Before and after Endovascular Aortic Repair in the Same Patients with Aortic Dissection: A Cohort Study of Four-Dimensional Phase-Contrast Magnetic Resonance Imaging

(1) Background: We used four-dimensional phase-contrast magnetic resonance imaging (4D PC-MRI) to evaluate the impact of an endovascular aortic repair (TEVAR) on aortic dissection. (2) Methods: A total of 10 patients received 4D PC-MRI on a 1.5-T MR both before and after TEVAR. (3) Results: The aortas were repaired with either a GORE TAG Stent (Gore Medical; n = 7) or Zenith Dissection Endovascular Stent (Cook Medical; n = 3). TEVAR increased the forward flow volume of the true lumen (TL) (at the abdominal aorta, p = 0.047). TEVAR also reduced the regurgitant fraction in the TL at the descending aorta but increased it in the false lumen (FL). After TEVAR, the stroke distance increased in the TL (at descending and abdominal aorta, p = 0.018 and 0.015), indicating more effective blood transport per heartbeat. Post-stenting quantitative flow revealed that the reductions in stroke volume, backward flow volume, and absolute stroke volume were greater when covered stents were used than when bare stents were used in the FL of the descending aorta. Bare stents had a higher backward flow volume than covered stents did. (4) Conclusions: TEVAR increased the stroke volume in the TL and increased the regurgitant fraction in the FL in patients with aortic dissection.


Introduction
Intramural hematoma, perforated aortic ulcers, and type A and type B aortic dissection (AD) have been described as acute aortic syndromes [1][2][3][4]. Patients with medically treated AD remain at significant risk for late adverse events. A recent study recognized that the increased aortic diameter, increased false lumen extent, and forming thrombosis within false lumen were strongly associated with late adverse events [5]. Thoracic endovascular aortic repair (TEVAR) has been used to reduce the growth of the dissecting aortic aneurysms in acute aortic syndrome. However, the effect of the TEVAR impact on hemodynamics is patient lying in a supine position. Our team performed anatomical scanning of blood vessels around the aortic dissection areas; three planes were scanned separately, and T2 turbo spin echo scanning was carried out with the following parameters: single-shot mode; time repetition (TR), shortest; echo time (TE), shortest; voxel size, 0.6 × 0.84 × 4 mm 3 ; the number of signals averaged (NSA), 1; scan duration, 1 min. Balanced turbo field echo scanning was also performed with identical settings, except the voxel size was instead 1.84 × 1.87 × 8 mm 3 . The axial area included the arch to the abdominal bifurcation level, the coronal area comprised the heart and aorta, and the oblique sagittal field included all aorta and parallel aortic arch. The two-dimensional images helped to understand the type and scope of aortic dissection and were the basis for subsequent 4D PC-MRI with the following parameters: three-dimensional turbo field echo (TFE); TR, shortest; TE, shortest; flip angle, 5 • ; voxel size, 2.25 × 2.25 × 3 mm 3 ; phase-contrast velocity, 120 cm/s; scan duration, 6.02 min. Imaging sections had to include the aortic arch and descending aorta. After scanning, the 4D images were used to determine the anatomical space occupied by the artery. Quantitative flow (QFlow) scanning was then performed on a plane perpendicular to the blood flow with the following parameters: scan technique, TFE PC; TR, shortest; TE, shortest; flip angle, 12 • ; slice thickness, 8 mm; field of view, 248 × 300; phase-contrast velocity, 200 cm/s; scan duration, 13 s while patients held their breath. Those parameters were used and images were captured without using a gadolinium-based contrast agent. We performed QFlow analysis by drawing the region of interest (ROI) on the false lumens and true lumens at the following vascular segments: the aortic root, aortic arch, descending aorta, abdominal aorta at the level of the diaphragm, and abdominal aorta between the level of the celiac trunk and superior mesenteric artery (SMA) (Figure 1). We set the flow direction from the heart to the legs as forwarding/positive flow. On the contrary, the flow direction from the legs to the heart was set as backward/negative flow. tient lying in a supine position. Our team performed anatomical scanning of blo around the aortic dissection areas; three planes were scanned separately, and spin echo scanning was carried out with the following parameters: single-shot m repetition (TR), shortest; echo time (TE), shortest; voxel size, 0.6 × 0.84 × 4 mm 3 ; th of signals averaged (NSA), 1; scan duration, 1 min. Balanced turbo field echo was also performed with identical settings, except the voxel size was instead 1.8 8 mm 3 . The axial area included the arch to the abdominal bifurcation level, th area comprised the heart and aorta, and the oblique sagittal field included all parallel aortic arch. The two-dimensional images helped to understand the type of aortic dissection and were the basis for subsequent 4D PC-MRI with the follo rameters: three-dimensional turbo field echo (TFE); TR, shortest; TE, shortest; 5°; voxel size, 2.25 × 2.25 × 3 mm 3 ; phase-contrast velocity, 120 cm/s; scan dura min. Imaging sections had to include the aortic arch and descending aorta. After the 4D images were used to determine the anatomical space occupied by the arte titative flow (QFlow) scanning was then performed on a plane perpendicular to flow with the following parameters: scan technique, TFE PC; TR, shortest; TE flip angle, 12°; slice thickness, 8 mm; field of view, 248 × 300; phase-contrast ve cm/s; scan duration, 13 s while patients held their breath. Those parameters w and images were captured without using a gadolinium-based contrast agent formed QFlow analysis by drawing the region of interest (ROI) on the false lu true lumens at the following vascular segments: the aortic root, aortic arch, d aorta, abdominal aorta at the level of the diaphragm, and abdominal aorta be level of the celiac trunk and superior mesenteric artery (SMA) (Figure 1). We se direction from the heart to the legs as forwarding/positive flow. On the contrary direction from the legs to the heart was set as backward/negative flow. Figure 1. Illustration of QFlow scanning and drawing the region of interest (ROI). The Q ning is performed at four levels to obtain two-dimensional images (perpendicular to bloo aortic curve). By drawing ROI on the vascular lumens (completely covering the true lume lumen), eight hemodynamic variables can be obtained for each ROI for the subsequen analysis. The flow direction to the head was set as positive flow. . The QFlow scanning is performed at four levels to obtain two-dimensional images (perpendicular to blood flow and aortic curve). By drawing ROI on the vascular lumens (completely covering the true lumen and false lumen), eight hemodynamic variables can be obtained for each ROI for the subsequent statistical analysis. The flow direction to the head was set as positive flow.
By drawing the ROI completely covering the vascular lumen, the computer could automatically generate analysis results of various variables. These variables include stroke volume (SV), forward flow volume (FFV), backward flow volume (BFV), regurgitant fraction (RF), absolute stroke volume (ASV), mean flux (MF, stroke distance (SD), and mean velocity (MV). All of the eight QFlow variables are shown as follows: 1.
Stroke volume, mL; The net volume of blood that passes through the contour of ROI during one cardiac cycle.

2.
Forward flow volume, mL; The volume of blood that passes through the contour of ROI in the positive direction (toward head direction) during one cardiac cycle.

3.
Backward flow volume, mL; The volume of blood that passes through the contour of ROI in the negative direction (toward foot direction) during one cardiac cycle.

4.
Regurgitant fraction, %; The fraction of the minor flow to the main flow that passes through the contour of ROI, automatically defined by the computer. 5.
Absolute stroke volume, mL; The absolute value of forwarding flow volume plus the absolute value of backward flow volume. 6.
Stroke distance, cm; The net distance that blood proceeds in the vessel during one cardiac cycle. 8.

Statistical Analysis
Continuous variables (age and QFlow measurements) were analyzed using an unpaired two-tailed Student's t test or one-way analysis of variance test, and discrete variables (sex, substance usage, comorbidities, and intervention history) were compared using a two-tailed Fisher's exact test. All statistical analyses were conducted using Data Analysis version 8.0 (Stata Corporation, College Station, TX, USA).

Results
Between April 2017 and July 2021, we enrolled 51 patients (all men; age: 39-56 years) whose aortic pathologies had been evaluated through 4D PC-MRI at a tertiary hospital. Among them, 10 underwent 4D PC-MRI before and after TEVAR. The time between the symptom onset of aortic dissection to the first MRI ranged from 7 days to 10 months. The 10 patients accepted endovascular aortic repair within three days after the first MRI and then arranged a second MRI for postoperative follow-up. The average time between the two MRIs was 215 days (range, 106-298 days). Regarding the patients' age, sex, comorbidities, aortic disease, TEVAR indication, previous relevant surgeries, stent type, and time between aortic dissection onset and intervention are listed in Table 1. Almost all of the patients were hypertensive; one had Guillain-Barré syndrome, two had polycystic kidney disease, and two had chronic renal insufficiency. Seven patients (Patients 1-7) received TEVAR for chronic dissecting aortic aneurysm with a graft stent (GORE TAG; W.L. Gore &Associates, Inc., Flagstaff, AZ, USA), and the other three (Patients 8-10) received a Zenith Dissection Endovascular Stent (Cook Medical LCC, Bloomington, IN, USA) for malperfusion syndrome after open repair of acute type A aortic dissection. One patient received superior mesentery artery revascularization with a Gore-covered stent, and one received carotid-carotid artery bypass to facilitate coverage of zone 1 in the aortic arch. All patients recovered uneventfully from TEVAR and then underwent postoperative 4D PC-MRI. Quantitative hemodynamic analysis was performed on all 10 patients before and after TEVAR. Table 2 demonstrates the QFlow measurements of the same 10 participants with aortic dissection before and after TEVAR. Figure 2 illustrates the stroke volume (SV), forward flow volume (FFV), backward flow volume (BFV), and a regurgitant fraction (RF) in the true and false lumens of aortic dissection before and after TEVAR. After TEVAR, the true lumen had higher SV than before TEVAR from the arch to the abdominal aorta. However, the SV of the false lumen decreased after TEVAR, mainly in the descending aorta. The increasing SV of the true lumen is primarily attributable to BFV augmentation in the descending and abdominal aorta. By contrast, FFV increased only in the aortic arch. After TEVAR, RF, which indicates a nonlaminar flow pattern, was higher in the false lumen and lower in the true lumen, mainly in the descending aorta, indicating that the true lumen had predominantly laminar flow after TEVAR. The nonlaminar flow was higher in the false lumen in the aortic arch after TEVAR.   Figure 3 displays the absolute SV, mean flux, SD, and mean velocity in the true and false lumens of aortic dissection before and after TEVAR. The mean flux exhibited a similar trend to that of the SV in both lumens. After TEVAR, the absolute SD increased in the true lumen, whereas the SD was nearly zero in the false lumen. The mean velocity was similar in both lumens after TEVAR. In conclusion, TEVAR increased the forward flow volume of the true lumen (TL). The SV of the false lumen primarily affected the descending aorta. TEVAR decreased the nonlaminar flow in the true lumen in the descending aorta but increased the RF in the false lumen, and the mean flux increased in the true lumen and decreased in the false lumen of the descending aorta. After TEVAR, the SD increased in the true lumen.  Figure 3 displays the absolute SV, mean flux, SD, and mean velocity in the true and false lumens of aortic dissection before and after TEVAR. The mean flux exhibited a similar trend to that of the SV in both lumens. After TEVAR, the absolute SD increased in the true lumen, whereas the SD was nearly zero in the false lumen. The mean velocity was similar in both lumens after TEVAR. In conclusion, TEVAR increased the forward flow volume of the true lumen (TL). The SV of the false lumen primarily affected the descending aorta. TEVAR decreased the nonlaminar flow in the true lumen in the descending aorta but increased the RF in the false lumen, and the mean flux increased in the true lumen and decreased in the false lumen of the descending aorta. After TEVAR, the SD increased in the true lumen.  Post-stenting quantitative flow analysis was performed to evaluate the impact on bare and covered stents ( Table 3, Figures 4 and 5). Covered stents (GORE TAG) caused greater reductions in the SV, backflow volume, and absolute SV than did bare stents in the false lumen of the descending aorta ( Figures 4A,C and 5A). Notably, bare stents led to higher backward flow than did the covered stents after TEVAR ( Figure 4C). The decrease in mean flux and mean velocity in the false lumen was similar between the covered and bare stents ( Figure 5B,D). The SD in the abdominal aorta was higher when covered stents were used than when bare stents were used ( Figure 5C). These findings are similar to the results of 4D flow visualizations (Supplementary Video S1).  Post-stenting quantitative flow analysis was performed to evaluate the impact on bare and covered stents ( Table 3, Figures 4 and 5). Covered stents (GORE TAG) caused greater reductions in the SV, backflow volume, and absolute SV than did bare stents in the false lumen of the descending aorta Figures 4A,C and 5A). Notably, bare stents led to higher backward flow than did the covered stents after TEVAR ( Figure 4C). The decrease in mean flux and mean velocity in the false lumen was similar between the covered and bare stents ( Figure 5B,D). The SD in the abdominal aorta was higher when covered stents were used than when bare stents were used ( Figure 5C). These findings are similar to the results of 4D flow visualizations (Supplementary Video S1).

Discussion
In this study, we observed the immediate hemodynamic impact upon the thoracic endovascular aortic repair by 4D phase-contrast MRI through the following parameters estimating true and false lumen of aortic dissection: stroke volume (SV), forward flow volume (FFV), backward flow volume (BFV), and regurgitant fraction (RF). To reduce interindividual variation, we compared the data in the identical patients before and after TEVAR (Figures 2 and 3). The SV was higher in the true lumen of patients with graft stents than in those with aortic dissection without intervention, and the RF, an indicator of nonlaminar flow, was higher in the false lumen than in the true lumen. Thus, TEVAR increased the forward flow volume of the true lumen (TL). The endovascular aortic stent reduced the nonlaminar flow in the true lumen. We also observed the increase in the regurgitant fraction in the false lumen after TEVAR; this result is similar to prior reports [27,28]. The mean flux increased in the true lumen and decreased in the false lumen of the descending aorta. After TEVAR, the SD increased in the true lumen, indicating more effective blood transport per heartbeat.
CT scanners with additional techniques include dual-energy CT and ECG gating manners improved the quality of obtained CTA aortic images [29][30][31][32]. These advances in the CTA dominated the surgical planning for TEVAR but also the post-TEVAR evaluation. However, in patients with impaired renal function or unstable renal flow due to malperfusion syndrome, contrast media may cause acute renal failure [33]. CTA also causes radiation exposure, and substantial accumulation of this radiation can occur, even in young patients [34][35][36][37]. Contrast-enhanced MRI demonstrated blood vessel pathology well with the administration of gadolinium-based contrast agents (GBCA), which shortens blood longitudinal relaxation (T1). This approach provides images with a high signal-to-noise ratio and high spatial resolution by two modes: single-phase and time-resolved MRA [38]. Single-phase MRA captures vascular images at a single point in time. Time-resolved MRA consists of acquiring multiple images of the volume following contrast injection. Blood flow is used as the intrinsic contrast agent, and the signal is based on an inflow effect. The vessels can be observed most clearly when they are orthogonal to the two-dimensional plane because in-plane vessels sometimes experience signal loss [36,37].
The new technique of 4D PC-MRI can, in a single scan, acquire flow information of the entire aortic volume over time [39]. In 4D PC-MRI or 4D flow MRI, the phase contrast, which encodes flow information in all three spatial directions within a large volumetric field of view, is acquired. Many hemodynamic parameters can be derived from these 4D flow data sets, including wall shear stress, pulse wave velocity, blood flow patterns with streamlines, and pressure differences. Pioneering laboratory research has demonstrated that 3.0-T 4D PC-MRI can be used to evaluate aortic dissection, with a focus on aneurysmal change [40,41]. The 4D PC MRI was then compared with the conventional CTA, with similar interexamination, interobserver, and intraobserver variability of these segmentations [42,43]. Recent 4D PC MRI studies have focused on false lumen pressure and the predicted growth in chronic type B aortic dissection [44,45]. They proposed false lumen flow fraction and maximum systolic flow deceleration rate inking to growth for dissection aortic aneurysm [46]. Researchers who conducted those studies did not identify significant limitations in reproducibility or repeatability that may affect measurements derived from 4D flow manners, which is consistent with our previous experience. We first applied 4D PC-MRI in a clinical setting; thus, 4D PC-MRI could provide similar information to that provided by CTA after open surgery for type A aortic dissections [17,46]. Moreover, 4D PC-MRI is also a reasonable imaging option for young patients and patients with poor renal function. However, the choice of stent affects further 4D PC-MRI evaluation. Imaging artifacts with 4D PC-MRI were minimal when nitinol-based endografts were used (GORE TAG and Cook Zenith Dissection Stents) [17]. Stainless steel endoprostheses should not be chosen if 4D PC-MRI is used as a follow-up modality; no such stent graft was used in the current study.
This study has some technical issues to be discussed. First, it revealed that stroke distance is more effective than stroke volume to reflect the hemodynamic difference after TEVAR. We hypothesize that this is because that stroke volume is more affected by the size of the vascular lumen. According to the algorithm, stroke distance is the net distance blood proceeds in the vessel during one cardiac cycle. Stroke volume is the net volume of blood that passes through the contour of ROI during one cardiac cycle. We observed that the vascular lumens (including both false and true lumens) were variable at a different vascular segment. This variability of luminal size at different vascular segments may affect the predictive power of stroke volume. Second, QFlow analysis revealed that regurgitation fractions in the true lumens are consistently small. However, the backward flow volume is large, and the forward flow volume is small in the true lumens of the descending and abdominal aorta. The regurgitant fraction was automatically calculated as the fraction of the minor flow (usually the flow toward the heart) to the main flow (usually the flow away from the heart) that passes through the contour of ROI of the two-dimensional QFlow scanning. The backward flow (negative direction, toward foot) is the main flow characteristic of true lumens at the descending and abdominal aorta. Thus, the regurgitation fractions are still small. Third, the stroke distance and mean velocity can be negative because that stroke distance and mean velocity reflect the "distance" (the flow direction to the head was set as positive flow) that blood proceeds in the vessels. On the contrary, absolute stroke volume and mean flux are positive because the absolute stroke volume was the absolute value of forwarding flow volume plus the absolute value of backward flow volume, and mean flux reflects the stroke amount.
We used bare stents only in patients with malperfusion syndrome after open repair of type A aortic dissection without a proximal covered stent on the secured proximal landing zone. The SD and backflow volume, although still being observed, were lower when bare stents were used than when covered stents were used ( Figures 4C and 5 C). Future studies should assess these hemodynamic parameters to explore their application in clinical practice, including prognostic prediction.
The cost of 4D PC MRI may be a concern in merging this diagnostic tool into daily clinical practice. No contrast medium is required for 4D PC-MRI; thus, it would cost little for our national health care system (<USD250 per examination). With the maturation of the radiologic team, this approach is less time consuming (processing time: 30 min), which enables its application for clinical practice.
Our MRI protocol performed QFlow scanning (perpendicular to blood flow and aortic curve) to obtain two-dimensional images, which contained phase-shifting information. By drawing ROI on the vascular lumens (completely covering the true lumen and false lumen), it can obtain hemodynamic variables for statistical analysis. We set the flow direction to the head as positive flow. On the contrary, the flow direction to the foot was set as negative flow. Thus, our result revealed that the net blood volume (stroke volume, SV) in the aortic root and aortic arch was mainly contributed by the forward flow volume (FFV; toward the head direction). On the other hand, the net blood volume (stroke volume) in descending aorta and abdominal aorta was mainly contributed by the backward flow volume (BFV; toward the foot direction) (Figure 2A-C and Figure 4A-C). This result is reasonable according to this study design and MRI protocol.

Study Limitations
In this study, we verified the clinical value of applying 4D PC-MRI to characterize aortic pathology. However, this study had some limitations. First, The QFlow measurements presented a large standard deviation, and most of the p-values are larger than 0.05, indicating no significant difference between groups. Second, this was a nonrandomized study with only a few patients. Further larger-scale randomized studies should be conducted. Third, although quantitative analysis can yield useful information for determining the optimal therapeutic strategy for complex aortic diseases, further studies on quantitative analysis and streamline computation are required, especially to evaluate the endoleak model and explore its other clinical applications.

Conclusions
As an approach that does not require the use of radiation or contrast media, 4D PC-MRI is a promising alternative modality for imaging aortic dissection. Moreover, this approach may be especially useful for aortic dissection diagnosis and treatment, especially in patients with malperfusion syndrome of visceral vessels, young patients, and patients with impaired renal function. TEVAR increased the SV in the true lumen and increased the RF in the false lumen in the patients enrolled in this study. Whether bare or covered stents are used can influence hemodynamic parameters in 4D PC-MRI.