Asian Cardiovasc Thorac Ann 2008;16:278-283
© 2008 Asia Publishing EXchange Ltd
Hemodynamic Evaluation of Aortic Regurgitation by Magnetic Resonance Imaging
Thomas Wittlinger, MD,
Omer Dzemali, MD,
Farhad Bakhtiary, MD,
Anton Moritz, PhD,
Peter Kleine, MD
Department of Thoracic and Cardiovascular Surgery, University Hospital, Frankfurt, Germany
For reprint information contact: Thomas Wittlinger, MD Tel: 49 69 6301 83315 Fax: 49 69 63013842 Email: thomaswittlinger{at}t-online.de, Department of Thoracic and Cardiovascular Surgery, University Hospital, Theodor-Stern Kai 7, 60590 Frankfurt, Germany.
 |
ABSTRACT
|
|---|
Magnetic resonance imaging was compared with echocardiography and angiography in determining the regurgitant volume in patients with aortic regurgitation. Forty patients were examined at 1.5 T. The regurgitant jet was located using a gradient-echo sequence. Cine measurements were performed to calculate left ventricular function. For flow evaluation, a velocity-encoded breath-hold phase-difference magnetic resonance sequence was used. The degree of aortic regurgitation assessed by magnetic resonance imaging agreed with that of angiography in 28 of 40 (70%) patients, and with the echocardiography result in 80%. Correlation between calculated stroke volume by magnetic resonance cine and flow measurements was very good (r > 0.9). Magnetic resonance imaging enables quick and reliable quantitative assessment of aortic regurgitant volume, and it might be the optimal technique for multiple follow-up studies and assessment of left ventricular function, leading to better evaluation of disease severity and optimization of the timing of valve surgery.
|
INTRODUCTION
|
|---|
Quantification of chronic aortic regurgitation (AR) is a difficult clinical problem and a prerequisite for optimal timing of surgical intervention. Every effort should be made to operate before serious left ventricular (LV) dysfunction occurs. The severity of AR has traditionally been estimated by contrast aortography, which is impractical for screening or serial follow-up. Angiography provides only semiquantitative assessment and correlates poorly with regurgitant volume (RV), especially in the presence of LV enlargement which dilutes LV opacification by the regurgitant aortic flow.1 Invasive calculation of RV and regurgitant fraction (RF) is affected by errors resulting from determination of cardiac output and quantification of LV volume. In Doppler echocardiography, AR is estimated by the size of the regurgitant jet in the LV cavity, the jet width in the LV outflow tract, and the pressure half-time measured by continuous-wave Doppler. Calculation of RV or RF is also possible by echocardiography because the total stroke volume through the aortic valve must equal forward stroke volume plus RV. This method is limited by the need for multiple measurements, and it assumes no regurgitation at the reference valves. Furthermore, physiological factors affect the appearance of the AR jet on Doppler interrogation, such as eccentricity and velocity of the jet, shape and size of the orifice, volume and pressure within the heart chambers, as well as heart rate, rhythm, and LV function.
Magnetic resonance imaging (MRI) is well suited to noninvasive imaging of the heart, and its role in heart valve disease is being evaluated. Semiquantitative grading by visualizing the AR jet has been compared to echocardiographic techniques, but quantitative measurements are now being explored. In contrast to previous studies with non-breath-hold velocity-encoded phase-difference sequences, we used a breath-hold phase-difference technique. In previous studies, this yielded accurate flow measurements in large vessels, coronary arteries, and bypass grafts.2 This study aimed to evaluate the accuracy of this MRI technique in quantifying RV in patients with chronic AR and to correlate the data with echocardiographic and angiographic results.
|
PATIENTS AND METHODS
|
|---|
Forty patients (16 women, 24 men) with a mean age of 60.6 years and mild (I°) to severe (IV°) AR were investigated by MRI, echocardiography, and cardiac catheterization. The etiology of AR was rheumatic fever in 12 patients, endocarditis in 8, aortic atherosclerotic disorders in 4, bicuspid valve in 7, aortic aneurysm in 4, and unknown in 5. Coronary artery disease was excluded in all patients. The mean LV ejection fraction (EF) was 58% ± 8%, and there was no concomitant disease except non-insulin-dependent diabetes mellitus in 6 patients and arterial hypertension in 9. Medications included beta blockers in 31 patients, angiotensin-converting enzyme inhibiters in 28, diuretics in 15, and oral antidiabetic medication in 5. Angiography and echocardiography were performed before MRI. Patients with atrial fibrillation and severe arrhythmias or unstable conditions were excluded. The study was approved by the local ethics committee. All patients were informed about the potential risks of MRI and gave their written consent.
Qualitative and quantitative assessments of AR were carried out by left and right heart catheterization 1–10 days before MRI. Angiograms were calibrated by exposing a standard grid on the radiographic film. Volumes were calculated by the method of Dodge and colleagues.3 For quantitative determination, the total stroke volume was assessed by cineangiography in 2 planes (right anterior oblique 30°, left anterior oblique 60°). The forward stroke volume was determined by thermodilution, and the difference between volumes yielded the RF. Regurgitant fraction was calculated as: (TSVangio – FSVFick)/TSVangio, where TSVangio is the total angiographic stroke volume, and FSVFick is the forward stroke volume (Fick’s principle). Qualitative assessment of the severity of AR was also performed by visual estimation of the concentration of contrast medium in the left ventricle, using the method of Seller and colleagues.4 AR I° is confinement of the jet to the LV outflow tract and disappearance at each systole. With AR II°, the contrast jet remains for more than one systolic contraction. III° is associated with continuous filling of the cavum between 2 or 3 heart cycles. IV° is opacification of the entire left ventricle at the first heart beat.
All patients underwent transthoracic Doppler echocardiography. Color-flow techniques included measurement of the maximal anteroposterior diameter (height) of the regurgitant jet at the junction of the LV outflow tract and the aortic annulus in parasternal long-axis view, and the maximum height of the LV outflow tract at the same location. Continuous Doppler-wave imaging of AR permits quantification of both the slope and pressure half-time. Regurgitant volume was calculated as: aortic flow mitral flow. Regurgitant fraction was calculated as RV/aortic flow. Based on these methods, the severity of AR was graded as I for mild, II for moderate, III for moderate to severe, and IV for severe.
Magnetic resonance imaging was carried out with a 1.5-T Magnetom Vision device (Siemens AG, Erlangen, Germany), with a 25-mT·m–1 gradient and 600-µsec rise time. All patients were investigated by electrocardiogram-gated MRI in the supine position with a phased-array body coil. Initially, a turbo fast low-angle shot sequence (repetition time/echo time 1100/2.3 ms, field of view 400 mm, flip angle 10°, 128 x 256 matrix, slice thickness 8 mm) was obtained in various planes (4 transverse, 1 coronal, 1 sagittal) during a breath-hold at deep inspiration. These images were used as localizers. A 2-dimensional T2-weighted breath-hold (end-inspiration) turbo spin-echo sequence, half-Fourier acquisition single-shot turbo spin-echo sequence (HASTE; effective echo time 44 ms, repetition time 800 ms, slice thickness 5 mm, field of view 230–350 mm, multislice technique with 7 slices, 176 x 256 matrix size) was planned on the coronal localizer for axial views. The sequence was triggered to every R wave, the trigger delay was 0 ms. All patients were imaged in axial, coronal, and sagittal planes. A breath-hold gradient-echo cine study in angulated coronal projection planned on a transversal HASTE image (LV outflow tract) was performed to visualize the direction and area of the regurgitant jet. In each patient, the same gradient-echo (cine) sequence was also performed in the short axis (planned on a transversal HASTE image perpendicular to the ventricular septum) to assess LV volume and EF (Figure 1
). For flow measurements, a breath-hold velocity-encoded phase-difference MR sequence was used ("through plane", segmented fast low-angle shot 2-dimensional sequence, repetition time/echo time 110/5 ms, velocity encoding 250 cm·sec–1). The duration of the breath-hold period was 15–25 sec. Flow measurements were performed in the ascending aorta by positioning the slice in the vicinity of the aortic valve. A repetition time of 110 ms allowed acquisition of 6–8 pairs of magnitude and corresponding phase images on the flow curve during the cardiac cycle. The total MRI examination time was approximately 30 min for all sequences.
View larger version (107K):
[in this window]
[in a new window]
|
Figure 1. Assessment of aortic regurgitant volume by magnetic resonance imaging flow measurement, and identification of the aortic regurgitant jet (arrow).
|
|
Magnetic resonance imaging data were analyzed by 2 independent and experienced observers who had no knowledge of the clinical AR grades of the patients. The cross-sectional area of the aorta was defined separately on each image magnitude by a region of interest. Within these individual regions of interest, the mean flow velocity was measured on a pixel-by-pixel basis. Using new work-in-progress software (Argus WIP 2.3, Siemens AG, Erlangen, Germany), the forward and reverse volumes and the net forward volume were determined, and RF was calculated. To avoid aliasing, velocity encoding could be shifted by the software program (e.g. from ± 250 to +400 to –100 cm·sec–1). A calculated RF of 0%–15% was graded I (mild), 16%–30% was graded II (moderate), 31%–50% was graded III (moderate to severe) and > 50% was graded IV (severe), according to the standard grading criteria.5 Data from the cine measurements were also analyzed by the Argus software to calculate LV function indices.
Echocardiography and angiography were performed by an experienced cardiologist not involved in the study, with no knowledge of the MRI results. Descriptive results were expressed as mean ± standard deviation for continuous variables, and as percentages for categorical variables. To correlate the different methods, the Spearman coefficient was used. The Pearson coefficient was used for samples with a normal distribution. To evaluate the influence of LV function, the chi-squared test was used. All calculations were performed with SAS/Stat statistical software version 8.2 (SAS Institute, Cary, NC, USA).
|
RESULTS
|
|---|
The mean RF values obtained by the 3 different methods are listed in Table 1. The mean RV assessed by each method is given in Table 2. There was good correlation among the 3 methods ( p < 0.001). When the patients were divided on the basis of good LV function (EF 
40%) or poor LV function (EF < 40%), there was a highly significant correlation between MRI and echocardiography data in those with good LV function (r = 8.01 – 0.9; Table 3). MRI data overestimated AR by one degree compared to echocardiography in 8 patients; in 7 of these, the degree of AR obtained by MRI was confirmed by angiography. In one patient, RF of 36% (AR III°) was determined by MRI, but AR II° was measured by both angiography and echocardiography. There was also good correlation between MRI and angiography regarding the degree of AR (r = 1.77 – 0.5); AR was overestimated by one grade in 7 patients, and underestimated by one grade in 5. In 9 of these 12 cases, the AR grade from MRI was confirmed by echocardiography. The AR grade from MRI was overestimated by one grade in 3 patients. LV function had a decisive influence on the quality of information given by echocardiography and angiography (Table 3). Each of the 3 patients in whom AR was overestimated had markedly reduced cardiac function (EF 26%–30%). When LV function is impaired, determination of RF by MRI is more accurate and clearly has less variability than angiography. This shows that more accurate determination of RF is possible using MRI in severe AR with reduced LV function (Table 3). Aortic regurgitation values obtained by MRI, echocardiography, and angiography correlated significantly when EF 40%. In the group with impaired LV function, only MRI (r = 0.934, p < 0.0001) showed significant correlation between AR degree, RF, and RV. With EF < 40%, correlation between the degree of AR determined by angiography and echocardiography was less significant, and there was no correlation with RV and RF. However, the relationship of AR severity with RF and RV using MRI was still significant with severely reduced LV function; thus MRI allows more exact evaluation of AR in these patients. Comparing AR assessed by MRI and echocardiography, only 1 patient had results outside 2 standard deviations; 97.5% of patients had results within 2 standard deviations. The reference level of the Bland-Altman test was < 95%, and the difference between the 2 methods was significant. Magnetic resonance imaging showed the lowest interobserver variability, with a kappa value of 0.811 compared to 0.585 with echocardiography and 0.641 with angiography.
|
DISCUSSION
|
|---|
For most patients with AR, a semiquantitative approach is adequate for therapy. The optimal timing of surgical intervention in cases of chronic AR remains controversial. Clearly, aortic valve replacement is appropriate in patients with severe AR and significant symptoms or signs of LV enlargement. Several studies have attempted to identify predictors of outcome and LV function after aortic valve replacement in symptomatic patients with AR.6 Some have followed asymptomatic patients with chronic AR prospectively to identify predictors of symptom occurrence or LV dysfunction. These studies established the principle that valve replacement is not needed until symptoms appear or there is evidence of LV dysfunction.6
The main criterion of an optimal diagnostic method is the capacity for serial evaluation of the severity of AR in addition to parameters of LV function. Magnetic resonance imaging is a noninvasive technique suitable for evaluating cardiac function and valve morphology.7–9 Initially, AR was assessed by the signal void on cine MR images or by measuring the difference between left and right ventricular stroke volumes.5,10,11 With the introduction of velocity-encoded phase-difference MRI, an accurate and direct measurement of blood flow velocity became possible. The principle depends on the accrued phase shift of spins moving along a magnetic field gradient.12 Chatzimavroudis and colleagues13 showed that with increasing slice distance from the aortic valve, a decrease in measured RV occurs due to an aortic compliance effect. The advantage of the breath-hold sequence is a fast examination time of approximately 30 sec. In contrast, the non-breath-hold techniques require 3–4 min, depending on heart rate. The benefit is 13–18 phases are acquired during the cardiac cycle, compared to 6–8 with the breath-hold sequence. The main disadvantages of the time-consuming non-breath-hold sequences are moving and breathing artefacts, especially in older or severely sick patients.
We acquire 6–8 measurements during one breath-hold period, which may result in a loss of 10% of the RR-interval. We do not believe that AR will be underestimated because of this, but there is a possibility that blood flow during breath-hold MRI may differ from physiological flow during normal breathing because breath holding can change intrathoracic pressure. However, several studies and our own experience indicate that breath-hold MR flow measurements using a small lung volume by shallow inspiration can provide exact flow quantification.14 In our opinion, the advantages of the breath-hold technique predominate, and the method can provide flow quantification close to the physiological flow. It is possible to determine RV with a high degree of accuracy, and this correlates well with the degree of AR determined by angiography and echocardiography. This offers the advantage of noninvasive serial determination of RF and RV in relation to LV function parameters. For grading the severity of disease, we found RF to be the best parameter. Determination of the anatomical configuration of the aortic valve is an important aspect of preoperative diagnosis. In adults, MRI has been shown to be superior to echocardiography for characterization of valve morphology, due to the limitation of an acoustic window in some patients.15
The degree of AR can be assessed using several echocardiographic techniques. The size or extent of the regurgitant jet within the LV, effective regurgitant orifice area, RF, and RV are distinct measures of the severity of AR. The effective regurgitant orifice area may be the most hemodynamically important parameter, but it is difficult to derive in most patients. The most common approach relies on the relationship between the size of the AR jet, visualized by color-flow imaging in several planes, and RV. The length of the jet conveys unreliable information about overall severity of AR. Perry and colleagues16 found a poor correlation between jet length and maximum jet area and the angiographic severity of AR. The proximal isovelocity surface area of a regurgitant color-flow jet can be useful for estimating valve insufficiency. Proximal isovelocity surface area is based on the hemodynamic principle of flow through a small circular orifice in a flat plate. The main theoretical limitation is that the regurgitant orifice is usually neither circular nor flat, and assessment of RV shows the relevant discrepancies. The proximal isovelocity surface area was not used because of technical difficulties that limit its application for the aortic valve. Advantages of echocardiographic evaluation of the regurgitant jet are fast and easy estimation of the degree of AR, and that no special analysis software is required. Disadvantages are that it allows only semiquantitative estimation, has poor correlation with actual AR, and eccentric jets are possibly underestimated. Eccentric jets may become entrained along the LV wall, which tends to alter their appearance and hence the perception of AR severity. Changes in gain, color scale, transducer frequency, and wall filters will affect the appearance of the jet, independent of degree of AR. Magnetic resonance imaging allows determination of AR even in patients with multiple valve disease, because RV determination is independent of mitral regurgitation. Echocardiographic and angiographic evaluation of RV and RF remains difficult or impossible in patients with coexisting mitral or tricuspid regurgitation.
Another advantage of MRI is the low interobserver variability. Engels and colleagues17 found better correlation with the clinical grade of valve disease in adults, using MRI flow measurements rather than echocardiography. During the MRI scan, special attention must be paid to locating the flow plane in the same position in all patients. The distance of the slice from the aortic root has a significant impact on results. We always place our slice just above the origin of the coronary arteries to obtain the best contour delineation of the aorta. Background phase-offset errors can lead to significant inaccuracies in flow measurements, although they may not be obvious on velocity map images. Even offset errors of 1% of the velocity encoding range set could cause large errors in RV. A possible solution is the phantom correction technique, but this time-consuming process is unacceptable.14 We start the flow measurements with a high level of suspicion, and a phantom correction acquisition is only performed if discrepancies are noticed between measured and expected results. Angiography can provide valuable hemodynamic information on the degree to which AR affects the cardiac chamber, but assessment of AR may be limited when based on a single projection and requires assumptions regarding jet geometry. The severity of AR may be overestimated if during ascending aortography, the catheter descends toward the aortic valve, thereby altering its function. Moreover, angiography is not the most suitable screening method as the optimal timing of surgery is determined mostly by LV parameters and these are more difficult to assess invasively. The critical parameters (LV end-diastolic/systolic diameters, EF) can be assessed with high reproducibility using echocardiography or MRI.
Magnetic resonance imaging is ideally suited to serial evaluation of valve regurgitation and LV dysfunction, and quantification of RV and RF is now readily performed with MRI. Despite such capability, there are no standard thresholds for qualitative assessment of AR. However, MRI is costly and time consuming, and arrhythmias or atrial fibrillation can interfere with electrocardiogram-gating and impair image quality. Unlike echocardiography, MRI cannot be performed in real time and shows lower spatial and temporal resolution. In our opinion, echocardiography is still the method of first choice to identify patients with AR. Our study has shown that MRI combined with flow measurement is a very accurate, robust, applicable, and simple method to identify AR and quantify cardiac function parameters. In addition, MRI gradient-echo techniques allow a 3-dimensional impression of the extent of the regurgitant jet within the left ventricle. In young patients, a complete cardiac evaluation, including MRI and coronary angiography, can be performed in a one-stop shop. Invasive determination of the severity of AR can no longer be considered the standard procedure because the method has several limitations.
It was concluded that MRI is a completely noninvasive technique that allows exact volume determination for clinical purposes. The severity of AR can be assessed either by calculation of RF or by analysis of the regurgitant jet extension by the gradient-echo technique (Figure 1
). Comparisons between purely quantitative measurements of aortic RV by MRI and qualitative assessment of AR by angiography or echocardiography showed good correlations. This suggests that MRI has great potential in reliably measuring the severity of AR in a quantitative manner, and it may provide additional anatomical and functional information leading to a better understanding of the disease.
|
REFERENCES
|
|---|
- Croft CH, Lipscomb K, Mathis K, Firth BG, Nicod P, Tilton G, et al. Limitations of qualitative angiographic grading in aortic or mitral regurgitation. Am J Cardiol 1984;53:1593–8.[Medline]
- Kreitner KF, Voigtlander T, Wittlinger T, Dahm M, Kalden P, Meyer J, et al. Flow quantification in coronary and bypass vessels with MR phase contrast technique. Radiologe 2000;40:143–9.[Medline]
- Dodge HT, Sandler H, Ballew DW, Lord JD Jr. The use of biplane angiocardigraphy for the measurement of left ventricular volume in man. Am Heart J 1960;60:762–8.[Medline]
- Sellers RD, Levy MJ, Amplatz K, Lillehei CW. Left retrograde cardioangiography in acquired cardiac disease: technic, indications and interpretations in 700 cases. Am J Cardiol 1964;14:437–47.[Medline]
- Globits S, Frank H, Mayr H, Neuhold A, Glogar D. Quantitative assessment of aortic regurgitation by magnetic resonance imaging. Eur Heart J 1992;13:78–83.[Abstract/Free Full Text]
- Bonow RO, Rosing DR, McIntosh CL, Jones M, Maron BJ, Lan KK, et al. The natural history of asymptomatic patients with aortic regurgitation and normal left ventricular function. Circulation 1983;68:509–17.[Free Full Text]
- Chatzimavroudis GP, Oshinski JN, Franch RH, Pettigrew RI, Walker PG, Yoganathan AP. Quantification of the aortic regurgitant volume with magnetic resonance phase velocity mapping: a clinical investigation of the importance of imaging slice location. J Heart Valve Dis 1998;7:94–101.[Medline]
- Gelfand EV, Hughes S, Hauser TH, Yeon SB, Goepfert L, Kissinger KV, et al. Severity of mitral and aortic regurgitation as assessed by cardiovascular magnetic resonance: optimizing with Doppler echocardiography. J Cardiovasc Magn Reson 2006;8:503–7.[Medline]
- Rominger MB, Kluge A, Bachmann GF. Biventricular MR volumetric analysis and MR flow quantification in the ascending aorta and pulmonary trunk for quantification of valvular regurgitation. Rofo 2004;176:342–9.[Medline]
- Wagner S, Auffermann W, Buser P, Lim TH, Kircher B, Pflugfelder P, et al. Diagnostic accuracy and estimation of the severity of valvular regurgitation from the signal void on cine magnetic resonance images. Am Heart J 1989;118:760–7.[Medline]
- Sechtem U, Pflugfelder PW, Cassidy MM, White RD, Cheitlin MD, Schiller NB, et al. Mitral or aortic regurgitation: quantification of regurgitant volumes with cine MR imaging. Radiology 1988;167:425–30.[Abstract/Free Full Text]
- Walker MF, Souza SP, Dumoulin CL. Quantitative flow measurement in phase contrast MR angiography. J Comput Assist Tomogr 1988;12:304–13.[Medline]
- Chatzimavroudis GP, Walker PG, Oshinski JN, Franch RH, Pettigrew RI, Yoganathan AP. Slice location dependence of aortic regurgitation measurements with MR phase velocity mapping. Magn Reson Med 1997;37:545–51.[Medline]
- Kilner PH, Gatehouse PD, Firmin DN. Flow measurements by magnetic resonance: a unique asset worth optimising. J Cardiovasc Magn Reson 2007;9:723–8.[Medline]
- Ley S, Eichhorn J, Ley-Zaporozhan, Ulmer H, Schenk JP, Kauczor HU, et al. Evaluation of aortic regurgitation in congenital heart disease: value of MR imaging in comparison to echocardiography. Pediatr Radiol 2007;37:426–36.[Medline]
- Perry GJ, Helmcke F, Nanda NC, Byard C, Soto B. Evaluation of aortic insufficiency by Doppler color flow mapping. J Am Coll Cardiol 1987;9:952–9.[Abstract]
- Engels G, Reynen K, Muller E, Wilke N, Bachmann K. Quantitative evaluation of aortic valve insufficiency in magnetic resonance tomography. Z Kardiol 1993;82:345–51.[Medline]
TOP
ABSTRACT
INTRODUCTION
