Abstract
Background: Patients with atrial fibrillation (AF) and heart failure (HF) have elevated left ventricular end-diastolic pressure in addition to decreased left atrial (LA) function, but there are few reports of useful prognostic indices that can be seen on echocardiography. In this study, we investigated the association between LA reservoir strain (LARS) and prognosis in this group of patients.
Methods and Results: We retrospectively enrolled patients with acute HF complicated by AF who were consecutively admitted to hospital between January 2014 and December 2018. A total of 320 patients (mean age 79±12 years, 163 women) were included in the analysis. During a median follow-up of 473 days, 92 cardiovascular deaths and 113 all-cause deaths occurred. In the multivariate analysis, LARS was an independent predictor of all-cause death (hazard ratio [HR] 0.94, 95% confidence interval [CI] 0.90–0.99, P=0.016). Multivariate analysis also showed that the patients in the lowest LARS tertile (<7.16%) had a significantly increased risk of cardiovascular death (HR 1.76, 95% CI 1.05–2.96; P=0.033) and all-cause death (HR 1.90, 95% CI 1.17–3.08; P=0.009) in comparison with patients in the highest LARS tertile (>10.52%).
Conclusions: We found a significant association between LARS and death in patients with AF and HF. Patients with reduced LARS had poor prognosis, suggesting the need for aggressive therapy to improve their LA dysfunction.
Atrial fibrillation (AF) and heart failure (HF) have synergistic negative effects by increasing left atrial (LA) pressure and decreasing cardiac output.1 Furthermore, 40% of patients with AF develop HF and vice versa, and mortality rates increase 2- to 3-fold in those with both conditions.2 Therefore, early screening to identify patients at high risk of death is essential. It has been reported that left ventricular (LV) systolic function does not affect prognosis in patients with AF and HF.3 The ratio of early diastolic peak flow velocity to early diastolic peak annular velocity (E/e’) correlates with LV end-diastolic pressure (LVEDP) and has been reported as a prognostic indicator in patients who have AF and HF with preserved ejection fraction (HFpEF).4 However, its usefulness is limited because tissue Doppler velocity is subject to error due to angle dependence and the effects of cardiac motion and tethering.5 Therefore, more useful prognostic indices are needed for patients with AF and HF.
Unlike conventional echocardiography, which evaluates blood flow and volume,6 2D speckle-tracking echocardiography can quantify myocardial deformation by measuring strain. The LA reservoir strain (LARS) has been reported as accurately assessing LA function by quantifying deformation of the LA during filling,7 and thus a useful indicator of diastolic dysfunction because it reflects the LVEDP.8 Furthermore, reduced LARS is associated with progressive structural and tissue remodeling9 and is reportedly an independent prognostic factor not only in various cardiac diseases but also in the general population.10–13 Patients with AF and HF, on the other hand, have reduced LA function and increased LVEDP, so their LARS is generally lower and less likely to differ among patients. Thus, LARS may have no prognostic impact in this population.14 Given that some patients hospitalized for acute HF have more progressive LA dysfunction and poor prognosis, in this study we sought to determine whether LARS is a prognostic factor in such patients.
Methods
Patients
The study had a retrospective observational design, and the patient information analyzed was based on the dataset used in our previously reported study,15 with the addition of further data from the electronic medical records. Briefly, we enrolled patients aged ≥18 years with acute HF complicated by AF who were consecutively admitted to the Department of Cardiology at the Center Hospital of the National Center for Global Health and Medicine (Tokyo, Japan) between January 2014 and December 2018. We excluded patients with acute coronary syndromes that could be complicated by AF in the acute phase, those for whom no transthoracic echocardiography was performed within 30 days before or after admission, and those with missing data. All patients were treated during hospitalization according to the then current Japanese Society of Cardiology guidelines for the treatment of acute HF16 and received extended follow-up from the date of admission until death or last contact or until March 2022. Follow-up events recorded included pulmonary vein isolation with catheter ablation, and cardiovascular and all-cause deaths.
The study was approved by the institutional review board of the National Center for Global Health and Medicine (NCGM-S-004598-00) and conducted in accordance with the Declaration of Helsinki. The requirement for written informed consent was waived by the institutional review board, and patients were given the opportunity to opt out of the study via a notice on the hospital website.
Echocardiography
Transthoracic echocardiography (TTE) was performed in the physiology laboratory by echocardiographers blinded to the clinical data, using an Aplio400®
or Artida®
ultrasound system (Toshiba Medical Systems, Otawara City, Japan) in M-mode and 2D with Doppler according to the protocol recommended by the guidelines.17 LV lumen measurements were obtained by the modified Simpson method in 2- and 4-chamber views, and the LVEF was calculated from the data obtained. LV weight was calculated by the modified Devereux method and divided by body surface area to determine the LV mass index. LA diameter was measured from the long-axis view, and LA volume was measured by the biplane area-length method and divided by body surface area to determine the LA volume index. Peak early diastolic flow velocity was measured by placing a sample volume of pulsed-wave Doppler at the tip of the mitral valve in the 4-chamber view. Peak early diastolic annular velocity was measured by placing a sample volume of pulsed-wave tissue Doppler at the mitral annulus septum in the 4-chamber view. All echocardiographic results were confirmed at the time by the attending cardiologist.
Strain Measurements
LARS was measured offline in the 2- and 4-chamber views and LV global longitudinal strain from the 2-, 3-, and 4-chamber views using AutoStrain in the 2D speckle-tracking software Image Arena®
(TOMTEC Imaging Systems, Unterschleissheim, Germany). We used QRS onset as the zero-reference point as well as the starting point for determining LARS. All segments of the LA and LV were analyzed with a resolution >45 frames. Endocardial boundaries were automatically traced by the software and manually corrected if tracing was inappropriate. Strain waveforms were displayed automatically, and the mean LARS was calculated as an absolute value from the 2- and 4-chamber views and used for analysis, as in previous studies.18 All strain measurements were performed by the same cardiologist (J.Y.) who was blinded to the clinical data.
Definitions
Patients’ characteristics, except for medications at discharge, were based on data collected at admission. Acute HF was defined as a diagnosis made by a cardiologist on admission based on the Framingham criteria.19 AF was defined as a cardiac rhythm with irregular RR intervals, loss of P waves or LA contraction waves, and fibrillatory waves, as confirmed by 12-lead ECG or TTE performed on admission. AF was defined as paroxysmal if it lasted <7 consecutive days, persistent if it lasted ≥7 consecutive days to <1 year, and permanent if it lasted ≥1 year. Cardiovascular death was defined as death from myocardial infarction (MI), HF, stroke (cerebral infarction or hemorrhage), or sudden cardiac death within 24 h of last being seen alive. Coronary artery disease was defined as a history of MI or angina pectoris treated with catheter intervention or bypass surgery. Valvular heart disease was defined as moderate or severe valvular disease according to clinical practice guidelines20 or a history of surgery for valvular disease.
Statistical Analysis
Continuous variables are shown as the mean±standard deviation or the median and 25th/75th percentiles (interquartile range), and categorical variables as the frequency and percentage. Given that the normal LARS value in patients with AF and HF is unknown, the patients were divided into LARS tertiles. Patient variables that were continuous were compared by one-way analysis of variance or the Kruskal-Wallis test, and those that were categorical were compared by Fisher’s exact test. Clinical factors that were significant when analyzed by tertile were examined in 2-group tests with P values corrected by the Bonferroni method. To identify correlation coefficients between LARS values and clinical factors, multiple regression analysis was performed using age, sex, and other clinical factors that were considered important and found to be significant in the single regression analysis. Kaplan-Meier curves for cardiovascular and all-cause deaths were drawn for each LARS tertile, and log-rank tests were performed. To identify independent risk factors for death, age, sex, and other clinical factors that were considered to be important and found to be significant in the univariate Cox analysis were included in the multivariate analysis model. Hazard ratios (HRs) and 95% confidence intervals (CIs) were calculated. The risk of death per tertile was compared by adjusting the multivariate models of the main analysis. Sensitivity analyses were performed by including the CHA2DS2-VASc score or LVEF, LA volume index, and E/e' as echocardiographic values in multivariate models with age, sex, and clinical factors that were significant in the main analysis. An additional sensitivity analysis was performed with patients who had echocardiography performed after admission. To evaluate the interaction with risk of death in the low LARS group with reference to the high LARS group, subgroup analysis was performed by dividing the 2 groups according to sex, paroxysmal AF and oral anticoagulant status, or clinically meaningful or median values for age, body mass index, estimated glomerular filtration rate, B-type natriuretic peptide (BNP), LVEF, and LA volume index. The intra-observer reproducibility of the strain measurements was determined by repeating the measurements in 30 randomly selected patients and calculating the intraclass correlation coefficient. For each strain analysis, the intra-examiner reliability of heart rate in each imaging section was determined by calculating the intraclass correlation coefficient. The statistical analyses were performed using R version 4.1.2 (The R Foundation for Statistical Computing, Vienna, Austria). All P values were two-tailed and considered statistically significant when <0.05.
Results
Patients’ Characteristics
After 68 exclusions, 320 patients (mean age 79±12 years, 163 women) were included in the analysis (Figure 1). The median duration of AF was 1.2 years; 11% (36/320) had paroxysmal AF and 48% (152/320) had permanent AF. The median BNP value was 623 pg/mL, and 52% of the patients (165/320) were in New York Heart Association (NYHA) functional class 4. Table 1 shows the patients’ characteristics according to LARS tertile (tertile 1, >10.52%; tertile 2, ≤10.52% to ≥7.16%; tertile 3, <7.16%). There were significantly more women and more patients with permanent AF in the low LARS group than in the high LARS group. Moreover, there were significantly fewer patients with chronic obstructive pulmonary disease (COPD) in the low LARS group. There were no significant differences in age, estimated glomerular filtration rate, hemoglobin, NYHA functional class, CHA2DS2-VASc score, or medications between the high and low LARS groups. Table 2 shows the TTE parameters. The median time from admission to echocardiography was 6 days after admission (IQR 3–9 days after admission). The mean heart rate during the echocardiographic examination was 86±20 beats/min. No patients were excluded due to poor image quality. LVEF was significantly lower and LA volume index, E/e', and LV global longitudinal strain were significantly higher in the low LARS group than in the high LARS group. Supplementary Table 1 shows the correlation coefficients between LARS values and clinical factors. Multiple regression analysis showed that LARS was significantly correlated with COPD, LA diameter, LA volume index, E/e', and LV global longitudinal strain.
Table 1. Patients’ Characteristics According to LARS Tertile
| LARS, % |
|
>10.52% |
7.16–10.52% |
<7.16% |
|
| LARS tertile |
Overall
(n=320) |
High
(n=106) |
Intermediate
(n=107) |
Low
(n=107) |
P value |
| Age, years |
79.0±11.7 |
78.9±11.0 |
78.1±12.5 |
79.9±11.5 |
0.50 |
| Female sex, n |
163 (50.9) |
46 (43.4) |
50 (46.7) |
67 (62.6)* |
0.011† |
| Weight, kg |
52.7±15.6 |
52.5±13.7 |
55.5±19.7 |
50.0±12.2# |
0.039† |
| Body mass index |
21.5±4.9 |
21.2±4.1 |
22.3±6.1 |
21.0±4.1 |
0.12 |
| Heart rate, beats/min |
97±30 |
96±30 |
96±30 |
97±29 |
0.96 |
| Blood pressure, mmHg |
| Systolic |
129.3±26.4 |
130.3±23.2 |
131.5±27.8 |
126.1±27.8 |
0.29 |
| Diastolic |
81.9±21.1 |
77.8±21.4 |
83.1±21.5 |
83.3±22.5 |
0.12 |
| NYHA functional class |
|
|
|
|
0.47 |
| II, n |
38 (11.9) |
17 (16.0) |
12 (11.2) |
9 (8.4) |
|
| III, n |
117 (36.6) |
40 (37.7) |
38 (35.5) |
39 (36.4) |
|
| IV, n |
165 (51.6) |
49 (46.2) |
57 (53.3) |
59 (55.1) |
|
| Duration of AF, years |
1.2
(0.0–5.4) |
0.4
(0.0–3.5) |
2.0
(0.0–6.3) |
2.2
(0.0–5.5) |
0.24 |
| Type of AF |
|
|
|
|
0.007† |
| Paroxysmal, n |
36 (11.2) |
16 (15.1) |
14 (13.1) |
6 (5.6) |
|
| Persistent, n |
132 (41.2) |
53 (50.0) |
39 (36.4) |
40 (37.4) |
|
| Permanent, n |
152 (47.5) |
37 (34.9) |
54 (50.5) |
61 (57.0)* |
|
| History of smoking, n |
136 (42.5) |
48 (45.3) |
46 (43.0) |
42 (39.3) |
0.67 |
| eGFR, mL/min/1.73 m2 |
46.4±20.9 |
50.0±21.6 |
45.0±21.0 |
44.1±19.9 |
0.092 |
| Hemoglobin, mg/dL |
12.2±2.3 |
12.1±2.2 |
12.3±2.3 |
12.3±2.5 |
0.77 |
| BNP, pg/mL |
623
(397–1,027) |
532
(340–858) |
661
(447–1,120) |
779
(429–1,143) |
0.004† |
| Coronary artery disease, n |
71 (22.2) |
21 (19.8) |
22 (20.6) |
28 (26.2) |
0.49 |
| Valvular heart disease, n |
101 (31.6) |
41 (38.7) |
27 (25.2) |
33 (30.8) |
0.11 |
| Hypertrophic cardiomyopathy, n |
7 (2.2) |
1 (0.9) |
1 (0.9) |
5 (4.7) |
0.23 |
| Dilated cardiomyopathy, n |
6 (1.9) |
0 (0.0) |
3 (2.8) |
3 (2.8) |
0.25 |
| Comorbidity |
| Hypertension, n |
236 (73.8) |
74 (69.8) |
80 (74.8) |
82 (76.6) |
0.50 |
| Dyslipidemia, n |
103 (32.2) |
28 (26.4) |
40 (37.4) |
35 (32.7) |
0.23 |
| Diabetes, n |
96 (30.0) |
27 (25.5) |
39 (36.4) |
30 (28.0) |
0.20 |
| COPD, n |
24 (7.5) |
14 (13.2) |
6 (5.6) |
4 (3.7)* |
0.023† |
| History of stroke, n |
44 (13.8) |
17 (16.0) |
12 (11.2) |
15 (14.0) |
0.58 |
| CHADS2 score |
3 (2–4) |
3 (2–3) |
3 (2–4) |
3 (2–4) |
0.60 |
| CHA2DS2-VASc score |
5 (4–6) |
4 (4–6) |
5 (4–6) |
5 (4–6) |
0.072 |
| Medication at discharge |
| OAC, n |
269 (84.1) |
88 (83.0) |
88 (82.2) |
93 (86.9) |
0.62 |
| Antiplatelets, n |
90 (28.1) |
31 (29.2) |
25 (23.4) |
34 (31.8) |
0.37 |
| Diuretics, n |
293 (91.6) |
96 (90.6) |
94 (87.9) |
103 (96.3) |
0.065 |
| Loop diuretics, n |
287 (89.7) |
95 (89.6) |
90 (84.1) |
102 (95.3)# |
0.023† |
| MRA, n |
146 (45.6) |
49 (46.2) |
48 (44.9) |
49 (45.8) |
0.99 |
| Thiazides, n |
27 (8.4) |
8 (7.5) |
6 (5.6) |
13 (12.1) |
0.25 |
| Tolvaptan, n |
39 (12.2) |
11 (10.4) |
19 (17.8) |
9 (8.4) |
0.098 |
| β-blockers, n |
235 (73.4) |
77 (72.6) |
80 (74.8) |
78 (72.9) |
0.95 |
| ACEI/ARB, n |
131 (40.9) |
46 (43.4) |
41 (38.3) |
44 (41.1) |
0.77 |
| SGLT2 inhibitors, n |
6 (1.9) |
0 (0.0) |
3 (2.8) |
3 (2.8) |
0.25 |
| Calcium-channel blockers, n |
128 (40.0) |
45 (42.5) |
50 (46.7) |
33 (30.8) |
0.047† |
| Antiarrhythmics, n |
35 (10.9) |
9 (8.5) |
9 (8.4) |
17 (15.9) |
0.16 |
| Statins, n |
89 (27.8) |
25 (23.6) |
35 (32.7) |
29 (27.1) |
0.33 |
Data are presented as the number (percentage), mean±standard deviation, or median (interquartile range). †P<0.05. *P<0.05 vs. tertile 1; #P<0.05 vs. tertile 2. ACEI, angiotensin-converting enzyme inhibitor; AF, atrial fibrillation; ARB, angiotensin receptor blocker; BNP, brain natriuretic peptide; COPD, chronic obstructive pulmonary disease; eGFR, estimated glomerular filtration rate; LARS, left atrial reservoir strain; MRA, mineralocorticoid receptor antagonist; NYHA, New York Heart Association; OAC, oral anticoagulants; SGLT2, sodium-glucose cotransporter 2.
Table 2. Echocardiographic Data According to LARS Tertile
| LARS, % |
|
>10.52% |
7.16–10.52% |
<7.16% |
|
| LARS tertile |
Overall
(n=320) |
Tertile 1
(n=106) |
Tertile 2
(n=107) |
Tertile 3
(n=107) |
P value |
| LVDd, mm |
49.8±9.5 |
49.5±8.5 |
49.6±10.1 |
50.3±9.8 |
0.80 |
| LVDs, mm |
37.9±10.8 |
36.0±8.9 |
38.1±11.4 |
39.6±11.8 |
0.057 |
| LVEDV, mL |
111.7±57.7 |
109.0±49.7 |
113.8±60.0 |
112.3±62.8 |
0.82 |
| LVESV, mL |
65.8±47.1 |
57.8±35.2 |
67.6±47.9 |
71.9±55.3 |
0.082 |
| LVEF, % |
44.9±13.9 |
48.8±12.4 |
44.7±13.6 |
41.4±14.7* |
<0.001† |
| IVSTd, mm |
10.4±2.0 |
10.3±1.7 |
10.3±1.7 |
10.6±2.4 |
0.50 |
| LVPWTd, mm |
10.4±1.6 |
10.3±1.4 |
10.4±1.6 |
10.6±1.9 |
0.33 |
| LVMI, g/m2 |
130.4±42.5 |
126.8±38.2 |
124.3±36.3 |
140.3±50.2# |
0.012† |
| LAD, mm |
49.2±9.2 |
48.6±10.0 |
48.4±7.5 |
50.8±9.7 |
0.11 |
| LAVI, mL/m2 |
89.7±56.1 |
80.1±42.4 |
83.2±39.5 |
105.7±75.8*,# |
0.001† |
| Mitral E velocity, cm/s |
109.1±36.0 |
111.6±36.7 |
106.3±32.6 |
109.4±38.6 |
0.55 |
| DcT, ms |
201.7±80.8 |
201.7±83.9 |
196.6±64.6 |
206.8±91.9 |
0.66 |
| Septal e' velocity, cm/s |
6.0±1.9 |
6.8±2.1 |
5.9±1.7* |
5.3±1.6* |
<0.001† |
| E/e' ratio |
19.7±8.8 |
17.6±7.2 |
19.4±8.3 |
22.1±10.1* |
0.001† |
| LVGLS, % |
−11.8±6.5 |
−14.5±5.8 |
−1.4±6.3* |
−9.1±6.2*,# |
<0.001† |
| LARS, % |
9.6±4.7 |
14.7±4.3 |
8.8±1.0* |
5.4±1.3*,# |
<0.001† |
Data are presented as the number (percentage) or as the mean±standard deviation. †P<0.05. *P<0.05 vs. tertile 1; #P<0.05 vs. tertile 2. DcT, deceleration time; E, early diastolic peak flow velocity; e', early diastolic peak annular velocity; IVSTd, interventricular septal thickness at end-diastole; LAD, left atrial diameter; LARS, left atrial reservoir strain; LAVI, left atrial volume index; LVDd, left ventricular dimension in diastole; LVDs, left ventricular dimension in systole; LVEDV, left ventricular end-diastolic volume; LVEF, left ventricular ejection fraction; LVESV, left ventricular end-systolic volume; LVGLS, left ventricular global longitudinal strain; LVMI, left ventricular mass index; LVPWTd, left ventricular posterior wall thickness at end-diastole.
During a median follow-up of 473 days, 8 patients underwent pulmonary vein isolation by catheter ablation. There were 92 cardiovascular deaths and 113 all-cause deaths, for an in-hospital mortality rate of 5% (16/320). The Supplementary Figure shows the types of cardiovascular deaths that occurred: 2 MI, 58 HF-related deaths, 9 strokes, and 23 sudden cardiac deaths. Figure 2A shows the Kaplan-Meier curves for cardiovascular death according to LARS tertile. At the 36-month follow-up, the survival rate was 71% (95% CI 59–80) in the high LARS group, 80% (95% CI 70–87) in the intermediate LARS group, and 46% (95% CI 33–58) in the low LARS group (P=0.002, log-rank test). Figure 2B shows the Kaplan-Meier curves for all-cause death in the 3 LARS groups at the 36-month follow-up: the survival rate was 68% (95% CI 56–77) in the high LARS group, 69% (95% CI 57–78) in the intermediate LARS group, and 41% (95% CI 29–53) in the low LARS group (P=0.004, log-rank test).
Table 3 shows the results of the Cox analysis of clinical characteristics predicting cardiovascular death. Multivariate analysis adjusted for age, sex, and factors that were significant in univariate analysis showed that LARS was not a significant independent prognostic indicator for cardiovascular death (HR 0.95, 95% CI 0.91–1.01, P=0.089). Hemoglobin was not included in the model because proportional hazards were not established. The risk of cardiovascular death was significantly higher in the low LARS group than in the high LARS group (HR 1.93, 95% CI 1.19–3.13, P=0.008; Figure 3); the results were consistent when adjusted for the multivariate model of Table 3 (HR 1.76, 95% CI 1.05–2.96, P=0.033).
Table 3. Risk Factors for Cardiovascular Death
| Risk factor |
Univariate analysis |
Multivariate analysis |
| HR |
95% CI |
P value |
HR |
95% CI |
P value |
| Age |
1.05† |
1.03–1.07† |
<0.001† |
1.03† |
1.00–1.06† |
0.023† |
| Female sex |
1.27 |
0.84–1.92 |
0.25 |
0.98 |
0.63–1.52 |
0.92 |
| Body mass index |
0.93† |
0.89–0.98† |
0.009† |
0.98 |
0.93–1.04 |
0.51 |
| eGFR |
0.98† |
0.96–0.99† |
<0.001† |
0.99† |
0.97–0.99† |
0.014† |
| Hemoglobin |
0.80† |
0.73–0.88† |
<0.001† |
|
|
|
| BNP (per 100 pg/mL) |
1.05† |
1.03–1.06† |
<0.001† |
1.03† |
1.01–1.05† |
0.007† |
| OAC |
0.24† |
0.14–0.40† |
<0.001† |
0.41† |
0.23–0.73† |
0.003† |
| Antiplatelets |
1.68† |
1.10–2.55† |
0.015† |
1.30 |
0.83–2.03 |
0.25 |
| β-blockers |
0.56† |
0.36–0.87† |
0.009† |
0.80 |
0.50–1.29 |
0.37 |
| ACEI/ARB |
0.87 |
0.57–1.31 |
0.50 |
|
|
|
| MRA |
0.75 |
0.49–1.14 |
0.17 |
|
|
|
| LVEF |
0.99 |
0.98–1.01 |
0.22 |
|
|
|
| LVMI |
1.00 |
0.99–1.01 |
0.061 |
|
|
|
| LAVI |
1.00 |
0.99–1.01 |
0.29 |
|
|
|
| E/e' ratio |
1.01 |
0.99–1.03 |
0.16 |
|
|
|
| LVGLS |
1.03 |
0.99–1.06 |
0.14 |
|
|
|
| LARS |
0.95 |
0.91–1.00 |
0.061 |
0.95 |
0.91–1.01 |
0.089 |
†P<0.05. CI, confidence interval; HR, hazard ratio. Other abbreviations as in Tables 1,2.
Table 4 shows the results of the Cox analysis performed to identify clinical characteristics that predict all-cause death. Multivariate analysis adjusted for age, sex, and factors that were significant in the univariate analysis showed that LARS was a significant independent prognostic indicator of all-cause death (HR 0.94, 95% CI 0.90–0.99, P=0.016). The risk of all-cause death was significantly higher in the low LARS group than in the high LARS group (HR 1.89, 95% CI 1.21–2.96, P=0.006; Figure 3). The results were consistent when adjusted for the multivariate model of Table 4 (HR 1.90, 95% CI 1.17–3.08; P=0.009).
Table 4. Risk Factors for All-Cause Death
| Risk factor |
Univariate analysis |
Multivariate analysis |
| HR |
95% CI |
P value |
HR |
95% CI |
P value |
| Age |
1.05† |
1.03–1.07† |
<0.001† |
1.02 |
0.99–1.05 |
0.078 |
| Female sex |
1.06 |
0.73–1.54 |
0.75 |
0.75 |
0.50–1.11 |
0.15 |
| Body mass index |
0.92† |
0.88–0.97† |
<0.001† |
0.97 |
0.92–1.02 |
0.23 |
| eGFR |
0.97† |
0.97–0.98† |
<0.001† |
0.99† |
0.98–0.99† |
0.010† |
| Hemoglobin |
0.80† |
0.74–0.88† |
<0.001† |
0.87† |
0.79–0.96† |
0.006† |
| BNP (per 100 pg/mL) |
1.05† |
1.03–1.06† |
<0.001† |
1.03† |
1.01–1.05† |
0.005† |
| OAC |
0.23† |
0.14–0.37† |
<0.001† |
0.45† |
0.26–0.77† |
0.004† |
| Antiplatelets |
1.61† |
1.10–2.36† |
0.014† |
1.22 |
0.81–1.82 |
0.34 |
| β-blockers |
0.56† |
0.38–0.84† |
0.005† |
0.87 |
0.56–1.36 |
0.55 |
| ACEI/ARB |
0.86 |
0.59–1.25 |
0.44 |
|
|
|
| MRA |
0.75 |
0.51–1.09 |
0.13 |
|
|
|
| LVEF |
0.99 |
0.98–1.01 |
0.22 |
|
|
|
| LVMI |
1.00 |
0.99–1.01 |
0.18 |
|
|
|
| LAVI |
1.00 |
0.99–1.01 |
0.34 |
|
|
|
| E/e' ratio |
1.00 |
0.98–1.02 |
0.83 |
|
|
|
| LVGLS |
1.02 |
0.99–1.06 |
0.18 |
|
|
|
| LARS |
0.95† |
0.91–0.99† |
0.024† |
0.94† |
0.90–0.99† |
0.016† |
†P<0.05. CI, confidence interval; HR, hazard ratio. Other abbreviations as in Tables 1,2.
Sensitivity analyses were performed for the models predicting all-cause death with addition of the CHA2DS2-VASc score in model 1 and LVEF, LA volume index, and E/e' as echocardiographic values in model 2 (Supplementary Table 2). In both models, the results were consistent with LARS being an independent prognostic factor. We also performed an additional sensitivity analysis with 305 patients who had echocardiography performed after admission, and the results were consistent with LARS being an independent prognostic factor (Supplementary Table 3).
Subgroup Analysis
Figure 4A,B shows the subgroup analyses of the HRs for cardiovascular and all-cause death, respectively, in the low LARS group with reference to the high LARS group. There was no statistically significant interaction between any of the subgroups, though the HR for cardiovascular death was higher in patients aged >80 years, those with BNP <500 pg/mL, those on oral anticoagulants, and those with LVEF ≥50%.
Reproducibility and Reliability of Strain Measurements
With respect to the reproducibility of strain measurements, the intra-observer correlation coefficient was 0.92 (95% CI 0.85–0.96) for LARS and 0.92 (95% CI 0.83–0.96) for LV global longitudinal strain, while that for the reliability of heart rate in each image section was 0.82 (95% CI 0.78–0.85) for LARS and 0.80 (95% CI 0.77–0.83) for LV global longitudinal strain.
Discussion
This study investigated the association between LARS and death in patients with AF and HF. LARS was the only independent predictor of all-cause death, but did not reach statistical significance for cardiovascular death. The mortality rate was significantly higher in the lowest LARS tertile than in the highest LARS tertile, and multivariate analysis showed a significant association with both cardiovascular and all-cause death. No medications other than oral anticoagulants were significantly associated with cardiovascular death, with more than one-third of patients dying after a median follow-up of 1 year. This high mortality rate has also been observed in previous cohort studies,21 suggesting that patients with coexisting AF and HF are a treatment-resistant group with a poor prognosis. To our knowledge, this is the first study to identify LARS as a prognostic factor in patients hospitalized for acute HF with AF.
Relationship Between LARS and Prognosis
LA strain comprises reservoir strain, conduit strain, and contractile strain, and LARS is the parameter most widely used to predict prognosis in patients with HF.18 It has been reported that LARS correlates more strongly with LVEDP than with the LA volume index or E/e'.22 The LA volume index reflects structural remodeling and the cumulative effect of the burden of chronic LVEDP, whereas LARS reflects functional remodeling and is an indicator of LVEDP at the time of measurement.23 Decreased LARS in patients with AF is also associated with LA fibrosis and is thought to be the result of the combined effects of atherosclerotic diseases such as hypertension, diabetes mellitus, and coronary artery disease.24 In one study, the mean LARS in the general population without heart disease was 37.4%,10 but that in the present study it was 9.6%, which is lower than in previous studies of HF and AF,14,24–26 suggesting that progression of LA remodeling was severe. Therefore, because LARS may be a useful index reflecting the severity of acute HF and LA dysfunction, a lower LARS may be a marker or cause of poor prognosis.
Prognostic Value of LARS in Patients With AF and HF
It has been suggested that LARS does not affect prognosis in patients with AF and HF.14 Acute and chronic elevation of LA pressure causes progressive fibrosis of the LA and structural remodeling.1 As the LA dilates, the amount of stretch required to accommodate the reservoir is relatively reduced,27 so LARS is generally lower and may be less likely to differ among patients. Furthermore, a lower LARS may be associated with an increased risk of stroke because of reduced LA function.28 Although all patients in our study were candidates for anticoagulation because of concomitant HF, it is possible that patients who were not prescribed oral anticoagulants had a worse prognosis regardless of their LARS value as a result of clinician selection bias or development of cerebral infarction. These findings suggest that the prognostic impact of LARS is reduced in patients with AF and HF, although the prognostic impact of LARS was unchanged in our sensitivity analysis adjusted for use of oral anticoagulants and the CHA2DS2-VASc score. Furthermore, the prognostic value of LARS was increased in the subgroup of patients using oral anticoagulants, albeit with no statistically significant interaction. Another study found a strong correlation between LARS and quality of life,29 which suggests LARS may be useful for identifying patients with progressive LA dysfunction because of coexisting AF and HF who are at high risk of death. However, in the present subgroups with LVEF ≥50% and BNP <500 pg/mL, the predictive value for cardiovascular death was higher in the low LARS group. Considering that LV systolic dysfunction may increase BNP and worsen prognosis more than diastolic dysfunction, LARS may be a useful prognostic indicator in patients with HFpEF without other useful indicators. Although a previous study indicated that LA dysfunction may predict adverse outcomes in patients with HFpEF and AF,26 the appropriate cutoff value for LARS is still unclear and awaits evaluation in large prospective studies. Other independent predictors of all-cause death include anemia and a low estimated glomerular filtration rate, which have been reported in previous studies of patients with AF.30 Given that low LARS has also been reported to predict poor renal function,31 it is possible that the all-cause mortality rate may have increased in our low LARS group as a result of worsening renal function during follow-up. Finally, LARS was not much lower in the patients with COPD, which has also been noted in previous studies of HFpEF.26 Furthermore, it has been reported that many patients with AF and COPD have an impaired right heart system owing to the effects of pulmonary heart disease, which may reduce right atrial reservoir strain but not LARS or LA volume index.25
Treatments for Patients With AF and HF
Our findings suggest that patients with AF and HF with reduced LARS are at high risk of death and should be considered for active therapeutic intervention. Activation of the renin-angiotensin-aldosterone system in patients with AF and HF may be involved in the pathogenesis of both diseases and contribute to their coexistence.32 There is modest evidence to suggest that angiotensin-converting enzyme inhibitors/angiotensin receptor blockers improve cardiovascular outcomes in this disease group, but reports on mineralocorticoid receptor antagonists are scarce.32 A previous meta-analysis found that β-blockers improved LVEF in patients with LVEF <50% but did not improve prognosis, regardless of LVEF.33 In contrast, treatment with an angiotensin receptor/neprilysin inhibitor (ARNI) or a sodium-glucose cotransporter 2 (SGLT2) inhibitor has been reported to have a positive effect in this group of patients.34 In the PARADIG-HF trial, cardiovascular mortality and HF hospitalization rates were significantly lower in patients who had HFrEF with or without AF who were treated with an ARNI than in those who were treated with enalapril.35 In a secondary analysis of the EMPA-REG OUTCOME trial,36 SGLT2 inhibitors significantly reduced the risks of all-cause and cardiovascular death in patients with type 2 diabetes and cardiovascular disease with or without AF. These results suggest that ARNI and SGLT2 inhibitors should be preferentially considered to improve prognosis in patients with AF and HF. However, when the patients in the present study were admitted, ARNI was not available in Japan and SGLT2 inhibitors were not covered for HF by the Japanese insurance system. Invasive treatments can be considered, including rhythm control, which has recently been reported to improve the long-term prognosis in patients with AF;37 ablation of the atrioventricular node and pacemaker implantation in patients who are refractory to antiarrhythmic drugs or catheter ablation;38 and closure of the LA appendage to prevent thromboembolism in patients who are contraindicated for anticoagulation therapy.39 The indications for these therapies should be considered early in patients with AF and HF in view of their poor prognosis, and careful follow-up is needed for patients who may have already missed the opportunity for treatment.
Study Limitations
There are several limitations to note. First, this was a single-center retrospective study of a small population. Furthermore, there were few pulmonary vein isolation procedures because the hospital did not become a catheter ablation facility until 2018. The differences between our results and those of related studies possibly reflect differences in study population, measurement environment, and treatment interventions. Nevertheless, confounding after discharge weakened the association with prognosis, which reinforces our findings. Second, echocardiography was performed within 30 days before or after admission, and thus in some cases may not reflect values from the acute phase of HF. However, the median time from admission to echocardiography was short, and the results were consistent in the sensitivity analysis of patients who had echocardiography performed after admission. Third, the echocardiographic protocol was not designed for LARS, which made it difficult to obtain accurate measurements in patients with a large LA and poor lateral wall resolution. Moreover, LARS was measured in 1 cardiac cycle, and its measurement in patients with heart rate >100 beats/min is not well established and therefore may not be sufficient for accurate measurement in patients with tachycardia and AF. However, in this study, heart rate was well controlled during the echocardiographic examination, the reliability of the heart rate in each image section was relatively high, and LARS was measured in both 4- and 2-chamber views, which increased the accuracy of our measurements. Furthermore, inaccuracy of LARS would result in underestimation of the association with prognosis, thus reinforcing the results of this study. Fourth, surveillance bias and detection bias cannot be completely excluded. On the one hand, death is characterized by a serious clinical course, and surveillance bias is unlikely to occur. On the other hand, the rate of cardiovascular death was higher in this study than in the general HF cohort,40 and more patients dropped out early. Patients at higher cardiovascular risk may have received preferential follow-up, which could have resulted in detection bias that increased the rate of cardiovascular death. Finally, we cannot exclude the possibility of unmeasured confounding factors.
Conclusions
We found a significant association between LARS and death in patients hospitalized for acute HF with AF, which suggested that patients with reduced LARS are refractory to therapy and have poor prognosis, indicating the need for aggressive treatment or careful follow-up. Larger studies are needed to clarify the optimal cutoff value of LARS for predicting prognosis, and prospective studies are needed to determine whether therapeutic interventions for patients with reduced LARS can improve LA dysfunction and their poor prognosis.
Disclosures
The authors declare that there are no conflicts of interest.
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
IRB Information
The study was approved by the institutional review board of the National Center for Global Health and Medicine (NCGM-S-004598-00).
Authors’ Contributions
J.Y.: Investigation, Data curation, Formal analysis, Strain analysis, Visualization, Writing – Original draft. M.M.: Writing – Reviewing and Editing. H. Hayama: Conceptualization, Methodology, Validation, Writing – Reviewing and Editing. M.Y.: Writing – Reviewing and Editing. H. Hara: Writing – Reviewing and Editing. Y.H.: Writing – Reviewing and Editing, Supervision. All authors read and approved the final manuscript.
Supplementary Files
Please find supplementary file(s);
https://doi.org/10.1253/circj.CJ-23-0238
References
- 1.
Carlisle MA, Fudim M, DeVore AD, Piccini JP. Heart failure and atrial fibrillation, like fire and fury. JACC Heart Fail 2019; 7: 447–456.
- 2.
Wang TJ, Larson MG, Levy D, Vasan RS, Leip EP, Wolf PA, et al. Temporal relations of atrial fibrillation and congestive heart failure and their joint influence on mortality. Circulation 2003; 107: 2920–2925.
- 3.
Taniguchi N, Miyasaka Y, Suwa Y, Harada S, Nakai E, Shiojima I. Heart failure in atrial fibrillation: An update on clinical and echocardiographic implications. Circ J 2020; 84: 1212–1217.
- 4.
Shin HW, Kim H, Son J, Yoon HJ, Park HS, Cho YK, et al. Tissue Doppler imaging as a prognostic marker for cardiovascular events in heart failure with preserved ejection fraction and atrial fibrillation. J Am Soc Echocardiogr 2010; 23: 755–761.
- 5.
Hoit BD. Left atrial size and function: Role in prognosis. J Am Coll Cardiol 2014; 63: 493–505.
- 6.
Mor-Avi V, Lang RM, Badano LP, Belohlavek M, Cardim NM, Derumeaux G, et al. Current and evolving echocardiographic techniques for the quantitative evaluation of cardiac mechanics: ASE/EAE consensus statement on methodology and indications: Endorsed by the Japanese Society of Echocardiography. J Am Soc Echocardiogr 2011; 24: 277–313.
- 7.
Vianna-Pinton R, Moreno CA, Baxter CM, Lee KS, Tsang TSM, Appleton CP. Two-dimensional speckle-tracking echocardiography of the left atrium: Feasibility and regional contraction and relaxation differences in normal subjects. J Am Soc Echocardiogr 2009; 22: 299–305.
- 8.
Thomas L, Marwick TH, Popescu BA, Donal E, Badano LP. Left atrial structure and function, and left ventricular diastolic dysfunction: JACC state-of-the-art review. J Am Coll Cardiol 2019; 73: 1961–1977.
- 9.
Vieira MJ, Teixeira R, Gonçalves L, Gersh BJ. Left atrial mechanics: Echocardiographic assessment and clinical implications. J Am Soc Echocardiogr 2014; 27: 463–478.
- 10.
Modin D, Biering-Sørensen SR, Møgelvang R, Alhakak AS, Jensen JS, Biering-Sørensen T. Prognostic value of left atrial strain in predicting cardiovascular morbidity and mortality in the general population. Eur Heart J Cardiovasc Imaging 2019; 20: 804–815.
- 11.
Iwahashi N, Gohbara M, Kirigaya J, Abe T, Horii M, Hanajima Y, et al. Prognostic significance of the combination of left atrial reservoir strain and global longitudinal strain immediately after onset of ST-elevation acute myocardial infarction. Circ J 2022; 86: 1499–1508.
- 12.
Carluccio E, Biagioli P, Mengoni A, Francesca Cerasa M, Lauciello R, Zuchi C, et al. Left atrial reservoir function and outcome in heart failure with reduced ejection fraction. Circ Cardiovasc Imaging 2018; 11: e007696.
- 13.
Santos ABS, Roca GQ, Claggett B, Sweitzer NK, Shah SJ, Anand IS, et al. The prognostic relevance of left atrial dysfunction in heart failure with preserved ejection fraction. Circ Heart Fail 2016; 9: e002763.
- 14.
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.
- 15.
Yamamoto J, Yamamoto M, Hara H, Hiroi Y. Relation between laxative use and risk of major bleeding in patients with atrial fibrillation and heart failure. Heart Vessels 2023; 38: 938–948.
- 16.
JCS Joint Working Group. Guidelines for treatment of acute heart failure (JCS 2011). Circ J 2013; 77: 2157–2201.
- 17.
Nagueh SF, Smiseth OA, Appleton CP, Byrd BF, Dokainish H, Edvardsen T, et al. Recommendations for the evaluation of left ventricular diastolic function by echocardiography: An update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging. J Am Soc Echocardiogr 2016; 29: 277–314.
- 18.
Sun BJ, Park JH. Echocardiographic measurement of left atrial strain: A key requirement in clinical practice. Circ J 2021; 86: 6–13.
- 19.
McKee PA, Castelli WP, McNamara PM, Kannel WB. The natural history of congestive heart failure: The Framingham Study. N Engl J Med 1971; 285: 1441–1446.
- 20.
Nishimura RA, Otto CM, Bonow RO, Carabello BA, Erwin JP, Guyton RA, et al. 2014 AHA/ACC guideline for the management of patients with valvular heart disease: A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol 2014; 63: e57–e185.
- 21.
Mamas MA, Caldwell JC, Chacko S, Garratt CJ, Fath-Ordoubadi F, Neyses L. A meta-analysis of the prognostic significance of atrial fibrillation in chronic heart failure. Eur J Heart Fail 2009; 11: 676–683.
- 22.
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.
- 23.
Cameli M, Mandoli GE, Loiacono F, Dini FL, Henein M, Mondillo S. Left atrial strain: A new parameter for assessment of left ventricular filling pressure. Heart Fail Rev 2016; 21: 65–76.
- 24.
van Mourik MJW, Artola Arita V, Lyon A, Lumens J, de With RR, van Melle JP, et al. Association between comorbidities and left and right atrial dysfunction in patients with paroxysmal atrial fibrillation: Analysis of AF-RISK. Int J Cardiol 2022; 360: 29–35.
- 25.
Goedemans L, Leung M, van der Bijl P, Abou R, Vo NM, Ajmone Marsan N, et al. Influence of chronic obstructive pulmonary disease on atrial mechanics by speckle tracking echocardiography in patients with atrial fibrillation. Am J Cardiol 2021; 143: 60–66.
- 26.
Weerts J, Barandiarán Aizpurua A, Henkens MTHM, Lyon A, van Mourik MJW, van Gemert MRAA, et al. The prognostic impact of mechanical atrial dysfunction and atrial fibrillation in heart failure with preserved ejection fraction. Eur Heart J Cardiovasc Imaging 2021; 23: 74–84.
- 27.
Smiseth OA, Baron T, Marino PN, Marwick TH, Flachskampf FA. Imaging of the left atrium: Pathophysiology insights and clinical utility. Eur Heart J Cardiovasc Imaging 2021; 23: 2–13.
- 28.
Sachdeva S, Desai R, Andi K, Vyas A, Deliwala S, Sachdeva R, et al. Reduced left atrial strain can predict stroke in atrial fibrillation: A meta-analysis. Int J Cardiol Heart Vasc 2021; 36: 100859.
- 29.
Cameli M, Sciaccaluga C, Loiacono F, Simova I, Miglioranza MH, Nistor D, et al. The analysis of left atrial function predicts the severity of functional impairment in chronic heart failure: The FLASH multicenter study. Int J Cardiol 2019; 286: 87–91.
- 30.
Hanna-Rivero N, Tu SJ, Elliott AD, Pitman BM, Gallagher C, Lau DH, et al. Anemia and iron deficiency in patients with atrial fibrillation. BMC Cardiovasc Disord 2022; 22: 204.
- 31.
Gan GCH, Bhat A, Kadappu KK, Fernandez F, Gu KH, Chen HHL, et al. Usefulness of left atrial strain to predict end stage renal failure in patients with chronic kidney disease. Am J Cardiol 2021; 151: 105–113.
- 32.
Koniari I, Artopoulou E, Mplani V, Mulita F, Alexopoulou E, Chourdakis E, et al. Atrial fibrillation in heart failure patients: An update on renin-angiotensin-aldosterone system pathway blockade as a therapeutic and prevention target. Cardiol J 2023; 30: 312–326.
- 33.
Cleland JGF, Bunting KV, Flather MD, Altman DG, Holmes J, Coats AJS, et al. Beta-blockers for heart failure with reduced, mid-range, and preserved ejection fraction: An individual patient-level analysis of double-blind randomized trials. Eur Heart J 2018; 39: 26–35.
- 34.
Farmakis D, Chrysohoou C, Giamouzis G, Giannakoulas G, Hamilos M, Naka K, et al. The management of atrial fibrillation in heart failure: An expert panel consensus. Heart Fail Rev 2021; 26: 1345–1358.
- 35.
McMurray JJV, Packer M, Desai AS, Gong J, Lefkowitz MP, Rizkala AR, et al. Angiotensin-neprilysin inhibition versus enalapril in heart failure. N Engl J Med 2014; 371: 132–133.
- 36.
Böhm M, Slawik J, Brueckmann M, Mattheus M, George JT, Ofstad AP, et al. Efficacy of empagliflozin on heart failure and renal outcomes in patients with atrial fibrillation: Data from the EMPA-REG OUTCOME trial. Eur J Heart Fail 2020; 22: 126–135.
- 37.
Kirchhof P, Camm AJ, Goette A, Brandes A, Eckardt L, Elvan A, et al. Early rhythm-control therapy in patients with atrial fibrillation. N Engl J Med 2020; 383: 1305–1316.
- 38.
Baudo M, D’Ancona G, Trinca F, Rosati F, di Bacco L, Curnis A, et al. Atrioventricular node ablation and pacing for atrial tachyarrhythmias: A meta-analysis of postoperative outcomes. Int J Cardiol 2022; 363: 80–86.
- 39.
Labori F, Bonander C, Persson J, Svensson M. Clinical follow-up of left atrial appendage occlusion in patients with atrial fibrillation ineligible of oral anticoagulation treatment: A systematic review and meta-analysis. J Interv Card Electrophysiol 2021; 61: 215–225.
- 40.
Desai AS, Jhund PS, Claggett BL, Vaduganathan M, Miao ZM, Kondo T, et al. Effect of dapagliflozin on cause-specific mortality in patients with heart failure across the spectrum of ejection fraction: A participant-level pooled analysis of DAPA-HF and DELIVER. JAMA Cardiol 2022; 7: 1227–1234.
