| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J Thorac Cardiovasc Surg 2005;130:456-463
© 2005 The American Association for Thoracic Surgery
Evolving Technology |
a Department of Radiology, Stanford University, Stanford, Calif.
b Department of Mechanical Engineering, Stanford University, Stanford, Calif.
c Department of Cardiothoracic Surgery, Stanford University, Stanford, Calif.
d Division of Cardiovascular Medicine, Stanford University, Stanford, Calif.
Read at the Thirtieth Annual Meeting of The Western Thoracic Surgical Association, Maui, Hawaii, June 23–26, 2004.
Received for publication April 27, 2004; revisions received August 11, 2004; accepted for publication August 20, 2004. * Address for reprints: Robert J. Herfkens MD, Stanford University, Department of Radiology, Lucas MRI/S Center, P263, Palo Alto, CA 94304. (Email: herfkens{at}stanford.edu).
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionOnline VideosReferences |
|---|
METHODS: Time-resolved 3-dimensional magnetic resonance imaging velocity mapping was performed to analyze pulsatile blood flow by using encoded 3-directional vector fields in the thoracic aortas of 10 volunteers and 12 patients after David reimplantation using a cylindrical tube graft (T. David I) and two versions of neosinus recreation (T. David-V and T. David-V-Smod). Aortic flow was evaluated by using 3-dimensional time-resolved particle traces and velocity vector fields reformatted onto 2-dimensional planes. Semiquantitative data were derived by using a blinded grading system (0–3: 0, none; 1, minimal; 2, medium; 3, prominent) to analyze the systolic vortex formation behind the cusps, as well as retrograde and helical flow in the ascending aorta.
RESULTS: Systolic vortices were seen in both coronary sinuses of all volunteers (greater in the left sinus [2.5 ± 0.5] than the right [1.8 ± 0.8]) but in only 4 of 10 noncoronary sinuses (0.7 ± 0.9). Comparable coronary vortices were detected in all operated patients. Vorticity was minimal in the noncoronary cusp in T. David-I repairs (0.7 ± 0.7) but was prominent in T. David-V noncoronary graft pseudosinuses (1.5 ± 0.6; P = .035). Retrograde flow (P = .001) and helicity (P = .028) were found in all patients but were not distinguishable from normal values in the T. David-V-Smod patients.
CONCLUSIONS: Coronary cusp vorticity was preserved after David reimplantation, regardless of neosinus creation. Increased retrograde flow and helicity were more prominent in T. David-V patients. These novel magnetic resonance imaging methods can assess the clinical implications of altered aortic flow dynamics in patients undergoing various types of valve-sparing aortic root replacement.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionOnline VideosReferences |
|---|
Interest in ascending aortic blood flow dates back to Leonardo da Vinci and was first explored by Bellhouse and Bellhouse
7
in 1968. Although studies investigating the mechanics of reimplanted aortic valves have been reported,
8
only 1 analysis of 3-directional blood flow in the thoracic aorta and the potential consequences of surgically altered aortic geometry and function has been published.
9
Magnetic resonance imaging (MRI) provides a noninvasive method for the accurate anatomic characterization of the heart and great vessels within a 3-dimensional (3D) data set.
10,11
In comparison to other imaging modalities, such as computed tomography, the intrinsic sensitivity of MRI with respect to flow and motion offers the unique capability to acquire cardiac-gated and spatially encoded functional information simultaneously with morphologic data in a single study.
12–19
The magnetic resonance (MR) method used simultaneously assessed morphology and blood flow in vivo within a user-defined 3D volume with high spatial resolution (1.2–3 mm). Data were acquired in an electrocardiogram-gated and respiratory-compensated single scan that yielded time-resolved 3-directional pulsatile blood flow throughout the entire thoracic aorta.
20,21
The acquired data set was analyzed by using 3D data reformatting and blood flow visualization tools.
22–24
Three different visualization tools (2-dimensional [2D] vector-fields, 3D streamlines, and 3D particle traces) that use different parts of the information available within the time-resolved 3D data sets were used and focused on flow features near the aortic valve cusps and in the ascending aorta.
The objective of this study was to demonstrate that 3D MR velocity mapping is a reliable technique for the qualitative and quantitative assessment and visualization of blood flow in the ascending aorta and that this technique can portray flow features such as vortex formation in the sinuses of Valsalva or helical aortic flow in volunteers and patients who have undergone valve-sparing aortic root replacement with the T. David reimplantation method. Of particular interest was comparing patients who received different types of valve reimplantation repairs, to analyze the effects of recreating graft neosinuses.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionOnline VideosReferences |
|---|
The T. David-V-Smod (Stanford modification) procedure used 2 (large proximal and small distal) grafts. One patient (T. David-V) had a bicuspid aortic valve. These studies were performed between 1 week and 92 months (mean, 30 months) after surgery. For comparison, image acquisition and postprocessing tools were evaluated in 10 healthy volunteers (4 men and 6 women; age range, 20–40 years; mean age, 27 years) with no history of cardiovascular disease. All studies were approved by Stanford University’s Institutional Review Board, and informed consent was obtained from all subjects.
Data Acquisition
A measurement technique often referred to as phase-contrast MRI, based on the intrinsic velocity sensitivity of the phase of the MR signal to motion, was used to assess blood-flow velocities. These 3D MR velocity mapping techniques were previously described and validated.
20,21
All experiments were performed on a 1.5-T MR system (Signa CV/I; GE, Milwaukee, Wis) by using a 4-element phased array torso coil. To resolve blood flow and cardiovascular anatomy as a function of cardiac phase (cine imaging), measurements were synchronized to the cardiac cycle by using electrocardiogram gating in combination with real-time parameter updates and were subsequently interpolated to reconstruct 20 images representing different phases of the cardiac cycle. Respiration control was performed by using bellows to estimate the respiratory position for each RR interval; these data were then used to select appropriate phase-encoding amplitudes such that the final data set appeared to be collected during a single respiratory cycle (Figure 1). As a result, the respiratory motion is spread out homogenously across the data matrix. This reduced image artifacts but did create some image blurring.
|
, 15°; bandwidth, ±62.5 kHz, velocity sensitivity, 200 cm/s along all 3 spatial directions; fractional field of view, (300–340 x 225–240)mm2; slab thickness, 83.2–96 mm; matrix, 256 x 144 x 32; and spatial resolution, (1.17–1.33 x 1.56-1.67 x 2.60-3.00) mm3. Four phase encodes per time frame in combination with 3-directional velocity encoding resulted in a temporal resolution (TRes) that was determined by the time needed to collect 16 k-space lines (TRes= 16TR). TRes, 75.1-80.7 ms, varied among subjects depending on the obliquity of the imaging slab and the echo (TE, 1.84-2.06 ms) and repetition (TR, 4.79-5.04 ms) times. Because 4 phase encodes were collected for each RR interval and the total number of raw data lines is fixed, the total acquisition time is heart rate dependent and ranged from 13.8 to 22.6 minutes. A routine contrast-enhanced MR angiogram (gadopentate dimeglumine [Magnevist]; 0.3 mmol/kg; injection rate, 2-3 mL/s intravenously) was performed. To improve image quality, the 4-dimensional flow scan was performed immediately after the MR angiogram to enhance the signal-noise ratio given the contrast agent remaining in the blood pool.
A 3D time-resolved steady-state free-precession series was acquired to improve the image quality of the underlying anatomic magnitude images.
25
3D phase-contrast and 3D time-resolved steady-state free-precession data sets were acquired with identical temporal and spatial resolution to provide equal imaging volumes. The resulting time series of 3D volumes was used to replace the anatomic images in the 3D MR velocity mapping data set to improve image quality and real-time navigation in the time-resolved 3D data set. All patients had a recent transthoracic echocardiogram in which aortic regurgitation (AR) was graded (0, none; 1, minimal; 2, medium; 3, prominent).
Blood-Flow Visualization
After reconstruction and preprocessing (fully automated noise filtering and eddy current correction
26
) of the acquired data, the entire time-resolved 3D MR velocity-mapping data set was loaded into a 3D visualization software package (EnSight; Computational Engineering International, Apex, NC). Blood flow through the ascending aorta was evaluated by using 3-directional velocity vector fields reformatted onto 2D planes transecting the ascending aorta, as well as 3D streamlines and time-resolved 3D particle traces originating at the level of the aortic valve. Velocity vector fields, positioned freely within the imaging volume, were used to analyze blood flow through the aortic valve, as well as vortical flow behind the aortic valve cusps.
To analyze distal ascending aortic retrograde and helical aortic flow, 3D streamlines and particle traces were generated by using equidistant emitter points from a 2D grid of variable size that could be positioned at any spatial location within the 3D data set. In contrast to streamlines, which depict traces along the instantaneous velocity vector field in a specific time, 3D particle traces use all of the functional information and reflect blood flow throughout the entire cardiac cycle. A schematic representation of this strategy for data acquisition, postprocessing analysis, and 3D blood flow visualization is shown in Figure 1.
Data Analysis
Quantitative analysis of systolic vortex formation behind individual aortic valve cusps (2D velocity vector fields), as well as late systolic and early diastolic retrograde and helical flow in the ascending aorta (time-resolved 3D particle traces), was performed. For all healthy subjects, blood-flow vortices were visualized and evaluated with color-coded velocity vector fields mapped onto different 2D planes transecting the aortic outflow and normal to the valve plane. For each aortic valve cusp, an individual plane was carefully positioned to portray the temporal and spatial evolution of aortic flow surrounding the respective cusp, including the flow in the corresponding sinus. The orientation of the planes was defined by using a systolic image including the opened aortic valve (Figure 2). For patients who underwent the T. David-I procedure, aortic graft neosinuses are nonexistent, and the planes were positioned according to the cusp geometry.
|
All data sets were presented in a blinded fashion, and the relevant flow features were graded (0, none; 1, minimal; 2, medium; 3, prominent) by 3 experienced radiologists. The variability of average scores is indicated as ±1 standard deviation. Data were compared by using repeated-measures analysis of variance (ANOVA) with a within-subjects factor of reader and a between-subjects factor of diagnosis (normal; type of repair) to determine significant differences. The factorial ANOVA model was used even though the data were an ordinal rather than interval scale, because there is no corresponding nonparametric test that takes advantage of the powerful multireader design.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionOnline VideosReferences |
|---|
= 0.38-0.50).
Vortex Formation
Examples of visualization of systolic aortic flow and vortex formation with color-coded velocity vector fields are shown in Figure 2. The images show systolic blood flow velocities behind the 2 coronary cusps for a healthy volunteer (top) and patients who underwent T. David-I (bottom left) and T. David-V (bottom right) aortic root replacement. Vortex formation (arrows) is evident in all 3 cases, even for those with a T. David-I repair without graft neosinuses. Systolic vortical flow was visualized in all 10 volunteers in both coronary sinuses of Valsalva and was most prominent for the LC cusp (score, 2.46 ± 0.53) compared with the RC cusp (score, 1.80 ± 0.83). Vortex formation was less obvious for the NC sinus (score, 0.67 ± 0.93), where vortices could be detected in only 4 of 10 healthy subjects (score, >0.5 where 
2 observers rated the vortex formation as minimal).
Comparable systolic vortices behind the coronary cusps were detected in all operated patients. Although no neosinuses were recreated in those who underwent the T. David-I procedure, reverse flow and vortex formation behind the 2 opened coronary cusps were clearly visible in all 4 patients (score: LC, 2.50 ± 0.58; RC, 2.00 ± 0.47), whereas NC vortex formation was minimal (score: 0.67 ± 0.72). For all 3 cusps, no significant differences in vortex formation were detected between volunteers and T. David-I patients.
Medium to prominent systolic vortical flow in the 2 coronary recreated graft neosinuses was also evident for all T. David-V (score: LC, 2.40 ± 0.60; RC, 1.87 ± 0.93) and T. David-V-Smod (score: LC, 1.89 ± 0.19; RC, 1.33 ± 0.33) patients. For the combined group of patients with either a T. David-V or a T. David-V-Smod reimplantation, however, vortex formation in the recreated NC graft pseudosinus was greater (score: NC, 1.54 ± 0.62; ANOVA: F1,16 = 5.33; P = .035) than in the volunteers. Comparison with T. David-I patients revealed similar vorticity in the 2 coronary sinuses, whereas NC vortex formation was amplified (score: NC, 0.67 ± 0.72; ANOVA: F1,10 = 4.85), but this was only possibly statistically significant (P = .052). Results regarding the quantitative vorticity analysis are summarized in Figure 3.
|
The analysis of helicity and retrograde flow in the distal ascending aorta is summarized in Tables 1 and 2. For all operated patients, a significant increase in helical flow (score: 2.58 ± 0.41; ANOVA: F1,20 = 5.59; P = .028) was observed compared with volunteers. An example of prominent helical aortic flow for a patient after T. David-V reimplantation is shown in Figure 4. Comparing individual surgical techniques with healthy subjects, helical flow was most prominent for the T. David-V patients (score: 2.80 ± 0.32; ANOVA: F1,13 = 1.20; P = .040). Helical flow in patients with a T. David-V-Smod (score: 2.22 ± 0.19) or a T. David-I (score: 2.58 ± 0.32) procedure, however, was not statistically different from that in volunteers (score= 2.03 ± 0.67).
|
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionOnline VideosReferences |
|---|
Results from the healthy volunteers indicated that systolic vortex formation was most prominent in the 2 coronary sinuses of Valsalva, whereas blood flow behind the NC cusp was seen only in 4 of 10 subjects. These observations differ from those recently reported by Kvitting and colleagues,
9
who detected vorticity in all 3 sinuses of 6 healthy subjects. Further, we are puzzled that the Linkoping group
9
saw more vorticity in the NC and RC sinuses of volunteers, whereas we observed more vorticity in the RC and LC cusps.
Valve-sparing aortic root replacement (David reimplantation) does not significantly affect the degree of vorticity behind the coronary aortic valve cusps, whether or not graft neosinuses were created. The patients who underwent T. David-V and T. David-V-Smod procedures exhibited readily detectable systolic vortex formation (with significantly more vortex formation in the NC sinus). Even patients who received the T. David-I procedure demonstrated substantial reverse flow and vorticity behind the opened coronary valve cusps. Therefore, it can be postulated that systolic vortex formation is not predominantly a consequence of the mechanical and geometrical properties of the sinuses of Valsalva but is more related to aortic blood flow and associated hemodynamics (eg, pressure differences across the aortic cusps). The recent MRI report from Linkoping
9
did not observe any vorticity in 2 T. David-I patients, whereas such was seen in all 4 of our T. David-I patients; this disparity may be explained in part by differences in the acquisition voxel volume (16 mm3 in their images vs 4.7-6.7 mm3), anisotropic voxel orientation, or other differences in image-acquisition methods. Further, the Linkoping group studied only streamline visualizations, whereas we used all 3 techniques (2D velocity vector fields, 3D streamlines, and 3D particle traces), which yielded similar findings with respect to vorticity. Although vortex formation was enhanced by the surgical creation of graft neosinuses in the T. David-V patients, normal vorticity was preserved even without sinus creation in the graft. In operated patients, considerably enhanced helical flow patterns could be identified with 3D particle traces in the distal ascending aorta, particularly in the T. David-V subgroup. Further studies are needed to determine whether such changes perturb aortic wall stress and possibly lead to deleterious long-term effects, such as dilatation of the aortic arch or risk of dissection. Our data indicate that the recent T. David-V-Smod modification of the T. David-V technique reduces helical aortic flow; hence, this operative technique both augments neosinus creation and reduces abnormal changes in distal aortic blood-flow characteristics. The small number of patients with a T. David-V-Smod repair analyzed in this study, however, makes this conclusion conjecture; future studies of more patients are needed.
Valve-sparing aortic root replacement significantly increases late systolic retrograde flow in the distal ascending aorta, and this was more prominent after a T. David-I procedure. Because the patients had no or only mild AR, no correlation between AR and retrograde flow could be established. This retrograde flow may represent energy loss.
Study limitations include the relatively small number of patients and the semiquantitative nature of the data analysis. Because MR velocity-mapping data permits absolute measurements of blood flow velocities, future work should include more quantitative analysis of data from larger patient populations in which vortex formation and helical and retrograde ascending aortic flow are characterized by numeric parameters directly derived from the velocity data. In addition, the temporal and spatial distribution of the velocities could be used to calculate impact forces on the aortic wall and the individual valve cusps themselves, which would help to assess the differences between the various types of David reimplantation procedures and how these might affect long-term valve durability.
The results of the statistical analysis in this study have to be interpreted with caution. Although comparison between all patients (or at least between larger subgroups, such as the combined T. David-V and T. David-V-Smod group) and healthy subjects provided reasonably large numbers, the results associated with smaller sample sizes may be less reliable. Our findings with respect to increased helicity, retrograde flow, and NC cusp vorticity are reliable, but less so for the comparisons between the smaller subgroups.
With respect to data acquisition, future work should be focused on reduction of the total acquisition time and/or improvements in temporal and spatial resolution by using parallel imaging techniques.
27
In addition, optimized respiratory control would improve image quality and reduce blurring, eg, degrading of spatial resolution in the 3D images.
In conclusion, the novel MRI methods presented here should prove to be useful to assess the clinical implications of increased vorticity and altered ascending aortic flow dynamics in patients undergoing various types of valve-sparing aortic root replacements.
|
Online Videos |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionOnline VideosReferences |
|---|
|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionOnline VideosReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  L. G. Svensson, M. Cooper, L. H. Batizy, and E. R. Nowicki Simplified David Reimplantation With Reduction of Anular Size and Creation of Artificial Sinuses Ann. Thorac. Surg., May 1, 2010; 89(5): 1443 - 1447. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Frydrychowicz, A. Berger, A. F. Stalder, and M. Markl Preliminary results by flow-sensitive magnetic resonance imaging after Tiron David I procedure with an anatomically shaped ascending aortic graft Interact CardioVasc Thorac Surg, August 1, 2009; 9(2): 155 - 158. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Katayama, N. Umetani, S. Sugiura, and T. Hisada The sinus of Valsalva relieves abnormal stress on aortic valve leaflets by facilitating smooth closure. J. Thorac. Cardiovasc. Surg., December 1, 2008; 136(6): 1528 - 1535. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Weigang, F. A. Kari, F. Beyersdorf, M. Luehr, C. D. Etz, A. Frydrychowicz, A. Harloff, and M. Markl Flow-sensitive four-dimensional magnetic resonance imaging: flow patterns in ascending aortic aneurysms Eur J Cardiothorac Surg, July 1, 2008; 34(1): 11 - 16. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Markl, C. Benk, D. Klausmann, A. F. Stalder, A. Frydrychowicz, J. Hennig, and F. Beyersdorf Three-dimensional magnetic resonance flow analysis in a ventricular assist device. J. Thorac. Cardiovasc. Surg., December 1, 2007; 134(6): 1471 - 1476. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. E. Muhs, H. J. M. Verhagen, M. G. Huddle, V. M. Pai, E. M. Hecht, and A. Dardik Theory, Technique, and Practice of Magnetic Resonance Angiography Vascular, December 1, 2007; 15(6): 376 - 383. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Grosse-Wortmann, A. Al-Otay, H. Woo Goo, C. K. Macgowan, J. G. Coles, L. N. Benson, A. N. Redington, and S.-J. Yoo Anatomical and Functional Evaluation of Pulmonary Veins in Children by Magnetic Resonance Imaging J. Am. Coll. Cardiol., March 6, 2007; 49(9): 993 - 1002. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Wetzel, S. Meckel, A. Frydrychowicz, L. Bonati, E.-W. Radue, K. Scheffler, J. Hennig, and M. Markl In Vivo Assessment and Visualization of Intracranial Arterial Hemodynamics with Flow-Sensitized 4D MR Imaging at 3T AJNR Am. J. Neuroradiol., March 1, 2007; 28(3): 433 - 438. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Bakhtiary, M. Schiemann, O. Dzemali, S. Dogan, V. Schachinger, H. Ackermann, A. Moritz, and P. Kleine Impact of Patient-Prosthesis Mismatch and Aortic Valve Design on Coronary Flow Reserve After Aortic Valve Replacement J. Am. Coll. Cardiol., February 20, 2007; 49(7): 790 - 796. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. C. Miller Valve-Sparing Aortic Root Replacement: Current State of the Art and Where Are We Headed? Ann. Thorac. Surg., February 1, 2007; 83(2): S736 - S739. [Full Text] [PDF]
|
|
|
|
|
|
T. N. Fitzgerald, A. Muto, F. A. Kudo, J. M. Pimiento, R. T. Constable, and A. Dardik Emerging Vascular Applications of Magnetic Resonance Imaging: A Picture is Worth More than a Thousand Words Vascular, November 1, 2006; 14(6): 366 - 371. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Ranga, O. Bouchot, R. Mongrain, P. Ugolini, and R. Cartier Computational simulations of the aortic valve validated by imaging data: evaluation of valve-sparing techniques Interact CardioVasc Thorac Surg, August 1, 2006; 5(4): 373 - 378. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Bakhtiary, M. Schiemann, O. Dzemali, T. Wittlinger, M. Doss, H. Ackermann, A. Moritz, and P. Kleine Stentless bioprostheses improve postoperative coronary flow more than stented prostheses after valve replacement for aortic stenosis J. Thorac. Cardiovasc. Surg., April 1, 2006; 131(4): 883 - 888. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
