Prognostic Impact of Echocardiographic Diastolic Dysfunction on Outcomes in Patients With Heart Failure With Preserved Ejection Fraction　― Insights From the PURSUIT-HFpEF Registry ―

Abstract

Background: Although diastolic dysfunction is important pathophysiology in heart failure with preserved ejection fraction (HFpEF), its prognostic impact in HFpEF patients, including those with atrial fibrillation (AF), remains to be elucidated.

Methods and Results: We included the data for 863 patients (321 patients with AF) registered in a prospective multicenter observational study of patients with HFpEF. Patients were divided into 3 groups according to the 2016 ASE/EACVI recommendations. The primary endpoint was a composite of all-cause death or HF rehospitalization. Median age was 83 years, and 55.5% were female. 196 (22.7%) were classified with normal diastolic function (ND), 253 (29.3%) with indeterminate (ID) and 414 (48.0%) with diastolic dysfunction (DD). The primary endpoint occurred more frequently in patients with DD than in those with ND or ID (log-rank P<0.001 for DD vs. ND, and log-rank P=0.007 for DD vs. ID, respectively). Taking ND as the reference, multivariable Cox regression analysis revealed that DD (hazard ratio (HR): 1.57, 95% confidence interval (CI):1.06–2.32, P=0.024) was independently associated with the composite endpoint, whereas ID (HR: 1.28, 95% CI: 0.84–1.95, P=0.255) was not. DD was associated with the composite endpoint in both patients with and without AF.

Conclusions: HFpEF patients classified with DD using the 2016 ASE/EACVI recommendations had worse clinical outcomes than those with ND or ID. DD may be considered a prognostic marker in patients with HFpEF regardless of AF.

The number of patients with heart failure with preserved ejection fraction (HFpEF) has increased dramatically with the aging of society.¹ Nevertheless, a specific and effective therapy for HFpEF has yet to be established.^2–5 It is well known that the pathophysiology of HFpEF is complex and heterogeneous,⁶ which makes the development of an effective therapy challenging. Appropriate classification of HFpEF patients based on pathophysiology is important in improving prognosis. Although assessment of diastolic function with Doppler echocardiography is still imprecise and controversial, and various criteria have been used in different studies,^7,8 left ventricular (LV) diastolic dysfunction may be a significant mechanism of HFpEF. However, results for the effect of diastolic dysfunction on prognosis in HFpEF among several substudies of randomized clinical trials (RCTs)^7,9,10 have been conflicting, due at least partially to heterogeneity in diagnostic criteria and the lack of a single reliable index of diastolic function. Moreover, those studies excluded patients with AF, because they used mitral inflow pattern, which is not available in AF, as criteria for diastolic dysfunction. Current diagnostic criteria of HFpEF^11,12 include the criteria for functional or structural alterations in LV diastolic function, which are not, however, criteria for diastolic dysfunction. Therefore, the effect and role of diastolic dysfunction in patients with HFpEF, including those with AF, who account for nearly one-third of HFpEF patients, remain to be elucidated. Recently, the American Society of Echocardiography (ASE) and European Association of Cardiovascular Imaging (EACVI) updated their algorithms for the evaluation of LV diastolic function by echocardiography, one of which is applicable even in patients with AF. In this study, we aimed to clarify the effect of diastolic dysfunction on prognosis in patients with HFpEF, including those with AF, using this algorithm. In addition, we examined the heterogeneity of cardiac geometry in patients with HFpEF.

Editorial p ????

Methods

The PURSUIT-HFpEF Registry

The PURSUIT-HFpEF study is a prospective, multicenter, observational study investigating the therapeutic procedures and prognosis of patients with HFpEF. The design of the study was described previously.¹³ Briefly, it is being conducted by Osaka University Hospital, in collaboration with 30 hospitals in the Kansai region of Japan (UMIN-CTR ID: UMIN000021831).¹³ The registry aims to collect a comprehensive range of patient data, including demographic data, blood samples, echocardiographic, therapeutic and prognostic information, for use in clarifying the pathophysiology of HFpEF and determining the prognostic factors. The registry enrolled HFpEF patients admitted due to decompensated HF based on the Framingham heart failure diagnostic criteria. HFpEF was diagnosed with the following criteria: (1) LV ejection fraction (LVEF) ≥50% and (2) NT-pro B-type natriuretic peptide (BNP) ≥400 pg/mL or BNP ≥100 pg/mL on admission. Written informed consent was given by each participating patient and the study protocol was reviewed and approved by the ethics committee of each participating hospital in accordance with the principles of the Declaration of Helsinki.

Study Subjects

We analyzed patients in the PURSUIT-HFpEF registry who were admitted from June 2016 to June 2020 and discharged alive. We excluded patients who died in hospital, as well as those with cardiac amyloidosis, sarcoidosis, and pulmonary arterial hypertension because of the rarity and unique pathophysiology of those conditions, and patients missing ≥2 echocardiographic parameters necessary for the classification of diastolic function. We included and analyzed patients with all cardiac rhythms, including AF, in this study.

Echocardiographic Examination

Transthoracic echocardiography (TTE) was performed at both admission and discharge, but we used the data from the discharge TTE in this study because they may be more appropriate and have a significant effect on the long-term prognosis of the study cohort than the admission data. Various parameters were selected: left atrial dimension (LAD), left atrial volume index (LAVI) (left atrial volume measurements were obtained in the apical 4-chamber view at end-systole), the ratio of early diastolic velocity on transmitral Doppler and early diastolic velocity of the mitral valve annulus obtained from tissue Doppler (E/e’), LV end-diastolic dimension (LVDd), LV end-systolic dimension (LVDs), interventricular septal thickness at end-diastole (IVSTd), LV posterior wall thickness at end-diastole (LVPWTd), LVEF, tricuspid annular plane systolic excursion (TAPSE), peak tricuspid regurgitation velocity (TRV) and others in accordance with the recommendations of the ASE.^14,15 In patients with AF, recordings of 5–7 consecutive beats were recommended. In addition, single-beat measurement of systolic or diastolic parameters for 1 beat occurring after 2 serial beats with average RR interval or 1 beat with an average Doppler-wave contour with an average velocity were also permitted in AF patients in accordance with previous studies.^16,17 All echocardiographic measurements were performed in a standardized fashion by well-trained echocardiographers at each corresponding institution.

Definition of Diastolic Dysfunction

Diastolic dysfunction (DD) was defined using the algorithm of the 2016 ASE/EACVI recommendations (Supplementary Figure),¹⁴ which we applied to all study patients including those with AF.¹⁴ The algorithm uses 4 criteria: (1) average E/e’ >14; (2) septal e’ velocity <7 cm/s or lateral e’ velocity <10 cm/s; (3) tricuspid regurgitation (TR) peak velocity >2.8 m/s; and (4) LAVI >34 mL/m². DD was considered in patients who had at least 3 of the parameters, normal diastolic function (ND) was considered if patients presented with only one or none of these criteria, and indeterminate diastolic function (ID) was considered with 2 criteria. For patients in whom 1 of the 4 parameters was missing and 3 other parameters were available, the algorithm was applied as follows: if 2 parameters did not meet the cutoff values, patients were classified as having ND; if 2 parameters met the cutoff values, patients were classified as having DD.^14,18–20

Classification of LV Geometry

LV geometry was defined by relative wall thickness (RWT) and LV mass index (LVMI). Calculation of RWT was performed with the formula: (2 × posterior wall thickness) / (LV internal diameter at end-diastole), LVM measurements were based on a linear method, and LVMI was derived from LVM indexed to body surface area.¹⁵ Classification of LV geometry was based on a previously reported recommendation¹⁵ and classified into 4 groups: normal geometry (RWT ≤0.42 and LVMI ≤115 g/m² in males or ≤95 g/m² in females), concentric remodeling (RWT >0.42 and LVMI ≤115 g/m² in males or ≤95 g/m² in females), eccentric hypertrophy (RWT ≤0.42 and LVMI >115 g/m² in males or >95 g/m² in females), and concentric hypertrophy (RWT >0.42 and LVMI >115 g/m² in males or >95 g/m² in females).¹⁵

Study Endpoints

The primary endpoint was a composite of all-cause death or rehospitalization for HF during the follow-up period. The secondary endpoints were each individual endpoint of all-cause death and rehospitalization for HF.

Statistical Analysis

The nonparametric Kruskal Wallis test was used for both normally and non-normally distributed continuous variables, and expressed as median [interquartile range: 25–75%]. Categorical variables were compared using the chi-square test or Fisher’s exact test where appropriate, and presented as frequency (percentages). Incidence rates of clinical outcomes according to diastolic function were calculated with the number of events divided by person-year. Kaplan-Meier curves were constructed for the composite endpoint and for each individual endpoint stratified by diastolic function. Statistical differences among the 3 groups were estimated by the log-rank test. Post-hoc test with Bonferroni correction was performed for any variables with significant P value by the Kruskal Wallis test, chi-square test or Fisher’s exact test, and log-rank test. Univariable and multivariable Cox regression analyses were conducted to investigate the effect of ID and DD on the study outcomes, using ND with a hazard ratio (HR) of 1.00 as the reference. Variables inserted into the multivariable models were any covariates with a P value <0.05 by univariable analysis or clinically relevant factors including age, female sex, hypertension, diabetes mellitus, AF at discharge, hemoglobin, estimated glomerular filtration rate, albumin, (log-transformed) NT-proBNP, hypertrophic cardiac geometry (eccentric or concentric hypertrophy), New York Heart Association (NYHA) classification ≥II, angiotensin-converting enzyme inhibitors/angiotensin-receptor blockers, and diuretics. In addition, a Cox regression model was used to test interaction, with adjustment for age and sex, to determine whether the relationship between the presence of DD and the composite endpoint varied in the subgroup of patients without or with AF. Univariable Cox regression analyses were performed to examine the association between each criterion for DD and each endpoint. A two-sided P value <0.05 was considered to indicate statistical significance, or P<0.05/3 for post-hoc test. All analyses were performed using IBM SPSS Statistics for Windows (version 26.0, Armonk, NY, USA).

Results

Patient Selection

Of 1,024 consecutive patients admitted from June 2016 to June 2020 and enrolled in the PURSUIT-HFpEF study, we excluded 16 cases of in-hospital death, 9 cases of cardiac amyloidosis and sarcoidosis, 6 cases of pulmonary arterial hypertension, and 130 cases of insufficient echocardiographic data (missing ≥2 parameters of diastolic function classification). Among the 130 cases, average E/e’ was missing in 125 cases (96.2%), septal e’ or lateral e’ velocity in 111 cases (85.4%), TR peak velocity in 79 cases (60.8%), and LAVI in 102 cases (78.5%). Finally, data for 863 patients (321 with AF: 37.2%) who were discharged alive with available data on echocardiography for the classification of diastolic function were used in our analysis (Figure 1). Characteristics of the excluded and included patient populations are presented in Supplementary Table 1. The number of adverse events was not significantly different between the included and excluded patients.

Figure 1.

Patient selection and classification. AF, atrial fibrillation; DF, diastolic function.

Patients’ Characteristics

Median age of all study subjects was 83 [77, 87] years and 55.5% were female. The number of patients with ND, ID and DD was 196 (22.7%), 253 (29.3%) and 414 (48.0%), respectively. Characteristics of each group of patients are described in Table 1. The frequency of AF at discharge was not significantly different among the 3 groups (Table 1). Baseline characteristics of the AF-only patients according to the groups of diastolic function were also examined (Supplementary Table 2).

Table 1. Baseline Patient Characteristics

	ND (n=196)	ID (n=253)	DD (n=414)	Total (n=863)	P value
Age, years	81 [75, 86]	83 [78, 87]*	84 [78, 88]*	83 [77, 87]	0.001
Female, n (%)	77 (39.3)	134 (53.0)*	268 (64.7)*,†	479 (55.5)	<0.001
Body mass index, kg/m2	21.1 [18.5, 24.0]	20.8 [18.8, 23.4]	22.3 [19.6, 24.7]*,†	21.4 [19.0, 24.1]	<0.001
SBP, mmHg	118 [105, 130]	118 [106, 130]	120 [108, 132]	119 [107, 131]	0.130
Heart rate, beats/min	72 [63, 81]	69 [61, 79]	70 [61, 79]	70 [61, 80]	0.116
Medical history, n (%)
Previous HF admission	29 (15.2)	56 (22.5)	119 (29.4)*	204 (24.1)	0.001
Hypertension	160 (81.6)	215 (85.0)	362 (88.1)	737 (85.7)	0.098
Diabetes mellitus	56 (28.7)	80 (31.9)	155 (37.7)	291 (34.0)	0.065
Dyslipidemia	64 (32.7)	109 (43.6)	185 (45.0)*	358 (41.8)	0.012
Ischemic heart disease	26 (13.4)	45 (18.1)	81 (20.1)	152 (18.0)	0.136
Chronic kidney disease	65 (33.5)	102 (40.6)	172 (41.8)	339 (39.6)	0.136
AF at discharge, n (%)	65 (33.2)	106 (41.9)	150 (36.2)	321 (37.2)	0.141
Pacemaker implantation, n (%)	14 (7.1)	27 (10.7)	30 (7.3)	71 (8.2)	0.245
NYHA classification ≥II, n (%)	117 (60.6)	164 (65.1)	263 (64.5)	544 (63.8)	0.577
Blood tests at discharge
Hemoglobin, g/dL	11.7 [10.3, 13.0]	11.7 [10.5, 13.2]	10.9 [9.6, 12.2]*,†	11.3 [10.1, 12.6]	<0.001
Creatinine, mg/dL	1.1 [0.8, 1.4]	1.1 [0.9, 1.4]	1.1 [0.9, 1.6]	1.1 [0.9, 1.5]	0.273
eGFR, mL/min/1.73 m2	47.5 [33.3, 59.8]	44.2 [30.5, 55.4]	40.2 [27.4, 51.7]*	41.8 [29.7, 54.8]	<0.001
Albumin, g/dL	3.4 [3.1, 3.7]	3.5 [3.2, 3.7]	3.3 [3.0, 3.7]†	3.4 [3.1, 3.7]	0.003
C-reactive protein, mg/dL	0.27 [0.12, 0.83]	0.22 [0.10, 0.62]	0.34 [0.14, 1.02]†	0.29 [0.11, 0.83]	0.009
NT-proBNP, pg/mL	670 [321, 1,620]	881 [435, 1,865]	1,387 [666, 3,224]*,†	1,070 [478, 2,426]	<0.001
HbA1c, %	5.8 [5.5, 6.5]	6.0 [5.6, 6.6]	6.0 [5.6, 6.6]	6.0 [5.6, 6.5]	0.143
Echocardiographic parameters
Left atrial dimension, mm	39 [35, 46]	44 [39, 48]*	45 [41, 50]*,†	44 [39, 49]	<0.001
LAVI, mL/m2 (modified Simpson)	31 [27, 42]	51 [39, 64]	54 [43, 71]	49 [36, 64]
Average E/e’	9.9 [8.2, 11.4]	10.8 [9.1, 12.7]	16.7 [14.2, 20.0]	12.5 [9.7, 16.7]
Septal e’ velocity, cm/s	6.1 [4.6, 7.8]	5.5 [4.3, 6.9]	5.0 [3.9, 5.9]	5.2 [4.0, 6.5]
Lateral e’ velocity, cm/s	8.3 [6.0, 11.0]	7.7 [6.0, 9.5]	6.6 [5.0, 8.0]	7.1 [5.6, 9.1]
TRPG, mmHg	23.0 [19.1, 27.0]	25.0 [21.0, 30.0]	31.7 [25.0, 37.0]	27.0 [22.0, 32.6]
LVDd, mm	44.0 [40.4, 49.0]	45.0 [41.0, 49.8]	46.0 [41.0, 50.2]*	45.0 [41.0, 50.0]	0.043
LVDs, mm	28.4 [25.2, 32.0]	30.0 [26.0, 32.4]	29.2 [26.0, 33.0]	29.0 [26.0, 32.5]	0.155
IVSTd, mm	9.1 [8.0, 11.0]	10.0 [8.7, 11.0]	10.0 [9.0, 11.1]*	10.0 [8.8, 11.0]	0.032
LVPWTd, mm	10.0 [8.2, 11.0]	9.3 [8.4, 10.9]	10.0 [9.0, 11.0]†	10.0 [8.9, 11.0]	0.012
LVEF (m-Simpson), %	61.0 [56.0, 64.9]	61.3 [55.7, 65.6]	61.0 [55.0, 66.0]	61.0 [55.4, 65.6]	0.958
LVMI, g/m2	96 [81, 114]	100 [81, 120]	107 [90, 130]*,†	102 [85, 123]	<0.001
RWT	0.42 [0.38, 0.50]	0.42 [0.37, 0.48]	0.43 [0.37, 0.50]	0.42 [0.37, 0.50]	0.312
TAPSE, mm	17.8 [15.0, 20.6]	17.3 [14.1, 20.4]	17.6 [14.6, 20.0]	17.5 [14.6, 20.3]	0.717
Average E/e’ >14, n (%)	0 (0)	24 (9.5)	302 (76.3)	326 (38.9)
Septal e’ velocity <7 cm/s or lateral e’ velocity <10 cm/s, n (%)	125 (63.8)	229 (90.5)	409 (98.8)	763 (88.4)
TR velocity >2.8 m/s, n (%)	6 (3.4)	41 (16.2)	187 (50.5)	234 (29.3)
LAVI >34 mL/m2, n (%)	45 (28.5)	212 (83.8)	359 (97.6)	616 (79.1)
No. of positive criteria from 2016 ASE/EACVI recommendations,a,14 n (%)					<0.001
0	20 (10.2)	0 (0)	0 (0)	20 (2.3)
1	176 (89.8)	0 (0)	0 (0)	176 (20.4)
2	0 (0)	253 (100)	71 (17.1)	324 (37.5)
3	0 (0)	0 (0)	257 (62.1)	257 (29.8)
4	0 (0)	0 (0)	86 (20.8)	86 (10.0)
HFA-PEFF score, n (%)					<0.001
1	6 (3.1)	0 (0)	0 (0)	6 (0.7)
2–4	120 (61.5)	73 (28.9)	84 (20.3)	277 (32.1)
≥5	69 (35.4)	180 (71.1)	330 (79.7)	579 (67.2)
Medications at discharge, n (%)
ACEI/ARBs	112 (57.1)	128 (50.6)	239 (57.7)	479 (55.5)	0.173
Calcium-channel blockers	92 (46.9)	114 (45.2)	233 (56.3)†	439 (50.9)	0.010
β-blockers	106 (54.1)	139 (55.2)	233 (56.3)	478 (55.5)	0.873
Diuretics	147 (75.0)	210 (83.0)	349 (84.3)*	706 (81.8)	0.018

*Significant vs. ND with Bonferroni correction post-hoc tests. ^†Significant vs. ID with Bonferroni correction post-hoc tests. Values are expressed as number (%), and median [interquartile range]. ^a(1) average E/e’ >14, (2) septal e’ velocity <7 cm/s or lateral e’ velocity <10 cm/s, (3) TR velocity >2.8 m/s, (4) LAVI >34 mL/m². ACEI, angiotensin-converting enzyme inhibitors; AF, atrial fibrillation; ARBs, angiotensin-receptor blockers; DD, diastolic dysfunction; eGFR, estimated glomerular filtration rate; HF, heart failure; ID, indeterminate diastolic function; IVSTd, interventricular septal thickness at end-diastole; LAVI, left atrial volume index; LVDd, left ventricular diameter at end-diastole; LVDs, left ventricular diameter at end-systole; LVEF, left ventricular ejection fraction; LVMI, left ventricular mass index; LVPWTd, left ventricular posterior wall thickness at end-diastole; ND, normal diastolic function; NYHA, New York Heart Association; RWT, relative wall thickness; SBP, systolic blood pressure; TAPSE, tricuspid annular plane systolic excursion; TRPG, tricuspid regurgitation peak gradient.

Association Between Diastolic Function and LV Structure

With respect to cardiac geometry, normal geometry was observed in 38% of patients with ND, 31% of those with ID, and 25% of those with DD. Concentric remodeling was 29%, 25%, and 18%, respectively, eccentric hypertrophy was 11%, 19% and 23%, respectively, and concentric hypertrophy was 22%, 25%, and 34%, for ND, ID and DD, respectively (P<0.001 for overall difference by chi-square test, Figure 2).

Figure 2.

Distribution of left ventricular geometries in 3 groups of diastolic function. (A) Normal diastolic function (ND), (B) indeterminate diastolic function (ID), (C) diastolic dysfunction (DD). P<0.001 for overall difference by chi-square test.

Prognosis According to Echocardiographic Diastolic Function

During a median follow-up period of 345 (18–672) days, Kaplan-Meier analyses showed that a composite of all-cause death or HF rehospitalization in DD was significantly higher than that in ND or ID, whereas that in ID was not significantly different from that in ND (Figure 3A). In addition, all-cause death in the DD group was significantly higher than that in ND, but there were no significant differences between DD and ID, or between ID and ND (Figure 3B). The prevalence of rehospitalization for HF in the DD group was significantly higher than that in ND, but there were no significant differences between DD and ID, or between ID and ND (Figure 3C). The detailed incidence rates of each event are shown in Table 2.

Figure 3.

Kaplan-Meier curves according to groups of diastolic function: (A) composite endpoint, (B) all-cause mortality, (C) HF rehospitalization. *P<0.05/3 with Bonferroni correction post-hoc tests. DD, diastolic dysfunction; ID, indeterminate diastolic function; ND, normal diastolic function.

Table 2. IRs of Clinical Outcomes According to Diastolic Function

Diastolic function	Person-year	Cases (n)	IR per 1,000 person-year
A composite of all-cause death or HF rehospitalization
ND	246.71	44	178
ID	261.15	67	257
DD	405.30	156	385
HF rehospitalization
ND	246.71	28	113
ID	261.15	48	184
DD	405.30	107	264
All-cause death
ND	277.05	20	72
ID	305.88	29	95
DD	508.64	77	151
Cardiac death
ND	277.05	7	25
ID	305.88	13	42
DD	508.64	32	63
Noncardiac death
ND	277.05	13	47
ID	305.88	16	52
DD	508.64	45	88

IR, incidence rate. Other abbreviations as in Table 1.

Association Between Echocardiographic Diastolic Function and Prognosis

By taking patients with ND as the reference, univariable Cox regression analysis revealed that the presence of DD was significantly associated with the composite endpoint, whereas that of ID was not (Table 3). Multivariable Cox regression analysis indicated that the presence of DD remained significantly associated with the composite endpoint, whereas the presence of ID was not (Table 3). Regarding each individual endpoint, the presence of DD was not associated with all-cause death (Supplementary Table 3), but was independently associated with HF rehospitalization (Supplementary Table 4).

Table 3. Univariable and Multivariable Cox Regression Analyses for the Composite Endpoint of All-Cause Death or HF Rehospitalization

Variable	Univariable			Multivariable
	HR	95% CI	P value	HR	95% CI	P value
Diastolic function
ND (reference)	1.00			1.00
ID	1.42	0.97–2.08	0.071	1.28	0.84–1.95	0.255
DD	2.09	1.50–2.93	<0.001	1.57	1.06–2.32	0.024
Age	1.04	1.02–1.06	<0.001	1.03	1.01–1.05	0.002
Sex (female vs. male)	1.18	0.92–1.51	0.186	0.96	0.72–1.28	0.770
Body mass index, kg/m2	0.98	0.95–1.01	0.108	–	–	–
SBP per 10 mmHg increase	1.02	0.95–1.10	0.541	–	–	–
Heart rate per 5 beats/min increase	1.05	1.00–1.09	0.052	–	–	–
Hypertension	0.98	0.69–1.38	0.902	0.91	0.60–1.38	0.672
Diabetes mellitus	1.20	0.93–1.53	0.157	1.15	0.87–1.52	0.323
Atrial fibrillation at discharge	1.25	0.98–1.60	0.069	1.18	0.89–1.57	0.244
Hemoglobin, g/dL	0.84	0.79–0.90	<0.001	0.93	0.86–1.01	0.098
eGFR, mL/min/1.73 m2	0.99	0.98–0.99	<0.001	1.00	0.99–1.01	0.946
Albumin, g/dL	0.49	0.38–0.64	<0.001	0.71	0.51–0.98	0.037
C-reactive protein, mg/dL	1.05	0.98–1.13	0.138	–	–	–
LogNT-proBNP, pg/mL	2.45	1.94–3.10	<0.001	2.08	1.50–2.88	<0.001
TAPSE, mm	0.99	0.96–1.02	0.364	–	–	–
Hypertrophic cardiac geometry	1.26	0.99–1.60	0.059	0.98	0.73–1.30	0.876
NYHA classification ≥II	1.49	1.15–1.93	0.003	1.05	0.78–1.42	0.742
ACEI/ARBs	0.81	0.64–1.03	0.088	1.00	0.76–1.32	0.981
Calcium-channel blockers	1.01	0.79–1.28	0.952	–	–	–
β-blockers	1.15	0.90–1.47	0.263	–	–	–
Diuretics	1.55	1.09–2.21	0.016	1.30	0.88–1.92	0.194

Adjusted variables: age, sex (female), hypertension, diabetes mellitus, atrial fibrillation at discharge, hemoglobin, eGFR, albumin, (log-transformed) NT-proBNP, hypertrophic cardiac geometry (eccentric or concentric hypertrophy), NYHA classification ≥II, ACEI/ARBs, and diuretics. CI, confidence interval; HR, hazard ratio. Other abbreviations as in Table 1.

We further examined the effect of DD on prognosis in patients with and without AF. In 321 patients with AF, Kaplan-Meier analysis and multivariable Cox regression analysis adjusted for age and sex demonstrated a higher risk for the composite endpoint in DD than in ND (Figure 4B). Similar findings were observed (Figure 4A) in 542 patients without AF. There was no significant interaction between the presence of DD and AF for the composite endpoint (P for interaction=0.996).

Figure 4.

Kaplan-Meier curves for the composite endpoint according to groups of diastolic function in patients without (A) and with (B) atrial fibrillation. *P<0.05/3 with Bonferroni correction post-hoc tests. P for interaction=0.996. DD, diastolic dysfunction; ID, indeterminate diastolic function; ND, normal diastolic function.

Prognostic Value of Each DD Criterion

Average E/e’ >14 and TR velocity >2.8 m/s were significantly associated with higher risk of the composite endpoint, all-cause death and HF rehospitalization. LAVI >34 mL/m² was significantly associated with the composite endpoint and HF rehospitalization but not with all-cause death. Septal e’ <7 cm/s or lateral e’ <10 cm/s was not associated with any endpoints (Table 4).

Table 4. Prognostic Value of Each Parameter of Diastolic Dysfunction by Univariable Cox Regression

	HR	95% CI	P value
Composite of all-cause death and HF rehospitalization
Average E/e’ >14	1.64	1.28–2.10	<0.001
Septal e’ <7 cm/s or lateral e’ <10 cm/s	1.12	0.78–1.62	0.537
TR velocity >2.8 m/s	1.73	1.34–2.23	<0.001
LAVI >34 mL/m2	1.90	1.32–2.74	0.001
All-cause death
Average E/e’ >14	1.81	1.26–2.60	0.001
Septal e’ <7 cm/s or lateral e’ <10 cm/s	1.35	0.76–2.40	0.308
TR velocity >2.8 m/s	1.97	1.36–2.84	<0.001
LAVI >34 mL/m2	1.44	0.87–2.39	0.157
HF rehospitalization
Average E/e’ >14	1.54	1.14–2.07	0.005
Septal e’ <7 cm/s or lateral e’ <10 cm/s	1.11	0.71–1.71	0.650
TR velocity >2.8 m/s	1.62	1.19–2.21	0.002
LAVI >34 mL/m2	2.33	1.46–3.72	<0.001

TR, tricuspid regurgitation. Other abbreviations as in Tables 1,3.

Discussion

In this study, we found that the presence of DD evaluated noninvasively using the 2016 ASE/EACVI recommendations was significantly associated with poor prognosis in HFpEF patients including those with AF. A certain prevalence of HFpEF patients showed ND (22.7%) or ID (29.3%). The association between DD and worse clinical outcomes was independent of major clinical confounding factors. Prognostic significance of DD evaluated with the algorithm was similarly observed even in patients with AF. This is the first report to clarify the effect of LV DD on the prognosis of HFpEF patients including those with AF using the 2016 ASE/EACVI recommendations,¹⁴ which are stricter criteria than the current diagnostic criteria of HFpEF.^11,12

Comparison With Previous Studies and Significance of This Study

Several previous substudies of RCTs have reported conflicting results for the relationship between DD and prognosis in HFpEF patients.^7,9 The CHARM echocardiographic substudy (CHARMES) found that moderate or severe DD diagnosed with the originally defined criteria was an independent predictor of cardiovascular death or hospitalization for HF.⁷ Although those results were similar to ours, their study population as well as the criteria to classify DD differed from the contemporary HFpEF criteria. Another substudy from I-PRESERVE revealed that DD was not associated with an increase in cardiovascular events.⁹ The study population consisted of patients with LVEF ≥45% and sinus rhythm only, unlike our study (LVEF ≥50%; both with and without AF), and the criteria for LV DD were also different. These differences in study population and criteria might explain the apparent discrepancy between our present and those previous results.

On the other hand, recently released diagnostic algorithms of HFpEF^11,12,21 include functional or structural alterations on echocardiography or different cutoff values of natriuretic peptide according to the presence of AF, whereas our current study did not. Instead, our criteria required hospitalization due to HF diagnosed using Framingham criteria, which may contribute to improving the accuracy of diagnosing HFpEF. Actually, >97% (843/863) of patients in this study had ≥1 functional or structural alteration, and all patients except 6 were at least diagnosed as intermediate or probable HFpEF by the HFA-PEFF score,¹² indicating that HFpEF patients diagnosed with the criteria used in this study and those using recent diagnostic algorithms are largely identical (Table 1). Regarding the definition of echocardiographic DD, the algorithm of the ASE/EACVI 2016 guideline used in this study requires more than half of 3–4 available variables to diagnose DD, which is stricter than the recent diagnostic algorithms. Collectively, although the criteria used for diagnosis of HFpEF and evaluation of diastolic function in this study were not identical to recent criteria, our study clarified the prognostic effect of echocardiographic DD diagnosed using authorized specific and more rigorous criteria in HFpEF patients who were defined with our criteria but almost fulfill the recent diagnostic algorithms.

Advantages of Using 2016 ASE Algorithm for Diagnosis of LV DD

Although the assessment of DD with Doppler echocardiography is still imprecise and controversial, an algorithm has been proposed. Recently updated 2016 ASE/EACVI algorithms for the evaluation of LV diastolic function by echocardiography¹⁴ have 2 parts: the first is for diagnosis of LV DD in patients with normal LVEF, and the second is for estimating left atrial pressure and grading LV DD, albeit only in patients with sinus rhythm. Several validation studies revealed that the second algorithm showed better prediction of left atrial pressure and prognosis in patients with various cardiac diseases including HF in sinus rhythm^18,22,23 than a previous guideline.²⁴ Lately, a single-center study reported that elevated left atrial pressure assessed by the second algorithm was associated with worse outcomes in HFpEF patients including both sinus rhythm and AF, while patients with AF were assessed by another newly proposed algorithm.¹⁷ In contrast, our study classified the diastolic function of HFpEF patients including those with AF using the first part of the 2016 ASE/EACVI algorithm (Supplementary Figure), and found that DD was associated with poor prognosis. Furthermore, a similar association was observed in both patients with and without AF at discharge. These findings suggest that we can assess the risk of DD in HFpEF patients regardless of cardiac rhythm using the abovementioned algorithm.

Significance of Patients With ID

In the present study, the frequency of patients with ID was ≈29.3%, and their prognosis was similar to that of patients with ND. The reported frequency of this group ranges from 11% to 23% in various cardiac diseases,^18,25 and their prognosis or diastolic function parameters are intermediate between normal diastolic function and diastolic dysfunction in patients with suspected HF.²⁵ Different patient characteristics, as well as a diversity of disease conditions, could explain this inconsistent frequency and prognosis of the indeterminate group. A more detailed characterization of this group is warranted.

Relationship Between Diastolic Function, Cardiac Structure and Prognosis

Although cardiac structural remodeling is common^26,27 and is an emerging pathophysiology of DD in HFpEF,²⁸ 25% of patients with DD showed normal LV structure in our study. Previous studies have not reported the proportion of patients with DD who have normal cardiac structure.^9,29 Our findings suggest that a certain percentage of patients develop DD without overt cardiac remodeling.

Our data also demonstrated that the hypertrophic cardiac geometries (concentric hypertrophy and eccentric hypertrophy) were dominant in patients with DD. However, multivariable Cox regression analysis showed that the prognostic effect of DD was independent of hypertrophic phenotype, and cardiac hypertrophy itself was not associated with adverse outcomes in our cohort. In contrast, Katz et al showed that LV hypertrophy was a significant prognostic indicator in HFpEF.³⁰ These findings suggest that complex relationships exist among diastolic dysfunction, cardiac remodeling, and prognosis.

HFpEF Patients With ND

We found that a substantial proportion of patients developed HFpEF despite having ND on echocardiography. One possibility is that these patients are in the early stage of the disease, whereas functional or structural changes in diastolic function occur at the later stage of disease progression as previously demonstrated.^31,32 A second possibility is that they might have DD or elevated LV filling pressure on exertion,³³ which could not be detected with echocardiographic tests at rest. Otherwise, other multiple extracardiac pathophysiologies, including endothelial dysfunction, arterial stiffness, autonomic dysfunction, renal dysfunction, global cardiovascular reserve dysfunction and others, may play a dominant role in the development of HFpEF in this group.^34–36

Clinical Implications

This study emphasizes the importance of evaluating diastolic function in patients with HFpEF as routine clinical practice in order to classify high-risk patients and provide better care. In addition, this study suggests that HFpEF patients presenting with normal diastolic function might be excluded from future clinical trials because they are low-risk patients who will require different therapeutic strategies from high-risk populations. Future RCTs should specifically recruit high-risk HFpEF patients requiring appropriate interventions, and may thereby show promising results and offer better insights into the management of this challenging disease.

Study Limitations

Several limitations of this observational study need to be acknowledged. First, both the number of patients and the event rate of clinical outcomes were relatively limited, and the follow-up period was rather short. Further investigation to confirm our results with a larger sample size is warranted. Second, although a decrease in LV systolic performance evaluated with LV global longitudinal strain is reportedly present in some patients with normal LVEF,³⁷ we did not examine it. Therefore, it is possible that some patients potentially had LV systolic dysfunction. Third, whether the same cutoff values for LAVI, e’ and E/e’ can be applied to patients with AF remain to be determined. Lastly, we did not have data regarding the exact cardiac rhythm during echocardiography; therefore, we divided patients with or without AF by the investigators’ clinical diagnosis at discharge. Our findings, therefore, need to be interpreted cautiously. Further investigation involving a more detailed and comprehensive assessment of cardiac performance may be useful in clarifying the mechanisms of our unique results.

Conclusions

In conclusion, HFpEF patients presenting with LV DD assessed using the 2016 ASE/EACVI recommendations had worse clinical outcomes than those with ND or ID among patients with HFpEF including those with AF. DD may therefore be considered a prognostic marker in patients with HFpEF regardless of the presence of AF.

Acknowledgments

The authors thank Sugako Mitsuoka, Masako Terui, Nagisa Yoshioka, Satomi Kishimoto, Kyoko Tatsumi and Noriko Murakami for their excellent assistance in data collection, data management and secretarial work, and Toshinari Onishi for technical advice on echocardiography.

Disclosures

S. Hikoso has received personal fees from Daiichi Sankyo Company, Bayer, Astellas Pharma, Pfizer Pharmaceuticals and Boehringer Ingelheim Japan, and received grants from Roche Diagnostics, FUJIFILM Toyama Chemical and Actelion Pharmaceuticals. N. Daisaku has received honoraria from Roche Diagnostics K.K. Y. Sakata has received personal fees from Otsuka Pharmaceutical, Ono Pharmaceutical, Daiichi Sankyo Company, Mitsubishi Tanabe Pharma Corporation and Actelion Pharmaceuticals, and received grants from Roche Diagnostic, FUJIFILM Toyama Chemical, Abbott Medical Japan, Otsuka Pharmaceutical, Daiichi Sankyo Company, Mitsubishi Tanabe Pharma Corporation and Biotronik. Y. Sakata is a member of Circulation Journal’s Editorial Team. Other authors have no conflicts of interest to disclose.

Sources of Funding

This work was supported by grants from Japan Society for the Promotion of Science KAKENHI (No. JP 17K09496) and Japan Agency for Medical Research and Development (No. JP16lk1010013).

IRB Information

Kansai Rosai Hospital Institutional Review Board (approval ID: 16co10 g), Kawachi General Hospital Ethics Committee (approval ID unavailable, but approved on 26 Apr 2016), Osaka Rosai Hospital Ethics Committee (approval ID: 28-5), Higashiosaka City Medical Center Institutional Review Board (approval ID: 02-0313), Osaka Prefectural Hospital Organization Osaka General Medical Center Institutional Review Board (approval ID: 28-2002), Hyogo Prefectural Nishinomiya Hospital Ethics Committee (approval ID: H28-3), Ikeda Municipal Hospital Ethics Committee (approval ID: 3280), Kawanishi City Hospital Institutional Review Board (approval ID: 28001), Rinku General Medical Center Ethics Committee (approval ID: 27-40), Saiseikai Senri Hospital Ethics Committee (approval ID: 280304), Yao Municipal Hospital Institutional Review Board (approval ID: H28-6), Kawasaki Hospital Ethics Committee (approval ID unavailable, but approved on 12 May 2016), Minoh City Hospital Ethics Committee (approval ID unavailable, but approved on 24 May 2016), National Hospital Organization Osaka National Hospital Second Institutional Review Board (approval ID: 16024), Kano General Hospital Ethics Committee (approval ID unavailable, but approved on 9 June 2016), Toyonaka Municipal Hospital Ethics Committee (approval ID: 2016-04-02), Kinan Hospital Ethics Committee 121, Japan Community Health Care Organization Osaka Hospital Ethics Committee (approval ID: 2016-2), Kobe Ekisaikai Hospital Ethics Committee (approval ID: 2016-3), Sakurabashi Watanabe Hospital Ethics Committee (approval ID: 16-15), Sumitomo Hospital Research Ethics Committee (approval ID: 28-01), Suita Municipal Hospital Institutional Review Board (approval ID: 2017-8), Kinki Central Hospital Ethics Committee (approval ID: 288), Osaka Police Hospital Institutional Review Board (approval ID: 593), Japan Community Health Care Organization Hoshigaoka Medical Center Institutional Review Board (approval ID: 1618), National Hospital Organization Osaka Minami Medical Center Institutional Review Board (approval ID: 28-3), Japan Community Health Care Organization Osaka Minato Central Hospital Ethics Committee (approval ID unvailable, but approved on 10 June 2016), Amagasaki Chuo Hospital Ethics Committee (approval ID unavailable, but approved on 1 Aug 2017), Otemae Hospital Institutional Review Board (approval ID: 2017-020), Osaka University Hospital Clinical Research Review Committee (approval ID: 15471), Osaka International Cancer Institute Institutional Review Board (No.20097).

Supplementary Files

Please find supplementary file(s);

http://dx.doi.org/10.1253/circj.CJ-21-0300

References

• 1.   Shimokawa H, Miura M, Nochioka K, Sakata Y. Heart failure as a general pandemic in Asia. Eur J Heart Fail 2015; 17: 884–892.
• 2.   Massie BM, Carson PE, McMurray JJ, Komajda M, McKelvie R, Zile MR, et al. Irbesartan in patients with heart failure and preserved ejection fraction. N Engl J Med 2008; 359: 2456–2467.
• 3.   Cleland JG, Tendera M, Adamus J, Freemantle N, Polonski L, Taylor J. The perindopril in elderly people with chronic heart failure (PEP-CHF) study. Eur Heart J 2006; 27: 2338–2345.
• 4.   Pitt B, Pfeffer MA, Assmann SF, Boineau R, Anand IS, Claggett B, et al. Spironolactone for heart failure with preserved ejection fraction. N Engl J Med 2014; 370: 1383–1392.
• 5.   Yusuf S, Pfeffer MA, Swedberg K, Granger CB, Held P, McMurray JJ, et al. Effects of candesartan in patients with chronic heart failure and preserved left-ventricular ejection fraction: The CHARM-Preserved Trial. Lancet 2003; 362: 777–781.
• 6.   Senni M, Paulus WJ, Gavazzi A, Fraser AG, Díez J, Solomon SD, et al. New strategies for heart failure with preserved ejection fraction: The importance of targeted therapies for heart failure phenotypes. Eur Heart J 2014; 35: 2797–2815.
• 7.   Persson H, Lonn E, Edner M, Baruch L, Lang CC, Morton JJ, et al. Diastolic dysfunction in heart failure with preserved systolic function: Need for objective evidence: Results from the CHARM Echocardiographic Substudy-CHARMES. J Am Coll Cardiol 2007; 49: 687–694.
• 8.   Bursi F, Weston SA, Redfield MM, Jacobsen SJ, Pakhomov S, Nkomo VT, et al. Systolic and diastolic heart failure in the community. JAMA 2006; 296: 2209–2216.
• 9.   Zile MR, Gottdiener JS, Hetzel SJ, McMurray JJ, Komajda M, Mckelvie R, et al. Prevalence and significance of alterations in cardiac structure and function in patients with heart failure and a preserved ejection fraction. Circulation 2011; 124: 2491–2501.
• 10.   Ohtani T, Mohammed SF, Yamamoto K, Dunlay SM, Weston SA, Sakata Y, et al. Diastolic stiffness as assessed by diastolic wall strain is associated with adverse remodelling and poor outcomes in heart failure with preserved ejection fraction. Eur Heart J 2012; 33: 1742–1749.
• 11.   Ponikowski P, Voors AA, Anker SD, Bueno H, Cleland JG, Coats AJ, et al. 2016 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure: The Task Force for the diagnosis and treatment of acute and chronic heart failure of the European Society of Cardiology (ESC)Developed with the special contribution of the Heart Failure Association (HFA) of the ESC. Eur Heart J 2016; 37: 2129–2200.
• 12.   Pieske B, Tschope C, de Boer RA, Fraser AG, Anker SD, Donal E, et al. How to diagnose heart failure with preserved ejection fraction: The HFA-PEFF diagnostic algorithm: A consensus recommendation from the Heart Failure Association (HFA) of the European Society of Cardiology (ESC). Eur Heart J 2019; 40: 3297–3317.
• 13.   Suna S, Hikoso S, Yamada T, Uematsu M, Yasumura Y, Nakagawa A, et al. Study protocol for the PURSUIT-HFpEF study: A prospective, multicenter, observational study of patients with heart failure with preserved ejection fraction. BMJ Open 2020; 10: e038294.
• 14.   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.
• 15.   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. J Am Soc Echocardiogr 2015; 28: 1–39.e14.
• 16.   Abe Y, Akamatsu K, Ito K, Matsumura Y, Shimeno K, Naruko T, et al. Prevalence and prognostic significance of functional mitral and tricuspid regurgitation despite preserved left ventricular ejection fraction in atrial fibrillation patients. Circ J 2018; 82: 1451–1458.
• 17.   Torii Y, Kusunose K, Yamada H, Nishio S, Hirata Y, Amano R, et al. Updated left ventricular diastolic function recommendations and cardiovascular events in patients with heart failure hospitalization. J Am Soc Echocardiogr 2019; 32: 1286–1297.
• 18.   Sato K, Grant ADM, Negishi K, Cremer PC, Negishi T, Kumar A, et al. Reliability of updated left ventricular diastolic function recommendations in predicting elevated left ventricular filling pressure and prognosis. Am Heart J 2017; 189: 28–39.
• 19.   Yamamoto J, Wakami K, Muto K, Kikuchi S, Goto T, Fukuta H, et al. Verification of echocardiographic assessment of left ventricular diastolic dysfunction in patients with preserved left ventricular ejection fraction using the American Society of Echocardiography and European Association of Cardiovascular Imaging 2016 recommendations. Circ Rep 2019; 1: 525–530.
• 20.   Chang SN, Sung KT, Huang WH, Lin JW, Chien SC, Hung TC, et al. Sex, racial differences and healthy aging in normative reference ranges on diastolic function in Ethnic Asians: 2016 ASE guideline revisited. J Formos Med Assoc, doi:10.1016/j.jfma.2020.12.024.
• 21.   Reddy YNV, Carter RE, Obokata M, Redfield MM, and Borlaug BA. A simple, evidence-based approach to help guide diagnosis of heart failure with preserved ejection fraction. Circulation 2018; 138: 861–870.
• 22.   Prasad SB, Lin AK, Guppy-Coles KB, Stanton T, Krishnasamy R, Whalley GA, et al. Diastolic dysfunction assessed using contemporary guidelines and prognosis following myocardial infarction. J Am Soc Echocardiogr 2018; 31: 1127–1136.
• 23.   Machino-Ohtsuka T, Seo Y, Ishizu T, Hamada-Harimura Y, Yamamoto M, Sato K, et al. Clinical utility of the 2016 ASE/EACVI recommendations for the evaluation of left ventricular diastolic function in the stratification of post-discharge prognosis in patients with acute heart failure. Eur Heart J Cardiovasc Imaging 2019; 20: 1129–1137.
• 24.   Nagueh SF, Appleton CP, Gillebert TC, Marino PN, Oh JK, Smiseth OA, et al. Recommendations for the evaluation of left ventricular diastolic function by echocardiography. J Am Soc Echocardiogr 2009; 22: 107–133.
• 25.   Sanchis L, Andrea R, Falces C, Poyatos S, Vidal B, Sitges M. Differential clinical implications of current recommendations for the evaluation of left ventricular diastolic function by echocardiography. J Am Soc Echocardiogr 2018; 31: 1203–1208.
• 26.   Shah AM. Ventricular remodeling in heart failure with preserved ejection fraction. Curr Heart Fail Rep 2013; 10: 341–349.
• 27.   Del Buono MG, Buckley L, and Abbate A. Primary and secondary diastolic dysfunction in heart failure with preserved ejection fraction. Am J Cardiol 2018; 122: 1578–1587.
• 28.   Jeong EM, Dudley SC Jr. Diastolic dysfunction. Circ J 2015; 79: 470–477.
• 29.   Shah AM, Shah SJ, Anand IS, Sweitzer NK, O’Meara E, Heitner JF, et al. Cardiac structure and function in heart failure with preserved ejection fraction: Baseline findings from the echocardiographic study of the Treatment of Preserved Cardiac Function Heart Failure with an Aldosterone Antagonist trial. Circ Heart Fail 2014; 7: 104–115.
• 30.   Katz DH, Beussink L, Sauer AJ, Freed BH, Burke MA, Shah SJ. Prevalence, clinical characteristics, and outcomes associated with eccentric versus concentric left ventricular hypertrophy in heart failure with preserved ejection fraction. Am J Cardiol 2013; 112: 1158–1164.
• 31.   Loai S, Cheng HM. Heart failure with preserved ejection fraction: The missing pieces in diagnostic imaging. Heart Fail Rev 2020; 25: 305–319.
• 32.   Ohara T, Ohte N, Little WC. Pulmonary hypertension in heart failure with preserved left ventricular ejection fraction: Diagnosis and management. Curr Opin Cardiol 2012; 27: 281–287.
• 33.   Borlaug BA, Kane GC, Melenovsky V, Olson TP. Abnormal right ventricular-pulmonary artery coupling with exercise in heart failure with preserved ejection fraction. Eur Heart J 2016; 37: 3293–3302.
• 34.   Shah SJ, Katz DH, Deo RC. Phenotypic spectrum of heart failure with preserved ejection fraction. Heart Fail Clin 2014; 10: 407–418.
• 35.   Pfeffer MA, Shah AM, and Borlaug BA. Heart failure with preserved ejection fraction in perspective. Circ Res 2019; 124: 1598–1617.
• 36.   Borlaug BA, Olson TP, Lam CS, Flood KS, Lerman A, Johnson BD, et al. Global cardiovascular reserve dysfunction in heart failure with preserved ejection fraction. J Am Coll Cardiol 2010; 56: 845–854.
• 37.   Kraigher-Krainer E, Shah AM, Gupta DK, Santos A, Clagget B, Pieske B, et al. Impaired systolic function by strain imaging in heart failure with preserved ejection fraction. J Am Coll Cardiol 2014; 63: 447–456.