The Relationship between Tumor Necrosis Factor Alpha and Left Ventricular Diastolic Function.

INTRODUCTION: Left ventricular (LV) diastolic dysfunction is a common condition. Tumor necrosis factor (TNF) alpha is an inflammatory cytokine that plays a role in the development of cardiac structural changes leading to LV diastolic dysfunction. The aim of this study was to examine the relationship between serum levels of TNF alpha levels and LV diastolic function.
METHODS: A case-control study that included 40 patients with echocardiographic evidence of LV diastolic dysfunction and 40 healthy controls. Standard transthoracic echocardiography was performed to assess LV and left atrial volumes, systolic and diastolic function according to the current recommendations. Serum TNF alpha levels were assessed using a specific enzyme-linked immunosorbent assay kit.
RESULTS: Mean serum TNF alpha level was significantly higher in the study group 3.48 ± 1.06 versus 1.22 ± 0.36 pg/ml in the control group, P < 0.001. It was also higher in patients with Grade 2 diastolic dysfunction (n = 16) 3.91 ± 1.21 versus 3.18 ± 0.86 pg/ml in those with Grade 1 diastolic dysfunction (n = 24), P = 0.03. TNF alpha showed a strong correlation with indexed left atrial volume (LAVI) in the study group but not in controls. Mean serum TNF showed a trend toward increase with worsening heart failure symptoms in the form of increased the New York Heart Association functional class.
CONCLUSION: Serum TNF alpha level is elevated in patients with LV diastolic dysfunction and is correlated to LAVI in such patients. Patients with Grade 2 diastolic dysfunction have higher serum levels of TNF alpha compared to those with Grade 1 diastolic dysfunction. TNF alpha levels increase with worsening heart failure symptoms. Copyright:

INTRODUCTION

Left ventricular (LV) diastolic dysfunction refers to changes in LV diastolic filling caused by impaired relaxation and abnormal distensibility of the LV myocardium. It is relatively common with advancing age and in the presence of hypertension, diabetes mellitus, obesity, and coronary artery disease. LV diastolic dysfunction is a predictor of future cardiovascular events and diastolic heart failure.[12]

LV diastolic dysfunction is commonly asymptomatic. LV diastolic pressure rises over time, leading to increase pulmonary venous pressure and LV stiffness. This causes further impairment of LV filling, thus increasing the left atrial pressure and pulmonary venous pressure. This increase in hydrostatic pressure may lead to pulmonary venous congestion and pulmonary edema. It is important to identify patients with asymptomatic diastolic dysfunction to start medical treatment and risk factor interventions that aim to stop or even reverse the progression of structural and functional heart abnormalities.[345] The structural changes occurring in patients with LV diastolic dysfunction involve an inflammatory process and later fibrosis. Perivascular inflammation and reactive fibrosis are responsible for the adverse prognostic effects on myocardial oxygen transport and are risk factors for the development of cardiac arrhythmias.[6]

Tumor necrosis factor (TNF) alpha is one of the cytokines involved in systemic inflammation and one of the acute phase reactants. It is mainly secreted by activated macrophages. The effects of TNF alpha on the heart depend on the amount and duration of exposure to TNF alpha. Short-term exposure is usually an adaptive response to cardiac stress while long-term exposure is maladaptive and may lead to decompensation.[78]

High levels of TNF alpha may lead to LV diastolic dysfunction, cardiomyopathy, and diastolic heart failure by promoting myocyte hypertrophy through the production of reactive oxygen intermediates in cardiac myocytes and inducing ventricular remodeling by promoting protein production in the extracellular matrix.[910]

The aim of this study was to examine the relationship between serum levels of TNF alpha levels and LV diastolic function.

METHODS

This was a case–control study conducted from September 2017 to March 2018. Approval of the Institutional Ethical Committee was obtained, and written informed consents were provided by all participants.

The study included eighty participants, 40 study patients and 40 controls who were consecutively selected from those presenting to our institution's echocardiography laboratory. Studied cases were adults with a confirmed diagnosis of LV diastolic dysfunction, while controls were adults with normal LV diastolic function. Diastolic function was assessed according to institutional protocols which follow most of the recommendations of the American Society of Echocardiography (ASE) and the European Association of Cardiovascular Imaging (EACVI) as detailed below.[1112]

Individuals were excluded from the study if they had any of the following criteria: being younger than 18 years old or older than 60 years old; having an LV ejection fraction (LVEF) <50%; a recent (<3 months) acute coronary syndrome, myocardial infarction, or cerebrovascular stroke; valvular heart disease of any etiology; end-stage renal failure; chronic liver disease; malignancy; acute and chronic pancreatitis; connective tissue disease; thyroid disease; or current use of medications that interfere with TNF alpha level such as TNF inhibitors and steroids.

History and baseline assessment

A thorough history and clinical examination were performed including identifying the New York Heart Association (NYHA) functional classification[13] for all participants. The weight was measured in kilograms and the height was measured in meters to calculate body mass index (BMI), which was calculated using the formula: BMI = Weight (kg)/height (m)2. Body surface area (BSA) was calculated using the Mosteller formula as follows: BSA = square root of ([height in cm × weight in kg]/3600).[14]

Transthoracic echocardiography

Standard transthoracic echocardiography was performed using a General Electric Vivid E9 machine (GE Canada) using an M4S probe that has a frequency range of 1.7–3.4 MHz. Participants were examined in the left lateral decubitus position with electrocardiographic gating.

Biplane Simpson's method from the apical 4- and 2-chamber views was used to measure LV end-diastolic volume, LV end-systolic volume, and LVEF by semi-automated tracing of the endocardial borders of the LV.[15]

Left ventricular diastolic function

To assess LV diastolic function, the following were estimated:

Septal and lateral e' velocities: Those were measured at the septal and lateral basal aspects of the LV just above the mitral annulus respectively form the apical four-chamber view using pulsed-wave tissue Doppler imaging. Average e' was calculated as the mean of both these measurements

Indexed left atrial volume (LAVI): The left atrial volume was measured by using the biplane method in dedicated apical 2- and 4-chamber views where left atrial length and transverse diameters were maximized. This measurement was divided by BSA to calculate LAVI

Transmitral pulsed-wave Doppler: From the apical four-chamber view guided by color flow imaging for optimal alignment with blood flow, a pulsed wave Doppler sample volume was placed between the tips of the mitral leaflets. E wave velocity, A wave velocity, E/A ratio, and deceleration time (DT) were measured. E/A ratio was measured again following Valsalva maneuver, where the patient was asked to perform forced expiration for 10 s with the mouth and nose closed. The change in transmitral E/A ratio during peak strain and following release was measured.

LV diastolic function was considered normal when septal e' ≥8 cm/s, lateral e' ≥10 cm/s, and LAVI <34 ml/m2. LV diastolic function was considered impaired when septal e' <8 cm/s, lateral e' <10 cm/s. Further grading of diastolic dysfunction was defined according to the following criteria: Grade 1: Mitral E/A ratio ≤0.8, DT >200 m, average E/e' ratio ≤8 cm/s, LAVI could be normal or ≥34 ml/m2, and Valsalva E/A change <0.5; Grade 2: Mitral E/A ratio >0.8 to <1.5, DT 160–200 m, average E/e' ratio 9–14 cm/s, LAVI ≥34 ml/m2, and Valsalva E/A change ≥0.5; Grade 3: Mitral E/A ratio >2, DT 160–200 m, average E/e' ratio >14 cm/s, LAVI ≥34 ml/m2, and Valsalva E/A change ≥0.5.

Tumor necrosis factor alpha estimation

A volume of 3 ml venous blood sample was drawn from each participant and centrifuged for separation of serum. The serum tube was labeled with a specific identification number for each subject and stored for the measurement of human serum TNF alpha level using a specific enzyme-linked immunosorbent assay kit (IBL International, Hamburg, Germany).[16]

Data management and statistical analysis

IBM Statistical Package for Social Sciences (SPSS) version 24.0 was used for data analysis (SPSS Inc., 2017, (IBM Corp., Armonk, NY: IBM Corp.). All variables were examined for normality distribution. The study power was more than 80. Descriptive statistics were presented as number and percentage for categorical variables and as mean ± standard deviation for continuous variables. Student's t-test and analysis of variances tests were used to compare means. Chi-square test was used to compare frequencies, and Fisher's exact test was used when the Chi-square was inapplicable. Correlations were analyzed using Pearson's correlation coefficient. The level of statistical significance was set at P ≤ 0.05.

RESULTS

Baseline demographic data and echocardiographic measurements

There were no differences between cases and controls regarding baseline demographics, anthropometric measurements, and clinical risk factors. As would be expected due to the study design, all echocardiographic measurements indicated the presence of LV diastolic dysfunction in the study group with normal LV diastolic function in the control group [Table 1].

Table 1

Baseline clinical and echocardiographic measurements

Variable	Study group (n=40)	Controls (n=40)	P
Clinical characteristics
Age (years)	43.65±9.64	45.88±8.44	0.28
Male gender, n (%)	32 (80)	29 (72.5)	0.60
Current smoker, n (%)	10 (25)	13 (32.5)	0.66
Hypertension, n (%)	11 (27.5)	16 (40)	0.34
Diabetes, n (%)	6 (15)	9 (22.5)	0.57
Dyslipidemia, n (%)	6 (15)	7 (17.5)	1.00
Chronic coronary syndrome	6 (15)	3 (7.5)	0.48
BMI (kg/m2)	30.95±4.70	29.52±3.98	0.145
BSA (m2)	1.90±0.20	1.83±0.17	0.096
Echocardiographic measurements
LVEDV (ml)	70.80±22.13	53.85±13.84	<0.001
LVESV (ml)	24.78±10.39	19.23±7.24	0.007
LVEF (%)	65.74±6.17	64.75±8.42	0.548
LAVI (ml/m2)	19.86±4.80	32.53±5.39	<0.001
E wave velocity (cm/s)	88.23±21.04	74.68±20.31	0.004
A wave velocity (cm/s)	60.45±13.97	86.80±23.68	<0.001
Trans-mitral E/A ratio	1.51±0.444	0.89±0.24	<0.001
Deceleration time (ms)	206.95±56.58	240.70±64.25	0.015
Septal e’ (cm/s)	13.25±3.03	6.43±1.38	<0.001
Lateral e’ (cm/s)	16.58±3.76	7.68±1.21	<0.001
Average e’ (cm/s)	14.91±3.11	7.05±1.15	<0.001
Average E/e’ ratio	6.11±1.84	10.90±3.67	<0.001

Continuous variables are expressed as mean and SD whereas categorical variables are expressed as number (percentage). BMI=Body mass index, BSA=Body surface area, LVEDV=Left ventricular end-diastolic volume, LVESV=Left ventricular end-systolic volume, LVEF=Left ventricular ejection fraction, LAVI=Indexed left atrial volume, SD=Standard deviation

Serum tumor necrosis factor alpha level

The mean serum TNF alpha level was significantly higher in the study group 3.48 ± 1.06 versus 1.22 ± 0.36 pg/ml in the control group, P < 0.001.

On further classifying the study group according to the grade of LV diastolic dysfunction, the mean serum TNF alpha level was higher in patients with Grade 2 diastolic dysfunction (n = 16) 3.91 ± 1.21 versus 3.18 ± 0.86 pg/ml in those with Grade 1 diastolic dysfunction (n = 24), P = 0.03.

The study group patients were also classified according to the NYHA functional class. The mean serum TNF showed a significant trend toward increase with increasing NYHA functional class, P (for trend) <0.0001. Using post hoc analysis a significant difference in serum TNF alpha was found when comparing NYHA Class 1 versus NYHA Class III, NYHA Class I versus NYHA Class IV, NYHA Class II versus NYHA Class III, NYHA Class II versus NYHA Class IV. However, there was no difference on comparing NYHA Class I versus NYHA Class II and difference did not reach significance when comparing NYHA Class III versus NYHA Class IV [Table 2].

Table 2

Serum tumor necrosis factor for different New York Heart Association functional classes in the study group

NYHA functional class	Mean serum TNF alpha (pg/ml)	P
NHYA Class I (n=17)	2.93±0.63	<0.001
NHYA Class II (n=12)	3.13±0.69
NHYA Class III (n=9)	4.48±0.98
NHYA Class IV (n=2)	5.60±0.42
Post hoc analysis of least significant difference		P
NYHA Class I versus. NYHA Class II		0.227
NYHA Class I versus NYHA Class III		<0.001
NYHA Class I versus NYHA Class IV		<0.001
NYHA Class II versus NYHA Class III		<0.001
NYHA Class II versus NYHA Class IV		<0.001
NYHA Class III versus NYHA Class IV		0.062

Continuous variables are expressed as mean and SD. TNF=Tumor necrosis factor, NYHA=New York Heart Association, SD=Standard deviation

The correlation of TNF alpha and echocardiographic parameters:

For study group patients, serum TNF alpha level showed a strong correlation with LAVI (r = 0.645, P < 0.0001). No significant correlations were found between TNF alpha and other echocardiographic parameters of systolic or diastolic function [Table 3].

Table 3

Correlation of serum tumor necrosis factor alpha level to echocardiographic parameters in the study group

Variable	Pearson’s correlation coefficient (r)	P
LVEDV	−0.079	0.624
LVESV	−0.206	0.203
LVEF	0.242	0.133
LAVI	0.645	<0.0001
E wave velocity	0.131	0.419
A wave velocity	−0.103	0.528
Trans-mitral E/A ratio	0.224	0.164
Deceleration time	0.015	0.926
Septal e’	0.178	0.469
Lateral e’	0.119	0.464
Average e’	0.133	0.415
Average E/e’ ratio	0.014	0.933

LVEDV=Left ventricular end-diastolic volume, LVESV=Left ventricular end-systolic volume, LVEF=Left ventricular ejection fraction, LAVI=Indexed left atrial volume

No correlations were found between TNF alpha and all echocardiographic parameters of systolic or diastolic function in controls [Table 4].

Table 4

Correlation of serum tumor necrosis factor alpha level to echocardiographic parameters in controls

Variable	Pearson’s correlation coefficient (r)	P
LVEDV	0.023	0.888
LVESV	−0.021	0.898
LVEF	0.059	0.716
LAVI	−0.145	0.374
E wave velocity	−0.192	0.235
A wave velocity	−0.189	0.241
Trans-mitral E/A ratio	0.039	0.807
Deceleration time	−0.083	0.612
Septal e’	0.195	0.227
Lateral e’	0.036	0.824
Average e’	0.117	0.473
Average E/e’ ratio	−0.244	0.13

LVEDV=Left ventricular end-diastolic volume, LVESV=Left ventricular end-systolic volume, LVEF=Left ventricular ejection fraction, LAVI=Indexed left atrial volume

DISCUSSION

The earliest relation between TNF alpha and the development of heart failure, both systolic and diastolic, was reported by Levine et al. in 1990 in their study that showed elevated serum levels of TNF alpha in patients with end-stage heart failure and cachexia.[17] During the same period, an experimental study performed on conscious dogs (n = 12) using recombinant human TNF alpha showed that TNF alpha impairs LV systolic and diastolic functions. Changes were noted 6 h after infusion and completely resolved after 24–72 h.[18]

Since then, it has been shown that increased TNF alpha levels in the circulation lead to the development of heart failure and LV diastolic dysfunction. Possible mechanisms include neurohormonal activation, cardiac myocyte hypertrophy, extracellular matrix protein production, and necrosis or apoptosis of cardiac myocytes.[1920212223]

The main findings of this study were that TNF alpha level in the circulation was significantly higher in patients with LV diastolic dysfunction of different grades compared to healthy controls with normal LV diastolic function. Serum TNF alpha level was found to be higher in patients with Grade 2 LV diastolic dysfunction compared to Grade 1. Serum TNF alpha levels also showed a positive correlation with LAVI in the study group but not in controls.

In addition, TNF alpha levels demonstrated a trend to be higher in patients with higher NYHA functional class, although they did not reach statistical significance on directly comparing each of Classes I and II, or Classes III and IV. This may be caused by the very small number of patients (n = 2) with NYHA functional Class IV. The trend toward a higher TNF alpha with higher NYHA class may be attributed to the net effect of TNF alpha on the cardiac function, which depends on the amount and duration of exposure to TNF alpha.

While findings of this study seem similar to previously reported data in other small studies, we believe this study is an important addition to the mounting evidence about the importance of TNF alpha in cardiac patients which is currently under extensive research along with other inflammatory cytokines as a possible novel treatment target in heart failure patients.[2425] In addition, we used the most current methods for the assessment of diastolic dysfunction according to the recommendations of the ASE and EACVI, which differ from those performed in some of the earlier studies to confirm that the conclusions they had reached still apply today.

A study on 41 patients undergoing coronary angiography reported that an increased serum level of TNF alpha was associated with LV diastolic dysfunction.[26] Another study on 184 patients with stable coronary artery disease and normal LV systolic function, in addition to, 39 healthy controls demonstrated that serum TNF alpha level was elevated and correlated with LV diastolic dysfunction.[27]

Several review articles reported a direct relation between serum levels of TNF alpha and the severity of LV diastolic dysfunction.[2829] In addition, the relationship between increasing levels of TNF alpha levels and worsening NYHA functional class was reported in a study on 47 patients with either severe aortic stenosis or mitral regurgitation, indicating that TNF alpha is elevated with chronic hemodynamic overload and early decompensated heart failure.[30]

A study involving 100 patients with heart failure with preserved ejection fraction (HFPEF), 100 patients with heart failure with reduced EF, and 50 healthy controls found that plasma levels of TNF alpha are elevated in patients with HFPEF compared to controls. In addition, they reported that in HFPEF patients, TNF alpha receptor 2 was associated with increasing grade of LV diastolic dysfunction and the severity of patient symptoms (NYHA functional class).[31]

A study performed in 2004 on 26 patients with diastolic heart failure and 10 healthy controls to define levels of TNF alpha and interleukin-6 in such patients found that the magnitude of change in TNF alpha levels is associated with NYHA functional class, left atrial volume and grade of diastolic dysfunction. The grade of diastolic dysfunction in that study was defined using changes in trans-mitral E/A ratio, which was the recommendation at that time. Such a correlation between TNF alpha and grade of diastolic dysfunction was not demonstrated in our study, which follows current recommendations of the ASE and EACVI.[32]

In a study on 38 female patients with rheumatoid arthritis and LV diastolic dysfunction, in addition to, 30 healthy female controls researchers reported a correlation between TNF alpha levels and each of LAVI and E/e' ratio. They also demonstrated that using the anti-TNF alpha agent (Infliximab) therapy for 3 months improved left atrial abnormalities, LAVI, and E/e' ratio.[33]

Another study comparing patients with echocardiographic evidence of LV diastolic dysfunction (n = 110) and matched controls (n = 50) found that serum TNF alpha level was higher in those with LV diastolic dysfunction compared to controls. In addition, significant correlations were found between TNF alpha and each of E/e' ratio and E/A ratio.[34]

Limitations of the current study are that it was a single-center study with a relatively small number of participants that could have influenced the results. We did not recruit any patients with Grade 3 LV diastolic dysfunction. LV m-mode measurements and LV mass were not assessed, which may have provided additional insight into the results. TNF alpha level is elevated in a wide array of clinical conditions and diseases. While we tried to eliminate any confounding factors in this study by carefully defining exclusion criteria and having statistically balanced cases and controls regarding other conditions that might affect TNF alpha levels in this population such as diabetes mellitus, hypertension, dyslipidemia, and chronic coronary syndromes, this cannot be guaranteed.

CONCLUSION

Serum TNF alpha level is elevated in patients with LV diastolic dysfunction and is correlated to LAVI in such patients. Patients with Grade 2 diastolic dysfunction have higher serum levels of TNF alpha compared to those with Grade 1 diastolic dysfunction. TNF alpha levels have a trend to increase with worsening heart failure symptoms.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.