Abstract
Background: Doppler echocardiography is the method of choice for diagnosis and evaluation of aortic stenosis. However, there are well-known limitations to this method in difficult-to-image patients. Flow acceleration in the left ventricular outflow tract (LVOT) can lead to overestimation of stroke volume (SV) and poor acoustic windows may impede the exact measurement of the LVOT. The present study aimed to evaluate the use of inert gas rebreathing (IGR)-derived SV in this situation. Patients and Methods: We replaced Doppler-derived SV measurements in the continuity equation (method A) by SV determined by IGR (method B) and by thermodilution during right heart catheterization (method C) to calculate the aortic valve area (AVA) in 21 consecutive patients with moderate or severe aortic stenosis. Results: Mean SV and AVA did not differ between methods at 72±21 ml and 0.71±0.2 cm2 (method A) vs. 66±18 ml and 0.67±0.21 cm2 (method B) vs. 64±15 ml and 0.67±0.21 cm2 (method C), respectively (all p-values >0.05). The mean difference and limits of agreement for AVA were 0.04±0.23 cm2 and -0.40 to 0.47 cm2 between methods A and B, 0.05±0.14 cm2 and -0.26 to 0.27 cm2 between A and C, and -0.05±0.23 cm2 and -0.45 to 0.35 cm2 between B and C, respectively (all p-values >0.05). Conclusion: The presented approach is a reliable method for the calculation of AVA and can add a diagnostic option for the use in difficult-to-image patients. Whereas the use of thermodilution is limited due to its invasive nature, IGR allows the fast and non-invasive determination of cardiac function at low cost.
In the elderly, the predominant valvular disease is calcified aortic stenosis due to focal calcification and thickening of the valvular cusps, resulting in impaired opening and blood flow across the aortic valve (1-3). According to current guidelines, Doppler echocardiography is the method of choice for diagnosis and evaluation of aortic stenosis, replacing invasive measurement of the pressure gradient across the affected valve (4-11). Applying the continuity equation to Doppler-derived data allows the aortic valve area (AVA) to be calculated with good correlation to invasive measurements. Measurement of the width of the left ventricular outflow tract (LVOT), the velocity time integral (VTI) of the pulsed wave in the LVOT and the VTI of the continuous wave (CW) Doppler signal through the stenotic valve are necessary. However, Doppler echocardiographic determination of aortic stenosis has also its limitations, particularly when the patient is difficult to image. A poor acoustic window, heavy calcification of the aortic valve or flow acceleration in the LVOT may impede accurate determination of the AVA. In these cases, AVA has to be independently validated by alternative methods. This can be done by echocardiographic planimetry of AVA (12, 13), by enhancing image quality with transpulmonary contrast agents (14), or by invasive calculation applying Gorlin's formula. Another possibility is to replace Doppler-derived stroke volume (SV) in the continuity equation by SV, determined by alternative methods (6, 15, 16). The traditional method for the latter approach is to determine SV by right heart catheterization. Its limitation lies in the invasive nature of the procedure and its risks and costs (17, 18). Recently, cardiac magnetic resonance imaging (CMR) has evolved into a new gold standard for the non-invasive assessment of cardiac function (19-22). In an earlier investigation evaluating the combination of Doppler-based measurements and CMR in the continuity equation, there was no statistically significant difference for mean AVA, moreover, only two patients were classified differently by the two methods (23). Although it overcomes the difficulties associated with echocardiographic application-alone in the continuity equation, its use is limited due to the high cost of personnel and equipment. A novel method for determining hemodynamic parameters such as SV is the inert gas rebreathing (IGR) method which is based on the Fick principle. The heart rate is registered by pulse oximetry and SV is calculated from the cardiac output (CO) measurement derived from the uptake of inert gas in the rebreathing system. The device we used (Innocor®, Innovision Ltd., Odense, Denmark) employs dinitrous oxide and sulfur hexafluoride as test gases, whose concentrations can be measured online by a photo-magnetoacoustic gas analyzer (24). The IGR method is readily available, reproducible, inexpensive and applicable at bedside. It has been thoroughly studied and compared to accepted invasive and non-invasive techniques (25) in recent years. The aim of our study was to replace Doppler-derived SV in the continuity equation by IGR-derived SV or thermodilution-derived SV and to compare the calculated AVAs.
Materials and Methods
Patients. A total of 29 consecutive patients with suspected or known moderate to severe aortic stenosis referred for Doppler-echocardiographic estimation of AVA were enrolled in our study. Exclusion criteria were poor acoustic window, presence of left-to right shunting, presence of more than mild aortic, tricuspid or mitral regurgitation, heart rate exceeding 100/min and heart rate difference greater than 10 bpm at the time of measurement between any method. Complete echocardiographic examination and IGR measurements were performed consecutively and promptly. Right heart catheterization was performed in 15 patients for clinical reasons, within a week of acquisition of the previous data. Heart rate was documented in every investigation. The study protocol was approved by the institutional Ethics Committee and is in accordance with the declaration of Helsinki. Informed consent was obtained from all patients
Echocardiography. All echocardiographic studies were performed and analyzed by two experienced physician sonographers (KH and DH) on a commercially available ultrasound system (Vivid7, GE Ultrasound, GE Healthcare, Chalfont St Giles, UK). Echocardiographic images were digitally recorded and analyzed offline on a workstation using commercial software (Echopac, GE Healthcare, Chalfont St Giles, UK). LVOT diameter was measured in the parasternal long-axis view according to the recommendations of current guidelines (11). The apical five-chamber view was used to acquire Doppler data in the LVOT. Peak transaortic velocity by continuous-wave Doppler was recorded from apical, right parasternal and suprasternal windows. For patients in sinus rhythm, a minimum of three measurements were obtained and the mean of these measurements was used for the further calculations. For patients in atrial fibrillation at least five consecutive measurements were averaged. SV was calculated as:
Equation 1
AVACE was calculated (Method A) according to the continuity equation (CE):
Equation 2
Severity of stenosis was defined as follows: severe stenosis, AVA<1.0 cm2; moderate stenosis, AVA 1.0 to 1.5 cm2; and mild stenosis, AVA>1.5 cm2.
IGR. Determination of CO using IGR is based on the Fick Principle, which relies on the observation that the total uptake of a substance by the peripheral tissues is equal to the product of the blood flow to the peripheral tissues and the arterial-venous concentration difference of the substance. Dinitric oxide (N2O, 0.5%) as a soluble gas and insoluble sulfur hexafluoride (SF6, 0.1%) were used as test gases, whose concentrations were measured online by the photo-magnetoacoustic gas analyzer of the IGR system (24). CO is the sum of pulmonary blood flow (PBF) and pulmonary shunt volume (QS):
Equation 3
PBF can be calculated from the slope of the regression line through the expiratory points of the logarithmic normalized soluble gas concentrations. For determining QS, the Fick Principle for oxygen can be applied to both pulmonary and systemic circulation. The following equations may be written:
Equation 4
Equation 5
where VO2=oxygen consumption [l/min], CaO2=arterial oxygen content [l STPD/l], CvO2=mixed venous oxygen content [l STPD/l], CcO2 =end-capillary oxygen content [l STPD/l], STPD=standard temperature, pressure, dry.
Assuming an end-capillary oxygen saturation of 98%, equations 3, 4 and 5 contain three unknown variables and can be solved into:
Equation 6
SVIGR was determined by:
Equation 7
AVAIGR was calculated (Method B) according to the continuity equation:
Equation 8
All IGR measurements were performed by JS and FT after demonstration without test gases to ensure correct execution of measurements. Rebreathing manoeuvre was started after a normal expiration at a breathing rate of 20/min (26). The volume of the rebreathing bag was calculated by the IGR system as a function of patient's gender, height and weight. Investigators were blinded to the results of either echocardiographic evaluation and IGR.
Right heart catheterization. Right heart catheterization was performed via the right femoral approach. A Swan-Ganz flow-directed thermodilution catheter (Edwards Lifesciences, Irvine, CA, USA) and a CO computer (Edwards Laboratories, Santa Ana, CA, USA) were used. The mean of three sequential measurements was used for patients in sinus rhythm and five measurements were acquired for patients in rate-controlled atrial fibrillation to determine COthermodilution and SVthermodilution.
AVAthermodilution was calculated (Method C) by the equation:
Equation 9
Statistical analysis. Continuous data were tested for normal distribution using the D'Agostino-Pearson test. Normally-distributed data are expressed as the mean±SD. Non-normally distributed data are presented as the median and 95% confidence interval. Comparisons of the results of the different methods were performed by standard paired Student t-test or Mann-Whitney test, respectively. The agreement between the methods was assessed using the methodology described by Bland and Altman (27). A p-value of <0.05 was considered significant for all comparisons.
Results
Baseline characteristics of the patient collective are shown in Table I. Out of 29 consecutive patients with moderate or severe aortic stenosis referred for echocardiographic evaluation, eight patients were excluded for the following reasons: heart rate difference >10 bpm between the diagnostic modalities (n=2), and poor quality of IGR (n=6), (see Figure 1). None of the patients had more than mild aortic, tricuspid or mitral regurgitation. Patients' age ranged from 43 to 91 years (median=81 years, inter quartile range=73-85 years). In 15 of the included patients right heart catheter data were available. Nineteen patients had sinus rhythm and two presented with rate-controlled atrial fibrillation. The mean heart rate did not significantly differ between methods (method A=74±13 bpm, method B=72±11 bpm, and method C=76±12 bpm). Detailed information on hemodynamic parameters is given in Table II. Using echocardiography and IGR, aortic stenosis was severe in 19 and moderate in two patients, right heart catheterization yielded severe stenosis in 14 and moderate in one case.
Comparison between standard continuity equation (Method A) and IGR (Method B). SV was not different between methods A and B (p=0.3). There was no statistically significant difference in mean AVA between the methods A and B. The mean difference was 0.04±0.23 cm2 and the limits of agreement were −0.40 to 0.47 cm2. The respective Bland-Altman plots are shown in Figure 2 (top). Compared to method A, grading of stenosis changed from moderate to severe in one patient and from severe to moderate in two.
Baseline characteristics of the patient collective.
Comparison between standard continuity equation (Method A) and right-heart catheterization (Method C). SV was not different between methods A and C (p=0.21). There was no statistically significant difference in mean AVA between method A and C. The mean difference was 0.05±0.14 cm2 and the limits of agreement were −0.26 to 0.27 cm2. Bland-Altman plots are shown in Figure 2 (middle). Compared to method A, grading of aortic stenosis changed from moderate to severe in one patient and from severe to moderate in one.
Comparison between IGR-derived AVA (Method B) and that derived from right heart catheterization (Method C). SV was not different between methods B and C (p=0.76). There was no statistically significant difference in mean AVA between method B and C. The mean difference was −0.05±0.23 cm2 and the limits of agreement were −0.45 to 0.35 cm2. Bland-Altman plots are shown in Figure 2 (bottom). Compared to method B, grading of aortic stenosis changed from moderate to severe in one patient and from severe to moderate in another. In seven patients, a marked difference for SV (defined as a difference greater than 15 ml) was noted between methods A and B. In six of these, Doppler-derived values were higher than IGR-derived values. In five of those patients, right heart catheterization data were available and the results lay in between measurements by method A and B.
Detailed overview of the patient collective and reasons for exclusion. IGR: Inert gas rebreathing; RHC: right heart catheterization.
Hemodynamic parameters by method.
Discussion
Echocardiography with calculation of AVA by means of Doppler echocardiography is the method of choice in clinical practice for the evaluation of aortic stenosis. It is non-invasive, readily-available and well-validated (4-8). Nevertheless, there are difficult-to-image patients in whom measurements may be affected by heavy calcification of the aortic valve, flow acceleration in the LVOT, or poor echogenicity. In this situation, overestimation of the numerator of the continuity equation might lead to an overestimation of the true AVA and, therefore, to misclassification of the patient (28-30). Consequently, alternative methods must be used to calculate AVA in these patients, evaluating the use of IGR-derived SV, as is the case in the study at hand. Replacing the numerator of the continuity equation by SV derived from the Simpson's biplane method of discs (28, 31, 32) may be of limited usefulness due to inaccuracies associated with geometric assumptions and plane positioning (12, 33), as well as poor acoustic window in the apical views. Transesophageal echocardiography as a semi-invasive method might be a feasible alternative in some patients, but it does not overcome the problem of poor visualization of the orifice area in severe valvular calcification (12, 33-35). CMR has become the ideal method in the determination of cardiac volumes and can be used to determine CO with great accuracy (19-22). It has already been shown to determine AVA with great accuracy and reproducibility (36). For the calculation of AVA, a combination of Doppler echocardiography and CMR has been described before and yielded reliable results when being compared to the standard procedure (23). However, signal void due to heavy calcification and turbulent flow across the valve as well as velocity aliasing, may be a source of inefficiency in CMR. It is also time consuming and not readily-available at point of care. We therefore studied a hybrid approach where SV was determined by the IGR method which is well-suited to bedside and ambulatory settings. IGR has been recently thoroughly investigated, yielding promising results in comparison to clinically-accepted methods (25, 37-41). A disadvantage of IGR is mandatory active collaboration of the patient, limiting its use in the elderly and severely ill. However, the method is robust even in the presence of atrial fibrillation and pulmonary diseases (42, 43), which are common comorbidities in patients with aortic stenosis. In our study, AVA determination with all three methods - Doppler echocardiography, IGR and thermodilution - was feasible, with a small range of error, as shown by the Bland-Altman plots. Right heart catheterization data was, however, not available in all patients of our small study population. The mean difference in AVA between the standard approach by echocardiography and IGR was only 0.04 cm2. Calculation of AVA was exact and rarely resulted in a change of aortic stenosis severity grading. Echocardiography yielded slightly higher values in the determination of SV and CO than IGR measurements and right heart catheterization, which gave very similar results. These differences were statistically not significant, suggesting that all three methods can be used interchangeably. The variations in the Doppler-derived SV may be explained by incorrect measurement of LVOT or determination of flow in the LVOT already in the acceleration zone before the stenotic valve. Invasive right heart catheterization, as well as non-invasive IGR measurements, showed less variation and a better correlation in regard to SV. The accuracy and reproducibility of our echocardiographic data were comparable to those reported in other published studies (29). Since IGR measurements can be performed rapidly and at bedside, this technique is being increasingly used in hospitals, as well as in the outpatient setting. Acquisition of transaortic VTI by echocardiography takes 5 to 10 minutes. The determination of SV by IGR will add another five minutes, resulting in a total time of 10 to 15 minutes. In contrast, measurement and image analysis of ventricular volumes by CMR requires at least 35 minutes, resulting in a total time of 40 to 45 minutes. The accuracy of our approach is comparable to using a combination of Doppler echocardiography and CMR described before (23), with a mean bias of 0.04 (limits of agreement −0.40 to 0.47) cm2 vs. 0.02 (0.32 to 0.36) cm2 against the standard continuity equation. Although it requires active collaboration, IGR is associated with low operating costs and requires only a few expendable items. Therefore our hybrid approach using IGR might add an option for grading aortic stenosis in selected patients with poor LVOT visualization or contraindications for other techniques. Due to its non-invasiveness, as well as its rapid and safe application, it might also be of utility in serial measurements for the follow-up evaluation of aortic valve stenosis. Although thermodilution yielded similar results in calculating the AVA, its widespread use for replacing the standard continuity equation seems to be at least limited due to its invasive nature, including risks and discomfort.
Top Standard continuity equation (Method A) vs. hybrid approach using IGR (Method B) with a mean bias of 0.04±0.23 cm2. Middle Standard continuity equation (Method A) vs. hybrid approach using right heart catheterization (Method C) with a mean bias of 0.05±0.14 cm2. Bottom: Hybrid approaches using inert gas rebreathing (Method B) vs. right heart catheterization (Method C) with a mean bias of −0.05±0.23 cm2. AVA: Aortic valve area; IGR: inert gas rebreathing.
Conclusion
Replacing Doppler-derived SV in the continuity equation by SV derived from IGR or thermodilution is a reliable method for the calculation of AVA and can add a diagnostic option for difficult-to-image patients. Whereas the use of thermodilution might be of limited use due to its invasive nature, IGR allows a fast and non-invasive determination of cardiac function at low cost.
Footnotes
-
↵* These Authors contributed equally to this study.
- Received July 18, 2012.
- Revision received September 25, 2012.
- Accepted September 27, 2012.
- Copyright © 2012 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved