4.       Introduction 

Head Up Tilt Table Testing (HUT) is used as a diagnostic tool for unexplained syncope [1-3] as well as for quantification of hemodynamic changes during orthostatic stress [4-6]. To demonstrate the influence of orthostatic stress on hemodynamics, continuous Heart Rate (HR) and Blood Pressure (BP) measurements during HUT are commonly used. Based on these two measurements predefined conditions as Orthostatic Hypotension (OH), Postural Orthostatic Tachycardia Syndrome (POTS) and Neurally Mediated Syncope (NMS) can be diagnosed [7,8]. For more extensive hemodynamic measurements (Stroke Volume (SV) and Cardiac Output (CO)) can be used. SV and CO can be measured by thermodilution (considered to be the gold standard) [9], but is limited by the invasive nature and therefore less useful during HUT. Other non-invasive techniques for the estimation of SVI and CI are transthoracic bio impedance [4,10], acetylene rebreathing [11], dye dilution [12] and Doppler Echocardiography (transthoracic apical imaging) [13-16]. Transthoracic apical imaging is not practical during HUT, due to the need for lateral positioning of the subjects for acquisition of the images. 

As an alternative, the Time-Velocity Integral of the aorta (VTI) using a suprasternal approach [17-19] has been used for stroke volume measurements. However, these studies used a limited number of subjects. 

Therefore, the aim of this study was to validate the suprasternal VTI Doppler derived stroke volumes by comparison of the stroke volumes with that of transthoracic Doppler echocardiography derived stroke volumes. 

For this purpose, 119 consecutive patients with Chronic Fatigue Syndrome/ Myalgic Encephalomyelitis (CFS/ME), undergoing HUT, were studied. The supine aortic VTI derived stroke volumes (prior to the start of the tilt) were compared with previously obtained stroke volumes by standard transthoracic echocardiograms. 

5.       Material and Methods 

5.1.  Patient selection 

Between November 2012 and December 2016 119 consecutive CFS/ME patients were studied in whom Orthostatic Intolerance (OI) assessment was part of the work-up. All underwent HUT using Nexfin recording for Heart Rate and Blood Pressure. Stroke Volume Indices (SVI) were calculated, using suprasternal continuous wave Doppler aortic VTI measurements (see below). For validation purposes in these subject’s continuous wave Doppler echocardiography derived VTI of the aortic valve was acquired by apical imaging. 

The study has been carried out in accordance with Declaration of Helsinki and was approved by the MEC of the Slotervaart hospital, Amsterdam, NL. 

5.2.  Doppler echocardiographic measurements 

VTI frames were obtained in the resting supine position. The aortic VTI was measured using a continuous wave Doppler pencil probe connected to a Vivid I machine (GE, Hoevelaken, NL) with the transducer positioned in the suprasternal notch. A maximal Doppler signal was assumed to be the optimal flow alignment. At least 2 frames of 6 seconds were obtained. Echo Doppler recordings were stored digitally. 

From a transthoracic echocardiogram the diameter of the left ventricular outflow tract was obtained as well as a VTI using continuous wave Doppler echocardiography of the aortic valve.

5.3.  Data Analysis 

For comparison with the transthoracic echocardiography SV only the VTI frames in the supine position, prior to tilting, were analysed. The time velocity integral was measured off line by manual tracing of at least 6 cardiac cycles, using the GE EchoPac post-processing software. This was done for both transthoracic as well as suprasternal acquired VTI’s by one operator (CMCvC). Stroke volumes were calculated from the time–velocity integral of the aortic valve, corrected for the aortic valve as described previously [20,21]. SVI was calculated by the equation: corrected LVOT cross-sectional area times the aortic VTI, divided by the Body Surface Area (BSA; DuBois formula). SVI’s of the separate cycles were averaged. 

6.       Statistical Analysis 

Data were analysed using SPSS version 21 (IBM). All continuous data were tested for normal distribution using the K_S test. Except for disease duration and BMI, all data were normally distributed, data are presented as means ± SD. Disease duration and BMI were presented as median and IQR. Groups were compared using Students T test for unpaired data, within group comparison by Students T test for paired data. Graphs were constructed using Graphpad Prism version 6.00, Graphpad software, La Jola California USA. Comparison of regression lines were performed using the Prism software. 

For measurement of intra- and interobserver variation 30 VTI aorta images were studied. The operator CMCvC measured the VTI data at least 3 months apart, being blinded for the first analysis. A second operator (FCV) measured the same VTI data being blinded for the name of the patients on the PC screen. This analysis was performed at least 6 months after the tilt study. The Intra-Class Correlation Coefficient (ICC) for intra and interobserver reliability was calculated with SPSS. 

7.       Results 

Baseline characteristics of ME/CFS patients and HV are presented in (Table 1).

79 patients had a normal heart rate and blood pressure response, 20 patients had POTS and 20 a classic or delayed OH. Mean age was 43±12 years, median duration of ME/CFS was 12 (IQR 6-20). Mean Systolic Blood Pressure (SPB) was 137±16 mmHg and mean Diastolic Blood Pressure (DBP) was 80±8 mmHg.

(Table 2) shows the hemodynamic and echocardiographic data of the ME/CFS patients. Heart Rate (HR), the Time Velocity Integral of the aorta (VTI) and Stroke Volume Index (SVI) were derived from both transthoracic echo (TTE) as well as during the supine phase of the Head Up Tilt Test (HUT), using suprasternal Doppler. HR was 73±11 bpm and 71±9 bpm for HUT and TTE respectively. VTI were 24±3 cm for both techniques and SVI was 37±5 ml/m² and 36±5 ml/m² for suprasternal and transthoracic measurement techniques.

(Figure 1) Shows the correlation between SVI measurements using suprasternal Doppler and transthoracic Doppler measurements. The correlation was highly significant and close to the line of identity with an equation of: Y=0.95^*X+1.12; r=0.92; p<0.0001 

(Table 3) Regression analysis of HUT hemodynamic results.

(Table 3) shows the regression analysis of the three different hemodynamic groups (Figure 2). The slopes of the three regression lines were not significantly different.

The ICC to assess intra-observer variation of the VTI determination was 0.98 and the ICC for inter-observer variation was 0.99. 

8.       Discussion 

The main finding of this study is that suprasternal Doppler for determination of Stroke Volume Index (SVI) and Cardiac Index (CI) is an accurate technique during HUT giving comparable results as SVI determination with transthoracic aortic Doppler measurements. 

One previous study validated the two techniques used in our study in 10 patients with coronary artery disease [22]. In this study a correlation similar to the line of identity, like in our study, was found. They also established that no systematic error was introduced with different observers. 

Limited studies are described in literature concerning comparisons of techniques, especially echo Doppler techniques. Bouchard et al. studied 41 patients undergoing cardiac catheterization and compared thermodilution determination of SVI/CI with echo Doppler measurements [18]. In 21 patients with adequate suprasternal and apical images, SVI measurements showed a similar correlation close to the line of identity as in our study. 

Spahn et al. studied 25 patients undergoing coronary artery bypass grafting surgery and compared CO results, using two invasive techniques (Fick and thermodilution) and 4 non-invasive measurements, being bio impedance, suprasternal Doppler measurements using Continuous Wave Doppler (CWD) and Pulsed Wave Doppler (PWD), as well as continuous wave Doppler at the aortic valve during Transoesophageal Echocardiography (TEE) [23]. Fick and thermodilution were highly correlated, the echo techniques were only available in 15 (TED and PWD) or 21 patients (CWD). The suprasternal CWD was the most reliable of the 3 echo techniques when compared to thermodilution with r values of 0.84 for CWD, 0.65 for PWD and 0.62 for TEE. For the determination of the aortic valve area, the surgical diameter of the ascending aorta 3-4 cm above the aortic valve was used. Limitations in alignment during transoesophageal echo (TEE) and higher velocities and limited sampling during PWD might be the reason for the poor r values. No comparison of the three echo Doppler techniques was available in this study. 

Espersen at al. compared CO measurements using Fick, CO2 rebreathing and suprasternal pulsed wave Doppler in 11 healthy volunteers [24]. Doppler measurements were accurate in the Fick range of 4.6-8.1 l/min, underestimated CO with higher than 8.1 l/min Fick values and overestimated CO with lower than 4.6 l/min Fick values. No comparison with transthoracic Doppler measurements were available.

Gardin et al. studied intra- interobserver and CO variability with a suprasternal pulsed wave Doppler technique in 10 healthy volunteers [17]. Serial recordings interpreted by the same observer showed reliably alteration in hemodynamics. Only pulsed wave suprasternal Doppler was studied, no comparison was made with continuous wave suprasternal Doppler or transthoracic apical Doppler measurements. 

9.       Limitations 

The limitations of Doppler SVI determination are that acquisition is time consuming, patient echo window dependent and investigator dependent. 

10.   Conclusion 

Using suprasternal Doppler imaging to acquire aortic VTI for stroke volume measurements in HUT as the more convenient technique compared to using transthoracic Doppler echocardiography gives reliable results when the two techniques are compared in CFS/ME patients.