Left Atrial Volumetric/Mechanical Coupling Index in Heart Failure With Preserved Ejection Fraction

Abstract

Background

Left atrial (LA) dysfunction is common in heart failure with preserved ejection fraction (HFpEF). The LA volumetric/mechanical coupling index (LACI), calculated as the LA volume index divided by the late diastolic mitral annular velocity (a′), is a recently advocated parameter reflecting LA contractile function based on the atrial Frank–Starling mechanism. An elevated LACI signifies impaired mechanical performance despite greater atrial volume, indicating volumetric/mechanical uncoupling. Its associations with echocardiographic indices and prognostic significance in HFpEF remain unclear.

Methods

This study included 562 patients from the PURSUIT‐HFpEF (Prospective Multicenter Observational Study of Patients With Heart Failure With Preserved Ejection Fraction) registry, a prospective multicenter registry for HFpEF. LACI was calculated and categorized into quartiles. Associations with echocardiographic indices and clinical outcomes were examined.

Results

LA volume index and late diastolic mitral annular velocity were negatively correlated (r=−0.26; P<0.001), reflecting impaired atrial Frank–Starling mechanism in this cohort. The median LACI was 6.44 (interquartile range, 4.16–9.69). Patients with elevated LACI had unfavorable backgrounds with increased filling pressure and diastolic dysfunction. During a median follow‐up of 40 months, 171 patients experienced cardiovascular death or rehospitalization for heart failure, with event incidence increasing across LACI quartiles (quartile 1, 26.4%; quartile 2, 39.0%; quartile 3, 46.1%; quartile 4, 46.4%; log‐rank, P=0.001). After multivariable adjustment, LACI remained a significant predictor of cardiovascular outcomes (hazard ratio, 1.18 per quartile increase [95% CI, 1.00–1.38]; P=0.045).

Conclusions

In patients with HFpEF, LA mechanical activity decreased with LA enlargement, suggesting volumetric/mechanical uncoupling inherent in this cohort. LACI may serve as a clinically relevant index for risk stratification of cardiovascular death or rehospitalization for heart failure in patients with HFpEF.

Registration

URL: https://www.clinicaltrials.gov; Unique identifier: UMIN000021831.

Nonstandard Abbreviations and Acronyms

HFpEF

heart failure with preserved ejection fraction

HFrEF

heart failure with reduced ejection fraction

LACI

left atrial volumetric/mechanical coupling index

LAVI

left atrial volume index

MAGGIC

Meta‐Analysis Global Group in Chronic Heart Failure

Clinical Perspective.

What Is New?

• The left atrial volumetric/mechanical coupling index (LACI), calculated by dividing the left atrial volume index by the late diastolic mitral annular velocity, is a measure that quantifies the left atrial contractile function derived from the atrial Frank–Starling curve.

• An elevated LACI indicates LA volumetric/mechanical uncoupling, which was prevalent in patients with heart failure with preserved ejection fraction; an elevated LACI was associated with unfavorable demographics, echocardiographic measurements, and an increased risk of cardiovascular death or rehospitalization for heart failure.

What Are the Clinical Implications?

• LACI assessment offers a simple and reliable method for evaluating the left atrial contractile function using standard Doppler echocardiography, making it feasible for routine clinical practice.

• The association between an elevated LACI and poor cardiovascular outcomes could be helpful in risk stratification and therapeutic management of patients with heart failure with preserved ejection fraction.

Left atrial (LA) dysfunction is common in patients with heart failure (HF) with preserved ejection fraction (HFpEF). ¹ , ² , ³ A hallmark of LA dysfunction is structural remodeling, ⁴ which occurs in response to chronic pressure and volume overload. Although LA size has prognostic value, it is influenced by left ventricular (LV) compliance and filling pressures. Therefore, LA size alone does not provide a pure reflection of LA function.

In contrast, the pathophysiological contribution and prognostic impact of LA mechanics have been noted in patients with HF. The diastolic tissue–Doppler myocardial velocity during the atrial contraction (a′) is an established measure of LA mechanics, but its specificity is low due to its dependence on the precontraction LA volume. Strain analysis has further enhanced the assessment of LA function, particularly its reservoir phase, by providing insights into filling dynamics. ⁵ , ⁶ , ⁷ , ⁸ , ⁹ , ¹⁰ However, this approach predominantly reflects LA compliance and volume handling rather than intrinsic contractile performance, limiting its ability to fully capture LA mechanical dysfunction.

The recently proposed LA volumetric/mechanical coupling index (LACI) aims to address some of these limitations by integrating LA volume and contractile parameters. ¹¹ , ¹² , ¹³ LACI, defined as the LA volume index (LAVI) divided by a′, is a novel marker of LA contractile function derived from the atrial Frank–Starling curve (Figure 1). ¹⁴ Although the clinical utility of LACI has been demonstrated in patients with HF with reduced ejection fraction (HFrEF) ¹² or floppy mitral valves, ¹³ its applicability and prognostic value in patients with HFpEF remain unclear. This study aimed to examine (1) the association between LACI and patient characteristics, (2) the correlation between LACI and echocardiographic measurements, and (3) the association between LACI and clinical outcomes in patients with HFpEF.

Figure 1. Physiological basis of LACI.

LACI is derived from the atrial Frank–Starling curve. ¹⁴ Originally, the horizontal and vertical axes of the curve represented the preatrial contraction LA volume and atrial stroke volume, respectively (left panel). To simplify this concept, an alternative curve has been created, using the LA volume index on the horizontal axis and the tissue Doppler imaging a′ on the vertical axis (right panel). According to this curve, the LA pump function (a′) should be proportionally enhanced in response to a larger volume (preload) for normal‐to‐moderately enlarged LA volumes (ie, volumetric/mechanical coupling). However, the Frank–Starling curve has been shown to have a descending limb for very large atrial volumes (ie, volumetric/mechanical uncoupling). LACI is a measure of how LA volumetric indices and LA mechanical indices are coupled or uncoupled. Based on the formula, LACI is the reciprocal of the slope of the atrial Frank–Starling curve. An elevated LACI indicate more pronounced uncoupling owing to reduced atrial contraction in the context of larger LA volumes. a′ indicates late diastolic mitral annular velocity; LA, left atrial; and LACI, left atrial coupling index.

Methods

The data that support the findings of this study are available from the corresponding author upon reasonable request.

The PURSUIT‐HFpEF Registry

The PURSUIT‐HFpEF (Prospective Multicenter Observational Study of Patients With Heart Failure With Preserved Ejection Fraction; UMIN‐CTR ID: UMIN000021831) is a prospective, multicenter, observational cohort study conducted in hospitalized patients with HFpEF. ¹⁵ The PURSUIT‐HFpEF registry was planned to include up to 1500 cases registered by Osaka University Hospital, in collaboration with 31 hospitals in Japan. This registry aimed to collect and record a comprehensive range of patient data, including demographics, blood samples, echocardiographic data, and therapeutic and prognostic information, to address multiple clinical questions regarding HFpEF. Consecutive patients with acute decompensated HFpEF were prospectively registered and agreed to be followed up for outcome data collection. Acute decompensated HF was diagnosed on the basis of the following criteria: (1) clinical symptoms and signs according to the Framingham Heart Study criteria and (2) serum NT‐proBNP (N‐terminal pro‐B‐type natriuretic peptide) levels ≥400 pg/mL or brain natriuretic peptide levels ≥100 pg/mL. Exclusion criteria included age<20 years, severe valvular disease (aortic stenosis, aortic regurgitation, mitral stenosis, or mitral regurgitation), acute coronary syndrome on admission, life expectancy <6 months due to the prognosis of noncardiac diseases, or a history of heart transplantation. The study collected detailed data on admission, at discharge, and 1 year after discharge, with additional follow‐up conducted annually for up to 5 years. All participants provided written informed consent, and the study protocol was approved by the ethics committee of each participating hospital. The study adhered to the ethical principles outlined in the Declaration of Helsinki.

Study Patients

In total, 1238 patients with HFpEF were registered between June 2016 and April 2022. After excluding 20 patients who died during hospitalization, follow‐up data after survival discharge were available for 1218 patients. Among them, 1176 (96.6%) had follow‐up data available for >1 year, ensuring that <10% were lost to follow‐up within 1 year after the final registration. Because the tissue Doppler imaging a′ velocities were not available for sustained atrial arrhythmias or atrial standstills, the corresponding patients (atrial fibrillation [AF], n=409; atrial flutter, n=18; unknown, n=217) were excluded from this study. Therefore, we enrolled 562 patients whose LAVI and a′ at discharge were available for this study.

Data Collection

Details of the data collection have been described elsewhere. ¹⁵ In brief, the basic patient characteristics, echocardiographic data, laboratory test results, and medication information were collected at admission, discharge, and each annual follow‐up time point. The present analysis focused on LACI measured from echocardiography at discharge. Therefore, we used laboratory and medication data at the time of discharge (after treatment for acute decompensated HF and in a stable condition).

Echocardiography

Comprehensive echocardiographic examinations were performed by trained cardiac sonographers according to the American Society of Echocardiography and European Association of Cardiovascular Imaging guidelines. ¹⁶ Recordings of at least 3 consecutive beats were performed, with mean values calculated for accuracy. The left ventricular end‐diastolic volume and left ventricular end‐systolic volume were measured using the biplane Simpson method from the apical 2‐ and 4‐chamber views. LAV was similarly calculated using the biplane Simpson method. The right ventricular (RV) diameter was measured in the middle third of the RV inflow using an RV‐focused apical 4‐chamber view. The tricuspid annular plane systolic excursion was assessed using M‐mode echocardiography. The diameter of the inferior vena cava was measured from a standard echocardiographic subcostal longitudinal view with patients in the supine position. All volume measurements were indexed to the body surface area.

Hemodynamic data were obtained using Doppler echocardiography. The stroke volume was estimated by measuring the LV outflow tract diameter and velocity‐time integral. Cardiac output was calculated by multiplying the heart rate and stroke volume. The cardiac index was calculated by dividing the cardiac output by the body surface area. The transmitral flow parameters, including the early (E) and late (A) diastolic filling velocities, E/A ratio, and E deceleration time, were recorded from an apical 3‐chamber view using conventional pulsed‐wave Doppler. The maximal velocities of the early (e′) and late (a′) diastolic mitral annulus motion were analyzed using tissue Doppler imaging, with the mean values calculated from the septal and lateral sides of the mitral annulus. The tricuspid regurgitation pressure gradient was calculated using the simplified Bernoulli equation.

Definition and End Points

LACI is a volumetric‐to‐mechanical coupling index and is defined as LAVI divided by a′ measured using tissue Doppler imaging at the medial mitral annulus. The unit of measurement is, therefore, mL×s/cm⁻¹ per m² of body surface area.

The primary end point of this study was a composite of rehospitalization for HF and cardiovascular death. The secondary end points included rehospitalization for HF, cardiovascular death, and noncardiovascular death. Patient follow‐up commenced at the time of hospital discharge, with all patients being followed up at their respective hospitals. Survival data were collected by dedicated coordinators and investigators through direct contact with patients and their physicians, either in hospitals or outpatient settings, and through telephone interviews with their families or via mail. This analysis included all available follow‐up data up to March 2024.

Statistical Analysis

Continuous variables are expressed as medians with interquartile ranges (IQRs). Categorical data are presented as numbers (percentages) unless otherwise specified. To evaluate the statistical significance across each group, the Pearson's χ² test was used for categorical variables, and the Kruskal–Wallis test was used for continuous variables.

LACI exhibited a skewed distribution (Kolmogorov–Smirnov test, P<0.001) (Figure S1), and analyses were conducted by dividing the patients into quartile groups or by applying a logarithmic transformation using base‐10 (log) to achieve a normal distribution. A trend analysis was performed using the Jonckheere–Terpstra test for continuous variables and the Cochran–Armitage trend test for categorical variables. Correlations between log‐transformed LACI and echocardiographic indices were assessed using Spearman's rank correlations. Linear regression models were applied descriptively to visualize these relationships.

For time‐to‐event analyses, patients were censored at the date of the last confirmed follow‐up or at the end of the observation period (March 2024), whichever occurred first. For cardiovascular‐specific outcomes, patients who died of noncardiovascular causes were censored at the date of death. The Kaplan–Meier method was used to estimate event rates, and comparisons across LACI quartiles were performed using the log‐rank test. Additionally, based on the previous study conducted in HFrEF, ¹² binary analyses stratified by LACI ≥6.0 versus <6.0 were conducted to assess clinical relevance in HFpEF. Furthermore, to explore outcome‐specific thresholds of LACI, we performed receiver operating characteristic curve analyses to identify the optimal cutoffs based on the Youden index. Univariable and multivariable Cox proportional hazards regression models were used to calculate the hazard ratios (HRs) and 95% CIs for LACI with respect to each end point. The multivariable model was adjusted for age, sex, body mass index, hemoglobin levels, estimated glomerular filtration rate, plasma NT‐proBNP levels, and E/e′. ¹⁷ , ¹⁸ , ¹⁹ The proportional hazards assumption for these variables was tested using Schoenfeld residuals, and no violations were detected. Restricted cubic spline analysis was performed to assess the associations between LACI and the composite end point.

Finally, the discriminative ability of LACI was assessed using the C‐statistic. The area under the receiver operating characteristic curves (AUROCs) of LACI, a′, and LAVI were compared to assess the predictive accuracy of discriminating the primary end points within 3 years. Moreover, the AUROC of LACI was compared with that of the Meta‐Analysis Global Group in Chronic Heart Failure (MAGGIC) score, ²⁰ , ²¹ a well‐validated score to predict prognosis in patients with chronic heart failure. To clarify the additional predictive value of LACI beyond the MAGGIC score, we compared the AUROC of the MAGGIC score alone with that of a combined model incorporating LACI (adding 1 point for each quartile increase), after confirming the absence of significant correlation between the two. A nested Cox regression model was also used to account for the time‐to‐event nature of the primary end point. Statistical analysis was performed using MedCalc version 23.0.2 (MedCalc Software, Ostend, Belgium), JMP version 13.2.1 (SAS Institute Inc., Chicago, IL), and R software (R Foundation for Statistical Computing, Vienna, Austria). A P value <0.05 was considered statistically significant.

Results

Association Between LAVI and a'

In this study, the scatter plot demonstrated a weak negative correlation between LAVI (median, 43 [IQR, 33–56 mL/m²]) and a′ (median, 6.9 cm [IQR, 5.3–8.5]; (r=−0.27; P<0.001), resulting in an overall LACI of 6.44 (IQR, 4.16–9.69). Based on the Frank–Starling theory and the concept of LACI, when the intercept was hypothetically set to 0, the approximate curve was parabolic in this cohort (Figure 2). Based on LACI, patients were categorized into quartiles (quartile 1, <4.16; quartile 2, 4.16–6.44; quartile 3, 6.45–9.69; and quartile 4, ≥9.70).

Figure 2. Association between LAVI and a′.

The scatter plot illustrates an association between LAVI (horizontal axis) and a′ (vertical axis) measured by tissue Doppler imaging. There is a weak negative correlation between the LAVI and a′ in the linear regression analysis (r=−0.26; P<0.001). However, based on the Frank–Starling principle, this relationship may be better conceptualized as nonlinear. When assuming minimal atrial contraction (a′=0) at minimal preload (LAVI=0)—that is, by hypothetically constraining the intercept to 0—the resulting curve approximates a parabolic shape, consistent with the atrial Frank–Starling mechanism. Notably, the curve has a descending limb at very large atrial volumes, as demonstrated by Anwar et al. ¹⁴ LACI indicates left atrial volumetric/mechanical coupling index; and LAVI, left atrial volume index.

Baseline Characteristics

The baseline characteristics of the study participants are presented in Table 1. Notably, a′ was not available during AF rhythm. The prevalence of AF was only 11% in the entire cohort, and AF was estimated to be paroxysmal. Several clinical differences were observed between the LACI quartiles. The Jonckheere–Terpstra trend test revealed that age, proportion of females, and body mass index increased with LACI, whereas heart rate decreased. Additionally, the prevalence of prior hospitalization for HF, hypertension, dyslipidemia, AF, and chronic kidney disease increased with LACI, whereas the prevalence of diabetes decreased. Laboratory data revealed a decline in renal function and an increase in NT‐proBNP levels as LACI increased. The prescription rates for β blockers, diuretics, and aldosterone antagonists also increased with increasing LACI quartiles.

Table 1.

Baseline Characteristics

	All n=562	Quartile 1: LACI <4.16 n=140	Quartile 2: LACI 4.16–6.44 n=141	Quartile 3: LACI 6.45–9.69 n=141	Quartile 4: LACI ≥9.70 n=140	P for trend
Clinical data
Age, y	82 (76–87)	80 (74–86)	82 (75–86)	83 (77–87)	83 (78–88)	0.003
Female sex	330 (59)	65 (46)	82 (58)	87 (62)	96 (69)	<0.001
Body mass index, kg/m2	21.4 (18.8–24.3)	20.0 (17.9–23.2)	21.5 (19.3–24.5)	21.9 (19.7–25.5)	21.5 (19.1–24.3)	0.012
Systolic blood pressure, mm Hg	122 (109–134)	120 (109–133)	125 (112–138)	125 (113–139)	117 (105–127)	0.085
Diastolic blood pressure, mm Hg	64 (57–72)	66 (59–72)	63 (58–73)	65 (56–72)	64 (57–71)	0.11
Heart rate, bpm	67 (60–76)	72 (63–80)	69 (62–76)	65 (58–76)	64 (60–73)	<0.001
NYHA class 3–4	24 (4.3)	4 (2.9)	6 (4.3)	8 (5.8)	6 (4.3)	0.53
Six‐min walk distance, m	250 (165–340)	261 (175–346)	259 (155–354)	217 (140–291)	290 (200–360)	0.80
Comorbidities
Prior HF hospitalizations	106 (19)	12 (8.8)	26 (19)	26 (18)	42 (31)	<0.001
Prior myocardial infarction	48 (9)	8 (6)	12 (9)	15 (11)	13 (9.4)	0.23
Hypertension	493 (88)	112 (80)	130 (93)	127 (90)	124 (89)	0.007
Diabetes	193 (35)	52 (37)	54 (39)	51 (36)	36 (26)	0.043
Dyslipidemia	281 (50)	58 (41)	72 (51)	76 (55)	75 (54)	0.031
Stroke	78 (14)	24 (17)	20 (14)	16 (11)	18 (13)	0.26
Atrial fibrillation	59 (11)	8 (5.7)	12 (8.5)	15 (11)	24 (17)	0.002
Chronic kidney disease	221 (40)	42 (30)	63 (45)	57 (41)	59 (42)	0.087
Peripheral artery disease	33 (6)	7 (5)	11 (8)	8 (6)	7 (5)	0.87
Chronic obstructive lung disease	33 (6)	12 (9)	8 (6)	7 (5)	6 (5)	0.14
Malignant disease	60 (11)	18 (13)	13 (9)	12 (9)	17 (13)	0.54
Laboratory data
Albumin, g/dL	3.4 (3.1–3.7)	3.3 (3.1–3.7)	3.3 (3.1, 3.7)	3.4 (3.0–3.6)	3.4 (3.1–3.7)	0.22
Hemoglobin, g/dL	11.0 (9.8–12.4)	11.1 (9.9–12.9)	10.9 (9.9, 12.3)	10.8 (9.8–11.8)	11.3 (10.0–12.5)	0.65
Creatinine, mg/dL	1.1 (0.9–1.6)	1.0 (0.8–1.4)	1.1 (0.9, 1.7)	1.1 (0.9–1.7)	1.2 (0.9–1.5)	0.037
eGFR, mL/min per 1.73 m2	40 (28–55)	49.5 (33.0–61.2)	39.4 (24.3, 55.6)	40.0 (25.4–51.2)	37.7 (29.7–49.2)	<0.001
Blood urea nitrogen, mg/dL	25.0 (18.3–34.0)	24.4 (17.0–31.8)	23.0 (18.0, 34.5)	26.0 (17.9–35.0)	27.0 (20.3–36.0)	0.012
NT‐proBNP, pg/mL	760 (350–1931)	467 (221–1213)	652 (333, 1390)	1055 (458–2423)	1060 (516–2412)	<0.001
C‐reactive protein, mg/dL	0.30 (0.13–0.80)	0.30 (0.13–1.02)	0.37 (0.14, 0.76)	0.34 (0.13–1.04)	0.24 (0.10–0.73)	0.19
Medications
ACE‐I or ARB	315 (56)	79 (56)	77 (55)	79 (56)	80 (56)	0.85
Calcium channel blocker	328 (58)	81 (58)	89 (63)	87 (62)	71 (51)	0.22
β Blocker	289 (51)	50 (36)	71 (50)	85 (60)	83 (59)	<0.001
Diuretics	443 (79)	96 (69)	109 (77)	114 (81)	124 (89)	<0.001
Aldosterone antagonist	191 (34)	39 (28)	47 (33)	49 (35)	56 (40)	0.035
Statin	223 (40)	47 (34)	60 (43)	67 (47)	49 (35)	0.62

Values are given as the median (IQR) or n (%). Age and comorbidities are indicated on admission and all the others at discharge. To assess the trend among quartiles, the Jonckheere–Terpstra and Cochran–Armitage tests were used for numerical and categorical variables, respectively. ACE indicates angiotensin‐converting enzyme; ARB, angiotensin receptor blocker; eGFR, estimated glomerular filtration rate; HF, heart failure; LACI, left atrial volumetric/mechanical coupling index; NT‐proBNP, N‐terminal pro‐B‐type natriuretic peptide; and NYHA, New York Heart Association.

Structural and Hemodynamic Correlates With LACI

The correlations between LACI and the echocardiographic structural and hemodynamic data are shown in Figures 3 and 4, respectively. For the structural correlates, a linear regression analysis revealed that the LV end‐diastolic volume index, LV end‐systolic volume index, and LV ejection fraction were not correlated with LACI. However, LAVI (r=0.80), indexed RV diameter (r=0.19), and inferior vena cava end‐expiratory diameter (r=0.24) were positively correlated with LACI (all P<0.001) (Figure 3). For the hemodynamic correlates, a linear regression analysis revealed that the CI and e′ were not correlated with LACI. In contrast, the E/A (r=0.56), E/e′ (r=0.39), and tricuspid pressure gradient (r=0.27) were positively correlated with LACI, and a′ (r=−0.75) was negatively correlated with LACI (Figure 4). The differences in LACI quartiles are described in Table S1. The results for the right heart catheterization partially performed in this cohort are described in Figure S2.

Figure 3. Correlations between LACI and echocardiographic structural measurements.

The scatter plot shows the association between the log‐transformed LACI (horizontal axis) and LVEDVI (A), LVESVI (B), LVEF (C), LAVI (D), indexed RV diameter (E), and IVC expiratory diameter (F). IVC indicates inferior vena cava; LACI, left atrial volumetric/mechanical coupling index; LAVI, left atrial volume index; LVEDVI, left ventricular end‐diastolic volume index; LVEF, left ventricular ejection fraction; LVESVI, left ventricular end‐systolic volume index; and RV, right ventricular.

Figure 4. Correlations between LACI and hemodynamic parameters.

The scatter plot shows the association between the log‐transformed LACI (horizontal axis) and cardiac index (A), E/A (B), a′ (C), e′ (D), E/e′ (E), and TRPG (F). A indicates late diastolic filling velocity; a′, late diastolic mitral annular velocity; E, mitral peak E‐wave velocity; e′, early diastolic mitral annulus velocity; LACI, left atrial volumetric/mechanical coupling index; and TRPG, tricuspid regurgitation pressure gradient.

Clinical Outcomes

During a median follow‐up of 40 months (IQR, 25–56 months), 185 (25.9%), 85 (15.1%), and 122 (21.7%) patients were rehospitalized for HF, died of cardiovascular causes, and died of noncardiovascular causes, respectively. As 48 patients experienced both events, 222 (39.5%) met the primary composite end point. The Kaplan–Meier analysis and the log‐rank test revealed that the incidence of the primary composite end point increased incrementally across the LACI quartiles (quartile 1 versus quartile 2 versus quartile 3 versus quartile 4, 26.4% versus 39.0% versus 46.1% versus 46.4%; log‐rank P=0.001; Figure 5A). Similar results were obtained at the end points for rehospitalization for HF (21.4% versus 31.2% versus 38.3% versus 40.7%; log‐rank P=0.002; Figure 5B) and cardiovascular death (7.9% versus 15.6% versus 14.9% versus 22.1%; log‐rank P=0.010; Figure 5C). Compared with the lowest quartile, higher LACI quartiles were associated with an increased risk of adverse cardiovascular events. Conversely, no difference was observed in the incidence of noncardiovascular death (25.0% versus 18.4% versus 23.4% versus 19.3%, log‐rank P=0.81; Figure 5D).

Figure 5. Kaplan–Meier survival curves stratified by LACI quartiles.

Kaplan–Meier event‐free survival curves are described at the end point of a composite of rehospitalization for HF and cardiovascular death (A), rehospitalization for HF (B), cardiovascular death (C), and noncardiovascular death (D). HF, indicates heart failure; and LACI, left atrial volumetric/mechanical coupling index.

Based on a previous study in HFrEF, ¹² a LACI threshold of 6.0 has been proposed as a clinically relevant cutoff. To evaluate its prognostic utility in HFpEF, we stratified patients according to LACI ≥6.0 versus <6.0. Kaplan–Meier analysis revealed that patients with LACI ≥6.0 had a significantly higher incidence of the primary composite end point than those with LACI <6.0 (45.4% versus 32.7%; log‐rank P=0.001; Figure 6A). Similar trends were observed for rehospitalization for HF (38.1% versus 26.9%; log‐rank P=0.003; Figure 6B) and cardiovascular death (18.5% versus 11.2%, log‐rank P=0.015; Figure 6C). No significant difference was observed in noncardiovascular death (20.9% versus 22.3%; log‐rank P=0.71; Figure 6D).

Figure 6. Kaplan–Meier survival curves stratified by LACI threshold of 6.0.

Kaplan–Meier event‐free survival curves are described at the end point of a composite of rehospitalization for HF and cardiovascular death (A), rehospitalization for HF (B), cardiovascular death (C), and noncardiovascular death (D). HF indicates heart failure; and LACI, left atrial volumetric/mechanical coupling index.

Cox Regression Analysis

The results of the univariable and multivariable Cox regression analysis are presented in Table 2 and Table S2. Even after adjusting for potential confounders, LACI remained significantly associated with the primary composite end point (hazard ratio [HR] per quartile increase, 1.15 [95% CI, 1.00–1.32]; P=0.048) and rehospitalizations for HF (HR, 1.20 [95% CI, 1.03–1.40]; P=0.021), but not with cardiovascular death (HR, 1.16 [95% CI, 0.92–1.46]; P=0.21) or noncardiovascular death (HR, 1.04 [95% CI, 0.85–1.26]; P=0.71). In contrast, the cutoff value of LACI ≥6.0 was not a significant predictor of these end points. The HR for the composite end point across the LACI spectrum was displayed using a restricted cubic spline curve (Figure 7). Notably, the HR increased approximately linearly within the IQR (LACI, 4.16–9.69), indicating a consistent escalation in risk with higher LACI.

Table 2.

Cox Regression Analysis

	Univariable		Multivariable*
	Crude HR (95% CI)	P value	Adjusted HR (95% CI)	P value
Composite end point
LACI, per quartile increase	1.26 (1.12–1.42)	<0.001	1.15 (1.00–1.32)	0.048
LACI ≥6.0 (cutoff from HFrEF study)	1.54 (1.18–2.02)	0.002	1.25 (0.92–1.71)	0.16
LACI >4.54 (data‐driven cutoff)	2.12 (1.52–2.97)	<0.001	1.62 (1.11–2.36)	0.012
Rehospitalization for heart failure
LACI, per quartile increase	1.29 (1.11–1.50)	<0.001	1.20 (1.03–1.40)	0.021
LACI ≥6.0 (cutoff from HFrEF study)	1.56 (1.16–2.10)	0.003	1.31 (0.93–1.84)	0.12
LACI >4.54 (data‐driven cutoff)	2.34 (1.60–3.43)	<0.001	1.81 (1.19–2.75)	0.006
Cardiovascular death
LACI, per quartile increase	1.34 (1.06–1.70)	0.015	1.16 (0.92–1.46)	0.21
LACI ≥6.0 (cutoff from HFrEF study)	1.74 (1.11–2.72)	0.016	1.17 (0.70–1.95)	0.55
LACI >7.77 (data‐driven cutoff)	2.45 (1.60–3.76)	<0.001	1.73 (1.04–2.89)	0.034
Noncardiovascular death
LACI, per quartile increase	0.97 (0.83–1.14)	0.74	1.06 (0.88–1.28)	0.54
LACI ≥6.0 (cutoff from HFrEF study)	0.98 (0.69–1.40)	0.92	1.24 (0.81–1.89)	0.32
LACI >3.62 (data‐driven cutoff)	0.70 (0.46–1.06)	0.089	0.82 (0.49–1.37)	0.45

HRs and 95% CIs were calculated for each variable using the Cox proportional hazards model. LACI was evaluated per quartile increase, using a fixed threshold of 6.0, and according to a data‐driven cutoff in multivariable models. The fixed threshold of 6.0 was derived from the previous HFrEF study, ¹² while the data‐driven cutoff was determined using the Youden index based on the C‐statistic for each outcome. E indicates peak early diastolic filling velocity; e′, early diastolic mitral annulus velocity; HFrEF, heart failure with reduced ejection fraction; HR, hazard ratio; and LACI, left atrial volumetric/mechanical coupling index.

^*Multivariable model was adjusted for age, sex, body mass index, hemoglobin levels, estimated glomerular filtration rate, plasma N‐terminal pro‐B‐type natriuretic peptide levels, and E/e′.

Figure 7. Adjusted hazard ratios for the primary composite end point across LACI.

The panel displays the results of an adjusted Cox regression using a restricted cubic spline model. The reference value (HR=1.0) was set at the median LACI (6.44). The model was adjusted for age, sex, body mass index, hemoglobin levels, estimated glomerular filtration rate, plasma NT‐proBNP levels, and E/e′. The shaded area represents the 95% CI. Within the interquartile range (4.16–9.69), the HR increases approximately linearly with higher LACI values. E/e′ indicates ratio of early transmitral flow velocity to early diastolic mitral annular velocity; HR, hazard ratio; LACI, left atrial volumetric/mechanical coupling index; and NT‐proBNP, N‐terminal pro‐B‐type natriuretic peptide.

Discrimination Ability of LACI

The discrimination ability of LACI for predicting the composite end point within 3 years was evaluated. As shown in Figure 8A, the AUROC of LACI was 0.613 (IQR, 0.571–0.654), which was significantly larger than the AUROC of LAVI alone (median, 0.581 [IQR, 0.538–0.622]) and was numerically larger than that of the a′ alone (median, 0.609 [IQR, 0.567–0.649]). Furthermore, the AUROC of LACI was similar to that of the MAGGIC score (median, 0.635 [IQR, 0.593–0.675]). Given the lack of correlation between the MAGGIC score and LACI (r=0.01; P=0.90), we created a combined model by adding 1 point to the MAGGIC score for each quantile increase in LACI. This combination significantly improved the AUROC compared with the MAGGIC score alone (Figure 8B). A nested Cox regression analysis also revealed the superiority of combining the MAGGIC score with LACI over the MAGGIC score alone (Figure 8C).

Figure 8. Comparison of discriminative ability.

The discriminative ability of LACI to predict the composite end point within 3 years was evaluated. A, The AUROC of LACI was compared with that of LAVI and a′. Although not statistically significant, the AUROC was numerically largest of the three. B, The AUROC of LACI was compared with that of the MAGGIC score, and both demonstrated similar levels of discrimination accuracy. When LACI was added to the MAGGIC score (1 point per quantile increase), the combined AUROC was significantly higher than that of the MAGGIC score alone. C, A nested Cox regression analysis revealed the statistical significance of adding LACI to the conventional MAGGIC score to predict composite end points. The adjusted covariates included age, sex, body mass index, hemoglobin, estimated glomerular filtration rate, plasma NT‐poBNP levels, and E/e′. AUROC indicates area under the receiver operating characteristic curve; E, peak early diastolic filling velocity; e′, early diastolic mitral annulus velocity; eGFR, estimated glomerular filtration rate; NT‐proBNP, N‐terminal pro‐B‐type natriuretic peptide; LACI, left atrial volumetric/mechanical coupling index; LAV, left atrial volume; MAGGIC, Meta‐Analysis Global Group in Chronic Heart Failure; and TDI, tissue Doppler imaging.

Discussion

Main Findings

LACI is a recently advocated index that reflects LA pump function but has not been systemically evaluated in patients with HFpEF. In this study, using data from the PURSUIT‐HFpEF registry, we investigated the association between LACI and patient characteristics, echocardiographic parameters, and its prognostic value in patients with HFpEF. Patients with elevated LACI had a more unfavorable background, multiple comorbidities, and impaired laboratory data. Echocardiographic data revealed that LACI was not associated with LV systolic parameters but was associated with LV diastolic parameters and LV filling pressure, including E/A, E/e′, and tricuspid pressure gradient. Cox regression analysis revealed that LACI had an incremental prognostic value, even after adjusting for several clinical variables. Considering the comparable discrimination ability of LACI to that of LAVI and a′ alone, further studies are needed to confirm the prognostic impact of this index. However, the findings indicate that LA pump dysfunction may be widely inherent and could actively affect clinical outcomes in patients with HFpEF.

LA Dysfunction and HFpEF

The left atrium is a crucial component of cardiac physiology. It collects blood into the heart and modulates LV filling during systole and diastole, respectively. In the early preclinical stages of HF, a compensatory increase in the LA volume and subsequent enhancement of LA pump function help maintain cardiovascular hemodynamics and cardiac output. However, as HFpEF progresses, LV relaxation and compliance worsen, and LV filling pressure increases with impaired diastolic function. The contribution of the LA booster pump to LV filling increases until the reserve limits of the LA preload are reached. Once the left atrium enlarges further and becomes dysfunctional with continuously elevated pressure, it loses its buffering capacity and primarily functions as a conduit, leading to mechanical failure. An increased LAVI is an established predictor of adverse events. ⁴ However, LA dysfunction before LA enlargement has recently been recognized, particularly in the early stages of HF.

LA strain imaging can reveal subtle changes in atrial function before overt structural changes occur, allowing for earlier identification of cardiac pathologies. LA function has been divided into 3 phases: reservoir, conduit, and booster pump. While strain imaging has advanced our understanding of LA function, it primarily emphasizes the reservoir phase ⁵ , ⁶ , ⁷ , ⁸ , ⁹ and has limited ability to isolate and quantify intrinsic contractile performance during the booster pump phase. This gap underscores the need for complementary indices such as LACI, which provides additional insights into LA mechanical coupling and active contractile function. ⁷ , ²² , ²³ Additionally, the assessment of LA strain requires specialized software and technical training for accurate measurements. Therefore, incorporating LACI to assess atrial pump function has been proposed to improve the detection of LA dysfunction. ⁹

LACI, a Measure of LA Volumes and Mechanics Coupling

LACI reflects LA contractile function and is derived from the atrial Frank–Starling curve (Figure 1). Initially, LA dysfunction and subsequent dilatation enhance conduit filling and contractile function through a compensatory mechanism. However, when atrial volumes exceed a critical threshold, the descending limb of the atrial Frank–Starling curve becomes evident, indicating volumetric/mechanical uncoupling. This uncoupling reflects the deterioration of the LA pump function relative to its dilatation.

In this study, the median LACI was 6.44 (IQR, 4.16–9.69), which was higher than that of the general population (2.8 [IQR, 2.1–3.8] ²⁴ and 2.3 [IQR, 1.8–3.0] ²⁵ ). A parabolic relationship was also observed between LAVI and a′ (Figure 2). These results suggested that LA dysfunction was prevalent in our study population. Echocardiographic data indicated a lack of correlation of LACI with LV chamber size and LV systolic function. However, a mild correlation was observed between LACI and E/A, E/e′, and tricuspid pressure gradient, highlighting the interplay between LA function and LV diastolic performance and filling pressure. After the mitral valve opens and the left atrium empties passively into the LV (the conduit phase), the left atrium actively empties into the left ventricle (the booster pump phase). ⁵ The mitral annular velocity during this phase was measured as the a′ velocity, modified by the LV compliance (E/A) and filling pressure (E/e′). In addition, RV dysfunction is sometimes associated with HFpEF. Therefore, RV and tricuspid annular dilatation can potentially affect atrioventricular geometry and function (a′). Because of these complex interactions, even LACI may not purely reflect LA contractile function. Nonetheless, LACI is a valuable physiological indicator based on the atrial Frank–Starling curve and can be used in daily clinical practice.

Prognostic Importance of an Increased LACI in Patients With HFpEF

This study revealed that LACI is independently and incrementally associated with cardiovascular prognosis in patients with HFpEF. However, the predictive value was limited, with an AUROC of 0.613 for predicting the composite end point within 3 years. This may be attributed to the significant proportion of noncardiovascular deaths in patients with HFpEF. In the present study, 207 patients (36.8%) died during the 40‐month follow‐up period, of which 122 (58.9%) died from noncardiovascular causes. LACI was not predictive of noncardiovascular events, and the high rate of noncardiovascular deaths likely reduced the statistical power to assess cardiovascular outcomes. Furthermore, most patients in this registry already exhibited impaired LA function, which may have attenuated the association between LACI and prognosis.

Although a threshold of LACI ≥6.0 has been proposed in HFrEF populations, this cutoff did not show predictive value for adverse outcomes in the present HFpEF cohort. Instead, exploratory analyses identified alternative thresholds (>4.54 for HF rehospitalization and >7.77 for cardiovascular death), suggesting that the risk of severe events may increase progressively with further deterioration of LA pump function. While LACI thresholds differ across disease contexts (≥6.0 in patients with HFrEF ¹² and >5.0 in patients with isolated floppy mitral valve ¹³ ), the consistent association between higher LACI and worse outcomes highlights the prognostic importance of LA dysfunction in patients with HF.

Importantly, the measurement of a′ necessary for the LACI calculation can be taken only during sinus rhythm. In previous HFpEF registries, the prevalence of AF varied between 15% and 41%, and the prognosis was significantly worse in patients with AF. ²⁶ , ²⁷ In this study of 1206 patients with follow‐up data, only 562 (46.6%) were included in the analysis. Thus, LACI cannot be used as a prognostic tool for cardiovascular outcomes in all patients with HFpEF. It is likely that LAVI still has significant prognostic value in patients with HFpEF and persistent AF.

From a different perspective, patients whose LACI is available may have less advanced stages of HFpEF. As the HFpEF stage advances with the worsening diastolic dysfunction, the LA structural remodeling progresses, and consequently, AF will develop and persist. The occurrence and continuation of AF will deteriorate hemodynamics and lead to further worsening of HF. ²⁸ However, based on this study, LA structural and mechanical alterations occur even before sustaining AF, partially mediating poor outcomes. An elevated LACI may reflect structural alterations in the LA, resulting from myocyte hypertrophy, necrosis, apoptosis, and fibrosis. These alterations may contribute to unfavorable cardiovascular prognosis in patients with elevated LACI.

Clinical Implications

Echocardiographic examination remains the cornerstone of HF diagnosis and evaluation. However, assessing atrial pump function is challenging. Typically, this assessment relies on strain measures, which are technically complex and time‐consuming to obtain. Additionally, interpreting these measures can be difficult when tracking the thin walls of the left atrium. Of remarkable importance for clinical practice, LACI can be obtained from a standard echocardiographic protocol (LAVI and a′) without the need for additional acquisition or postprocessing software. Therefore, calculating LACI based on an echocardiogram at discharge may be useful for risk stratification in patients with HFpEF. Furthermore, while LACI alone has limited predictive power for cardiovascular events, its effectiveness is comparable to that of the MAGGIC score. Combining these 2 measures can enhance risk stratification for patients with HFpEF.

Limitations

Several limitations should be acknowledged in this study. First, potential selection bias could not be avoided, as the data in this registry were derived from institutions specializing in cardiology and HF management. Second, the results were derived from a Japanese HFpEF cohort, which somewhat differs from the HFpEF cohorts in Western countries. ²⁹ The extrapolation of the findings of this study to other HFpEF populations should be validated in future studies. Third, treatment was left to the attending physician's discretion. Although AF ablation potentially affects LA function, it was not recorded in the PURSUIT‐HFpEF registry. Furthermore, because this study was designed before 2016, data on using Na‐glucose cotransporter 2 inhibitors were inadequate. Fourth, interobserver differences are inevitable in echocardiography measurements. Fifth, the LACI was not a statistically significant predictor for cardiovascular death in this study, and the superiority of its discrimination accuracy over that of LAVI and a′ alone was also unclear. Studies with greater statistical power may be necessary to clarify this issue. In addition, given the high incidence of noncardiovascular deaths in HFpEF, the observed associations may have been attenuated. Future studies applying competing risk models such as the Fine–Gray method may help clarify this effect. Finally, the prognostic thresholds for LACI remain undefined and require validation across diverse populations. Establishing these thresholds is essential for integrating LACI into clinical practice and ensuring its utility in risk stratification and managing patients with HFpEF.

Conclusions

In the PURSUIT‐HFpEF registry, LA mechanical activity decreased with enlargement of the left atrium, suggesting volumetric/mechanical uncoupling commonly inherent in patients with HFpEF. An elevated LACI reflected functional alterations of LA contractile function and was associated with unfavorable baseline characteristics with diastolic dysfunctions and elevated LV filling pressures. An elevated LACI was associated with worse cardiovascular outcomes, suggesting the prognostic importance of LA contractile function in patients with HFpEF.

Sources of Funding

This work was funded by Roche Diagnostics K.K. and Fuji Film Toyama Chemical Co. Ltd.

Disclosures

Dr Tanaka received personal fees from Daiichi Sankyo, Bayer, Boehringer Ingelheim Japan, AstraZeneca, Johnson & Johnson K.K., and Medtronic. Dr Iwakura received personal fees from Daiichi Sankyo, Bayer, Boehringer Ingelheim Japan, and Bristol‐Myers Squibb. Dr Hikoso received grants from Roche Diagnostics, FUJIFILM Toyama Chemical, TOA EIYO, and Bristol‐Myers Squibb; and received personal fees from Daiichi Sankyo, Bayer, Astellas Pharma, Pfizer Pharmaceuticals, Novartis Pharmaceuticals, Kowa Company, Otsuka Pharmaceutical, AstraZeneca, Eli Lilly Japan, Ono Pharmaceutical, TOA EIYO, Kyowa Kirin, and Boehringer Ingelheim Japan. Dr Sotomi has received grants from Roche Diagnostics, FUJIFILM Toyama Chemical, Abbott Medical Japan, Biosense Webster, Boston Scientific Japan, Bristol‐Myers Squibb, TERUMO, TOA EIYO, and NIPRO, and received personal fees from Abbott Medical Japan, Amgen Astellas BioPharma, and Bristol‐Myers Squibb. Dr Sakata received grants from Roche Diagnostic, FUJIFILM Toyama Chemical, Bristol‐Myers Squibb, Biosense Webster, Inc., Abbott Medical Japan, Otsuka Pharmaceutical, Daiichi Sankyo Company, Mitsubishi Tanabe Pharma Corporation, Astellas Pharma, Kowa Company, Boehringer Ingelheim Japan, and Biotronik; and received personal fees from Otsuka Pharmaceutical, Ono Pharmaceutical, Daiichi Sankyo, Mitsubishi Tanabe Pharma Corporation, AstraZeneca, and Actelion Pharmaceuticals. The remaining authors have no disclosures to report.

Supporting information

Data S1

Tables S1–S2

Figures S1–S2