OUP user menu

Myocardial systolic and diastolic consequences of left ventricular mechanical dyssynchrony in heart failure with normal left ventricular ejection fraction

Daniel A. Morris, Amalia Vaz Pérez, Florian Blaschke, Hermann Eichstädt, Cemil Özcelik, Wilhelm Haverkamp
DOI: http://dx.doi.org/10.1093/ehjci/jes042 556-567 First published online: 29 February 2012

Abstract

Aims The purpose of this study was to test the hypothesis that left ventricular (LV) mechanical dyssynchrony deteriorates the longitudinal systolic and diastolic function of the left ventricle (LV) in patients with heart failure with a normal LV ejection fraction (HFNEF).

Methods and results In patients with HFNEF and in a control group consisting of asymptomatic patients with LV diastolic dysfunction [LVDD], matched by age, gender, and LV ejection fraction, we assessed the global longitudinal systolic (global strain), diastolic [global early-diastolic strain rate (SRe)], and synchronous (standard deviation of time-to-peak systolic strain) function of the LV by two-dimensional speckle-tracking echocardiography using a 18-segment LV model. A total of 325 patients were included (85 with HFNEF and 240 with asymptomatic LVDD). Patients with HFNEF had a significant impairment of the longitudinal systolic and diastolic function of the LV as compared with the control group. Concerning the pathophysiological mechanisms linked to these findings, we found that HFNEF patients with asynchronous LV contractions had significantly more impaired longitudinal systolic and diastolic LV function (global strain −14.76 ± 3.44%, global SRe 0.79 ± 0.24 s−1) than patients without asynchronous LV contractions (global strain −18.57 ± 3.10%, global SRe 1.06 ± 0.32 s−1; all P < 0.0001). Accordingly, in HFNEF patients with LV mechanical dyssynchrony the rates of LV longitudinal systolic and diastolic dysfunction were 64 and 70%, respectively, whereas these rates were significantly lower (19.5 and 41.3%), respectively, in patients without asynchronous LV contractions. In addition, HFNEF patients with LV mechanical dyssynchrony presented higher LV filling pressures and worse NYHA functional class than those with normal LV contractions.

Conclusion In patients with HFNEF, LV mechanical dyssynchrony is associated with an important impairment of the longitudinal systolic and diastolic function of the LV. Therefore, the restoration of asynchronous LV contractions could help to improve and/or correct both the systolic and the diastolic longitudinal dysfunction of the LV in HFNEF and thereby improve the symptomatology of these patients.

  • Dyssynchrony
  • Systolic
  • Diastolic
  • Dysfunction
  • Heart failure

Introduction

Heart failure (HF) with a normal left ventricular (LV) ejection fraction (HFNEF) has long been considered a disease characterized by LV diastolic dysfunction (LVDD) and a normal LV myocardial systolic function.1 However, recently numerous investigations have evidenced that in these patients the myocardial systolic function of the left ventricle (LV) is not preserved.2,3 In this respect, with the development of new echocardiographic techniques (tissue Doppler imaging and two-dimensional speckle-tracking) several researches showed that despite a normal LV ejection fraction (LVEF) patients with HFNEF have impaired LV longitudinal systolic function as well as LV longitudinal systolic dyssynchrony.2,3 Nonetheless, despite these recent studies,2,3 the myocardial mechanisms contributing to LV longitudinal systolic dysfunction as well as the consequences of LV mechanical dyssynchrony in patients with HFNEF remain poorly understood.

Several studies have demonstrated that both the systolic and the diastolic longitudinal function of the LV are severely deteriorated by asynchronous LV contractions in patients with systolic HF.46 However, in the setting of diastolic HF it remains unclear if LV mechanical dyssynchrony contributes to an impairment of the longitudinal systolic and diastolic function of the LV. Therefore, determining whether asynchronous LV contractions contribute to a longitudinal systolic and diastolic dysfunction of the LV in HFNEF could provide new insights and potential therapeutic targets for this complex disease for which, so far, no effective therapies exist.711

The purpose of this study was to test the hypothesis that asynchronous LV contractions contribute to a significant impairment of the longitudinal systolic and diastolic function of the LV in patients with HFNEF. With the aim of validating this hypothesis, we analyzed the longitudinal systolic, diastolic, and synchronous function of the LV using two-dimensional speckle-tracking echocardiography in patients with HFNEF and in a control group consisting of asymptomatic patients with LVDD.

Methods

Study population

We enrolled consecutive patients ≥18 years with signs or symptoms of HF, evidence of LVDD, and LVEF >50% by transthoracic echocardiography, according to the diagnostic criteria of HFNEF and of LV diastolic dysfunction of the European Association of Echocardiography (EAE).12,13 We also studied a control group consisting of asymptomatic patients with LVDD without a history of HFNEF [in accordance with the diagnostic criteria of LVDD of the EAE,13 i.e. septal e′ mitral annular peak velocity <8 cm/s, or lateral e′ mitral annular peak velocity <10 cm/s, or maximal left atrial volume index (LAVI) ≥34 ml/m2]. Patients with HFNEF and subjects with asymptomatic LVDD were matched by age, gender, and LVEF (matching 1:2, i.e. 1 HFNEF patient/2 control patients). Three conditions were necessary for the diagnosis of HFNEF: (i) the presence of signs or symptoms of congestive HF [dyspnea (NYHA class ≥II), pulmonary rales, pulmonary oedema, bilateral lower extremity oedema, hepatomegaly, or fatigue]; (ii) the presence of normal LV systolic function (LVEF >50% by Simpson's method); and (iii) the evidence of LV diastolic dysfunction (septal e′ mitral annular peak velocity <8 cm/s, or lateral e′ mitral annular peak velocity <10 cm/s, or LAVI ≥34 ml/m2).12,13 We included consecutive in- and out-patients admitted in the department of cardiology (Campus Virchow-Klinikum) of the Charité University Hospital from August 2009 until April 2010 (some of these patients were also included in previous studies of our research group).14,15 The institutional review board approved this research project and informed consent was obtained from all subjects.

The selection of exclusion criteria in this study was based on the consensus of experts on HFNEF and on LV diastolic function of the EAE.12,13 In order to avoid reversible causes of myocardial dysfunction, patients with active coronary artery disease (CAD) were excluded from this study, i.e. patients with unstable angina or non-ST-segment elevation myocardial infarction (NSTEMI) without revascularization or with revascularization within the last 72 h, patients with ST-segment elevation acute myocardial infarction (STEMI) within the last 30 days, subjects waiting for coronary artery bypass graft (CABG) or within 90 days postoperatively, subjects with chronic stable angina, and patients with evidence of myocardial ischaemia assessed by stress echocardiography. Moreover, with the purpose of excluding causes of dyspnoea or myocardial dysfunction other than HFNEF, patients with the following characteristics were excluded from this study: (i) primary or secondary pulmonary hypertension of causes other than isolated LVDD or HFNEF; (ii) severe pulmonary disease defined as pulmonary pathology with supplemental oxygen requirement; (iii) severe kidney disease defined as estimated glomerular filtration rate (GFR) <30 ml/min/1.72 m2 for at least 3 months, history of renal transplantation, or severe acute renal failure with dialysis requirement; (iv) severe chronic liver disease or history of liver transplantation; (v) congenital heart disease; (vi) pericardial disease characterized by moderate or severe pericardial effusion (echo-free space in end-diastole ≥10 mm) or constrictive pericarditis; (vii) cardiomyopathy; (viii) valvular heart disease defined as mild, moderate, or severe mitral or aortic stenosis; moderate or severe non-functional mitral or tricuspid regurgitation; and moderate or severe aortic regurgitation [according to the diagnostic criteria of the guidelines for the management of patients with valvular heart disease of the American College of Cardiology (ACC)].16 Furthermore, to avoid underestimations of myocardial and mitral annular measurements, patients with valvular heart surgery, mitral annular calcification (≥5 mm), cardiac pacing, and poor two-dimensional quality in one or more than one myocardial segments of the LV (not suitable for analysis by two-dimensional speckle-tracking echocardiography in apical four-chamber, two-chamber, and long-axis views) were also excluded from this study. Moreover, to avoid mistakes in the myocardial measurements of the LV due to variability of R–R interval, patients with atrial or ventricular arrhythmias at the time of inclusion in the study were also excluded.

Transthoracic echocardiography

All patients were examined at rest in the left lateral decubitus position using a Vivid-7 (GE-Healthcare, Horten, Norway) ultrasound system followed by an offline analysis using EchoPac version 110.1.0 workstation (namely, measurements by two-dimensional speckle-tracking echocardiography). The echocardiographic measurements and analyses were performed by experienced echocardiographers blinded to each other's results. LV diameters, LV volumes, LV mass and grading of LV hypertrophy (LVH), LVEF (Simpson's method), LV midwall fractional shortening (m-FS), LA volume, LV filling pressures, and LV diastolic function and grading of LVDD were assessed as recommended by the EAE.13,17 All echocardiographic measurements using speckle-tracking (mean frame rates of 68.2 ± 8.4 frames/s), Doppler, and conventional two-dimensional echocardiography were calculated as the average of three measurements.

Two-dimensional speckle-tracking echocardiography

The analyses by two-dimensional speckle-tracking echocardiography were performed offline and blinded to the clinical characteristics of the patients. The measurements of LV longitudinal systolic strain and LV longitudinal early-diastolic strain rate (SRe) were performed at basal, mid, and apical levels in the apical four-chamber, two-chamber, and long-axis views (i.e. 18 segments of the LV).2,14,15,18 The average value of LV longitudinal peak negative systolic strain (during LV systole) and LV longitudinal peak positive early-diastolic SRe (during LV diastole) from 18 LV segments was named LV global longitudinal systolic strain (LV Strain) and LV global longitudinal early-diastolic strain rate (LV SRe), respectively.2,14,15,18

LV longitudinal systolic dyssynchrony

The measurements of LV longitudinal systolic dyssynchrony were evaluated by two-dimensional speckle-tracking echocardiography using the same 18-segment LV model utilized for the aforementioned measurements of LV global systolic strain.1921 The time-to-peak systolic longitudinal strain (Ts) was measured as the interval from the onset of the QRS to the peak negative systolic of LV longitudinal strain throughout the cardiac cycle, including post-systolic contractions.1922 The parameters of LV longitudinal systolic dyssynchrony were assessed in the following way:1922 Ts-SD (strain dyssynchrony index) = standard deviation (SD) of Ts from 18 LV segments; Ts-18-max = maximal difference in Ts between any 2 of the 18 segments of the LV. Moreover, with the purpose of assessing the percentage of abnormally delayed or inefficient LV contractions [i.e. LV segments with a maximal longitudinal contraction occurring after aortic valve closure (AVC)] in patients with LV mechanical systolic dyssynchrony, we measured the post-systolic index (PSI).23,24 The PSI was calculated in each LV segment as: PSI = 100 × (peak strain after AVC − end systolic strain)/peak strain after AVC];23,24 measuring end-systolic strain at the moment of AVC, which was automatically determined with the software and ultrasound system used in this study.1923 In addition, in order to quantify the proportion of abnormally delayed or inefficient LV contractions in the whole LV, we averaged the value of the PSI from 18 LV segments to obtain a global PSI.

LV longitudinal diastolic dyssynchrony

The measurements of LV longitudinal diastolic dyssynchrony were evaluated by two-dimensional speckle-tracking echocardiography using the same 18-segment LV model utilized for the above-mentioned measurements of LV global SRe.14,15,18 The time-to-peak early-diastolic longitudinal strain rate (Tse) was measured as the interval from the onset of the QRS to the peak positive early-diastolic of LV longitudinal strain rate during LV diastole. The parameters of LV longitudinal diastolic dyssynchrony were assessed in the following way: Tse-SD = standard deviation of Tse from 18 LV segments; Tse-18-max = maximal difference in Tse between any 2 of the 18 segments of the LV.

Echocardiographic criteria and range of values in healthy subjects

In a group of healthy subjects, we analysed the range of values of the above-mentioned parameters of LV longitudinal systolic and diastolic function and synchronicity in order to determine the 95% confidence interval (CI) of these measurements (calculated as ±1.96 SDs from the mean of this population), which was utilized to define a longitudinal systolic or diastolic dysfunction and mechanical dyssynchrony of the LV (i.e. values < or >95% CI from healthy subjects, as appropriate). Healthy subjects were defined as all those individuals with a normal echocardiogram according to the diagnostic criteria of the EAE13,17 and the ACC16 and without history or presence of one or more of the following findings: cardiovascular disease; pathology with known cardiovascular involvement; kidney, liver, or lung disease; and medications with known cardiovascular effects. A total of 106 healthy subjects (42.2 ± 11.7 years, 62.8% women, LVEF 62.4 ± 5.4%) were included and the values of the myocardial measurements of the LV in this group were: LV strain: −20.61 ± 2.63%; LV SRe: 1.53 ± 0.29 s−1; Ts-SD (strain dyssynchrony index): 27.4 ± 11.7 ms; global PSI (parameter of abnormally delayed or inefficient LV contractions): 2.10 ± 2.14%; Tse-SD: 29.8 ± 12.5 ms. Clinical characteristics and conventional echocardiographic assessments of this cohort of healthy subjects are shown in more detail in the Supplementary data online.

Statistical analysis

Continuous data are presented as mean ± SD and dichotomous data in percentage. Differences in continuous variables between groups (comparisons of two groups) were assessed using unpaired Student's t-test only, because all data were normally distributed (the Kolmogorov–Smirnov test was used to test for normal distribution). Categorical variables were compared by χ2 test and Fisher exact-test as appropriate. Comparisons between three or more groups were assessed by one-way analysis of variance (ANOVA). Comparisons between groups adjusted for other co-variables were performed by post hoc ANOVA-analysis using the Scheffe's method. The relationship between continuous variables was analysed using simple linear regression analysis. Selection of independent variables for the prediction of LV longitudinal systolic, diastolic, and asynchronous dysfunction was performed using a forward stepwise multivariate analysis and a logistic regression analysis. With the purpose of determining the reproducibility of LV myocardial measurements, we analyzed the intra- and inter-observer variability on 20 randomly selected patients using a Bland–Altman analysis. All statistical analyses were performed with SAS® 9 (SAS Institute, NC, USA). Differences were considered statistically significant when P < 0.05.

Results

Patient clinical characteristics

A total of 384 subjects met the eligibility criteria during the study period (128 with HFNEF and 256 with asymptomatic LVDD). However, 59 patients (43 with HFNEF and 16 with asymptomatic LVDD) could not be enrolled, because of poor two-dimensional quality in one or more than one segments of the LV not suitable for analysis by two-dimensional speckle-tracking echocardiography (n = 28), severe kidney disease (n = 2), cardiac pacing (n = 1), severe chronic liver disease (n = 1), NSTEMI within the last 72 h (n = 2), STEMI within the last 30 days (n = 2), CABG within the last 90 days (n = 1), evidence of myocardial ischaemia assessed by stress echocardiography (n = 2), mild aortic stenosis (n = 1), and atrial or ventricular arrhythmias at the time of inclusion in the study (n = 19). Thus, 325 patients were finally studied and analysed (85 with HFNEF and 240 with asymptomatic LVDD). Clinical characteristics of HFNEF patients and control subjects are shown in Table 1.

View this table:
Table 1

Clinical characteristics and LV measurements

HFNEF (n = 85)Asymptomatic LVDD(n = 240)P value
Clinical characteristics
 Age (years)70.4 ± 10.168.5 ± 8.90.1383
 Women [n (%)]85 (44.7)84 (35)0.1129
 Body mass index (kg/m2)28.7 ± 4.927.4 ± 4.00.0538
 Haemoglobin (g/dl)13.3 ± 1.613.5 ± 1.60.3391
 GFR (ml/min/1.73 m2)68.1 ± 23.374.8 ± 20.70.0161
 Hypertension [n (%)]85 (100)197 (82.1)<0.0001
 Type 2 diabetes [n (%)]30 (35.3)54 (22.5)0.0206
 Obesity [n (%)]26 (30.5)25 (10.4)<0.0001
 History of CAD [n (%)]52 (61.2)85 (35.4)<0.0001
 Systolic blood pressure (mm Hg)138.1 ± 20.2132.5 ± 20.90.0734
 Diastolic blood pressure (mm Hg)80.4 ± 11.978.8 ± 12.30.3732
 Heart rate (beats/min)70.1 ± 9.471.4 ± 9.90.2884
 QRS duration (ms)106.3 ± 18.1107.4 ± 18.10.6295
 QRS >120 ms [n (%)]11 (12.9)37 (15.4)0.5356
LV conventional measurements
 LV ejection fraction (%)60.4 ± 7.161.9 ± 6.50.0572
 LV end-diastolic volume index (ml/m2)45.3 ± 12.342.1 ± 11.30.0880
 LV mass index (g/m2)122.6 ± 28.3106.2 ± 24.00.0001
 Severe LVH [n (%)]27 (31.7)10 (4.2)<0.0001
 Septal e′ mitral annular peak velocity (cm/s)4.5 ± 1.46.0 ± 1.4<0.0001
 Lateral e′ mitral annular peak velocity (cm/s)6.4 ± 1.58.1 ± 1.7<0.0001
 Mitral E/e′ (average septal–lateral) ratio17.3 ± 6.110.8 ± 4.2<0.0001
LV measurements by speckle-tracking
 LV global longitudinal systolic strain (LV strain; %)−15.85 ± 3.84−18.12 ± 3.00<0.0001
 LV longitudinal systolic dysfunction (LV strain > −16%) [n (%)]43 (50.6)53 (22.0)<0.0001
 LV global longitudinal early-diastolic strain rate (LV SRe; s−1)0.84 ± 0.251.04 ± 0.30<0.0001
 LV longitudinal diastolic dysfunction (LV SRe <0.95 s−1) [n (%)]54 (63.5)104 (43.3)0.0002
  • Data are expressed as mean ± SD or counts and percentages.

  • Severe LVH, LV mass index ≥122 g/m2 in women or ≥149 g/m2 in men. Signs or symptoms of HFNEF, dyspnoea 100% [NYHA functional class II (n = 64), class III (n = 14), class IV (n = 7)], cardiogenic pulmonary oedema 51.8%, bilateral lower extremity oedema 47%, chronic stable HF 48.2%, and acute HF (first presentation or as decompensation of chronic HF) 51.8%.

Longitudinal contractile, diastolic, and synchronous function of the LV in HFNEF

Patients with HFNEF presented significantly more decreased LV longitudinal contractile function than those with asymptomatic LVDD and healthy subjects (all P < 0.0001, P value adjustment for age <0.0001, Table 1 and Figure 1). In addition, patients with HFNEF showed significantly more impaired parameters of LV longitudinal systolic dyssynchrony than asymptomatic LVDD patients and healthy subjects, even after adjustment for age (all P < 0.0001, Table 2 and Figures 1 and 2). Furthermore, when we assessed the longitudinal relaxation and diastolic synchronicity of the LV, while patients with HFNEF had a significant impairment of the longitudinal diastolic function of the LV (Table 1), in the parameters and rates of LV longitudinal diastolic dyssynchrony there were no significant differences compared with subjects with asymptomatic LVDD (Table 2).

View this table:
Table 2

Assessment of the longitudinal systolic and diastolic dyssynchrony of the LV

HFNEF (n = 85)Asymptomatic LVDD (n = 240)P value
LV longitudinal systolic dyssynchrony
 Ts-18-max (ms)178.9 ± 64.1157.6 ± 66.20.0106
 Ts-SD (ms)54.8 ± 17.748.6 ± 19.30.0091
 Ts-SD (strain dyssynchrony index) >50 ms58.8%44.2%0.0201
LV longitudinal diastolic dyssynchrony
 Tse-18-max (ms)153.9 ± 52.3147.2 ± 50.00.3092
 Tse-SD (ms)44.7 ± 16.144.6 ± 16.50.9460
 Tse-SD >54.5 ms27.1%27%0.9881
  • Data are expressed as mean ± SD or percentages.

  • Ts, time-to-peak longitudinal systolic strain; Ts-SD, standard deviation of Ts from 18 segments of the LV; Ts-18-max, maximal difference in Ts between any 2 of the 18 segments of the LV; Tse, time-to-peak early-diastolic strain rate; Tse-SD, standard deviation of Tse from 18 segments of the LV; Tse-18-max, maximal difference in Tse between any 2 of the 18 segments of the LV.

Figure 1

Example showing a significant impairment of the longitudinal systolic and synchronous function of the LV in a patient with HFNEF compared with a healthy subject and an asymptomatic patient with LVDD. Note also in the patient with HFNEF (top figure) the strong association of abnormally delayed and asynchronous LV contractions [measured by the global post-systolic index (PSI) and the strain dyssynchrony index (Ts-SD), respectively] with dysfunctional LV segments (assessed by LV systolic strain) which is marked with a darker blue, yellow-orange, and lighter red colour, respectively.

Figure 2

Example showing a HFNEF patient with asynchronous LV contractions, with consequent impairment of the longitudinal systolic function of the LV (top figure; note that delayed asynchronous segments lead to inefficient LV contractions, i.e. LV segments with a maximal longitudinal contraction occurring after aortic valve closure (AVC). In contrast, at the below figure is shown a patient with HFNEF without asynchronous LV contractions, who therefore does not present an impaired LV longitudinal systolic function.

Myocardial contractile consequences of LV mechanical systolic dyssynchrony in HFNEF

In the myocardial analyses that we have performed to assess the potential contractile consequences of LV mechanical systolic dyssynchrony in HFNEF, we observed that HFNEF patients with LV longitudinal systolic dyssynchrony had significantly more impaired LV longitudinal contractile function than patients with a normal LV synchronous function (Table 3 and Figure 2). In addition, HFNEF patients with LV mechanical systolic dyssynchrony were characterized by having abnormally delayed or inefficient LV contractions (Table 3 and Figures 1 and 2). Besides, we observed that LV longitudinal systolic dyssynchrony was significantly and inversely linked to the longitudinal systolic function of the LV (Table 4).

View this table:
Table 3

Myocardial consequences of LV mechanical dyssynchrony in patients with HFNEF

Dyssynchrony Ts-SD >50 ms (n = 50)Non-dyssynchrony Ts-SD ≤50 ms (n = 169)P value
LV longitudinal contractile function
 LV global longitudinal systolic strain (%)−14.76 ± 3.44−18.57 ± 3.10<0.0001
 LV midwall fractional shortening (%)17.94 ± 4.4720.00 ± 4.910.0097
 Inefficient LV contractions72%21.3%<0.0001
 LV longitudinal systolic dysfunction64%19.5%<0.0001
LV longitudinal diastolic function
 LV global longitudinal early-diastolic SRe (s−1)0.79 ± 0.241.06 ± 0.32<0.0001
 Average septal–lateral e′ mitral annular velocity (cm/s)3.21 ± 1.044.33 ± 1.16<0.0001
 Average septal–lateral a′ mitral annular velocity (cm/s)7.08 ± 2.369.22 ± 2.24<0.0001
 LV longitudinal diastolic dysfunction70%41.3%0.0003
LV filling pressures
 Mitral E/e′ average septal–lateral ratio26.6 ± 9.417.0 ± 6.7<0.0001
 LAVI >34 ml/m284%27.8%<0.0001
  • Data are expressed as mean ± SD or percentages. The results represent the analyses of HFNEF patients with LV mechanical dyssynchrony (i.e. strain dyssynchrony index [Ts-SD] >50 ms; n = 50) in comparison with patients without LV mechanical dyssynchrony (i.e. strain dyssynchrony index [Ts-SD] ≤ 50 ms; n = 169).

  • LV systolic dyssynchrony, strain dyssynchrony index [Ts-SD] >50 ms; LV longitudinal systolic dysfunction, LV global longitudinal systolic strain > −16%; Inefficient LV contractions, global post-systolic index >6.30%; LV longitudinal diastolic dysfunction, LV global longitudinal early-diastolic strain rate <0.95 s−1. LV filling pressures were assessed by the mitral E/e′ average septal–lateral ratio, i.e. the ratio of early-diastolic mitral inflow peak velocity by pulsed-wave (PW) Doppler [E] to early-diastolic mitral annular [e′] (average septal–lateral) peak velocity using spectral PW tissue Doppler imaging (DTI). a′, late-diastolic mitral annular (average septal–lateral) peak velocity using spectral PW DTI. LAVI, maximal left atrial volume index.

View this table:
Table 4

Interrelations of LV mechanical dyssynchrony with the longitudinal systolic and diastolic function of the LV

Strain dyssynchrony indexLV systolic strainLV diastolic SRe
VariablesrPrPrP
LV global longitudinal systolic strain (%)0.48<0.0001aN/AN/A0.54<0.0001c
LV global longitudinal diastolic SRe (s−1)0.34<0.0001a0.54<0.0001bN/AN/A
Average e′ mitral annular velocity (cm/s)0.38<0.0001a0.38<0.0001b0.40<0.0001c
Ts-SD (ms)N/AN/A0.48<0.0001b0.34<0.0001c
Tse-SD (ms)0.34<0.0001a0.240.00010.42<0.0001c
Global PSI (%)0.66<0.0001a0.57<0.0001b0.34<0.0001c
QRS duration (ms)0.200.00020.090.08780.080.1285
LV mass index (g/m2)0.32<0.0001a0.34<0.0001b0.38<0.0001c
Age (years)0.020.69600.080.11630.160.0034
Pulse pressure (mmHg)0.010.85160.010.85670.050.4966
LV midwall fractional shortening (%)0.28<0.00010.34<0.00010.32<0.0001
  • N/A, not applicable; SRe, peak early diastolic strain rate; Ts, time-to-peak longitudinal systolic strain; Ts-SD, standard deviation of Ts from 18 segments of the LV; Tse, time-to-peak early-diastolic strain rate; Tse-SD, standard deviation of Tse from 18 segments of the LV; PSI, post-systolic index; LV systolic strain, LV global longitudinal systolic strain; LV diastolic SRe, LV global longitudinal diastolic SRe. Data correspond to analysis of all patients (85 HFNEF patients and 240 asymptomatic patients with LVDD).

  • aIndependent predictor of strain dyssynchrony index by multivariate analysis.

  • bIndependent predictor of LV global longitudinal systolic strain by multivariate analysis.

  • cIndependent predictor of LV global longitudinal diastolic SRe by multivariate analysis.

  • The variables with significant correlations by univariate analysis were included in the multiple regression model.

LV mechanical systolic dyssynchrony and its influence on the longitudinal diastolic function of the LV

Between the effects of LV mechanical systolic dyssynchrony on the longitudinal diastolic function of the LV, we found that HFNEF patients with asynchronous LV contractions presented a significant impairment of the diastolic function of the LV (Table 3). In relation to these findings, LV filling pressures were significantly higher in HFNEF patients with asynchronous LV contractions than in patients with normal LV contractions (Table 3). In line, we found that the grade of impairment of LV diastolic function was significantly linked to the severity of the longitudinal systolic dyssynchrony of the LV (Table 5). In addition, we observed that LV longitudinal systolic dyssynchrony was significantly and inversely linked to the longitudinal diastolic function of the LV (Table 4).

View this table:
Table 5

Grade of impairment of LV diastolic function linked to the severity of the longitudinal systolic and asynchronous dysfunction of the LV

LV diastolic dysfunction
Normal (n = 106)Mild (n = 44)Moderate (n = 67)Severe (n = 214)P-ANOVA
Parameter of LV mechanical systolic dyssynchrony (Ts-SD, ms)27.4 ± 11.748.2 ± 16.246.4 ± 18.951.9 ± 19.5<0.0001
Parameter of abnormally delayed LV contractions (global PSI, %)2.10 ± 2.145.42 ± 4.695.80 ± 5.056.91 ± 5.70<0.0001
Parameter of LV longitudinal systolic function(global strain, %)−20.61 ± 2.63−18.14 ± 2.61−18.02 ± 3.19−17.25 ± 3.56<0.0001
  • Data are expressed as mean ± SD.

  • Normal, corresponding to normal healthy subjects, i.e. with a normal LV diastolic function (namely, septal e′ mitral annular peak velocity ≥8 cm/s, lateral e′ mitral annular peak velocity ≥10 cm/s, and LAVI < 34 ml/m2); LV diastolic dysfunction (LVDD), septal e′ mitral annular peak velocity <8 cm/s, or lateral e′ mitral annular peak velocity <10 cm/s, or LAVI ≥34 ml/m2; Grade I or mild LVDD, corresponding to patients with the above-mentioned criteria of LVDD more: mitral E/A ratio <0.8, or mitral deceleration time (DT) >200 ms, or mitral E/e′ average septal–lateral ratio ≤8; Grade II or moderate LVDD, corresponding to patients with the aforementioned criteria of LVDD more: mitral E/A ratio of 0.8–1.5, or DT of 160–200 ms, or mitral E/e′ average septal–lateral ratio of 9 to 12; Grade III or severe LVDD, corresponding to patients with the above-mentioned criteria of LVDD more: mitral E/A ratio ≥2, or DT <160 ms, or mitral E/e′ average septal–lateral ratio of ≥13. Ts-SD (strain dyssynchrony index), standard deviation of time-to-peak longitudinal systolic strain from 18 LV segments; global PSI, global post-systolic index; global strain, LV global longitudinal systolic strain.

Clinical and echocardiographic variables linked to the longitudinal systolic and diastolic dysfunction and mechanical dyssynchrony of the LV

In the analysis of the clinical and echocardiographic variables that could be associated with changes of the systolic and diastolic function and synchronicity of the LV, we found that comorbidities such as diabetes, hypertension, obesity, history of CAD, and severe LVH were significant predictors of both systolic and diastolic myocardial abnormalities of the LV (Table 6). Furthermore, in relation to the aforementioned effects of asynchronous LV contractions on the systolic and diastolic function of the LV, we found that LV mechanical systolic dyssynchrony (i.e. strain dyssynchrony index >50 ms) was an important factor linked to the longitudinal systolic and diastolic dysfunction of the LV (Table 6).

View this table:
Table 6

Clinical and echocardiographic variables linked to elevated LV filling pressures, symptomatic status, and to the longitudinal systolic, diastolic, and asynchronous dysfunction of the LV

LV systolic dysfunctionLV diastolic dysfunctionLV systolic dyssynchronyElevated LV filling pressuresNYHA functional class ≥2
VariablesORPORPORPORPORP
Strain dyssynchrony index >50 ms3.8<0.00011.80.0081N/AN/A2.60.00351.80.0209
LV global systolic strain > −16%N/AN/A4.8<0.00015.8<0.00012.10.01842.80.0001
Abnormally delayed LV contractionsa5.4<0.00011.80.00907.1<0.00012.10.02642.10.0033
LV global diastolic SRe <0.95 s−14.8<0.0001N/AN/A3.5<0.00011.80.04222.20.0020
Type 2 diabetes1.80.02691.30.22291.20.51691.50.27011.90.0184
Hypertension3.40.00661.20.54851.90.06451.60.22252.00.0122
Obesity1.40.37521.20.51421.60.11721.50.24402.7<0.0001
History of CAD2.50.00012.10.00102.20.00052.00.03542.8<0.0001
Severe LVH2.40.04362.90.03433.0<0.00011.90.03882.4<0.0001
QRS >120 ms1.60.09260.90.95561.70.03500.90.74890.80.5974
  • LV systolic dysfunction, LV global longitudinal systolic strain > −16%; LV systolic dyssynchrony, strain dyssynchrony index [Ts-SD] >50 ms; LV diastolic dysfunction, LV global longitudinal early-diastolic strain rate <0.95 s−1; elevated LV filling pressures, mitral E/e′ average septal–lateral ratio ≥13.

  • LV global systolic strain, LV global longitudinal systolic strain; LV global diastolic SRe, LV global longitudinal early-diastolic strain rate. N/A, not applicable; OR, odds ratio.

  • The odds ratio in these analyses represents the ratio of the odds of an event (namely, the first row of the table, e.g. LV systolic dysfunction) occurring in one group to the odds of it occurring in another group (i.e. the opposite variable, e.g. diabetes vs. non-diabetes or strain dyssynchrony index >50 ms vs. strain dyssynchrony index ≤50 ms). Data correspond to analysis of all patients (85 HFNEF patients and 240 asymptomatic patients with LVDD).

  • aAbnormally delayed LV contractions = global post-systolic index >6.30%.

Symptomatic status and its relationship with asynchronous LV contractions

In the analyses that we have performed to assess the link between the symptomatic status of the patients and the changes in the systolic synchronicity of the LV, we found that the functional capacity during exercise (i.e. NYHA functional class) was significantly altered in patients with asynchronous LV contractions compared with those with normal LV contractions (strain dyssynchrony index >50 vs. ≤50 ms = NYHA class 2.4 ± 0.6 vs. 1.2 ± 0.6; P < 0.0001). In relation to these findings, the asynchronous function of the LV was an important predictor of elevated LV filling pressures and impaired functional capacity during exercise (i.e. NYHA functional class ≥2; Table 6).

Reproducibility

The variability of the myocardial measurements of the LV was not of clinical significance (intra- and inter-observer variability (respectively): mean absolute differences: LV strain = 0.48 ± 0.44 and 0.52 ± 0.46%; LV SRe = 0.04 ± 0.02 and 0.05 ± 0.03 s−1; strain dyssynchrony index = 4.18 ± 4.23 and 4.48 ± 4.18 ms; global PSI = 0.22 ± 0.28 and 0.28 ± 0.32%; interclass correlation coefficient: LV strain = 0.988 and 0.985; LV SRe = 0.986 and 0.984; strain dyssynchrony index = 0.974 and 0.967; and global PSI = 0.987 and 0.985).

Discussion

In the present study we have performed a comprehensive assessment of the systolic, diastolic, and synchronous function of the LV in patients with HFNEF and in a control group consisting of asymptomatic patients with LVDD. Using two-dimensional speckle-tracking echocardiography at rest we have demonstrated that the longitudinal systolic and diastolic function of the LV is severely altered in patients with HFNEF. Concerning the possible pathophysiological mechanisms linked to these findings, we found that asynchronous LV contractions were significantly associated with the longitudinal systolic and diastolic dysfunction of the LV in HFNEF and thereby linked to the symptoms of these patients.

Comorbidities, LV subendocardial fibrosis, and longitudinal systolic, diastolic, asynchronous dysfunction of the LV in HFNEF

Like in other reports,2,25 HFNEF patients in the present study were characterized by having high rates of comorbid conditions such as type 2 diabetes, obesity, hypertension, history of CAD, and severe LVH. Several studies have demonstrated that these comorbid disorders are characterized by causing interstitial fibrosis of the LV by means of diverse mechanisms,2630 affecting primarily the subendocardial layer of the LV and consequently the longitudinal systolic and diastolic function and synchronicity of the LV.18,2738 In line with these reports, in our study we found that these comorbidities were linked to both the systolic and diastolic longitudinal dysfunction and mechanical dyssynchrony of the LV. Therefore, one might expect that in a setting with high rates of these comorbidities such as HFNEF, the grade of impairment of the longitudinal systolic and diastolic function and mechanical dyssynchrony of the LV would be elevated. In this regard, in the present study which is characterized by HFNEF patients with a high prevalence of the aforementioned comorbid conditions, we found elevated rates of LV longitudinal systolic and diastolic dysfunction as well as of LV longitudinal systolic dyssynchrony. According to these findings, recent studies also evidenced a significant alteration of the longitudinal contractile and synchronous function of the LV in patients with HFNEF.2,3 However, despite these recent reports,2,3 the myocardial mechanisms contributing to the longitudinal systolic dysfunction of the LV as well as the consequences of LV mechanical dyssynchrony in HFNEF remain poorly understood.

New insights in LV myocardial systolic dysfunction in HFNEF

In the present study characterized by a large cohort of subjects with HFNEF, we found several myocardial abnormalities that could help to understand the mechanisms and consequences of an asynchronous and dysfunctional LV in HFNEF. In this regard, we observed that asynchronous LV contractions lead to a significant myocardial systolic dysfunction of the LV in patients with HFNEF. Among the possible underlying mechanisms that could explain these findings, the subendocardial fibrosis of the LV could have an important role. Recent studies of basic and clinical research have demonstrated the pivotal importance of LV subendocardial fibrosis in the delay of the longitudinal contraction of the LV as well as in the development of LV longitudinal systolic dyssynchrony.3941 In addition, previous research in the setting of systolic HF have demonstrated that delayed asynchronous LV contractions deteriorate the longitudinal systolic function of the LV because the maximal longitudinal shortening of these delayed and asynchronous LV segments occurs with the aortic valve closed.42,43 In line, we found that HFNEF patients with LV mechanical systolic dyssynchrony had a high proportion of abnormally delayed or inefficient LV contractions. In this regard, we consider that in subjects with HFNEF the subendocardial fibrosis of the LV, as a possible consequence of the above-mentioned comorbidities, could induce an asynchronous delay of the longitudinal contraction of the LV. Consequently, this asynchronous contractile delay could lead the maximal longitudinal contraction in LV segments to occur after aortic valve closure (i.e. inefficient LV contractions) causing thus an impairment of the longitudinal systolic function of the LV.

Asynchronous LV contractions and its influence on the longitudinal diastolic function of the LV

In the present study we have demonstrated the strong influence of asynchronous LV contractions on the longitudinal diastolic dysfunction of the LV in patients with HFNEF. Between the possible mechanisms that could explain these findings, we believe that delayed asynchronous LV contractions principally affect the early LV diastole, with consequent impairment of the passive filling of the LV and increase of LV filling pressures. In this regard, we observed that HFNEF patients with asynchronous LV contractions had significantly more impaired LV early diastolic function as well as higher LV filling pressures than those with normal LV contractions. These findings are in agreement with recent reports which also demonstrated the important association of delayed and asynchronous LV contractions with the deterioration of the early LV diastolic function and the passive filling of the LV.4,5,44

Potential mechanisms that could link asynchronous LV contractions with the symptoms of HFNEF

The mechanisms linking LV longitudinal systolic dyssynchrony with the symptoms of patients with HF have been well studied in systolic HF.46,41,43,45 However, the role of the longitudinal systolic dyssynchrony in the symptomatology of HFNEF has not been elucidated. In the current study in a large cohort of patients with HFNEF we demonstrated that asynchronous LV contractions deteriorate the systolic and diastolic longitudinal function of the LV with consequent increase of LV filling pressures and worsening of functional capacity during exercise. In addition, we showed that the mechanical systolic dyssynchrony of the LV was an important predictor of both systolic and diastolic longitudinal dysfunction of the LV as well as of elevated LV filling pressures and impaired functional capacity (i.e. NYHA functional class). These findings are consistent with recent reports in hypertensive patients and in systolic HF,46,41,43,45 which also demonstrated that asynchronous LV contractions contribute to a significant alteration in the diastolic and systolic longitudinal function of the LV with consequent increase of LV filling pressures and worsening of NYHA functional class.46,41,43,45 Hence, we consider that LV longitudinal systolic dyssynchrony contributes to the symptomatology of patients with HFNEF due to a deterioration of the systolic and diastolic longitudinal function of the LV.

Clinical perspectives

Isolated LV diastolic dysfunction (i.e. abnormalities of LV diastolic stiffness and relaxation with a normal LVEF) has long been considered the main mechanism underlying HFNEF.1 Based on this pathophysiological evidence, several clinical trials have been conducted with the aim of restoring LV diastolic dysfunction in subjects with HFNEF and thereby improve the prognosis of these patients.811 However, none of these treatments has been shown to decrease the mortality of patients with HFNEF.811 For this reason, new pathophysiological paradigms should arise with the goal of discovering new therapeutic targets in this disease.7

In the present study using measurements at rest with two-dimensional speckle-tracking echocardiography we have demonstrated that HFNEF is characterized by both diastolic and systolic longitudinal dysfunction of the LV which may be a consequence of asynchronous LV contractions. Thus, we have shown that HFNEF is not only a pathophysiological process of isolated LV diastolic dysfunction. Therefore, we consider that in subjects with HFNEF, treatments destined to improve LV diastolic dysfunction without restoring the contractile and asynchronous longitudinal dysfunction of the LV are unlikely to improve the symptoms and/or the prognosis of these patients.

Furthermore, with the perspective of assessing new therapeutic pathways in HFNEF, on the basis of the present study it is important to note that asynchronous LV contractions are involved in the pathogenesis of a dysfunctional LV in patients with HFNEF. These findings are encouraging, since asynchronous LV contractions are amenable to correction by cardiac resynchronization therapy (CRT) with consequent enhancement of the longitudinal contractile and diastolic function of the LV.5,6,43,45 Hence, echocardiographic parameters of LV systolic dyssynchrony (as we have used) could be of great importance for guiding and/or evaluating a possible treatment with CRT in patients with HFNEF. Nonetheless, it should be noted that so far no study evaluating the effects of CRT on morbidity and/or mortality or the improvement of symptoms in subjects with HFNEF has been published. Moreover, there is also a lack of research regarding the prognostic significance of LV systolic dyssynchrony in these patients. Therefore, we consider that future clinical trials should address first, the prognosis value of the systolic dyssynchrony of the LV in HFNEF, and second, whether medical-pharmacological treatments and/or CRT provide a benefit in HFNEF patients with an asynchronous LV.

Likewise, it should be noted the important clinical contribution of the myocardial analyses in healthy subjects, in which data concerning LV measurements using two-dimensional speckle-tracking are limited. In this regard, in a large cohort we established the myocardial systolic and diastolic values corresponding to the 95% CI from healthy subjects, which can be used in different settings as cutoff points to define a longitudinal systolic and diastolic global dysfunction and mechanical dyssynchrony of the LV (i.e. values < or >95% CI from healthy subjects, respectively).

Limitations

Our study has diverse limitations. Our analyses were limited by the lack of invasive haemodynamic data regarding LVDD. Nevertheless, several studies have demonstrated high sensitivity, specificity, and accuracy of lateral-septal e′ mitral annular peak velocity and LAVI to determine LVDD in patients with HFNEF.46,47 In addition, these non-invasive measurements for the assessment of LVDD are currently recommended by the consensus of experts or guidelines in LV diastolic function and in HFNEF of the EAE.12,13 Furthermore, it is important to point out that in this work we did not perform an analysis of the variables under study using exercise echocardiography, which recently has shown that it might have an interesting role in HFNEF.48

Another study limitation is that our cohort of HFNEF was principally characterized by patients in NYHA functional class II. Therefore, future studies with larger numbers of patients in NYHA functional class III–IV are needed to validate the findings of this study. Additionally, it is also worth mentioning that the results of the present study can be only applied to patients with the inclusion and exclusion criteria of this work, i.e. patients in sinus rhythm with compensated or decompensated HF, with a normal LVEF, and without valvular heart disease or severe kidney, liver, or pulmonary disease. Extrapolating these findings to a group of patients with different characteristics might be subject to errors.

Conclusions

In patients with HFNEF, the longitudinal systolic and diastolic dysfunction of the LV is common and associated with asynchronous LV contractions. Furthermore, our findings suggest that asynchronous LV contractions through an impairment of the longitudinal systolic and diastolic function of the LV lead to an increase of LV filling pressures with consequent worsening of functional capacity during exercise in patients with HFNEF. Therefore, on the basis of these findings, the restoration of asynchronous LV contractions could help to improve the symptomatology of patients with HFNEF. Thus, our research provides new pathophysiological insights and thereby potential therapeutic pathways for this complex disease for which, so far, no effective therapies exist.

Acknowledgements

The authors thank the patients and staff of the Department of Echocardiography, Charité University Hospital (Campus Virchow-Klinikum) for their participation in this project.

Conflict of interest: none declared.

References

View Abstract