From the Western Vascular Society
Feasibility of endovascular repair of ascending aortic pathologies as part of an FDA-approved physician-sponsored investigational device exemption Ali Khoynezhad, MD, PhD,a Carlos E. Donayre, MD,b Irwin Walot, MD,b Matthew C. Koopmann, MD,b George E. Kopchok, BS,c and Rodney A. White, MD,b Los Angeles and Torrance, Calif Objective: Endovascular treatment of ascending aortic lesions has been reported, but to date, no FDA-approved studies have been conducted to define feasibility and the use of endografts in this particular location or to analyze the critical factors involved. Methods: Patients were consented for entry into an FDA-approved physician-sponsored investigational device exemption study to investigate the outcome of those with ascending aortic pathologies. These patients were suitable according to the instructions for use for endovascular repair with a Valiant Captivia (Medtronic, Inc, Minneapolis, Minn) thoracic stent graft, a device designed specifically for deployment in the ascending aorta. All patients had sequential gated-cardiac computed tomography scans, with data being entered into the VQI Complex TEVAR software (West Lebanon, NH). All procedures were performed in a hybrid room, with the capability to convert to an open repair to ensure maximal patient protection. The first five patients constituted the feasibility study, with continued enrollment based on initial results and submission of an annual report to the FDA. Results: Thirty-nine patients were screened, and six patients were entered into the physician-sponsored investigational device exemption study. Although there was no early mortality, there was one late death. All patients had sequential computed tomographies and cardiac echocardiograms with no evidence of migration, one type 1a endoleak, one postoperative stroke, and regression of the aortic lesions in the excluded aortic segment. Conclusions: In this feasibility study, the preliminary evaluation of endovascular treatment for ascending aortic pathologies demonstrates uniform accuracy of deployment and secure fixation up to 17.5 months of follow-up. There is positive remodeling of the excluded aortic segments similar to surveillance studies involving the descending aorta. (J Vasc Surg 2016;63:1483-95.)
Pathologies of the aorta have been associated with very high morbidity and mortality rates and remain the top sixth leading cause of death in elderly men.1 Many patients with aortic issues are extreme high-risk or nonsurgical candidates due to comorbid conditions. Some elderly patients may have concomitant valvular heart From the Division of Cardiothoracic Surgery, Cedars-Sinai Medical Center, Los Angelesa; the Department of Surgery, David Geffen UCLA School of Medicine, University of California, Los Angeles, UCLA School of Medicine,b and Los Angeles Biomedical Research Institute,c Harbor-UCLA Medical Center, Torrance. Clinical trial number NCT02201589. This study does include off-label drugs and/or devices. The specific name of the product discussed is the Medtronic Valiant PS-IDE Stent Graft. Author conflict of interest: A.K., C.E.D, and R.A.W. are consultants for Medtronic Vascular. Presented in Scientific Session III at the Thirtieth Annual Meeting of the Western Vascular Society, Maui, Hawaii, September 21, 2015. Correspondence: Ali Khoynezhad, MD, PhD, FACS, Professor of Cardiovascular Surgery, Director of Aortic Surgery, Cedars-Sinai Medical Center, 127 San Vicente Blvd, Ste #3306, Los Angeles, CA 90048 (e-mail: [email protected]). 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. 0741-5214 Copyright Ó 2016 by the Society for Vascular Surgery. Published by Elsevier Inc. http://dx.doi.org/10.1016/j.jvs.2015.12.029
disease (mostly calcific aortic valve stenosis). Treatment options for the nonsurgical patient population are very limited. Anti-impulsive/antihypertensive therapy and other adjunct pharmacotherapy are the only available options and unfortunately have limited results. There are currently no transcatheter or other less-invasive treatment options for this patient cohort. The success of endovascular stent graft repair of the descending thoracic aorta has led to the application of these devices for the management of thoracic aortic diseases in other locations such as the aortic arch and ascending aorta.2,3 Off-label use of a descending thoracic stent graft in the ascending aorta has been described since 2000 from outside of the United States,4 while Ihnken and colleagues reported the first off-label use in the United States in 2004.5 There are a series of case reports mostly involving high-risk patients with high-risk features or pseudoaneurysms of the ascending aorta undergoing endovascular repair.6-14 The purpose of this study is to investigate the outcome of selected patients with ascending aortic pathologies including type A aortic dissection, retrograde type A aortic dissection, intramural hematoma, penetrating ulcer, or pseudoaneurysm with isolated diseases affecting the aorta between the sinotubular junction and the innominate artery orifice (with no involvement of the aortic valve and the 1483
JOURNAL OF VASCULAR SURGERY June 2016
1484 Khoynezhad et al
Table I. Secondary observations
Table II. Inclusion and exclusion criteria
Early observations d Successful delivery and deployment of the stent graft d Coverage of lesion and/or proximal entry tear d Aortic remodeling based on serial imaging d SAE d Rupture
Inclusion criteria d Ascending thoracic aortic pathology affecting the area between the sinus of Valsalva and the innominate artery orifice (with no involvement of the aortic valve) including:
Late observations d Aortic remodeling at 6- and 12-month visit (defined below) d Secondary procedures within 12 months d Continuing or new FL perfusion d SAEs within 12 months d Rupture within 12 months d All-cause mortality within 12 months Aortic remodeling as measured by Change from baselinea in the maximum TL diameter over the length of the stent graft Change from baselinea in the maximum FL diameter over the length of the stent graft Change from baselinea in the maximum total thoracic aortic diameter FL thrombosis over the length of the stent graft CTA, Computerized tomography angiogram; FL, false lumen; SAE, serious adverse events; TL, true lumen. a The predischarge CTA will be used as baseline to assess aortic remodeling. If discharge occurs after 1-month follow-up, the 1 month follow-up CTA will be used for baseline purposes.
aortic root). These patients underwent thoracic endovascular aortic repair (TEVAR) with the Valiant Stent Graft device (Medtronic, Santa Rosa, Calif) as part of a physician-sponsored investigational device exemption (PSIDE) approved by the FDA. METHODS Study design. This was a prospective trial of the Valiant PS-IDE device (Medtronic) to determine the feasibility of successful implantation, as indicated by exclusion of the thoracic lesion and graft patency at implant, time of discharge, and 1, 6, and 12 months following implantation, and to determine the proportion of patients who die or experience adverse events during and after the implantation. The primary end points were all-cause early mortality, reinterventions, surgical conversions, aortic remodeling (defined as any reduction of aortic aneurysm volume or aortic diameter stabilization or reduction), and postprocedural strokes. Secondary end points are detailed in Table I. The study protocol was approved by the institutional review boards at Harbor-UCLA Medical Center and LA Biomed Research Institute in Torrance, Calif, and Cedars-Sinai Medical Center in Los Angeles, Calif, and was registered at ClinicalTrials.gov (identifier: NCT02201589) with Drs Khoynezhad and White as joint principal investigators. Enrollment. Investigators will assess potential subjects who are diagnosed with type A thoracic aortic dissection, retrograde type A thoracic aortic dissection, intramural hematoma, penetrating ulcer or pseudoaneurysm of the ascending thoracic aorta for their suitability for enrollment in the clinical
Type A thoracic aortic dissection Retrograde type A thoracic aortic dissection B Intramural hematoma B Penetrating ulcer B Pseudoaneurysm d One cm or greater proximal and distal landing zones in the ascending aorta d Ascending aorta diameter between 28-44 mm d American Society of Anesthesiologists Class IV d Candidates for endovascular repair B B
Exclusion criteria d Pregnancy d Age less than 21 years of age d Presence of a condition that threatens to infect the stent graft/aortic valve prosthesis d Allergy to the stent graft material d Inability to obtain informed consent d Expected survival less than one year
study. The patient enrollment goal is 20 patients in five-patient increments. The results of each five-patient group will be analyzed and reported prior to enrolling further patients as a risk-mitigating strategy. Study subject accrual is planned to occur over approximately 5 years at the intervals previously cited and will be under surveillance for life even following study completion. Suitable candidates for inclusion comprised adult male or female patients who agreed to follow-up. In addition, they signed a consent approved by the FDA and institutional review board. To be considered a candidate for endovascular repair, the patient must have had an aforementioned aortic pathology of the ascending thoracic aorta affecting the area between the sinotubular junction and the innominate artery orifice with no involvement of the aortic valve. The patient must also have at least 1 cm of suitable landing zones both proximally and distally to the dissection or diseased area, with an ascending aorta diameter between 28 mm and 44 mm and must a be high-risk surgical candidate with an American Society of Anesthesiologists (ASA) score of IV (Table II). The following patients were excluded: (1) pregnant or pediatric patients (younger than 21 years of age); (2) patients who have a condition that threatens to infect the stent graft/aortic valve prosthesis; (3) patients with allergies to the stent graft material; (4) patients who fail to sign informed consent; and (5) patients with an expected survival less than 1 year. If a patient also has a concomitant descending thoracic aortic dissection or aneurysm that fits the criteria of PS-IDE, the candidate may be offered endovascular therapeutics as part of that PS-IDE in a separate operative setting (Table II). Device and delivery system. The Valiant PS-IDE consists of a modified Valiant Thoracic Stent Graft placed by Medtronic’s Captivia Delivery System and is intended to
JOURNAL OF VASCULAR SURGERY Volume 63, Number 6
be used for the endovascular treatment of pathologies of the ascending aorta. For the purposes of this study, the Valiant Thoracic Stent Graft has been modified from the original configuration to fit the shorter length and wider diameter of the ascending aorta. The modifications to the design are discussed in the information to follow. Two configurations of the Valiant PS-IDE Stent Graft were proposed: (1) a proximal FreeFlo taper and distal closed web and (2) a proximal closed-web design with bare springs distal. To treat the patient population proposed in this study, only one modification was made to the Valiant Stent Graft, consisting of the elimination of stent rings in grafts with shorter lengths. The proximal FreeFlo design was identical to the Valiant Thoracic Stent approved in the U.S. for the descending aorta with the exception of the shorter lengths (5-, 7-, and 9-cm lengths). Fig 1 illustrates the intended device placement. The proximal closed-web configuration with a distal bare spring was available in lengths of 5, 7, and 9 cm. Fig 2 details all device sizes and configurations. The delivery system configuration was also identical to that of the Valiant Captivia Delivery System. A nontip capture device was used for the proximal closed-web design, whereas the tip capture system was used for the FreeFlo configuration. Both delivery systems included hydrophilic coating to aid in delivery. All other aspects of the Valiant Stent Graft remained unchanged, and no changes to the Captivia Delivery System were made. Proposed Valiant Stent Graft sizes included 30-mm to 46-mm diameters. Imaging. All patients had electrocardiogram-gated cardiac, 64-slice computed tomography (CT) scans with SmartPrep (Boston, Mass) and acquisition at end-systole: precontrast CT chest (3-mm cuts), followed by CT angiography neck, chest, abdomen/pelvis (1 mm) (starting images from the base of the skull to the femoral head). A CT with and without contrast and echocardiogram 30 days after the device is implanted, CT with and without contrast at 6 and 12 months postprocedure and once a year from 2 years through 5 years after the implant will be performed. A trained technician generates a 3D model of the aortic root including the ascending aorta, left ventricle, and any thrombotic or calcified plaques present as individual objects from gated, enhanced CT axial slices. Preoperative work-up. The following will take place prior to the implant procedure, which includes (1) signing of the informed consent, (2) collecting demographic data, (3) medical history, (4) vital signs (heart rate, blood pressure, respiratory status), (5) pulse evaluation (left and right radial, dorsalis, and/or tibial artery), (6) pregnancy test (females of child-bearing potential), and other clinical laboratory tests (serum creatinine levels), (7) CT angiography or magnetic resonance angiography of the chest/ abdomen/pelvis with and without contrast, and (8) an echocardiogram. Additional tests may be ordered at the investigator’s discretion (diffusion-weighted magnetic resonance imaging of the head, neuropsychological evaluation, and/or transcranial Doppler). Procedure. The details of the operative procedure have been described in detail in previous publications.2,3,15
Khoynezhad et al 1485
Fig 1. Proposed Valiant PS-IDE Stent Graft placement in the ascending aorta.
In brief, bilateral surgical femoral arterial and venous access is provided. The arterial system is used for angiogram and intravascular ultrasound (IVUS) evaluation, while the contralateral side is used for device delivery. The venous access sites are used for ventricular pacing and heart-lung machine access in case of emergency. After initial triple imaging (angiogram, IVUS, and echocardiogram), the proper device is selected and deployed under ventricular pacing in the desired anatomic position. Postoperative triple imaging confirms the proper position of the device and lack of endoleak. Subsequently, the catheters are removed, and the access vessels are repaired. RESULTS Primary and secondary end points. In the last 2 years, 39 patients were screened, and 6 patients were entered in the PS-IDE and followed up for an average of 17.5 months (range, 10-23 months). Reasons for screening failure were lack of minimal proximal or distal landing zone in 23, poor access vessels in 7, and aortic rupture/lost to followup in 3 patients. Among the patients entered in the study, there were no intraoperative or early ruptures or early conversions to open ascending repair. Furthermore, there were
JOURNAL OF VASCULAR SURGERY June 2016
1486 Khoynezhad et al
Fig 2. Size representation of the Valiant PS-IDE Stent Graft.
no intraoperative or early mortality. Patient No. 3 required a total arch replacement and frozen elephant repair due to de novo ulceration in the mid-aortic arch. This patient died 16 weeks after the index TEVAR due to unrelated causes. Successful delivery and deployment of the Valiant PS-IDE ascending stent graft was achieved in all subjects. Patient No. 2, who had a history of previous stroke, suffered a periprocedural lacunar infarct, from which he recovered promptly with no clinically evident residuals. The same patient had also a type Ia endoleak due to incomplete coverage of the aortic lesion proximally that remains stable on surveillance imaging. Similarly, patient No. 6 who has thrombosed most of the pseudoaneurysm, has a small residual perfusion (type Ia), which so far has been stable as shown in surveillance
imaging. Table III is a summary of patient indications of operation, complications, and residual patient deficits. Table IV is an overview of volumetric data acquired via 3D reconstruction by M2S in the surveillance imaging, summarizing three variables: (1) ascending maximum sac diameter, (2) volume of the ascending aorta from the sinotubular junction to the origin of the innominate artery, and (3) the angle between a centerline of the ascending aorta and the intersection of a line from the apex of the left ventricle to the center of the aortic valve. Table IV displays the same parameters for enrolled patients. Review of patients and technical observations at follow-up. Patient No. 1, a 77-year-old male with a history of atrial fibrillation, pacemaker implantation,
JOURNAL OF VASCULAR SURGERY Volume 63, Number 6
Khoynezhad et al 1487
Table III. Summary of indications of operation, endografts deployed, device proximal and distal diameters and length, complications and residual patient effects Patient, No.
Indication
No. Endo-grafts
Device size, mm
Complication(s)
Residual effects
1 2
Penetrating ulcer Pseudoaneurysm
1 1
38 x 38 x 60 L 36 x 36 x 80 L
3
Type A dissection based on penetrating aortic ulcer Chronic type A dissection Pseudoaneurysm Pseudoaneurysm
1
40 x 40 x 60 L
2 1 1
46 x 46 x 80 L 46 x 46 x 40 L Aortic valve leaflet entrapment None 34 x 34 x 60 L None None 40 x 40 x 60 L None Proximal pseudoaneurysm
4 5 6
Wire perforation left ventricle None Lacunar infarct, type 1 Mild residual endoleak symptoms De novo aortic ulcer requiring Late death frozen elephant trunk repair
Table IV. Overview of volumetric data
Patient No.1 Ascending maximum sac Volume Angle Patient No. 2 Ascending maximum sac Volume Angle Patient No. 3 Ascending maximum sac Volume Angle Patient No. 4 Ascending maximum sac Volume Angle Patient No. 5 Ascending maximum sac Volume Angle Patient No. 6 Maximum size diameter Volume Angle
Baseline
1 Month
6 Months
1 Year
37.2 60.5 150
39.8 70.1 144
39.8 76 145.3
40.4 71 143.8
54.7 101.1 145.8
54.4 113.3 132.8
diameter, mm
46 105.8 147.5
47.4 111 143
50.8 110.6 132.6 At 2 months 54.9 NA 143.2
diameter, mm
57.6 188.3 124
57.3 186.1 141.6
58.2 196.1 155.8
37.4 63.7 125
37.6 60.6 129.4
37.5 63.2 138.3
X
94.1 447.9 133.3
90.7 402.3 131.4
93.5 404.2 125.8
X
diameter, mm
diameter, mm
diameter, mm
peripheral vascular disease, zone II TEVAR for symptomatic penetrating aortic ulcer and intramural hematoma, underwent ascending stent grafting in November 2013. He had an intraoperative pericardial effusion with no sign of cardiac tamponade detected by transesophageal echocardiogram. This pericardial effusion was stable after reversal of heparin, followed by transthoracic echo, and was found to regress by postoperative day 7. The investigators opted to use an Amplatz Extra-Stiff Wire guide with a 3-mm radius tip for future subjects instead of the Meier wire that has a shorter flexible tip. This strategy would reduce the risk for future left ventricular injury that may have been the origin of this patient’s pericardial effusion. The 1-year CT scan of patient No. 1 is shown in Fig 3. The ascending aortic penetrating ulcer is excluded with
Decrease
Stable
Increase
X X X X X X
58.7 198.7 NA
X X X
X
firm fixation of the device in the deployed position. He is doing clinically well with no adverse events as seen on follow-up examination. Patient No. 2 is an 84-year-old female with prohibitive surgical risk including previous stroke, history of coronary artery bypass grafting and mitral valve repair, and postoperative sternal wound infection. She presented with an enlarging pseudoaneurysm at the surgical cannulation site and underwent ascending TEVAR. Although preoperative interrogation revealed a proper landing zone, the stent graft was not able to completely seal proximally. The patient’s composite serial CTs (Fig 4) are consistent with a type Ia endoleak. The inadequate seal in this patient was due to horizontal aortic anatomy (significant angulation of the centerline ascending aorta to the long axis of the
1488 Khoynezhad et al
ventricle) that did not allow complete coaxial placement of the stent graft at the sinotubular junction of the ascending aorta (Fig 5). Given the proximity of the left main coronary artery orifice, the investigators decided not to place another device. Patients with a horizontal aorta may best be treated by a transapical approach for the same deployment advantages that have been obtained with transapical aortic valve replacements. However, the investigators did not have FDA approval for such an approach in this patient. This patient was extubated and ambulatory within 24 hours. She had a postoperative stroke while talking to her family that resulted in a left lacunar infarct. This left her with right upper extremity weakness and significant speech and swallowing difficulty. Her symptoms improved significantly with physical and speech therapy. At this point, the investigators opted to observe the pseudoaneurysm. M2S calculations of the ascending maximum sac diameter (52 mm on the pretreatment 9/19/2013 CT, 54.7 at 1 day postoperative, and 50.8 on the 4/8/14 postprocedure CT) showed a decrease in this parameter. The volume was slightly higher (106.8 cc at pretreatment, 101.1 cc at 1 day posttreatment, and 110.6 cc at the 19-week postprocedure CT). Patient No. 3 is an 88-year-old female with a history of hypertension and deep vein thrombosis after hip replacement, who presented with a DeBakey type I aortic dissection based on a penetrating ulcer of the ascending aorta. She also had some intramural hematoma that went down to the thoracoabdominal aorta. The dissection did not surpass the sinotubular junction; therefore, she was offered an ascending TEVAR as part of the PS-IDE. In order to have a proper distal landing zone according to the PS-IDE protocol, a retroesophageal carotid-carotid bypass was performed. IVUS interrogation and aortogram postdeployment showed excellent stent graft vessel apposition with exclusion of the aortic ulcer (Fig 6). The patient did not require coverage of the innominate artery, and therefore, the carotid-carotid bypass eventually thrombosed due to competitive flow. She had postoperative swallowing issues but was cleared for regular diet. Due to inadequate intake, she ultimately required a percutaneous gastrostomy tube placement and was discharged. Approximately 7 weeks after the index procedure, the patient presented to another facility with complaints of shortness of breath and interscapular pain. A CT of the chest, abdomen, and pelvis revealed a new penetrating near-perforated mid-aortic arch ulcer and an enlarged intramural hematoma in the descending thoracic aorta (Figs 7 and 8). Patient No. 3 was transferred to CedarsSinai Medical Center, underwent a total arch replacement and a stent graft implantation of the descending thoracic aorta (frozen elephant trunk repair), and had a replacement of her ascending aorta with an aortic valve resuspension. She was extubated the next day and was transferred to floor on postoperative day 3. Her postsurgical course was complicated atrial fibrillation and bilateral acute thrombus in the jugular, subclavian, innominate, and axillary veins requiring anticoagulation with warfarin.
JOURNAL OF VASCULAR SURGERY June 2016
Fig 3. Patient No. 1 at 1-year follow-up, showing the device without migration or endoleak.
The patient was discharged to a rehabilitation center on postoperative day 18. She died at home 16 weeks past the index operation from gastrointestinal bleeding while being on warfarin. Patient No. 4 is an 84-year-old male with chronic localized type A aortic dissection that may have been due to his previous open-heart operation. Upon deployment of the stent graft, there was evidence of aortic regurgitation followed by hypotension. An angiogram confirmed the entrapment of the aortic cusp resulting in aortic insufficiency. Prompt distal repositioning of the stent graft using a compliant balloon normalized the aortic valve function assessed by transesophageal echocardiogram and the hemodynamics. Imaging for patient No. 4 showed firm fixation of the device and no endoleaks (Fig 9). The patient did very well postoperatively without any complications and was sent home in stable condition on postoperative day 4. Patient No. 5 was a 67-year-old male with a history of TEVAR of the descending thoracic aorta for aneurysmal pathology and an ascending pseudoaneurysm. He underwent uneventful TEVAR and had an uncomplicated postoperative course (Fig 10). He was discharged home on postoperative day 2 and had done well in follow-up.
JOURNAL OF VASCULAR SURGERY Volume 63, Number 6
Khoynezhad et al 1489
Fig 4. Composite serial computed tomography (CT) postprocedure (patient No. 2), showing a consistent proximal type 1 endoleak (arrow), with no enlargement of the root over the first year.
Fig 5. Conformity of the device to the aortic curve (patient No. 2).
Patient No. 6 is a 52-year-old male with a complicated history including ascending aortic repair for reported syphilitic aortitis in 2010. He has subsequently undergone two additional reinterventions for aortic root replacement. The last operation included placement of a mechanical heart
valve. At the time of entry into the study, the patient was found to have a large anterior mediastinal pseudoaneurysm at the Dacron graft of the ascending aorta. The vascular prosthesis was found to be very mobile along its entire length on IVUS with near collapse of the graft during
1490 Khoynezhad et al
JOURNAL OF VASCULAR SURGERY June 2016
Fig 6. Ascending aortogram of patient No. 3 prior to (left) and at follow-up (right), showing exclusion of the entry site of the intramural hematoma.
Fig 7. Computed tomography (CT) angiogram at 7 weeks (patient No. 3) showing stability of the ascending graft with exclusion of the original penetrating ulcer (left, arrow), but with progression of the intramural hematoma in the descending thoracic aorta with suspected contained rupture (right, arrows).
JOURNAL OF VASCULAR SURGERY Volume 63, Number 6
Khoynezhad et al 1491
Fig 8. An expanding intramural hematoma in the arch (patient No. 3) identified as endoleak (arrow) on an official read. This has otherwise been interpreted as expansion of intramural hematoma outside the device.
diastole. Following successful identification of the graft dehiscence site and prior to the endograft exclusion, a carotid-carotid bypass was performed to provide additional landings during the endograft deployment as needed. Fig 11 demonstrates the thrombosis of the pseudoaneurysm over the first 6 months with development of a pseudoaneurysm at the distal attachment site once the ascending graft stabilized and the ascending portion of the aneurysm reformed. Subsequent imaging of the pseudoaneurysm has demonstrated it is currently stable (Fig 12), and the investigators plan to intervene and cover this segment in the near future. DISCUSSION Safe and effective implementation of TEVAR in the treatment of ascending aortic pathologies remains one of the last bastions for less invasive and endovascular therapy of the aorta.2,3 The alternative strategy in patients entered in this PS-IDE would have been a (redo) sternotomy, open-heart surgery with hypothermic circulatory arrest to replace the entire ascending aorta with or without a
hemi-arch replacement. Furthermore, all these patients would not have been candidates for a zone 0 debranching operation, given the fact that there was no portion of the ascending aorta to perform the ascending innominate bypass. Therefore, most aortic specialists believe that ascending TEVAR may represent a disruptive technology in high-risk patients, as it fulfills an unmet need. Stenting of the ascending aorta presents unique challenges, including negotiating the curvature of the aortic arch, proximal fixation close to the aortic valve and coronary ostia, distal fixation that may impinge on the innominate artery, hemodynamic forces, the risk of cardio-aortic injury, and retrograde dissection. Some of the major complications of this procedure are novel to the endovascular specialist experienced in aortic endografting and include perforation of the left ventricle, injury and dissection of the aortic root, and occlusion of the coronary arteries.2,3 Furthermore, the ascending aorta is on the average a centimeter larger than the descending thoracic aorta and about one-fourth in length. These physiologic and anatomic characteristics of the ascending aorta have ramifications in stent
1492 Khoynezhad et al
JOURNAL OF VASCULAR SURGERY June 2016
Fig 9. Composite serial computed tomography (CT) postprocedure (patient No. 4) showing firm fixation of the device and no endoleaks.
Fig 10. Exclusion of an aortic pseudoaneurysm (patient No. 5); 6-month computed tomography (CT) shows regression of lesion (arrows).
JOURNAL OF VASCULAR SURGERY Volume 63, Number 6
Khoynezhad et al 1493
Fig 11. Imaging of patient No. 6 demonstrates thrombosis of the pseudoaneurysm over the first 6 months postindex procedure, with development of a pseudoaneurysm at the distal attachment site once the ascending graft stabilized and the ascending portion of the aneurysm reformed.
graft design that should entail shorter devices with a large diameter for proximal and distal landing zone fixation. In addition, there is a large variation in aortic diameter between the cardiac cycles in the ascending aorta. Based on 4-D evaluation of our patients, we found that the variation in size is about 10% in the descending thoracic aorta, while it is 15% in the ascending aorta. Therefore, the sizing strategy in this PS-IDE was more generous for nondissection patients (closer to 20%), while in patients with acute aortic dissection or intramural hematoma, the principal investigators aimed for 5% oversizing. The significant differences in the aortic compliance will likely have ramifications in stent graft fixation for devices with poor radial force, although fixation of the Valiant devices using conventional sizing parameters was stable in the midterm follow-up in this study. This PS-IDE required a 1-cm landing zone and may be criticized for such a short landing zone predisposing patients to type 1 endoleaks. The rationale behind this is the limited “real estate” in the ascending aorta. This is in contrast to patients with aortic arch and descending aortic pathology, in whom the landing zone may be moved more proximally to prolong the landing zone. Some of the patients screened in this PS-IDE had a length less than 6 cm between the sinotubular junction and the innominate artery orifice. The traditional 2 cm for a landing zone is great when achievable, but it is unrealistic in many patients in this PS-IDE. Given the lack of severe tortuosity in the ascending aorta and the need for limited landing zone
coverage in nonaneurysmal patients, the principal investigators feel that the 1-cm landing zone may be the solution for most patients. Long-term data are needed to substantiate this claim. The cohort enrolled in this PS-IDE was a heterogeneous patient population including those with acute and chronic type A aortic dissection, penetrating aortic ulcers, and pseudoaneurysms. We believe this is the strength of this PS-IDE, as it will give us insight about TEVAR treatment in various aortic pathologies in long-term follow-up. Although it is difficult to make predictions, the principal investigators feel the “low-hanging fruit” comprises anatomically suitable high-risk patients with ascending aortic pathologies that most cardiac surgeons are hesitant to operate on: patients with acute type A aortic dissections and pseudoaneurysms in a redo setting in those with a high Society of Thoracic Surgeons operative risk score. Ascending TEVAR has a realistic chance to slash the early mortality rate into half for these clinically challenging patients. Although this may represent about 5% to 10% of the current ascending aortic pathology cohort, most aortic specialists believe it can make an impact on a significant number of patients. Deployment in the ascending aorta tends to orient the stent graft obliquely due to tortuosity of the aortic arch. This issue is procedure related rather than device related. The principal investigators have been experiencing this issue also with patients undergoing zone 0 or zone 1
JOURNAL OF VASCULAR SURGERY June 2016
1494 Khoynezhad et al
Fig 12. Subsequent imaging of the pseudoaneurysm has demonstrated stability at the moment, and investigators have elected to observe this for future investigation.
endografting outside of this IDE. However, in patients with a “horizontal aorta” (patients with the horizontal proximal ascending aorta attributable to tortuosity), this malorientation can significantly reduce the accuracy of the deployment and result in a type I endoleak. This was certainly the case for patient No. 2, in whom the investigators were unable to solve the type I endoleak during the index procedure. We believe that performing TEVAR via transapical access (through the apex of the left ventricle) can improve the accuracy of the device deployment in the outer curve of the ascending aorta. This approach has been used extensively in transcatheter heart valve replacement. Transapical access is the shortest route to the ascending aorta, thereby significantly increasing the accuracy of the deployment. More importantly, transapical (antegrade) delivery of the stent graft allows for coaxial placement of the stent graft in the ascending aorta, thereby reducing type I endoleaks. As a feasibility study, this PSIDE has allowed the principal investigator to troubleshoot common issues with the procedure and improve patients’ safety and outcomes for future subjects. There are multiple limitations to this study. First, it represents a highly selected patient population and thereby represents a feasibility study. The principal investigators screened 39 patients and enrolled 6 patients. The results of this study may not be generalizable to those patients
with the combination of ascending and proximal arch or total arch pathologies. Furthermore, new generations of stent grafts are needed for the ascending aorta to meet the hemodynamic and anatomic needs of the ascending aorta. The described design of the Valiant Stent Graft in the ascending aorta represents a significant improvement in our armamentarium, but future improvements adding branched stent grafts for supplemental arch treatment may ease the delivery, deployment, and fixation. The principal investigators were also bound by the FDA to offer a transfemoral approach for the first six patients, some of whom would have benefited from a transapical approach due to a horizontal aorta. CONCLUSIONS Preliminary evaluation of the Valiant PS-IDE at midterm follow-up shows a uniform accuracy of deployment and secure fixation up to 1 year with positive signs of remodeling of the excluded aortic segments. Increased operator experience, better patient selection, and alternative access routes should reduce the endoleak rate in future patients. Transapical access, which has already been granted by the FDA, may reduce complications associated with extremely tortuous and acutely angled aortas that are found in some patients. Long-term results are needed to further substantiate these findings.
JOURNAL OF VASCULAR SURGERY Volume 63, Number 6
AUTHOR CONTRIBUTIONS Conception and design: AK, CD, RW Analysis and interpretation: AK, CD, IW, MK, GK, RW Data collection: AK, CD, IW, MK, GK, RW Writing the article: AK, CD, IW, MK, GK, RW Critical revision of the article: AK, CD, RW Final approval of the article: AK, CD, RW Statistical analysis: AK, RW Obtained funding: Not applicable Overall responsibility: AK
REFERENCES 1. Acierno LJ. The history of cardiology. London and New York: Parthenon Publishing Group, Cambridge University Press (CUP); 1994. 2. Shah A, Bombien R, Khoynezhad A. Thoracic endovascular aortic repair: are we approaching total endovascular solutions for thoracic aortic disease? Multimed Man Cardiothorac Surg 2014;2014. 3. Muehle A, Chou D, Shah A, Khoynezhad A. Experiences with ascending aortic endografts using FDA IDE approved devices. What has been done, what lesions can be treated and what challenges remain. J Cardiovasc Surg (Torino) 2015;56:11-8. 4. Dorros G, Dorros AM, Planton S, O’Hair D, Zayed M. Transseptal guidewire stabilization facilitates stent-graft deployment for persistent proximal ascending aortic dissection. J Endovasc Ther 2000;7: 506-12. 5. Ihnken K, Sze D, Dake MD, Fleischmann D, Van der Starre P, Robbins R. Successful treatment of a Stanford type A dissection by percutaneous placement of a covered stent graft in the ascending aorta. J Thorac Cardiovasc Surg 2004;127:1808-10.
Khoynezhad et al 1495
6. Chan YC, Cheng SW. Endovascular management of Stanford type A (ascending) aortic dissection. Asian Cardiovasc Thorac Ann 2009;17: 566-7. 7. Heye S, Daenens K, Maleux G, Nevelsteen A. Stent-graft repair of a mycotic ascending aortic pseudoaneurysm. J Vasc Interv Radiol 2006;17:1821-5. 8. Hope TA, Markl M, Wigström L, Alley MT, Miller DC, Herfkens RJ. Comparison of flow patterns in ascending aortic aneurysms and volunteers using four-dimensional magnetic resonance velocity mapping. J Magn Reson Imaging 2007;26:1471-9. 9. Kolvenbach RR, Karmeli R, Pinter LS, Zhu Y, Lin F, Wassiljew S, et al. Endovascular management of ascending aortic pathology. J Vasc Surg 2011;53:1431-7. 10. Senay S, Alhan C, Toraman F, Karabulut H, Dagdelen S, Cagil H. Endovascular stent-graft treatment of type A dissection: case report and review of literature. Eur J Vasc Endovasc Surg 2007;34:457-60. 11. Szeto WY, Moser WG, Desai ND, Milewski RK, Cheung AT, Pochettino A, et al. Transapical deployment of endovascular thoracic aortic stent graft for an ascending aortic pseudoaneurysm. Ann Thorac Surg 2010;89:616-8. 12. Zhang H, Li M, Jin W, Wang Z. Endoluminal and surgical treatment for the management of Stanford type A aortic dissection. Eur J Cardiothorac Surg 2004;26:857-9. 13. Zimpfer D, Czerny M, Kettenbach J, Schoder M, Wolner E, Lammer J, et al. Treatment of acute type A dissection by percutaneous endovascular stent-graft placement. Ann Thorac Surg 2006;82:747-9. 14. Lin PH, Kougias P, Huynh TT, Huh J, Coselli JS. Endovascular repair of ascending aortic pseudoaneurysm: technical considerations of a common carotid artery approach using the Zenith aortic cuff endograft. J Endovasc Ther 2007;14:794-8. 15. Muehle A, Shah A, Khoynezhad A. Thoracic endovascular aortic repair in the ascending aorta. Innovations (Phila) 2015;10:363-7. Submitted Sep 30, 2015; accepted Dec 21, 2015.
Feasibility of endovascular repair of ascending aortic pathologies as part of an FDA-approved physician-sponsored investigational device exemption Ali Khoynezhad, MD, PhD,a Carlos E. Donayre, MD,b Irwin Walot, MD,b Matthew C. Koopmann, MD,b George E. Kopchok, BS,c and Rodney A. White, MD,b Los Angeles and Torrance, Calif Objective: Endovascular treatment of ascending aortic lesions has been reported, but to date, no FDA-approved studies have been conducted to define feasibility and the use of endografts in this particular location or to analyze the critical factors involved. Methods: Patients were consented for entry into an FDA-approved physician-sponsored investigational device exemption study to investigate the outcome of those with ascending aortic pathologies. These patients were suitable according to the instructions for use for endovascular repair with a Valiant Captivia (Medtronic, Inc, Minneapolis, Minn) thoracic stent graft, a device designed specifically for deployment in the ascending aorta. All patients had sequential gated-cardiac computed tomography scans, with data being entered into the VQI Complex TEVAR software (West Lebanon, NH). All procedures were performed in a hybrid room, with the capability to convert to an open repair to ensure maximal patient protection. The first five patients constituted the feasibility study, with continued enrollment based on initial results and submission of an annual report to the FDA. Results: Thirty-nine patients were screened, and six patients were entered into the physician-sponsored investigational device exemption study. Although there was no early mortality, there was one late death. All patients had sequential computed tomographies and cardiac echocardiograms with no evidence of migration, one type 1a endoleak, one postoperative stroke, and regression of the aortic lesions in the excluded aortic segment. Conclusions: In this feasibility study, the preliminary evaluation of endovascular treatment for ascending aortic pathologies demonstrates uniform accuracy of deployment and secure fixation up to 17.5 months of follow-up. There is positive remodeling of the excluded aortic segments similar to surveillance studies involving the descending aorta. (J Vasc Surg 2016;63:1483-95.)
Pathologies of the aorta have been associated with very high morbidity and mortality rates and remain the top sixth leading cause of death in elderly men.1 Many patients with aortic issues are extreme high-risk or nonsurgical candidates due to comorbid conditions. Some elderly patients may have concomitant valvular heart From the Division of Cardiothoracic Surgery, Cedars-Sinai Medical Center, Los Angelesa; the Department of Surgery, David Geffen UCLA School of Medicine, University of California, Los Angeles, UCLA School of Medicine,b and Los Angeles Biomedical Research Institute,c Harbor-UCLA Medical Center, Torrance. Clinical trial number NCT02201589. This study does include off-label drugs and/or devices. The specific name of the product discussed is the Medtronic Valiant PS-IDE Stent Graft. Author conflict of interest: A.K., C.E.D, and R.A.W. are consultants for Medtronic Vascular. Presented in Scientific Session III at the Thirtieth Annual Meeting of the Western Vascular Society, Maui, Hawaii, September 21, 2015. Correspondence: Ali Khoynezhad, MD, PhD, FACS, Professor of Cardiovascular Surgery, Director of Aortic Surgery, Cedars-Sinai Medical Center, 127 San Vicente Blvd, Ste #3306, Los Angeles, CA 90048 (e-mail: [email protected]). 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. 0741-5214 Copyright Ó 2016 by the Society for Vascular Surgery. Published by Elsevier Inc. http://dx.doi.org/10.1016/j.jvs.2015.12.029
disease (mostly calcific aortic valve stenosis). Treatment options for the nonsurgical patient population are very limited. Anti-impulsive/antihypertensive therapy and other adjunct pharmacotherapy are the only available options and unfortunately have limited results. There are currently no transcatheter or other less-invasive treatment options for this patient cohort. The success of endovascular stent graft repair of the descending thoracic aorta has led to the application of these devices for the management of thoracic aortic diseases in other locations such as the aortic arch and ascending aorta.2,3 Off-label use of a descending thoracic stent graft in the ascending aorta has been described since 2000 from outside of the United States,4 while Ihnken and colleagues reported the first off-label use in the United States in 2004.5 There are a series of case reports mostly involving high-risk patients with high-risk features or pseudoaneurysms of the ascending aorta undergoing endovascular repair.6-14 The purpose of this study is to investigate the outcome of selected patients with ascending aortic pathologies including type A aortic dissection, retrograde type A aortic dissection, intramural hematoma, penetrating ulcer, or pseudoaneurysm with isolated diseases affecting the aorta between the sinotubular junction and the innominate artery orifice (with no involvement of the aortic valve and the 1483
JOURNAL OF VASCULAR SURGERY June 2016
1484 Khoynezhad et al
Table I. Secondary observations
Table II. Inclusion and exclusion criteria
Early observations d Successful delivery and deployment of the stent graft d Coverage of lesion and/or proximal entry tear d Aortic remodeling based on serial imaging d SAE d Rupture
Inclusion criteria d Ascending thoracic aortic pathology affecting the area between the sinus of Valsalva and the innominate artery orifice (with no involvement of the aortic valve) including:
Late observations d Aortic remodeling at 6- and 12-month visit (defined below) d Secondary procedures within 12 months d Continuing or new FL perfusion d SAEs within 12 months d Rupture within 12 months d All-cause mortality within 12 months Aortic remodeling as measured by Change from baselinea in the maximum TL diameter over the length of the stent graft Change from baselinea in the maximum FL diameter over the length of the stent graft Change from baselinea in the maximum total thoracic aortic diameter FL thrombosis over the length of the stent graft CTA, Computerized tomography angiogram; FL, false lumen; SAE, serious adverse events; TL, true lumen. a The predischarge CTA will be used as baseline to assess aortic remodeling. If discharge occurs after 1-month follow-up, the 1 month follow-up CTA will be used for baseline purposes.
aortic root). These patients underwent thoracic endovascular aortic repair (TEVAR) with the Valiant Stent Graft device (Medtronic, Santa Rosa, Calif) as part of a physician-sponsored investigational device exemption (PSIDE) approved by the FDA. METHODS Study design. This was a prospective trial of the Valiant PS-IDE device (Medtronic) to determine the feasibility of successful implantation, as indicated by exclusion of the thoracic lesion and graft patency at implant, time of discharge, and 1, 6, and 12 months following implantation, and to determine the proportion of patients who die or experience adverse events during and after the implantation. The primary end points were all-cause early mortality, reinterventions, surgical conversions, aortic remodeling (defined as any reduction of aortic aneurysm volume or aortic diameter stabilization or reduction), and postprocedural strokes. Secondary end points are detailed in Table I. The study protocol was approved by the institutional review boards at Harbor-UCLA Medical Center and LA Biomed Research Institute in Torrance, Calif, and Cedars-Sinai Medical Center in Los Angeles, Calif, and was registered at ClinicalTrials.gov (identifier: NCT02201589) with Drs Khoynezhad and White as joint principal investigators. Enrollment. Investigators will assess potential subjects who are diagnosed with type A thoracic aortic dissection, retrograde type A thoracic aortic dissection, intramural hematoma, penetrating ulcer or pseudoaneurysm of the ascending thoracic aorta for their suitability for enrollment in the clinical
Type A thoracic aortic dissection Retrograde type A thoracic aortic dissection B Intramural hematoma B Penetrating ulcer B Pseudoaneurysm d One cm or greater proximal and distal landing zones in the ascending aorta d Ascending aorta diameter between 28-44 mm d American Society of Anesthesiologists Class IV d Candidates for endovascular repair B B
Exclusion criteria d Pregnancy d Age less than 21 years of age d Presence of a condition that threatens to infect the stent graft/aortic valve prosthesis d Allergy to the stent graft material d Inability to obtain informed consent d Expected survival less than one year
study. The patient enrollment goal is 20 patients in five-patient increments. The results of each five-patient group will be analyzed and reported prior to enrolling further patients as a risk-mitigating strategy. Study subject accrual is planned to occur over approximately 5 years at the intervals previously cited and will be under surveillance for life even following study completion. Suitable candidates for inclusion comprised adult male or female patients who agreed to follow-up. In addition, they signed a consent approved by the FDA and institutional review board. To be considered a candidate for endovascular repair, the patient must have had an aforementioned aortic pathology of the ascending thoracic aorta affecting the area between the sinotubular junction and the innominate artery orifice with no involvement of the aortic valve. The patient must also have at least 1 cm of suitable landing zones both proximally and distally to the dissection or diseased area, with an ascending aorta diameter between 28 mm and 44 mm and must a be high-risk surgical candidate with an American Society of Anesthesiologists (ASA) score of IV (Table II). The following patients were excluded: (1) pregnant or pediatric patients (younger than 21 years of age); (2) patients who have a condition that threatens to infect the stent graft/aortic valve prosthesis; (3) patients with allergies to the stent graft material; (4) patients who fail to sign informed consent; and (5) patients with an expected survival less than 1 year. If a patient also has a concomitant descending thoracic aortic dissection or aneurysm that fits the criteria of PS-IDE, the candidate may be offered endovascular therapeutics as part of that PS-IDE in a separate operative setting (Table II). Device and delivery system. The Valiant PS-IDE consists of a modified Valiant Thoracic Stent Graft placed by Medtronic’s Captivia Delivery System and is intended to
JOURNAL OF VASCULAR SURGERY Volume 63, Number 6
be used for the endovascular treatment of pathologies of the ascending aorta. For the purposes of this study, the Valiant Thoracic Stent Graft has been modified from the original configuration to fit the shorter length and wider diameter of the ascending aorta. The modifications to the design are discussed in the information to follow. Two configurations of the Valiant PS-IDE Stent Graft were proposed: (1) a proximal FreeFlo taper and distal closed web and (2) a proximal closed-web design with bare springs distal. To treat the patient population proposed in this study, only one modification was made to the Valiant Stent Graft, consisting of the elimination of stent rings in grafts with shorter lengths. The proximal FreeFlo design was identical to the Valiant Thoracic Stent approved in the U.S. for the descending aorta with the exception of the shorter lengths (5-, 7-, and 9-cm lengths). Fig 1 illustrates the intended device placement. The proximal closed-web configuration with a distal bare spring was available in lengths of 5, 7, and 9 cm. Fig 2 details all device sizes and configurations. The delivery system configuration was also identical to that of the Valiant Captivia Delivery System. A nontip capture device was used for the proximal closed-web design, whereas the tip capture system was used for the FreeFlo configuration. Both delivery systems included hydrophilic coating to aid in delivery. All other aspects of the Valiant Stent Graft remained unchanged, and no changes to the Captivia Delivery System were made. Proposed Valiant Stent Graft sizes included 30-mm to 46-mm diameters. Imaging. All patients had electrocardiogram-gated cardiac, 64-slice computed tomography (CT) scans with SmartPrep (Boston, Mass) and acquisition at end-systole: precontrast CT chest (3-mm cuts), followed by CT angiography neck, chest, abdomen/pelvis (1 mm) (starting images from the base of the skull to the femoral head). A CT with and without contrast and echocardiogram 30 days after the device is implanted, CT with and without contrast at 6 and 12 months postprocedure and once a year from 2 years through 5 years after the implant will be performed. A trained technician generates a 3D model of the aortic root including the ascending aorta, left ventricle, and any thrombotic or calcified plaques present as individual objects from gated, enhanced CT axial slices. Preoperative work-up. The following will take place prior to the implant procedure, which includes (1) signing of the informed consent, (2) collecting demographic data, (3) medical history, (4) vital signs (heart rate, blood pressure, respiratory status), (5) pulse evaluation (left and right radial, dorsalis, and/or tibial artery), (6) pregnancy test (females of child-bearing potential), and other clinical laboratory tests (serum creatinine levels), (7) CT angiography or magnetic resonance angiography of the chest/ abdomen/pelvis with and without contrast, and (8) an echocardiogram. Additional tests may be ordered at the investigator’s discretion (diffusion-weighted magnetic resonance imaging of the head, neuropsychological evaluation, and/or transcranial Doppler). Procedure. The details of the operative procedure have been described in detail in previous publications.2,3,15
Khoynezhad et al 1485
Fig 1. Proposed Valiant PS-IDE Stent Graft placement in the ascending aorta.
In brief, bilateral surgical femoral arterial and venous access is provided. The arterial system is used for angiogram and intravascular ultrasound (IVUS) evaluation, while the contralateral side is used for device delivery. The venous access sites are used for ventricular pacing and heart-lung machine access in case of emergency. After initial triple imaging (angiogram, IVUS, and echocardiogram), the proper device is selected and deployed under ventricular pacing in the desired anatomic position. Postoperative triple imaging confirms the proper position of the device and lack of endoleak. Subsequently, the catheters are removed, and the access vessels are repaired. RESULTS Primary and secondary end points. In the last 2 years, 39 patients were screened, and 6 patients were entered in the PS-IDE and followed up for an average of 17.5 months (range, 10-23 months). Reasons for screening failure were lack of minimal proximal or distal landing zone in 23, poor access vessels in 7, and aortic rupture/lost to followup in 3 patients. Among the patients entered in the study, there were no intraoperative or early ruptures or early conversions to open ascending repair. Furthermore, there were
JOURNAL OF VASCULAR SURGERY June 2016
1486 Khoynezhad et al
Fig 2. Size representation of the Valiant PS-IDE Stent Graft.
no intraoperative or early mortality. Patient No. 3 required a total arch replacement and frozen elephant repair due to de novo ulceration in the mid-aortic arch. This patient died 16 weeks after the index TEVAR due to unrelated causes. Successful delivery and deployment of the Valiant PS-IDE ascending stent graft was achieved in all subjects. Patient No. 2, who had a history of previous stroke, suffered a periprocedural lacunar infarct, from which he recovered promptly with no clinically evident residuals. The same patient had also a type Ia endoleak due to incomplete coverage of the aortic lesion proximally that remains stable on surveillance imaging. Similarly, patient No. 6 who has thrombosed most of the pseudoaneurysm, has a small residual perfusion (type Ia), which so far has been stable as shown in surveillance
imaging. Table III is a summary of patient indications of operation, complications, and residual patient deficits. Table IV is an overview of volumetric data acquired via 3D reconstruction by M2S in the surveillance imaging, summarizing three variables: (1) ascending maximum sac diameter, (2) volume of the ascending aorta from the sinotubular junction to the origin of the innominate artery, and (3) the angle between a centerline of the ascending aorta and the intersection of a line from the apex of the left ventricle to the center of the aortic valve. Table IV displays the same parameters for enrolled patients. Review of patients and technical observations at follow-up. Patient No. 1, a 77-year-old male with a history of atrial fibrillation, pacemaker implantation,
JOURNAL OF VASCULAR SURGERY Volume 63, Number 6
Khoynezhad et al 1487
Table III. Summary of indications of operation, endografts deployed, device proximal and distal diameters and length, complications and residual patient effects Patient, No.
Indication
No. Endo-grafts
Device size, mm
Complication(s)
Residual effects
1 2
Penetrating ulcer Pseudoaneurysm
1 1
38 x 38 x 60 L 36 x 36 x 80 L
3
Type A dissection based on penetrating aortic ulcer Chronic type A dissection Pseudoaneurysm Pseudoaneurysm
1
40 x 40 x 60 L
2 1 1
46 x 46 x 80 L 46 x 46 x 40 L Aortic valve leaflet entrapment None 34 x 34 x 60 L None None 40 x 40 x 60 L None Proximal pseudoaneurysm
4 5 6
Wire perforation left ventricle None Lacunar infarct, type 1 Mild residual endoleak symptoms De novo aortic ulcer requiring Late death frozen elephant trunk repair
Table IV. Overview of volumetric data
Patient No.1 Ascending maximum sac Volume Angle Patient No. 2 Ascending maximum sac Volume Angle Patient No. 3 Ascending maximum sac Volume Angle Patient No. 4 Ascending maximum sac Volume Angle Patient No. 5 Ascending maximum sac Volume Angle Patient No. 6 Maximum size diameter Volume Angle
Baseline
1 Month
6 Months
1 Year
37.2 60.5 150
39.8 70.1 144
39.8 76 145.3
40.4 71 143.8
54.7 101.1 145.8
54.4 113.3 132.8
diameter, mm
46 105.8 147.5
47.4 111 143
50.8 110.6 132.6 At 2 months 54.9 NA 143.2
diameter, mm
57.6 188.3 124
57.3 186.1 141.6
58.2 196.1 155.8
37.4 63.7 125
37.6 60.6 129.4
37.5 63.2 138.3
X
94.1 447.9 133.3
90.7 402.3 131.4
93.5 404.2 125.8
X
diameter, mm
diameter, mm
diameter, mm
peripheral vascular disease, zone II TEVAR for symptomatic penetrating aortic ulcer and intramural hematoma, underwent ascending stent grafting in November 2013. He had an intraoperative pericardial effusion with no sign of cardiac tamponade detected by transesophageal echocardiogram. This pericardial effusion was stable after reversal of heparin, followed by transthoracic echo, and was found to regress by postoperative day 7. The investigators opted to use an Amplatz Extra-Stiff Wire guide with a 3-mm radius tip for future subjects instead of the Meier wire that has a shorter flexible tip. This strategy would reduce the risk for future left ventricular injury that may have been the origin of this patient’s pericardial effusion. The 1-year CT scan of patient No. 1 is shown in Fig 3. The ascending aortic penetrating ulcer is excluded with
Decrease
Stable
Increase
X X X X X X
58.7 198.7 NA
X X X
X
firm fixation of the device in the deployed position. He is doing clinically well with no adverse events as seen on follow-up examination. Patient No. 2 is an 84-year-old female with prohibitive surgical risk including previous stroke, history of coronary artery bypass grafting and mitral valve repair, and postoperative sternal wound infection. She presented with an enlarging pseudoaneurysm at the surgical cannulation site and underwent ascending TEVAR. Although preoperative interrogation revealed a proper landing zone, the stent graft was not able to completely seal proximally. The patient’s composite serial CTs (Fig 4) are consistent with a type Ia endoleak. The inadequate seal in this patient was due to horizontal aortic anatomy (significant angulation of the centerline ascending aorta to the long axis of the
1488 Khoynezhad et al
ventricle) that did not allow complete coaxial placement of the stent graft at the sinotubular junction of the ascending aorta (Fig 5). Given the proximity of the left main coronary artery orifice, the investigators decided not to place another device. Patients with a horizontal aorta may best be treated by a transapical approach for the same deployment advantages that have been obtained with transapical aortic valve replacements. However, the investigators did not have FDA approval for such an approach in this patient. This patient was extubated and ambulatory within 24 hours. She had a postoperative stroke while talking to her family that resulted in a left lacunar infarct. This left her with right upper extremity weakness and significant speech and swallowing difficulty. Her symptoms improved significantly with physical and speech therapy. At this point, the investigators opted to observe the pseudoaneurysm. M2S calculations of the ascending maximum sac diameter (52 mm on the pretreatment 9/19/2013 CT, 54.7 at 1 day postoperative, and 50.8 on the 4/8/14 postprocedure CT) showed a decrease in this parameter. The volume was slightly higher (106.8 cc at pretreatment, 101.1 cc at 1 day posttreatment, and 110.6 cc at the 19-week postprocedure CT). Patient No. 3 is an 88-year-old female with a history of hypertension and deep vein thrombosis after hip replacement, who presented with a DeBakey type I aortic dissection based on a penetrating ulcer of the ascending aorta. She also had some intramural hematoma that went down to the thoracoabdominal aorta. The dissection did not surpass the sinotubular junction; therefore, she was offered an ascending TEVAR as part of the PS-IDE. In order to have a proper distal landing zone according to the PS-IDE protocol, a retroesophageal carotid-carotid bypass was performed. IVUS interrogation and aortogram postdeployment showed excellent stent graft vessel apposition with exclusion of the aortic ulcer (Fig 6). The patient did not require coverage of the innominate artery, and therefore, the carotid-carotid bypass eventually thrombosed due to competitive flow. She had postoperative swallowing issues but was cleared for regular diet. Due to inadequate intake, she ultimately required a percutaneous gastrostomy tube placement and was discharged. Approximately 7 weeks after the index procedure, the patient presented to another facility with complaints of shortness of breath and interscapular pain. A CT of the chest, abdomen, and pelvis revealed a new penetrating near-perforated mid-aortic arch ulcer and an enlarged intramural hematoma in the descending thoracic aorta (Figs 7 and 8). Patient No. 3 was transferred to CedarsSinai Medical Center, underwent a total arch replacement and a stent graft implantation of the descending thoracic aorta (frozen elephant trunk repair), and had a replacement of her ascending aorta with an aortic valve resuspension. She was extubated the next day and was transferred to floor on postoperative day 3. Her postsurgical course was complicated atrial fibrillation and bilateral acute thrombus in the jugular, subclavian, innominate, and axillary veins requiring anticoagulation with warfarin.
JOURNAL OF VASCULAR SURGERY June 2016
Fig 3. Patient No. 1 at 1-year follow-up, showing the device without migration or endoleak.
The patient was discharged to a rehabilitation center on postoperative day 18. She died at home 16 weeks past the index operation from gastrointestinal bleeding while being on warfarin. Patient No. 4 is an 84-year-old male with chronic localized type A aortic dissection that may have been due to his previous open-heart operation. Upon deployment of the stent graft, there was evidence of aortic regurgitation followed by hypotension. An angiogram confirmed the entrapment of the aortic cusp resulting in aortic insufficiency. Prompt distal repositioning of the stent graft using a compliant balloon normalized the aortic valve function assessed by transesophageal echocardiogram and the hemodynamics. Imaging for patient No. 4 showed firm fixation of the device and no endoleaks (Fig 9). The patient did very well postoperatively without any complications and was sent home in stable condition on postoperative day 4. Patient No. 5 was a 67-year-old male with a history of TEVAR of the descending thoracic aorta for aneurysmal pathology and an ascending pseudoaneurysm. He underwent uneventful TEVAR and had an uncomplicated postoperative course (Fig 10). He was discharged home on postoperative day 2 and had done well in follow-up.
JOURNAL OF VASCULAR SURGERY Volume 63, Number 6
Khoynezhad et al 1489
Fig 4. Composite serial computed tomography (CT) postprocedure (patient No. 2), showing a consistent proximal type 1 endoleak (arrow), with no enlargement of the root over the first year.
Fig 5. Conformity of the device to the aortic curve (patient No. 2).
Patient No. 6 is a 52-year-old male with a complicated history including ascending aortic repair for reported syphilitic aortitis in 2010. He has subsequently undergone two additional reinterventions for aortic root replacement. The last operation included placement of a mechanical heart
valve. At the time of entry into the study, the patient was found to have a large anterior mediastinal pseudoaneurysm at the Dacron graft of the ascending aorta. The vascular prosthesis was found to be very mobile along its entire length on IVUS with near collapse of the graft during
1490 Khoynezhad et al
JOURNAL OF VASCULAR SURGERY June 2016
Fig 6. Ascending aortogram of patient No. 3 prior to (left) and at follow-up (right), showing exclusion of the entry site of the intramural hematoma.
Fig 7. Computed tomography (CT) angiogram at 7 weeks (patient No. 3) showing stability of the ascending graft with exclusion of the original penetrating ulcer (left, arrow), but with progression of the intramural hematoma in the descending thoracic aorta with suspected contained rupture (right, arrows).
JOURNAL OF VASCULAR SURGERY Volume 63, Number 6
Khoynezhad et al 1491
Fig 8. An expanding intramural hematoma in the arch (patient No. 3) identified as endoleak (arrow) on an official read. This has otherwise been interpreted as expansion of intramural hematoma outside the device.
diastole. Following successful identification of the graft dehiscence site and prior to the endograft exclusion, a carotid-carotid bypass was performed to provide additional landings during the endograft deployment as needed. Fig 11 demonstrates the thrombosis of the pseudoaneurysm over the first 6 months with development of a pseudoaneurysm at the distal attachment site once the ascending graft stabilized and the ascending portion of the aneurysm reformed. Subsequent imaging of the pseudoaneurysm has demonstrated it is currently stable (Fig 12), and the investigators plan to intervene and cover this segment in the near future. DISCUSSION Safe and effective implementation of TEVAR in the treatment of ascending aortic pathologies remains one of the last bastions for less invasive and endovascular therapy of the aorta.2,3 The alternative strategy in patients entered in this PS-IDE would have been a (redo) sternotomy, open-heart surgery with hypothermic circulatory arrest to replace the entire ascending aorta with or without a
hemi-arch replacement. Furthermore, all these patients would not have been candidates for a zone 0 debranching operation, given the fact that there was no portion of the ascending aorta to perform the ascending innominate bypass. Therefore, most aortic specialists believe that ascending TEVAR may represent a disruptive technology in high-risk patients, as it fulfills an unmet need. Stenting of the ascending aorta presents unique challenges, including negotiating the curvature of the aortic arch, proximal fixation close to the aortic valve and coronary ostia, distal fixation that may impinge on the innominate artery, hemodynamic forces, the risk of cardio-aortic injury, and retrograde dissection. Some of the major complications of this procedure are novel to the endovascular specialist experienced in aortic endografting and include perforation of the left ventricle, injury and dissection of the aortic root, and occlusion of the coronary arteries.2,3 Furthermore, the ascending aorta is on the average a centimeter larger than the descending thoracic aorta and about one-fourth in length. These physiologic and anatomic characteristics of the ascending aorta have ramifications in stent
1492 Khoynezhad et al
JOURNAL OF VASCULAR SURGERY June 2016
Fig 9. Composite serial computed tomography (CT) postprocedure (patient No. 4) showing firm fixation of the device and no endoleaks.
Fig 10. Exclusion of an aortic pseudoaneurysm (patient No. 5); 6-month computed tomography (CT) shows regression of lesion (arrows).
JOURNAL OF VASCULAR SURGERY Volume 63, Number 6
Khoynezhad et al 1493
Fig 11. Imaging of patient No. 6 demonstrates thrombosis of the pseudoaneurysm over the first 6 months postindex procedure, with development of a pseudoaneurysm at the distal attachment site once the ascending graft stabilized and the ascending portion of the aneurysm reformed.
graft design that should entail shorter devices with a large diameter for proximal and distal landing zone fixation. In addition, there is a large variation in aortic diameter between the cardiac cycles in the ascending aorta. Based on 4-D evaluation of our patients, we found that the variation in size is about 10% in the descending thoracic aorta, while it is 15% in the ascending aorta. Therefore, the sizing strategy in this PS-IDE was more generous for nondissection patients (closer to 20%), while in patients with acute aortic dissection or intramural hematoma, the principal investigators aimed for 5% oversizing. The significant differences in the aortic compliance will likely have ramifications in stent graft fixation for devices with poor radial force, although fixation of the Valiant devices using conventional sizing parameters was stable in the midterm follow-up in this study. This PS-IDE required a 1-cm landing zone and may be criticized for such a short landing zone predisposing patients to type 1 endoleaks. The rationale behind this is the limited “real estate” in the ascending aorta. This is in contrast to patients with aortic arch and descending aortic pathology, in whom the landing zone may be moved more proximally to prolong the landing zone. Some of the patients screened in this PS-IDE had a length less than 6 cm between the sinotubular junction and the innominate artery orifice. The traditional 2 cm for a landing zone is great when achievable, but it is unrealistic in many patients in this PS-IDE. Given the lack of severe tortuosity in the ascending aorta and the need for limited landing zone
coverage in nonaneurysmal patients, the principal investigators feel that the 1-cm landing zone may be the solution for most patients. Long-term data are needed to substantiate this claim. The cohort enrolled in this PS-IDE was a heterogeneous patient population including those with acute and chronic type A aortic dissection, penetrating aortic ulcers, and pseudoaneurysms. We believe this is the strength of this PS-IDE, as it will give us insight about TEVAR treatment in various aortic pathologies in long-term follow-up. Although it is difficult to make predictions, the principal investigators feel the “low-hanging fruit” comprises anatomically suitable high-risk patients with ascending aortic pathologies that most cardiac surgeons are hesitant to operate on: patients with acute type A aortic dissections and pseudoaneurysms in a redo setting in those with a high Society of Thoracic Surgeons operative risk score. Ascending TEVAR has a realistic chance to slash the early mortality rate into half for these clinically challenging patients. Although this may represent about 5% to 10% of the current ascending aortic pathology cohort, most aortic specialists believe it can make an impact on a significant number of patients. Deployment in the ascending aorta tends to orient the stent graft obliquely due to tortuosity of the aortic arch. This issue is procedure related rather than device related. The principal investigators have been experiencing this issue also with patients undergoing zone 0 or zone 1
JOURNAL OF VASCULAR SURGERY June 2016
1494 Khoynezhad et al
Fig 12. Subsequent imaging of the pseudoaneurysm has demonstrated stability at the moment, and investigators have elected to observe this for future investigation.
endografting outside of this IDE. However, in patients with a “horizontal aorta” (patients with the horizontal proximal ascending aorta attributable to tortuosity), this malorientation can significantly reduce the accuracy of the deployment and result in a type I endoleak. This was certainly the case for patient No. 2, in whom the investigators were unable to solve the type I endoleak during the index procedure. We believe that performing TEVAR via transapical access (through the apex of the left ventricle) can improve the accuracy of the device deployment in the outer curve of the ascending aorta. This approach has been used extensively in transcatheter heart valve replacement. Transapical access is the shortest route to the ascending aorta, thereby significantly increasing the accuracy of the deployment. More importantly, transapical (antegrade) delivery of the stent graft allows for coaxial placement of the stent graft in the ascending aorta, thereby reducing type I endoleaks. As a feasibility study, this PSIDE has allowed the principal investigator to troubleshoot common issues with the procedure and improve patients’ safety and outcomes for future subjects. There are multiple limitations to this study. First, it represents a highly selected patient population and thereby represents a feasibility study. The principal investigators screened 39 patients and enrolled 6 patients. The results of this study may not be generalizable to those patients
with the combination of ascending and proximal arch or total arch pathologies. Furthermore, new generations of stent grafts are needed for the ascending aorta to meet the hemodynamic and anatomic needs of the ascending aorta. The described design of the Valiant Stent Graft in the ascending aorta represents a significant improvement in our armamentarium, but future improvements adding branched stent grafts for supplemental arch treatment may ease the delivery, deployment, and fixation. The principal investigators were also bound by the FDA to offer a transfemoral approach for the first six patients, some of whom would have benefited from a transapical approach due to a horizontal aorta. CONCLUSIONS Preliminary evaluation of the Valiant PS-IDE at midterm follow-up shows a uniform accuracy of deployment and secure fixation up to 1 year with positive signs of remodeling of the excluded aortic segments. Increased operator experience, better patient selection, and alternative access routes should reduce the endoleak rate in future patients. Transapical access, which has already been granted by the FDA, may reduce complications associated with extremely tortuous and acutely angled aortas that are found in some patients. Long-term results are needed to further substantiate these findings.
JOURNAL OF VASCULAR SURGERY Volume 63, Number 6
AUTHOR CONTRIBUTIONS Conception and design: AK, CD, RW Analysis and interpretation: AK, CD, IW, MK, GK, RW Data collection: AK, CD, IW, MK, GK, RW Writing the article: AK, CD, IW, MK, GK, RW Critical revision of the article: AK, CD, RW Final approval of the article: AK, CD, RW Statistical analysis: AK, RW Obtained funding: Not applicable Overall responsibility: AK
REFERENCES 1. Acierno LJ. The history of cardiology. London and New York: Parthenon Publishing Group, Cambridge University Press (CUP); 1994. 2. Shah A, Bombien R, Khoynezhad A. Thoracic endovascular aortic repair: are we approaching total endovascular solutions for thoracic aortic disease? Multimed Man Cardiothorac Surg 2014;2014. 3. Muehle A, Chou D, Shah A, Khoynezhad A. Experiences with ascending aortic endografts using FDA IDE approved devices. What has been done, what lesions can be treated and what challenges remain. J Cardiovasc Surg (Torino) 2015;56:11-8. 4. Dorros G, Dorros AM, Planton S, O’Hair D, Zayed M. Transseptal guidewire stabilization facilitates stent-graft deployment for persistent proximal ascending aortic dissection. J Endovasc Ther 2000;7: 506-12. 5. Ihnken K, Sze D, Dake MD, Fleischmann D, Van der Starre P, Robbins R. Successful treatment of a Stanford type A dissection by percutaneous placement of a covered stent graft in the ascending aorta. J Thorac Cardiovasc Surg 2004;127:1808-10.
Khoynezhad et al 1495
6. Chan YC, Cheng SW. Endovascular management of Stanford type A (ascending) aortic dissection. Asian Cardiovasc Thorac Ann 2009;17: 566-7. 7. Heye S, Daenens K, Maleux G, Nevelsteen A. Stent-graft repair of a mycotic ascending aortic pseudoaneurysm. J Vasc Interv Radiol 2006;17:1821-5. 8. Hope TA, Markl M, Wigström L, Alley MT, Miller DC, Herfkens RJ. Comparison of flow patterns in ascending aortic aneurysms and volunteers using four-dimensional magnetic resonance velocity mapping. J Magn Reson Imaging 2007;26:1471-9. 9. Kolvenbach RR, Karmeli R, Pinter LS, Zhu Y, Lin F, Wassiljew S, et al. Endovascular management of ascending aortic pathology. J Vasc Surg 2011;53:1431-7. 10. Senay S, Alhan C, Toraman F, Karabulut H, Dagdelen S, Cagil H. Endovascular stent-graft treatment of type A dissection: case report and review of literature. Eur J Vasc Endovasc Surg 2007;34:457-60. 11. Szeto WY, Moser WG, Desai ND, Milewski RK, Cheung AT, Pochettino A, et al. Transapical deployment of endovascular thoracic aortic stent graft for an ascending aortic pseudoaneurysm. Ann Thorac Surg 2010;89:616-8. 12. Zhang H, Li M, Jin W, Wang Z. Endoluminal and surgical treatment for the management of Stanford type A aortic dissection. Eur J Cardiothorac Surg 2004;26:857-9. 13. Zimpfer D, Czerny M, Kettenbach J, Schoder M, Wolner E, Lammer J, et al. Treatment of acute type A dissection by percutaneous endovascular stent-graft placement. Ann Thorac Surg 2006;82:747-9. 14. Lin PH, Kougias P, Huynh TT, Huh J, Coselli JS. Endovascular repair of ascending aortic pseudoaneurysm: technical considerations of a common carotid artery approach using the Zenith aortic cuff endograft. J Endovasc Ther 2007;14:794-8. 15. Muehle A, Shah A, Khoynezhad A. Thoracic endovascular aortic repair in the ascending aorta. Innovations (Phila) 2015;10:363-7. Submitted Sep 30, 2015; accepted Dec 21, 2015.
