga K, et al. LA reservoir strain: A sensitive parameter for estimating LV filling pressure in patients with preserved EF. Int J Cardiovasc Imaging 2021; 37: 2707–2716.
• 21.   Hoit BD. Left atrial size and function: Role in prognosis. J Am Coll Cardiol 2014; 63: 493–505.
• 22.   Matsuda Y, Toma Y, Ogawa H, Matsuzaki M, Katayama K, Fujii T, et al. Importance of left atrial function in patients with myocardial infarction. Circulation 1983; 67: 566–571.
• 23.   Chu AA, Wu TT, Zhang L, Zhang Z. The prognostic value of left atrial and left ventricular strain in patients after ST-segment elevation myocardial infarction treated with primary percutaneous coronary intervention. Cardiol J 2021; 28: 678–689.
• 24.   Galderisi M, Cosyns B, Edvardsen T, Cardim N, Delgado V, Di Salvo G, et al. Standardization of adult transthoracic echocardiography reporting in agreement with recent chamber quantification, diastolic function, and heart valve disease recommendations: An expert consensus document of the European Association of Cardiovascular Imaging. Eur Heart J Cardiovasc Imaging 2017; 18: 1301–1310.
• 25.   Iwahashi N, Kirigaya J, Abe T, Horii M, Toya N, Hanajima Y, et al. Impact of three-dimensional global longitudinal strain for patients with acute myocardial infarction. Eur Heart J Cardiovasc Imaging 2020; 22: 1413–1424.
• 26.   Tsutsui H, Ide T, Ito H, Kihara Y, Kinugawa K, Kinugawa S, et al. JCS/JHFS 2021 Guideline focused update on diagnosis and treatment of acute and chronic heart failure. J Card Fail 2021; 27: 1404–1444.
• 27.   Lumens J, Prinzen FW, Delhaas T. Longitudinal strain: “Think Globally, Track Locally”. JACC Cardiovasc Imaging 2015; 8: 1360–1363.
• 28.   Tschöpe C, Kasner M. Can speckle-tracking imaging improve the reliability of echocardiographic parameters for outcome evaluation in clinical trials? Eur Heart J 2014; 35: 605–607.
• 29.   Tsujiuchi M, Ebato M, Maezawa H, Ikeda N, Mizukami T, Nagumo S, et al. The prognostic value of left atrial reservoir functional indices measured by three-dimensional speckle-tracking echocardiography for major cardiovascular events. Circ J 2021; 85: 631–639.