Fig 1. Calculation of angulation. A, Three-dimensional reconstruction image. B, Illustration of vector inner product

Transcription

1 The impact of bird-beak configuration on aortic remodeling of distal arch pathology after thoracic endovascular aortic repair with the Zenith Pro-Form TX2 thoracic endograft Hung-Lung Hsu, MD, a Chun-Ku Chen, MD, b Po-Lin Chen, MD, a I-Min Chen, MD, a,c Chiao-Po Hsu, MD, PhD, a,c Chih-Wen Chen, MD, a,c and Chun-Che Shih, MD, PhD, a,c Taipei, Taiwan Objective: Structural changes and incomplete endograft apposition to the aortic arch (bird-beak configuration) after thoracic endovascular aortic repair are poorly understood. The aim of this study was to analyze the morphologic changes, conformability, and angulation factors in patients who underwent stainless steel-based stent graft repair of thoracic aortic pathology. Methods: From March 2011 to March 2012, the study enrolled 19 patients with aortic pathology requiring proximal fixation in zones 2 and 3 who underwent stent graft repair using Zenith Pro-Form TX2 stent grafts (Cook Medical, Bloomington, Ind). For comparison, another 19 patients who received Zenith Z-Trak stent grafts were selected from December 2009 to February Chest computed tomography scans were analyzed at baseline, and then at 1, 6, and 12 months postoperatively. Arch angulation and bird-beak configuration were evaluated according to sealing zones of attachment by Aquarius intuition software (TeraRecon, San Mateo, Calif). Results: The treated diseases included chronic type B aortic dissection in 17 patients and degenerative aneurysms in 21. Significant arch angle transformation was noted at the zone 2 level between the Pro-Form and Z-Trak treated groups ( vs ; P [.033) and left subclavian artery level ( vs ; P [.031) during 1 year of follow-up. The bird-beak configuration was detected in six patients (32%) in the Pro-Form group and in 11 (58%) in the Z-Trak group (P [.096) at 1 month, and in six (32%) in the Pro-Form group and in 14 (74%) in the Z-Trak group (P [.022) at 12 months. The mean bird-beak angle was significantly less in Pro-Form-treated patients at 1 month (5 6 9 vs ; P [.019) and at 1 year ( vs ; P [.033). In the Pro-Form platform, a preoperative zone 2 angle <151.1 was a better estimation of the presence of a postoperative bird-beak configuration, with a sensitivity of 86% and specificity of 83%. Conclusions: Aortic remodeling after stainless steel stent graft repair of thoracic aortic pathology is a continuous process. Significant arch angle transformation was discovered over the zone 2 and left subclavian artery levels. TX2 Pro-Form stent grafts improved arch conformation after 1 year of follow-up. Furthermore, in the patients with dissection, a preoperative distal arch angle of zone 2 was predictive of postoperative bird-beak configuration, regardless of whether they were treated with a Pro-Form stent graft. (J Vasc Surg 2014;59:80-8.) Thoracic endovascular aortic repair (TEVAR) is a safe and efficacious option in the treatment of thoracic aortic diseases. 1 Successful TEVAR requires adequate sealing and fixation to avoid stent graft migration, collapse, or From the Division of Cardiovascular Surgery, Department of Surgery a and Department of Radiology, b Taipei Veterans General Hospital; and the Institutes of Clinical Medicine, National Yang-Ming University School of Medicine. c This work was supported by grants NSC B MY3 and B-065-MY3 from the National Science Council, Taiwan, and V100C-018 and V101-C-162 from Taipei Veterans General Hospital, Taiwan. This work was also assisted in part by the Division of Experimental Surgery of the Department of Surgery, Taipei Veterans General Hospital, Taipei, Taiwan. Author conflict of interest: none. Reprint requests: Dr Chun-Che Shih, Division of Cardiovascular Surgery, Taipei Veterans General Hospital, #201, Section 2, Shih-Pai Rd, Taipei 112, Taiwan, ROC ( ccshih@vghtpe.gov.tw). The editors and reviewers of this article have no relevant financial relationships to disclose per the JVS policy that requires reviewers to decline review of any manuscript for which they may have a conflict of interest /$36.00 Copyright Ó 2014 by the Society for Vascular Surgery. endoleak formation. Anatomic complexities of the aortic arch are important factors in the failure of stent graft treatment. A highly angulated or curved arch prevents the stent graft from being able to conform appropriately. The lack of apposition to the aortic wall along the lesser curve results in the bird-beak configuration, a wedge-shaped gap between the stent graft and the aortic wall. Studies have shown this incomplete endograft apposition is related to a higher risk of endoleak formation, 2 especially in zone 2 or 3 aortic pathology. The aim of this study was to assess remodeling of aortic arch angulation and morphologic portending factors of bird-beak configuration on distal aortic pathology after TEVAR with Zenith TX2 Pro-Form stent grafts (Cook Medical, Bloomington, Ind), which are designed specifically to address the common problem of proximal apposition of thoracic stent grafts. METHODS From March 2011 to March 2012, 19 patients with aortic pathology requiring proximal fixation in zones 2 and 3, who underwent successful TEVAR with Zenith Pro-Form TX2 thoracic stent grafts (Cook Medical, 80

2 Volume 59, Number 1 Hsu et al 81 Fig 1. Calculation of angulation. A, Three-dimensional reconstruction image. B, Illustration of vector inner product ð ȧ $ ḃ ¼jȧ jj ḃ j cos qþ. Fig 2. Calculation of bird-beak angulation. A, Straight multiplanar reconstruction image. B, The cross-section view shows that the tip of the graft and that the distance (D) from the tip of the stent graft to the aortic wall is 7.55 mm. C, The projecting length (L) of the bird-beak configuration was measured along the aortic wall until the stent graft attached to the wall without a gap. D, Illustration of tan F ¼ D/L. Bloomington, Ind), were included and monitored for at least 12 months. For comparison, another 19 patients with aortic pathology requiring proximal fixation in zones 2 and 3 who received Zenith Z-Trak stent grafts were selected from December 2009 to February The patients were categorized into two treatment groups: Pro-Form and Z-Trak. The Pro-Form consisted of 10 patients with chronic type B aortic dissection and nine with thoracic aortic aneurysm, and in the Z-Trak group, there were seven and 12 patients, respectively. Computed tomography scans were performed at baseline and at 1, 6, and 12 months after endografting with a standard protocol of 3.0-mm cuts. The Digital Imaging and Communications in Medicine (DICOM) data were used for analysis, and aortic arch angulation and birdbeak configuration were determined by computer-assisted measurements. Aortic arch angulation. Three-dimensional reconstruction software, Aquarius intuition 4.4 (TeraRecon, San Mateo, Calif), was used to measure the aortic arch

3 82 Hsu et al January 2014 Table I. Patient demographics and aortic pathology Variable a All (N ¼ 38) Pro-Form (n ¼ 19) Z-Trak (n ¼ 19) P Age, years Age $80 years 11 (29) 5 (26) 6 (37).721 Female 5 (13) 1 (5) 4 (21).150 Coronary artery disease b 6 (16) 3 (16) 3 (16) >.99 Hypertension 23 (61) 11 (58) 12 (63).584 Smoking 16 (42) 10 (53) 6 (32).189 COPD 2 (5) 1 (5) 1 (5) >.99 Renal insufficiency c 11 (29) 7 (37) 4 (21).283 Stroke 6 (16) 3 (16) 3 (16) >.99 Peripheral artery disease 3 (8) 2 (11) 1 (5).547 Diabetes mellitus 5 (13) 2 (11) 3 (16).631 Body mass index, kg/m Hyperlipidemia 3 (8) 2 (11) 1 (5).547 Aortic pathology Chronic type B dissection 17 (45) 10 (53) 7 (37).328 Aneurysm 21 (55) 9 (47) 12 (63).328 Zone 2 34 (90) 18 (95) 16 (84).290 Zone 3 4 (10) 1 (5) 3 (16).290 Landing zone diameter, mm Proximal Distal Mean aortic diameter, mm Graft oversizing, % COPD, Chronic obstructive pulmonary disease. a Continuous data are shown as mean 6 standard deviation and categoric data as number (%). b Receiving antiplatelet agents for positive cardiac thallium scan, percutaneous coronary intervention, or coronary artery bypass graft surgery. c Chronic kidney disease $stage 3 (glomerular filtration rate <60 ml/min/1.73 m 2 ). angulation. The median centerline was computed from the sinotubular junction of the aortic root to the diaphragm level and divided into zones 1 to 3, according to the definition of zones of attachment. 3 The center of the sinotubular junction was designated as the reference point (0, 0, 0), and the software generated three-dimensional coordinates of each point in 1-mm intervals along the median centerline. The angulation of each point (q) was calculated in 30-mm ranges using vector inner products ( ȧ ḃ ¼jȧjj ḃ j cos q, where j ȧ j and j ḃ j were 15 mm in length along the median centerline and q was the angle between vector a and vector b; Fig 1). The origins of the innominate artery, left common carotid artery, and left subclavian artery (LSA) were used as landmarks to define zones of attachment and to map the median centerline. The angles at the levels of these landmarks were also recorded and analyzed. The sharpest angle in each landing section represented the angle of the attachment zone. Bird-beak angle. The bird-beak angles of the stent grafts were calculated on straight multiplanar reconstruction images. Under cross-section view, the distance (D) from the proximal end of the stent graft to the aortic wall can be measured. The projecting length (L) of the bird-beak configuration was measured along the aortic wall until the stent graft attached to the wall without a gap. The angle of the bird-beak configuration (F) was defined by the following formula: tan F ¼ D/L, F ¼ tan 1 (D/L) (Fig 2). Aortic diameters. The diameter of the proximal or distal landing zone was calculated by the mean of maximal and minimal transverse diameters. The average diameter of the aorta was defined as mean of diameters of the proximal and distal landing zones. All of the morphologic data were analyzed using Excel 2010 software (Microsoft Corp, Redmond, Wash) and compared at baseline and at 1, 6, and 12 months postoperatively between the Pro-Form and the Z-Trak groups. Continuous variables are expressed as mean 6 standard deviation and compared with Student t-test. Categoric variables were tested with c 2 and Fisher exact tests. A receiver-operating characteristic curve (ROC) analysis was used to determine the cutoff values of the significant factors. Values of P <.05 were considered to indicate statistical significance. Statistical analyses were performed with SPSS 18.0 software (SPSS Inc, Chicago, Ill), and ROC curves were plotted by MedCalc software (MedCalc Software, Ostend, Belgium). RESULTS Patient demographics are listed in Table I. All patients underwent TEVAR #14 days after diagnosis, and no operations were performed under emergency conditions. No open surgical conversions were necessary at the time of the initial operation or during the follow-up period. There was no type I endoleak or graft collapse at 1 year of followup. One patient with a Pro-Form stent and one patient with Z-Trak stent had type II endoleak from the LSA, which were detected by computed tomography scan at 1 month. Both were patients with aortic dissection, and the endoleaks were resolved gradually by 1 year.

4 Volume 59, Number 1 Hsu et al 83 Fig 3. Angle remodeling with each point presented as the mean angle degree with the standard error (range bars) for the (A) Pro-Form and (B) Z-Trak stents. INM, Innominate artery; LCCA, left common carotid artery; LSA, left subclavian artery. The progressive angle remodeling is shown in Fig 3. Angulation changes were not evident at 1 month and 6 months in either group. However, after 1 year of follow-up, these changes were evident in zone 2 ( at baseline vs at 12 months; P ¼.041) and at the level of the LSA ( at baseline vs at 12 months; P ¼.043) in the Z-Trak group but not in the Pro-Form group. Significant angulation transformation in Pro-Form and the Z-Trak groups were noted in zone 2 ( vs , respectively; P ¼.033) and at the level of the LSA ( vs ; P ¼.031) at 12 months (Fig 4). Significant changes of angulation between the Pro-Form and the Z-Trak stent grafts were also observed in zone 2 ( vs 4 6 6, respectively; P ¼.049) and at the LSA level ( vs 6 6 9, respectively; P ¼.042) at 1 year. The bird-beak configuration was detected in six patients (32%) in the Pro-Form and in 11 (58%) in the Z-Trak (P ¼.096) groups at 1 month. At 12 months, the Pro-Form group had the same six patients (32%), but the Z-Trak had three additional patients, totaling 14 patients (74%) with bird-beak (P ¼.022). The mean bird-beak angle was significantly smaller in the Pro-Form group than in the Z-Trak group at 1 month (5 6 8 vs ; P ¼.019) and at 1 year ( vs ; P ¼.033; Fig 5). The morphologic parameters between the groups with and without bird-beak were compared in the dissection or aneurysm group. No significant differences were noted in those factors in the aneurysm group. In the dissection group, however, the mean preoperative angle of zone 2 was significantly smaller in patients with a bird-beak configuration ( vs ; P ¼.003; Table II). We also evaluated these factors on the Pro-Form platform, and only the preoperative angle of zone 2 was significant (P ¼.009; Table III). The ROC curves analysis showed was the cutoff value for the preoperative angle of zone 2 to estimate the presence of a postoperative bird-beak configuration (sensitivity, 86%; specificity, 83%; Fig 6). DISCUSSION TEVAR for aortic arch pathology often involves a proximal landing zone in angulated anatomy. Conformation to the aortic arch poses a challenge, especially when the angle is acute and sharp. Manually fitting a circle to the aortic arch with an approximation of its radius 4 or with an approximation of angle at the landing zone 5 is subjective, and these techniques do not fully reflect the complexity

5 84 Hsu et al January 2014 Fig 5. Comparison of bird-beak angle is shown at different periods of follow-up for the Pro-Form and Z-Trak stents. Each point presented is the mean angle degree with the standard deviation (range bar). Table II. Comparisons of morphologic parameters between the dissection group patients with and without bird-beak Variable a Bird-beak (n ¼ 8) No bird-beak (n ¼ 9) P b Fig 4. Significant angulation transformation is shown between the Pro-Form group and Z-Trak group for (A) zone 2 and (B) the left subclavian artery (LSA). Each point presented is the mean angle degree with the standard deviation (range bar). and variety of aortic angulation. The luminal centerlinebased quantitation of aortic angulation we used in the present study provides a more objective assessment. In contrast to the curvature index used in previous studies, 6-8 the quantitative analysis in our study provided the angle of the intended location. Furthermore, using triangle and vector formulas, our data could also be transformed into a curvature index. For example, in our model corresponded to a curvature of 0.2 cm 1 or a radius of 5.0 cm, a threshold for calculation in the study of Ueda et al. 6 The results of this study showed a continuous and dynamic remodeling process of the aortic angle after TEVAR. The angulation in the Z-Trak group was more unsteady than that in the Pro-Form group, especially in the zone 2 and LSA levels of the distal arch. This was also reflected by the significant angle discrimination in zone 2 and the LSA between these two generations of Zenith thoracic endografts. Although the mechanism was not clear, we hypothesize that when a stent graft has to land at a distal aortic arch, the inner curve of the aortic arch in zone 2 or 3 acts as a fulcrum over which the sealing segment hinges at the first or second interstent junction. 9 With the interaction between the wind-sock effect of the aortic blood flow and the rigidity of the proximal stent graft, the leading edge of the stent graft will be lifted off the inner curve, resulting in the bird-beak configuration. The upward force Stent graft Pro-Form Z-Trak 4 3 Preoperative angle, Innominate Zone LCCA Zone LSA Zone Landing zone diameter, mm Proximal Distal Mean aortic diameter, mm Graft oversizing, % LCCA, Left common carotid artery; LSA, left subclavian artery. a Categoric data are shown as number and continuous data as mean 6 standard deviation. b Continuous variables were tested with the Student t-test, and categoric variables were tested with Fisher exact tests. produced by the blood flow on the outer edge of the sealing segment and the straightening of the stent graft will then flatten the distal aortic arch, increase the risk of malposition to the arch, and enlarge the angle of the bird-beak configuration. This may explain the sequential angulation changes of zone 2 and LSA levels, the increased incidence of bird-beak formation, and enlarged angle of the bird-beak configuration in the Z-Trak group during the follow-up period in this study (Figs 7 and 8). In agreement with previous reports, 9,10 the newly modified deployment system of the Pro-From TX2 improved conformation to the inner curve of the arch, resulting in a lower incidence of the bird-beak configuration in the present study. A smaller angle of bird-beak configuration and lower incidence of bird-beak formation

6 Volume 59, Number 1 Hsu et al 85 Table III. Comparisons of morphologic parameters between patients with and without bird-beak in the Pro- Form group Variable a Bird-beak (n ¼ 6) No bird-beak (n ¼ 13) P b Group Dissection Aneurysm 2 7 Preoperative angle, Innominate Zone LCCA Zone LSA Zone Landing zone diameter, mm Proximal Distal Mean aortic diameter, mm Graft oversizing, % LCCA, Left common carotid artery; LSA, left subclavian artery. a Categoric data are shown as number and continuous data as mean 6 standard deviation. b Continuous variables were tested with the Student t-test, and categoric variables were tested with Fisher exact tests. were also noted during the 1-year follow-up period. Lee et al 9 found similarly favorable results, with a shorter gap distance between the leading endograft edge and the inner curve of the arch in Pro-Form patients. In the current study, the distance from the proximal end of the stent graft to the aortic wall (D) was similar to the gap distance described by Lee et al. 9 The D of the Pro-Form group in our study was significantly smaller than that of the Z-Trak group ( mm vs mm; P ¼.036) after 1 month. However, no significant differences were seen at 6 months ( mm vs mm; P ¼.236) or 12 months ( mm vs mm; P ¼.056). In addition, the analysis of projecting length (L) revealed that L was longer in the Z-Trak group at 1 month ( mm vs mm; P ¼.215), at 6 months ( mm vs mm; P ¼.622), and at 12 months ( mm vs mm; P ¼.211). This indicates that the length of protrusion of the proximal end of the endograft was longer in the Z-Trak group for a given bird-beak angle. In other words, there was a shorter attachment of the proximal sealing segment to the inner curve of the aortic arch in the patients with a bird-beak configuration in the Z-Trak group. According to the report by Ueda et al 2 and the present study, patients with a birdbeak configuration and a Z-Trak stent are at a higher risk of endoleak formation. The encouraging Pro-Form results can be attributed to the modification of the deployment system with a diameter-reducing tie that plays an important role in achieving a better apposition to the aortic arch and serves two major functions. 9 First, it forces the proximal sealing stent lying along the inner curve of the arch to partially intussuscept Fig 6. Receiver operating characteristic (ROC) curve and the area under the curve (AUC) are shown for the preoperative angle of zone 2 in the dissection group. within the second stent, thereby allowing the terminal edge of the endograft to lie more parallel to the proximal down slope of an angulated proximal landing zone. Second, it constrains the proximal endograft segment more effectively so that wind-socking is further minimized, making the endograft more stable during deployment and positioning. Previous studies have shown that a greater curvature of the aortic arch increases the risk of type I endoleaks 7 and that a bird-beak configuration increases the risk of endoleak formation after TEVAR. 2 However, the relationship between aortic angulation and bird-beak formation is not yet known. After comparing the parameters of angulation and aortic diameters, our results showed that the preoperative angle of zone 2 was significantly different between patients with and without bird-beak configuration in the dissection group but not in aneurysm group. To investigate whether the aortic diameters were a factor in these two pathologies, we compared the preoperative zone 2 angle and the aortic diameters in patients with bird-beak configuration between dissection and aneurysm groups. The data showed that the preoperative angle of zone 2 was significantly smaller in dissection patients (Table IV). This means that the distal aortic arch angle does play an important role in the formation of bird-beak in dissection. The actual mechanism was not clear, and no suitable model is available to explain and evaluate the bird-beak formation between dissection and aneurysm. Among the possible factors that need further investigation are the dynamics of blood flow, true lumen expansion, false lumen regression or thrombosis, distal communication of true and false lumen, and aneurysm sac thrombosis. For the Pro-Form stent landing in zone 2 or 3, the preoperative angle of zone 2 was also significantly smaller in patients with bird-beak configuration. We tried to determine the estimated cutoff value, and was selected with a sensitivity of 86% and a specificity of 83%. The

7 86 Hsu et al January 2014 Fig 7. Follow-up images after thoracic endovascular aortic repair (TEVAR) with the Pro-Form stent graft show a stable distal arch angle and proper apposition of the stent graft to the inner curve of aortic arch. A, A sagittal-view computed tomography image shows the thoracic aorta. B, Three-dimensional reconstruction image shows the median centerline (green line) and the distal arch angle (red angle). C, Straight multiplanar reconstruction image shows the median centerline (green line) and apposition of the stent graft to the inner curve of the aortic arch (red circle). positive likelihood ratio was 5.0 and suggests that a patient with a preoperative zone 2 angle of <151.1 who receives Pro-Form endografting will have increased chance of having a bird-beak formation by a factor of 5.0. To investigate whether the negative aspects with Z-Trak were simply delayed in Pro-Form recipients, or even different at a later time, a longer period of followup with adequate sample size is necessary to assess the changes. With limited data, it is hard to recognize the trends of the bird-beak angle in either group. During the 2 years of follow-up until now, data for 10 patients in the Z-Trak group were collected, and seven of them had bird-beak configuration. They were all among the 14 patients with bird-beak at 12 months, and the mean angle of bird-beak was , which was not significantly different than the angle at 12 months (P ¼.822). Five patients in the Pro-Form group were also evaluated, and there was also no new bird-beak formation. Among them, two had bird-beak configuration with a mean angle of 8 6 4, but this also not significant to the angle at 12 months (P ¼.817). Six patients were found to have enlarged bird-beak during follow-up at 1 year. Two of these six patients were from the Pro-Form group and the rest were from the Z-Trak group. The increased angle was not significantly different between the Pro-Form (5 6 0 ) and

8 Volume 59, Number 1 Hsu et al 87 Fig 8. Sequential images of a Z-Trak stent graft with bird-beak configuration show the enlarged distal arch angle (red angle) and worsening apposition of the stent graft to the inner curve of the aortic arch (red circle). A, A sagittal-view computed tomography image shows the thoracic aorta. B, Three-dimensional reconstruction image shows the median centerline (green line) and the distal arch angle (red angle). C, Straight multiplanar reconstruction image shows the median centerline (green line) and apposition of the stent graft to the inner curve of aortic arch (red circle). Z-Trak (7 6 7 ; P ¼.556) groups. We assessed the morphologic data and found no significant differences. However, it is noteworthy that the projecting length (L) of the bird-beak configuration was longer in the group with an enlarged ( mm) compared with a nonenlarged ( mm; P ¼.06) angle. According to the computational model by Pasta et al, 11 a longer protrusion length of stent graft leads to an increased transmural pressure gradient across the protruded wall of that stent graft, and that would portend endograft enfolding. We think this mechanism also caused the expansion of the birdbeak angle in our present study. The present study provides several new insights into aortic remodeling after TEVAR. The aortic arch angulation changes with time, and the transformation varies in different zones. Accurate assessment of the trend of the aortic remodeling is crucial for the prevention of complications and in the design of stent grafts. When treating aortic dissection with a stent graft landing on zone 2 or 3, the angulation of the distal arch plays an important role in

9 88 Hsu et al January 2014 Table IV. Comparisons of preoperative zone 2 angle and aortic diameter in bird-beak between the dissection and aneurysm groups Variable a Bird-beak in dissection (n ¼ 8) Bird-beak in aneurysm (n ¼ 9) P b Preoperative zone 2 angle, Landing zone diameter, mm Proximal Distal Mean aortic diameter, mm Graft oversizing, % a Data are shown as mean 6 standard deviation. b Student t-test. bird-beak formation. We encourage consideration of this factor when planning the treatment. Although the Pro-Form endograft can provide better conformability on proximal fixation in the distal aortic arch, we believe that special caution should be exercised when a patient presents with a small distal arch angle, such as #151.1 in our model. When considering endografting in these patients, careful preoperative planning is critical, and regular follow-up is needed to detect development of the bird-beak configuration. The limitations of this study include the nonrandomized nature, the limited number of patients, the relatively short follow-up time, and the use of stainless steel stent grafts from one manufacturer. In addition, we only included aortic pathology involving zone 2 or 3 and only analyzed remodeling of the aortic arch. The findings in the present study should not be applied to lesions or remodeling in other portions of the aorta. Dissection and aneurysm were both included in the present report; however, more specific studies on the basis of dissection or aneurysms are required. Finally, the quantitative analysis in this study has not been validated. Further investigations are needed and underway. CONCLUSIONS Angulation remodeling after stainless steel stent graft repair of the thoracic aorta is a continuous process. For aortic pathology requiring proximal fixation in zones 2 and 3, a significant arch angle transformation was discovered in the distal aortic arch. Pro-Form stent grafts improved the arch conformation during the 1-year follow-up period. In patients with dissection, a preoperative angle of zone 2 seemed to be predictive of postoperative bird-beak formation. AUTHOR CONTRIBUTIONS Conception and design: HH, CK, CS Analysis and interpretation: HH, CK Data collection: HH, PL, IM, CP Writing the article: HH Critical revision of the article: HH, CW, CS Final approval of the article: CS Statistical analysis: HH Obtained funding: CS Overall responsibility: CS REFERENCES 1. Ishida M, Kato N, Hirano T, Cheng SH, Shimono T, Takeda K. Endovascular stent graft treatment for thoracic aortic aneurysms: shortto midterm results. J Vasc Interv Radiol 2004;15: Ueda T, Fleischmann D, Dake MD, Rubin GD, Sze DY. Incomplete endograft apposition to the aortic arch: bird-beak configuration increases risk of endoleak formation after thoracic endovascular aortic repair. Radiology 2010;255: Fillinger MF, Greenberg RK, McKinsey JF, Chaikof EL. Society for Vascular Surgery Ad Hoc Committee on TRS. Reporting standards for thoracic endovascular aortic repair (TEVAR). J Vasc Surg 2010;52: , e Sternbergh WC 3rd, Money SR, Greenberg RK, Chuter TA, Zenith I. Influence of endograft oversizing on device migration, endoleak, aneurysm shrinkage, and aortic neck dilation: results from the zenith multicenter trial. J Vasc Surg 2004;39: Bowman JN, Silverberg D, Ellozy S, Teodorescu V, Poblete H, Marin M, et al. The role of anatomic factors in predicting success of endovascular repair of thoracic aortic aneurysms. Vasc Endovasc Surg 2010;44: Ueda T, Takaoka H, Raman B, Rosenberg J, Rubin GD. Impact of quantitatively determined native thoracic aortic tortuosity on endoleak development after thoracic endovascular aortic repair. AJR Am J Roentgenol 2011;197:W Nakatamari H, Ueda T, Ishioka F, Raman B, Kurihara K, Rubin GD, et al. Discriminant analysis of native thoracic aortic curvature: risk prediction for endoleak formation after thoracic endovascular aortic repair. J Vasc Interv Radiol 2011;22: e2. 8. Tillich M, Bell RE, Paik DS, Fleischmann D, Sofilos MC, Logan LJ, et al. Iliac arterial injuries after endovascular repair of abdominal aortic aneurysms: correlation with iliac curvature and diameter. Radiology 2001;219: Lee WA, Martin TD, Hess PJ Jr, Beaver TM, Klodell CT. First United States experience of the TX2 Pro-Form thoracic delivery system. J Vasc Surg 2010;52: Melissano G, Civilini E, Bertoglio L, Logaldo D, Chiesa R. Initial clinical experience with the modified Zenith Pro-Form TX2 thoracic endograft. J Endovasc Ther 2010;17: Pasta S, Cho JS, Dur O, Pekkan K, Vorp DA. Computer modeling for the prediction of thoracic aortic stent graft collapse. J Vasc Surg 2013;57: Submitted May 3, 2013; accepted Jul 17, 2013.