Relation between E/e’ ratio and NT-proBNP levels in elderly patients with symptomatic severe aortic stenosis

Background Symptoms in the elderly patients with severe aortic stenosis (AS) and co-morbidities seem to lack in specificity. Therefore, objective parameters for increased left ventricular(LV) filling pressures are needed. The aim of this study was to investigate the correlation between the septal, lateral and average E/e’ ratio and the value of the N-terminal pro-hormone of brain natriuretic peptide (NT-proBNP). Methods Two-hundred-fifty consecutive symptomatic patients (mean age 80 ± 8 years, 52 % men) with severe AS underwent transthoracic echocardiography and NT-proBNP measurement. Results In the overall population the septal E/e’ (r = 0,459, r2 = 0,21, P <0,0001), lateral E/e’ (r = 0,322, r2 = 0,10, P <0,0001), and the average E/e’ (r = 0,432, r2 = 0,18, P <0,0001) were all significantly correlated to NT-proBNP. After the exclusion of patients with confounders (more than mild aortic or mitral regurgitation, severe renal dysfunction, obesity or severe COPD) the septal E/e’ (r = 0,584, r2 = 0,34, P <0,0001), lateral E/e’ (r = 0,377, r2 = 0,14, P <0,0001), and the average E/e’ (r = 0,487, r2 = 0,24, P <0,0001) were all significantly better correlated to NT-proBNP. In obese patients no significant correlations were seen. Previous bypass surgery did not alter the correlations. Conclusions In elderly patients with severe symptomatic AS there is a significant correlation between the E/e’ ratio and NT-proBNP, in particular after exclusion of confounders. The correlation was best for the septal E/e’ ratio and was preserved in patients with a history of bypass surgery.


Introduction
Aortic stenosis (AS) is the most common valvular heart disease in the Western world and its prevalence is expected to increase due to the aging of the population [1]. Current recommendations warrant the presence of symptoms in order to indicate aortic valve replacement [2]. However, in elderly patients symptoms related to increased left ventricular (LV) filling pressures are less specific, because of the normal aging process and the multiple comorbidities [3]. One objective parameter of increased LV filling pressures is the echocardiographic ratio of the transmitral E-wave velocity and early diastolic velocity of the myocardium, the so called E/e' ratio [4][5][6][7]. Although the value of this ratio may be less reliable in elderly AS patients with important annular calcifications [7], the E/e' was proven to be a strong predictor of mortality in non-operated elderly patients with AS, especially in conjunction with the N-terminal pro-hormone of brain natriuretic peptide (NT-proBNP) [8]. This hormone reflects the total burden of the disease on the LV, and has been demonstrated to have a good prognostic value in the setting of severe AS [9][10][11][12][13][14][15][16].
This study sought to assess the relation between the E/e' ratio and NT-proBNP. More specifically, the values of the septal versus lateral myocardial wall derived E/e' ratio were investigated in detail.

Patient population
After exclusion of patients with an aortic or mitral valvular prosthesis, mitral stenosis or significant areas of focal LV akinesia (defined as an akinetic region extending to at least 3 segments, involving the basal septum and/or the basal lateral wall) the study included 250 consecutive subjects (mean age 80 ± 8 years, 52 % men) with symptomatic severe AS, defined as an aortic valve area less than 1 cm 2 and/or less than 0.6 cm 2 /m 2 and available NT-proBNP level who were treated by transcatheter aortic valve replacement (TAVI). The study was approved by the Medical Ethical Committee of the Erasmus MC. All patients gave their informed consent.

Transthoracic echocardiography
Two-dimensional echocardiography was performed using a Philips iE33 system (Philips, Best, The Netherlands) with the patient in the left lateral decubitus position. Echocardiographic studies were performed by an independent experienced echocardiographer, blinded to the patient's clinical and biochemical status. All echocardiograms were saved as video loops or still frames in a digital database and were reanalyzed by an experienced cardiologist (MS). The LV ejection fraction cut-off value of 50 % was calculated using the biplane modified Simpson rule. The mitral inflow velocity profile (E/A wave ratio, E wave deceleration time) was assessed with pulse-wave Doppler. The early diastolic velocities at the septal and lateral annulus level were assessed with pulse-wave Tissue Doppler from a standard apical 4-chamber view.

Blood sampling
Venous blood samples were drawn from patients with AS within 30 min of the Doppler echocardiography study from an antecubital vein into EDTA acid Vacutainer test tubes (Mediost BV, Doesburg, The Netherlands) after 30 min of supine rest. Samples were placed immediately on ice, and plasma separation was performed at 4°C. For NT-proBNP determination, an electrochemiluminescence immunoassay (ProBNP Elecsys, Roche Diagnostics GmbH, Mannheim, Germany) was used. For the evaluation of renal function, serum creatinine levels were determined and glomerular filtration rate (GFR) was calculated using the 2009 CKD-EPI creatinine equation [17].

Definitions of possible confounders
Impaired ejection fraction was defined as <50 %. Obesity was defined as a BMI (body mass index) above 30 kg/m [2,18]. Severe COPD was considered in GOLD class III or IV [19].
Severe impairment of the renal function was defined as GFR less than 30 ml/min/1,73 m 2 , present for more than 3 months, as estimated by 2009 CKD-EPI creatinine equation [17]. Aortic and mitral regurgitation were assessed semi quantitatively according to current guidelines for the evaluation of native valves [20].
Significant coronary artery disease was defined as >50 % stenosis in at least one coronary artery.
Pulmonary hypertension was considered when the estimated pulmonary pressure derived from Doppler tracings of the tricuspid insufficiency was above 40 mmHg.

Statistical analysis
Continuous variables are presented as means (± SD) if normally distributed, or otherwise by geometric means for natriuretic peptide levels. Because of the very large range of values of the natriuretic peptide and the abnormal distribution of this variable, the log10 of the NT proBNP was used in the analysis. Categorical variables are presented as frequencies and percentages. Differences between similar continuous variables were assessed by the paired-samples t test. Multivariable linear regression was performed in order to identify the possible confounders. Correlations were computed using Pearson's method, and graphically represented with linear regression lines whenever appropriate. For nonlinear parameters, a best-fit regression line was traced, using multiple nonlinear regression equations and choosing the most statistically significant model that fitted the data. A two-sided p value less than 0.05 was used for declaring statistical significance. All statistical analyses were performed with SPSS 21.0 software (SPSS Inc, Chicago, IL, USA).
In contrast, the level of linear correlation with the mitral E/A ratio did not change (r = 0,255, r 2 = 0,07, P = 0,01) and the E-wave deceleration time was not at all correlated to NT-proBNP (r = −0,065, r 2 = 0,04, P = 0,5). As seen in Fig. 3, the best-fit correlation lines were described by quadratic equations, with r 2 = 0,12, P = 0,002 for the E/A ratio and non-significant results on all models for the E-wave deceleration time.
The value of the NT proBNP in obese patients was lower, with a geometrical mean of 94 pmol/l, 95 % CI (81,142), compared to 217 pmol/L in the non-obese patients (P =0.027). In the other subgroups (COPD, renal dysfunction) a specific analysis was not possible because of the limited number of patients.

Discussion
In this study in elderly patients with severe symptomatic AS the main findings were: 1) there is a significant correlation between the E/e' ratio and NT-proBNP, in particular in patients without obesity, severe renal dysfunction, severe COPD, or significant left-sided valvular regurgitation, 2) the correlation was best for the septal E/e' ratio, and 3) this latter correlation was preserved in patients with a history of CABG. In elderly patients with severe AS, clinical symptoms are difficult to ascertain and less specific because of the normal aging process and the multiple comorbidities usually present in the AS patients referred for TAVI. In order to help the decision for aortic valve intervention, an objective parameter of increased LV filling pressure seems important in their evaluation. The E/é ratio is the main stay of echocardiography to indirectly assess LV filling pressure [21,22]. The current guidelines recommend the use of the average E/é ratio derived from the septal and lateral wall [22,23]. However, evidence for the use of the average E/e' ratio is lacking and the specific accuracy of the septal or lateral E/e' is still debated. Each of the two E/e' ratios has its specific benefits. Translational movement of the heart may less affect septal wall velocity and certainly Doppler beam angle errors are less likely to occur when septal wall velocities are measured. Also, it was shown that the septal E/e' ratio best correlated to LV filling pressures in subjects with preserved systolic LV function [24] and in particular in the elderly [25]. On the other hand, Nagueh et al. claimed that "the lateral E/e' ratio was easier to quantify" [26,27]. Obviously, a septal or lateral localized myocardial wall infarction will negatively influence the respective measurements. Therefore, patients with such abnormalities were excluded from our study.
Our study population has some important unique features. The patients were elderly with extensive aortic and mitral annular calcifications and a significant number of these patients had coronary artery disease and a history of previous CABG. Despite these characteristics a good correlation between the E/e' ratio and NT-proBNP was found. The best results were seen for the septal E/é ratio with a correlation coefficient of 0.459. Exclusion for confounders (obesity, severe renal dysfunction, severe COPD, or significant left-sided valvular regurgitation) resulted in even better correlations. The best improvement was seen for the septal E/é ratio with a correlation coefficient of 0.584.
The septal E/é ratio could be superior due to the aforementioned reasons but also because of the extensive posterior mitral annular calcification that may have resulted in more problems in assessing the lateral annulus, which is a part of the posterior annular ring [28]. We had expected that a previous CABG with its influence on septal function [29,30] would affect the accuracy of the septal E/é ratio. Surprisingly, a history of CABG did not affect the correlation of the septal E/é ratio with NT-proBNP.

Comparison to other studies and clinical implications
The level of correlation found between E/E' and NT-proBNP in our study in elderly patients with AS is quite similar to correlations published in the literature in general cardiac patients [31][32][33]. In AS little data are available [16,34]. There is only a very small study on 29 middle-aged (65 ± 12 years) severe AS patients [16]. In that paper weak correlations were described between E/e' and NT-proBNP with only significant results seen for the septal E/e'. Because of the evidence in the literature [16,25], as well as the results of the present study and the specific practical benefits, the use of septal E/e' ratio rather than the lateral or average E/e' should be encouraged, until other sound data are available from prospective large-scale trials such as the EURO-FILLING study [35].
In the global population included here, the correlation between the surrogate markers of LV filling pressures remains moderate. Only about 44 % of the symptomatic elderly patients with severe aortic stenosis displayed strong correlations. From a practical clinical standpoint, care should be taken in interpreting the results of the usual diastolic function parameters in the presence of the confounders defined in this study: more than mild aortic/mitral regurgitation, severe renal impairement, severe COPD, obesity and left ventricular dysfunction.
Apart from the negative impact of impaired EF on the correlation between E/e' and NT-proBNP, the influence of obesity was striking. Even for the septal E/e' ratio no significant correlation was seen, probably because of the well-known paradoxically lower NT-proBNP levels than in normal weight patients for the same filling pressures [36]. In such patients we can only speculate that the E/e' ration better reflects LV filling pressures.
These findings could have an impact on the costs of medical care, since adding a NT proBNP assay to the routine evaluation is relatively expensive [37] and not covered by medical insurances, while E/e' ratio can be measured repetitively during each echocardiographic examination, without any supplemental risk or cost for the patient.

Study limitations
As mentioned above, the population included in this study was carefully selected by a "heart team" according to present guidelines on the management of valvular heart disease [20]. That means first of all that they had to have clinical symptoms in association with a severe aortic stenosis in order to be considered for TAVI. An asymptomatic control group with severe aortic stenosis could have better demonstrated the role of echocardiographic data in the decision making. Such a group is difficult to constitute in an elderly population with several comorbidities.
The cross-sectional nature of the study could be a limitation in itself. However, data collection in this population was very rigorous, according to a very strict and prospectively designed protocol.
Although it demonstrates a correlation between echocardiographic and biological markers of elevated left ventricular filling pressures, it is possible that the cutoff value of 15 for the septal E/e' is too sensitive (practically all the patients having an E/E' ratio above this value, with a mean of 20 ± 9), or that all the selected patients had chronically elevated LVEDP, as reflected by the mean value of NT proBNP = 217pmol/l. Unfortunately, a better cutoff point could not be determined from our data, because the LVEDP was measured in all patients during the initial phase of the intervention, before TAVI, but already on general anesthesia, which invariably leads to dramatic changes in pressures.

Conclusion
In elderly patients with severe symptomatic AS there is a significant correlation between the E/e' ratio and NT-proBNP, in particular after exclusion of confounders. The correlation was best for the septal E/e' ratio and was preserved in patients with a history of CABG.