Endovascular treatment of thoracoabdominal aortic aneurysm using physician-modified endografts

From the Western Vascular Society

Endovascular treatment of thoracoabdominal aortic aneurysm using physician-modified endografts Matthew P. Sweet, MD, Benjamin W. Starnes, MD, and Billi Tatum, RN, CRCC, Seattle, Wash Objective: To report an initial experience with physician-modified thoracic endografts for endovascular treatment of thoracoabdominal aortic aneurysm (TAAA). Methods: Single-center cohort study of the treatment of TAAA using a physician-modified fenestrated thoracic endograft for patients deemed to be at high risk of open repair. The cohort includes 21 patients in a prospective physician-sponsored U.S. Food and Drug Administration-approved investigational device exemption study and three patients treated outside the investigational device exemption. The procedure involves physician modification of a Cook TX2 thoracic stent graft with reinforced fenestrations. Branch stents were iCast balloon expandable stents. Treatment success was defined as successful aneurysm exclusion with freedom from permanent organ system dysfunction and return to preoperative level of independent functional status. Results: Twenty-four consecutive patients were treated. Twenty-one patients (88%) met the endpoint of treatment success at a mean of 11 months follow-up. One patient (4%) died within 30 days due to complications of spinal cord injury (SCI). One patient (4%) died 4 months postoperatively after a prolonged recovery from surgery. One other patient (4%) is alive 13 months after operation with permanent SCI. One renal reintervention has been required. No device failures have occurred. Conclusions: Early-term data suggest that physician-modified fenestrated thoracic endografts can be used to safely and effectively treat TAAA in patients at high risk of open repair. Physician-modified devices perform similarly to commercially manufactured grafts in terms of treatment success, SCI, perioperative death, and clinical outcome at short-term follow-up. Physician modification is immediately available and allows for a high level of customizability. Procedure success is contingent upon careful preoperative planning, patient selection, experienced providers, and a high volume center. (J Vasc Surg 2015;62:1160-7.)

Thoracoabdominal aortic aneurysms (TAAAs) are a complex management challenge. A decision about operative repair is a balance between the risk of surgery and the risk of aneurysm rupture. Open aortic replacement has served as the standard of care and has undergone major improvements over the last 20 years.1 Centers of excellence have reported good results in retrospective, unmonitored case series.1-3 Larger population-based datasets, however, reveal that those results are not seen outside such centers.4,5 Thus far, survival has been used as the primary outcome measure of these case series. Few data are available about postoperative functional status and disability resulting from the operative repair. Significant morbidity

From the Division of Vascular Surgery, Department of Surgery, University of Washington. Author conflict of interest: M.P.S. has received sponsored travel by Cook Medical. B.W.S. has been a paid consultant for Cook Medical. Presented at the plenary session of the Twenty-ninth Annual Meeting of the Western Vascular Society, Coronado, Calif, September 20-23, 2014. Correspondence: Matthew P. Sweet, MD, Department of Surgery, University of Washington Medical Center, 1959 NE Pacific St, Box 356410, Seattle, WA 98195 (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 2015 by the Society for Vascular Surgery. Published by Elsevier Inc. http://dx.doi.org/10.1016/j.jvs.2015.05.036

1160

is common following open repair, with only 52% of patients having a “good” outcome at 1 year in one large study from a center of excellence.6 Given the morbidity of open repair, many patients are deemed ineligible for repair and are left untreated.7 Endovascular repair of infrarenal and thoracic aortic aneurysms carries a lower risk of perioperative death and major morbidity compared with open repair.8,9 That benefit comes with less definitive long-term effectiveness and a higher incidence of aortic reintervention. It is hoped that endovascular therapy of TAAA may yield effective aneurysm exclusion with preservation of functional status, thereby preserving quality as well as duration of life. Furthermore, less invasive therapy may make more patients eligible for treatment. Several centers in the U.S. and around the world have demonstrated excellent results using a commercially manufactured stent graft in both custom and off-the-shelf designs in patients deemed to be at high risk of open repair.10-12 The devices used in these reports are not widely available in the U.S., and as such, physicians have sought alternative means for endovascular treatment of TAAA. One such technique utilizes individual surgeon modification of commercial endografts, called fenestratedbranched devices, to treat these aneurysms.13-16 Similar techniques have been used to treat juxta-renal aneurysms with excellent results.13,16,17 Few data are available for physician-modified endografts for TAAA, as these procedures have been done mostly outside of U.S. Food and

JOURNAL OF VASCULAR SURGERY Volume 62, Number 5

Sweet et al 1161

Drug Administration (FDA)-approved studies and have gone unreported. Uncertainty remains about the safety, effectiveness, and durability of this therapy. This study reports the initial consecutive case series of physician-modified endovascular grafts for treatment of TAAA done under a FDA-approved physician-sponsored investigational device exemption (IDE). METHODS Study design. This initial single-center, nonrandomized, prospective clinical trial was conducted under a physician-sponsored IDE-approved by the FDA and the University of Washington Human Subjects Division (study name: B-TEVAR IDE). The study is registered with www. clinicaltrials.gov. Given the experimental nature of this technique, study approval was limited to high-risk patients. All subjects were deemed to be at high risk of open repair in the judgment of the principle investigator. Furthermore, all subjects were evaluated by a second physician to confirm that they were at high risk of open repair. The second physician review was performed by a staff anesthesiologist or surgeon in the Cardiothoracic Intensive Care Unit. High-risk status was based on both anatomic and physiologic criteria. Anatomic criteria included reoperative aortic surgery, other prior nonaortic abdominal or thoracic surgery, and obesity. Physiologic criteria included advanced age, limited preoperative functional status, any advanced organ system dysfunction, or a clinical assessment of frailty. No prespecified criteria were used to define high risk. Rather, a subjective assessment of these various factors was used in each individual patient by physicians experienced in the care of these patients. Anatomic study inclusion criteria include: TAAA size >5.5 cm, and/or symptomatic, and/or rapidly expanding; anatomy suitable for the device with #5 target visceral vessels (mesenteric and/or renal) measuring 4 to 10 mm in diameter; proximal seal/fixation of a minimum 2.5 cm of healthy thoracic aorta; and suitable distal seal in the infrarenal aorta or iliac vessels. Exclusion criteria include: free rupture with hemodynamic instability; ongoing infection; connective tissue disease diagnosis; and life expectancy of <1 year despite successful aneurysm exclusion. All patients provided informed written consent to participate and understood the study involved off-label physician modification of the endograft. Study follow-up consisted of physical examination and laboratory assessment of renal function as well as computed tomography (CT) imaging. Visits occurred at 1, 6, and 12 months postoperatively and annually thereafter. Postoperative imaging consists of a contrast-enhanced arterialphase CT angiogram with fine cuts performed on the same schedule. For subjects with compromised renal function, a noncontrast CT scan with fine cuts as well as aortic and branch duplex was used. Imaging review was performed by the principle investigator. A Clinical Events Committee comprised of physicians uninvolved in the

Fig 1. Intraoperative photograph of the completed physicianmodified endograft showing the reinforced fenestrations and diameter reducing ties. B-TEVAR, Branched thoracic endovascular aortic repair.

study adjudicated all major adverse events. Data was audited for accuracy and completeness by an external reviewer on a monthly basis. Patient population. Between November 2011 and April 2015, 24 patients with TAAA underwent endovascular repair using the physician-modified TX2 endograft at the University of Washington. Twenty-one patients were treated as a part of the B-TEVAR IDE study. Three patients treated outside the IDE study presented with symptomatic aneurysms. Two patients were treated prior to approval of the IDE. One patient was treated on a compassionate use basis concurrent with the IDE as he did not meet study inclusion criteria. Specifically, the patient had advanced heart failure and chronic renal insufficiency and presented with a symptomatic TAAA and was not thought to have a >1 year life expectancy. The same technique and follow-up protocol was used for the patients treated within and outside the IDE. All patients were started on aspirin and statin medications unless contraindicated. Devices. The physician-modified graft utilizes Cook TX2 (Cook Medical, Bloomington, Ind) devices. Detailed preoperative anatomic planning is done according to a prescribed process using a 3D workstation (TeraRecon Inc, Foster City, Calif). The device is unsheathed on a back table under sterile conditions. Reinforced fenestrations are created with Atrium SST PTFE (Atrium Medical Corp, Hudson, NH) and Fibered Platinum coils (Boston Scientific, Marlborough, Mass). Permanent and temporary diameter reducing ties are created using Gore (W. L. Gore Inc, Flagstaff, Ariz) and Chromic Gut sutures, respectively (Fig 1). Once the modifications are complete, the device is reconstrained in its original delivery sheath. Graft modification takes 2 hours and is completed before the patient is put under anesthesia. The modified device consists of three segments: the proximal fixation/seal zone, the peri-visceral fenestrated segment, and the distal seal/fixation zone. The proximal and distal segments are constrained with temporary reducing ties to allow for device manipulation for

1162 Sweet et al

JOURNAL OF VASCULAR SURGERY November 2015

Fig 2. A, Completion 3D volume-rendered images from a postoperative computed tomography (CT) angiogram. This patient had a prior thoracic endovascular aortic repair (TEVAR) with celiac embolization and coverage that developed a type Ib endoleak and was treated with a three-vessel branched thoracic endovascular aortic repair (B-TEVAR) device. B, Completion 3D volume rendered CT images. This patient had staged repair of an extent I thoracoabdominal aortic aneurysm (TAAA). C, Sagittal images from before and after B-TEVAR procedure showing an in-situ infra-renal Cook Zenith Flex device with supra-renal struts. The second images shows the completion scan after B-TEVAR with the modified device sitting within the main body of the Zenith Flex device.

fenestration alignment. The perivisceral fenestrated segment is constrained with permanent diameter reducing ties when needed to narrow the graft several mm smaller than the aortic flow channel. This device construct has been narrowed as small as 16 mm in diameter at the level of open fenestrations. The perivisceral segment of the device is designed not to have wall apposition with the aorta. Deliberate space (offset) is created between the fenestrated stent graft and the visceral vessels to facilitate target vessel cannulation and allow for ongoing target vessel perfusion during device deployment. Given this offset, aneurysm exclusion and device integrity are contingent upon sealing between the fenestrated aortic device and the covered branch iCast stent grafts (Atrium Medical Corp, Hudson, NH). Additional, unmodified, aortic devices are used to establish proximal and distal fixation and seal as needed.

The unmodified devices could include one or more of the following: Cook Zenith TX2, Cook Zenith Flex, and Endologix AFX (Endologix, Inc, Irvine, Calif). A complete technical description of the procedure is beyond the scope of this manuscript. Briefly, the proximal and distal fixation is achieved with unmodified devices as needed. The modified device is then partially deployed with temporary and permanent diameter-reducing ties to allow for device rotation and fenestration alignment. Once the device is aligned and partially deployed, the superior mesenteric artery (SMA) is cannulated through its fenestration, and the device is completely deployed from its delivery system. The SMA stent is then deployed, the other target vessels are then selectively catheterized, and the branch stents are deployed one at a time. Initially, selection of two target vessels (SMA and either renal) through the fenestrations was performed prior to releasing

JOURNAL OF VASCULAR SURGERY Volume 62, Number 5

the device. With improved device offset and procedural experience, the SMA is now the only vessel routinely selected prior to device deployment. The temporary ties are released after adjacent fenestrations are selected. The branches consist of iCast balloon-expandable covered stents. In some cases, a self-expanding uncovered Zilver stent (Cook Medical) was deployed to smooth the transition into a tortuous target vessel. All procedures were conducted from femoral access. Brachial access is allowed within the IDE. Femoral access is preferred by the principle investigator as it avoids instrumentation of the arch and for ease of branch cannulation (Fig 2). Definititions. TAAAs were classified according to the Crawford classification as modified by Safi.18 Spinal cord injuries (SCIs) were classified according a standardized scoring system.10 The primary study outcome was the percentage of patients achieving successful treatment. This was defined as a composite endpoint of successful aneurysm exclusion (absence of type I or III endoleak) with return to preoperative functional state and freedom from SCI or any major organ system failure, aneurysm rupture, or death. Branch loss, if not resulting in other patient-related complication, was not considered a treatment failure. Change in renal function was defined as a drop in estimated glomerular filtration resulting in a change of chronic kidney disease (CKD) classification as defined by the National Kidney Foundation.19 RESULTS A total of 24 patients (18 men and 6 women) with TAAA were treated with fenestrated-branched physicianmodified endovascular stent grafts. A total of 88 branches were created. Demographics, medical comorbidities, and operative details are reported in Tables I-III. Complete aneurysm exclusion was achieved in all patients (100%). Twenty-one patients (88%) met criteria for the composite endpoint of successful treatment: return to preoperative functional state and complete aneurysm exclusion at mean 11 months (range, 1-41 months) follow-up. Two patients (8%) died. One patient (4%) died on the 22nd postoperative day due to complications of spinal cord ischemia as described below. Another patient (4%) died 4 months after repair. Her operation was complicated by a subarachnoid hemorrhage due to cerebrospinal fluid drainage, severe postimplantation syndrome with multisystem organ failure, and temporary dialysis. She recovered, went home in frail condition, and developed bacterial endocarditis due to a toe infection 4 months postoperation. One other patient (4%) is alive 13 months postprocedure with aneurysm exclusion but has permanent SCI. There were no cases of stroke or limb ischemia. No patients have been lost to follow-up. High fluoroscopy doses were frequently required, but these were given in multiple different views, and no skin injury has occurred for any patient. Eleven patients (46%) were treated with the B-TEVAR device in a single operative setting. The other 13 patients (54%) had more than one surgical procedure performed

Sweet et al 1163

Table I. Demographics Age, years Men ASA class 3 4 TAAA extent I II III IV V Diameter, cm Symptomatic Prior aortic operation

73 (59-88) 18 (75) 18 (75) 6 (25) 4 5 7 7 1 6.6 3 12

(17) (21) (29) (29) (4) (5.5-9.1) (13) (50)

ASA, American Society of Anesthesiologists; TAAA, thoracoabdominal aortic aneurysm. Continuous data are presented as mean (range) and categoric data as number (%).

Table II. Medical comorbidities High blood pressure Hyperlipidemia Coronary artery disease Congestive heart failure History of smoking Diabetes mellitus Chronic renal insufficiency (eGFR < 60) Chronic obstructive pulmonary disease History of stroke

22 20 13 2 12 3 7 10 2

(92) (83) (54) (8) (50) (13) (29) (42) (8)

eGFR, Estimated glomular filtration rate. Categoric data are presented as number (%).

Table III. Operative details Number of branches Estimated blood loss, mL Length of aortic coverage, cm % of aortic coverage: left common carotid artery to aortic bifurcation Contrast usage, mL Fluoroscopy time, minutes Total dose, Gy

3.7 232 36 74

(2-4) (50-600) (23-57) (46-99)

173 (60-200) 61 (35-128) 5.6 (3.5-8.6)

Continuous data are presented as mean (range).

as a part of their overall treatment. Five patients (21%) were treated with planned staged aortic coverage, with placement of a proximal thoracic endovascular aortic repair (TEVAR) device extending from the proximal seal zone to the celiac artery. Fenestrated-branched graft implantation occurred at a mean of 8 months (range, 2.5-14 months) following the initial TEVAR procedure. The second stage was intended to be done within 6 to 12 weeks of the initial stage, although the specific time interval was individualized based on the size/acuity of the aneurysm as well as by the subject’s recovery from the first-stage procedure. One patient delayed proceeding with the second stage due to new back pain after the initial operation. Two other subjects had interval sac regression of the larger thoracic

1164 Sweet et al

component of their aneurysm and delayed the second stage for personal reasons. Two subjects had the second stage performed as scheduled at 2 and 3 months following the initial procedure. All five patients went on to have successful repair. One additional patient delayed the second stage for personal reasons. She then had a stroke 6 months after her initial stage and has deferred proceeding with repair until she fully recovers. Multiple other adjunctive procedures were performed in a staged fashion. Carotid-subclavian revascularization was performed in four patients (17%). Two subjects (8%) had prestenting of the renal arteries as they were covered by the bare suprarenal fixation after prior endovascular aneurysm repair (EVAR). This was done to ensure that the suprarenal stent would not impede complete deployment of the iCast stent graft, as incomplete deployment of the iCast branching stent could compromise seal at the fenestration. Two patients (8%) had coil embolization of a replaced hepatic and an accessory renal artery arising from the aorta that were too small for branch revascularization. Four patients (17%) required multiple procedures to create branches to the renal arteries. One of these procedures was successful using endovascular means and the other three are discussed below. One patient with an extent II TAAA due to dissection required replacement of the external iliac with reimplantation of the internal iliac in an effort to preserve internal iliac flow for SCI prevention. Twelve patients (50%) had undergone prior aortic operation by another provider in the remote past. Target vessel stenting was successful in 84 of 88 intended target vessels (95%). All four failed branches were to renal target arteries. There was fenestration malalignment to both renal arteries in one patient due to eccentric mural atheroma. After extensive endovascular efforts, surgical iliorenal bypasses were performed, and a cuff was used to cover the fenestrations. Two others patient had small and tortuous renal arteries arising from very large (8 and 9 cm) extent II aneurysms that made selective catheterization and stenting very difficult. Due to excessive sheath manipulation in one patient, a distal renal artery dissection occurred that subsequently bled, requiring coil embolization of the vessel and occlusion of the fenestration. In the other patient, a branching stent was deployed, but there was inadequate distal seal resulting in dislodgment of the distal end of the branching stent. This resulted in an endoleak in the early postoperative setting. Attempts to extend the branch were unsuccessful, and due to the size of the aneurysm (>9 cm), the branch was occluded to ensure sac exclusion. Branch occlusion was accomplished by deploying an iCast stent graft into the sac, flaring it to seal it to the fenestration, and then deploying an Amplatzer Occluder plug into the iCast. In all three patients with failure of renal branch creation, complete aneurysm exclusion was achieved. Among the two patients with renal fenestration occlusion, one went from CKD stage 2 to 3A, and the other from stage 2 to 3B. Neither has clinical renal dysfunction. One other patient suffered a renal artery dissection that resulted in occlusion of a branch vessel and infarction of the

JOURNAL OF VASCULAR SURGERY November 2015

superior pole of the kidney. This patient had deterioration of her renal function from CKD stage 3A to stage 3B. One other patient had a wire injury to a branch of the SMA resulting in a mesenteric hematoma. This patient fully recovered without adverse sequelae. In total, there were three downstream branch artery injuries. All patients required unilateral large bore (>12F) femoral access. Bilateral large bore access was used in 14 (58%), and the other 10 (42%) patients were treated with unilateral large bore and contralateral small bore (#12F) access. The use of unilateral large bore access has become standard as our experience has grown, with nine of the last 11 procedures done in this manner. This evolution of practice has come with improved case planning with greater offset, and greater procedural experience. High brachial access was used for one patient in an unsuccessful attempt to salvage a dissected renal artery as described above. Percutaneous access was used in all cases except one patient who had a pre-existing femoral pseudoaneurysm due to cardiac catheterization. Four femoral artery cut-downs were required at case completion. Three cases of femoral cut-down occurred in patients with prior large bore sheath access; two for staged procedures and one for a prior EVAR done in the remote past. Following deployment of the modified endograft (20-22F), the access site is downsized to 16 to 18F, allowing improved ipsilateral pelvic and leg perfusion during completion of the procedure. The use of a contralateral 12F sheath may also improve pelvic and lower extremity circulation by allowing increased internal iliac flow. There were no access-related complications, including leg or pelvic ischemia or need for access site reintervention. Nineteen patients (79%) were discharged to home, and all have returned to their baseline functional status. Four (17%) were discharged to a skilled nursing facility. Of those four, two have returned home at their baseline functional status after a short stay at the facility; one patient returned home in frail condition, developed bacterial endocarditis, and expired 4 months postoperation; and one patient is permanently paralyzed due to a presumed embolic spinal cord infarction. This diagnosis was based on the extensive mural atheroma in his distal descending thoracic aorta, his modest length of aortic coverage, his low spinal pressures, and the fact that he showed no improvement despite aggressive drainage and blood pressure augmentation immediately postoperation. That subject is alive 13 months following surgery and has trace lower extremity motor function and sensation intact to the thighs. One patient (4%) died in the hospital on postoperative day 22 of complications of spinal cord ischemia. This patient was 82 years old with an extent II TAAA. He developed SCI 6 hours postoperation and was treated with blood pressure support and drainage of cerebrospinal fluid without neurologic improvement. He developed delirium while in the intensive care unit, aspirated, and his family then withdrew support. No patients have experienced aneurysm rupture or sac expansion at mean 11 months (range, 1-41 months) follow-up. All branches have remained patent. One renal

JOURNAL OF VASCULAR SURGERY Volume 62, Number 5

branch developed radiographic evidence of in-stent nonocclusive thrombus when the patient stopped taking his aspirin. This was successfully treated with repeat stenting. One type II endoleak has been seen at the 30-day followup and is being observed. One other type III endoleak seen at the 30-day follow-up had resolved by the 6-month follow-up, and the patient’s aneurysm sac has regressed. No other endoleaks have been observed. No stent migration, component separation, stent fracture, distal embolization, or other branch complication has been identified on follow-up imaging. At last follow-up, renal function had deteriorated by one CKD stage in four patients (17%). Temporary dialysis was required for one patient. No patients have required permanent dialysis. No other late aneurysm-related complications have occurred. DISCUSSION Early-term results suggest that physician-modified endografts are a safe and effective treatment for selected patients at high risk of open repair of TAAA. Although the procedure is done using minimally invasive techniques, it remains a major physiologic stress and has significant risks of morbidity and mortality. All patients have had complete aneurysm exclusion with acceptable 4% perioperative and 8% 1-year mortality rates. These are comparable to the results published in large series using a commercially manufactured device. In these series, 30-day mortality has ranged from 2.6% to 10% with 1-year mortality ranging from 13% to 15.6%.10-12,20 Failure of target vessel stenting occurred in 5% of cases, but all of these were renal arteries that were successfully treated at planned reintervention. The device offset allowed for continued renal artery perfusion throughout the process, and only two target arteries were lost. Improvement in planning is expected with growing clinical experience. Commercial devices with branches, adjustable fenestrations, and/or preloaded wires may also help reduce these events.21 In this initial series, 88% of subjects had treatment success, using a composite, patient-centered outcome, with complete aneurysm exclusion and return to preoperative functional status and no loss of organ system function. This result has not been reported routinely. The group at the University of California San Francisco has reported that this was achieved in 90% of their cases.11 It is important to discriminate between successful aneurysm exclusion and major loss of function. Prior case series after open repair have shown that permanent functional deficits are common after technically successful open repair.6 As we progress toward value-based treatments, preservation of independent functional status will become an ever more important outcome metric. Furthermore, many patients with life-threatening medical conditions value their independent functional status and quality of life over prolongation of life.22 Preservation of independent functional status also has significant impact on the overall cost of care.23 SCI occurred in three (13%) subjects, with one complete recovery and permanent SCI in two (8%). These results are commensurate with what has been reported by

Sweet et al 1165

large volume centers using commercially manufactured endografts. The incidence of any (temporary or permanent) SCI has ranged from 22% to 32%, with permanent SCI affecting 4% to 10%.10-12,24,25 The risk of SCI following endovascular treatment of TAAA remains a primary limitation of this technique. Centers of excellence in open repair have reported levels of permanent SCI as low as 3% to 5%.26,27 It is not clear that those results are achievable in lower-volume centers, however. Endovascular treatment requires more length of aortic coverage than does an open repair, as proximal and distal seal zones for endografts are longer than is required for a sewn anastomosis. Furthermore, the inability to reimplant intercostal/lumbar vessels is another limitation of endograft technology as it now exists. The physician-modified model has advantages and disadvantages. The advantages include maximal adaptability to anatomy, immediate availability, and the ability to work within a very small aortic flow channel (16 mm or larger). The disadvantages include the complex training required to develop facility with the technique, the potential for manufacturing “defects” or contamination that may be less likely to occur in a commercial manufacturing setting, the need for absolute precision in case planning, and the unproven long-term durability. In comparison to the t-Branch device (Cook Medical), the B-TEVAR IDE device can fit aortic diameters down to 16 mm (in contrast to 25 mm) in the peri-visceral aorta; it can have branches up to 10 mm in diameter (in contrast to 8 mm); and it can be done through femoral access, often only one side requiring a sheath over 12F.28 Commercial fenestrated devices will likely be able to accommodate similar anatomy as this construct, although the anatomic limitations of that device remain unreported, and the manufacturing delay remains 6 to 8 weeks. As commercial branched endografts become available, the physician-modified construct is likely to be needed for a minority of patients with uncommon anatomic features that are not compatible with off-theshelf designs and/or those who cannot delay treatment while a commercial custom device is constructed. TAAA is thought to occur in similar frequency among men and women. Six patients (25%) in this study are women. Larger series using commercial devices report that 65% to 93% of their cohorts were male.10-12 Conversely, among a cohort of patients deemed ineligible for repair, 56% were female.7 These data suggest that, for this construct as with commercial devices for TAAA, women are less likely to meet inclusion criteria, and future developments will need to address this shortcoming. With the initiation of the IDE study, procedural standardization has been crucial to facilitate case planning and build the team of collaborators to provide optimal care. This team “learning curve” has occurred in the setting of an institution with a high-volume TEVAR and juxta-renal fenestrated EVAR, a dedicated cardiovascular intensive care unit, and a dedicated cardiovascular anesthesia service. Furthermore, the conduct of the procedures under the auspices of a physician-sponsored IDE has required significant

JOURNAL OF VASCULAR SURGERY November 2015

1166 Sweet et al

institutional investment. The study requires that multiple stakeholders accept what is a very high-risk undertaking. There is substantial institutional risk that must be mitigated through careful review processes. While many clinicians accept the premise that a no-option patient might accept a very high-risk procedure, institutional stake holders may perceive this differently as there is ever greater scrutiny of in-hospital patient outcomes and resource utilization. Patient selection has proven to be one of the more challenging elements to the study. Anatomic criteria delimiting the boundaries of tortuosity, target vessel stenosis, and access challenges are poorly defined. Other important anatomic risk factors include diffuse atheromatous disease of the aorta and prediction of stent graft lie within an eccentric flow channel or large aneurysm. As with iliac disease in routine EVAR, the presence of stenosis, calcification, or tortuosity can often be managed without undue risk. When multiple anatomic risk factors are present in combination, however, procedural risk rises exponentially. Determination of anatomic suitability remains a subjective assessment by the operating surgeon. Five procedures (21%) were performed with deliberately staged aortic coverage. There are two reasons for staging aortic coverage, SCI prevention and management of complex anatomy. Staged repair has been shown in a contemporary larger series to be associated with a reduced risk of SCI.25 Staging was used for extent I and II TAAA in this series and has become the preferred method for eligible patients at their high-volume center. Staging of aortic coverage may reduce SCI by two mechanisms. First, coverage of some intercostal vessels in a sequential fashion appears to allow for improved collateral circulation in animal models.29 Furthermore, endovascular stent graft repair of aneurysms is associated with an acute inflammatory response, often described as the “postimplantation” syndrome. This is likely due to platelet activation on the Dacron fabric as well as consumptive coagulopathy in the excluded aneurysm sac.30 This response correlates with length of coverage, and as such, staged repair may mitigate the severity of that response by decreasing the inflammatory stimulus into two smaller doses. As we saw in subject #002, the acute inflammatory response ultimately led to the patients’ demise despite a technically successful procedure. This inflammatory response can cause severe hemodynamic lability, which is a major risk for SCI in the postoperative setting. The second advantage of staging repairs is to facilitate branch creation in complex anatomy. In the setting of a large aneurysm or very eccentric mural atheroma, it can be difficult to determine where the stent graft will sit within the aortic flow channel. Staging proximal components can facilitate that by clarifying the position of the proximal devices. A staged approach requires management of patient expectations in terms of the time course of therapy and does leave the patient at risk of interval rupture between stages. Which patients are most likely to benefit from staged aortic coverage remains uncertain. In this study, staging has been used for eligible patients who need in excess of

15 to 20 cm of coverage above the celiac artery (ie, extent I-III). Patients ineligible for staging of aortic coverage include those with an aneurysm at imminent risk of rupture (ie, large size and/or symptomatic) and those with hostile ilio-femoral access, where repeated access may incur additional morbidity, although iliac conduit could be done in a staged fashion in such patients if needed. Patients who are elderly and frail may be better able to tolerate several less physiologically stressful procedures than younger/fitter patients, although this remains unproven. Thus far, no evident adverse events have occurred due to staging, although femoral cut-down was more common among these patients. Type III endoleaks were frequently seen at case completion while the patient was heparinized. These appeared to arise from the physician modifications themselves, likely suture-hole bleeding at the fenestration or diameter-reducing ties, not from the fenestration/branch interface itself as they were not seen on individual branch angiograms. In all cases but one, they had resolved at 1month follow-up, and the other case resolved by the 6month follow-up. Thus far, there have been no late type I or III endoleaks, no sac expansion, and no graft failures, albeit at short-/medium-term follow-up. Longer term follow-up is required to determine if these intraoperative findings are of clinical significance. The limitations of this study include the small cohort size and relatively short-/medium-term follow-up. As the study is ongoing, longer term follow-up and a larger cohort size will be essential to demonstrate the long-term safety and effectiveness of the therapy. Furthermore, this study is essentially a single-center and single surgeon experience, so the results cannot be generalized outside of that setting. The results achieved with this specific graft construct also cannot be generalized to other physician-modified designs, such as those with stent graft branches sewn to the primary aortic graft. Thus far there are insufficient data to make meaningful comparisons between different graft designs. The role of physician-modified endografts will be clarified once commercially manufactured grafts are widely available. Premanufactured commercial devices appear to have a “few sizes fits most” concept. Commercially manufactured custom devices are currently in use and have excellent results, but they require 6 to 8 weeks for manufacture. As such, there may be a role for physician-modified devices when those devices are available, although it is expected to be for a small subset of urgent/symptomatic patients whose anatomy is not amenable to an off-the-shelf commercial device. The minimal case volume of physician modification required to maintain proficiency may not be sustainable when commercial devices are widely available. CONCLUSIONS Totally endovascular treatment of TAAA using a physician-modified thoracic endograft is safe and effective at early term follow-up and performs commensurately with commercially manufactured grafts. Although it can be done using minimally invasive means, it remains a

JOURNAL OF VASCULAR SURGERY Volume 62, Number 5

high-risk procedure. B-TEVAR is highly adaptable to complex anatomy and can be done with current commercially available devices. This treatment may provide a safe and effective means of treatment of TAAA among patients at high risk of open repair. Longer term follow-up is essential to prove long-term effectiveness.

Sweet et al 1167

14.

15.

AUTHOR CONTRIBUTIONS

16.

Conception and design: MS, BS, BT Analysis and interpretation: MS, BS, BT Data collection: MS, BT Writing the article: MS Critical revision of the article: MS, BS, BT Final approval of the article: MS, BS, BT Statistical analysis: MS Obtained funding: Not applicable Overall responsibility: MS

17.

REFERENCES 1. Lancaster RT, Conrad MF, Patel VI, Cambria MR, Ergul EA, Cambria RP. Further experience with distal aortic perfusion and motor evoked potential monitoring in the management of extent I-III thoracoabdominal aortic aneurysms. J Vasc Surg 2013;58:283. 2. Acher C, Wynn M. Outcomes in open repair of the thoracic and thoracoabdominal aorta. J Vasc Surg 2010;52:3S-9S. 3. Lemaire SA, Price MD, Green SY, Zarda S, Coselli JS. Results of open thoracoabdominal aortic aneurysm repair. Ann Cardiothorac Surg 2012;1:286-92. 4. Rigberg DA, McGory ML, Zingmond DS, Maggard MA, Agustin M, Lawrence PF, et al. Thirty-day mortality statistics underestimate the risk of repair of thoracoabdominal aortic aneurysms: a statewide experience. J Vasc Surg 2006;43:217-23. 5. Cowan JA, Dimick JB, Henke PK, Huber TS, Stanley JC, Upchurch GR. Surgical treatment of intact thoracoabdominal aortic aneurysms in the United States: hospital and surgeon volume-related outcomes. J Vasc Surg 2003;37:1169-74. 6. Rectenwald JE, Huber TS, Martin TD, Ozaki CK, Devidas M, Welborn MB, et al. Functional outcome after thoracoabdominal aortic aneurysm repair. J Vasc Surg 2002;35:640-7. 7. Hansen PA, Richards JM, Tambyraja AL, Khan LR, Chalmers RT. Natural history of thoraco-abdominal aneurysm in high-risk patients. Eur J Vasc Endovasc Surg 2010;39:266-70. 8. Lederle FA, Freischlag JA, Kyriakides TC, Matsumura JS, Padberg FT, Kohler TR, et al; OVER Verterans Affairs Cooperative Study Group. Long-term comparison of endovascular and open repair of abdominal aortic aneurysm. N Engl J Med 2012;367:1988-97. 9. Matsumura JS, Melissano G, Cambria RP, Dake MD, Mehta S, Svensson LG, et al. Five-year results of the thoracic endovascular aortic repair with the Zenith TX2. J Vasc Surg 2014;60:1-10. 10. Greenberg RK, Lu Q, Roselli EE, Svensson LG, Moon MC, Hernandez AV, et al. Contemporary analysis of descending thoracic and thoracoabdominal aneurysm repair: a comparison of endovascular and open techniques. Circulation 2008;118:808-17. 11. Reilly LM, Rapp JH, Grenon M, Hiramoto JS, Sobel J, Chuter TA. Efficacy and durability of endovascular thoracoabdominal aortic aneurysm repair using the caudally directed cuff technique. J Vasc Surg 2012;56:53-64. 12. Guillou M, Bianchini A, Sobocinski J, Maurel B, D’elia P, Tyrrell M, et al. Endovascular treatment of thoracoabdominal aortic aneurysms. J Vasc Surg 2012;56:65-73. 13. Fillinger M. Branched-fenestrated and chimney-type endovascular repairs for juxtarenal and thoracoabdominal aortic aneurysms. Presented

18. 19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

at the 2012 Vascular Annual Meeting, National Harbor, Md, June 79, 2012. Oderich GS, Fatima J, Gloviczki P. Stent graft modification with minicuff reinforced fenestrations for urgent repair of thoracoabdominal aortic anuerysms. J Vasc Surg 2011;54:1522-6. Scali ST, Waterman A, Feezor RJ, Martin TD, Hess PJ, Huber TS, et al. Treatment of acute visceral aortic pathology with fenestrated/ branched endovascular repair in high-surgical risk patients. J Vasc Surg 2013;58:56-65. Schanzer A, Thompson P, Baril D, Robinson WP, Simons JP, Aiello FA, et al. Lessons learned from a single center’s experience developing a complex endovascular fenestrated/branched aortic program and obtaining a physician-sponsored FDA Investigational Device Exemption (IDE). Presented at the 2014 Vascular Annual Meeting, Boston, Mass, June 5-7, 2014. Starnes BW, Tatum B. Early report from an investigator-initiated investigational device exemption clinical trial on physician-modified endovascular grafts. J Vasc Surg 2013;58:311-7. Safi HJ. How I do it: thoracoabdominal aortic aneurysm graft replacement. Cardiovasc Surg 1999;7:607-13. National Kidney Foundation. K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification. Am J Kidney Dis 2002;39(Suppl 1):S1-266. Muhs BE, Verhoeven EL, Zeebregts CJ, Tielliu IF, Prins TR, Verhagen HJ, et al. Mid-term results of endovascular aneurysm repair with branched and fenestrated endografts. J Vasc Surg 2006;44:9-15. Mastracci TM, Greenberg RK, Eagleton MJ, Hernandez AV. Durability of branches in branched and fenestrated endografts. J Vasc Surg 2013;57:926-33. Higginson IJ, Gomes B, Calanzani N, Gao W, Bausewein C, Daveson BA, et al; on behalf of Project PRISMA. Priorities for treatment, care and information if faced with serious illness: a comparative population-based survey in seven European countries. Palliat Med 2014;28:101-10. Vaislic M, Vaislic C, Alsac JM, Benjelloun A, Chocron S, Unterseeh T, et al. Economic impacts of treatment for type II or III thoracoabdominal aortic aneurysm in the United States. Res Cardiovasc Med 2014;3:e9568. Dias NV, Sonesson B, Kristmundsson T, Holm H, Resch T. Shortterm outcome of spinal cord ischemia after endovascular repair of thoracoabdominal aortic aneurysms. Eur J Vasc Endovasc Surg 2015;49:403-9. O’Callaghan A, Mastracci TM, Eagleton MJ. Staged endovascular repair of thoracoabdominal aortic aneurysms limits incidence and severity of spinal cord ischemia. J Vasc Surg 2015;61:347-54.e1. Acher CW, Wynn MM, Mell MW, Tefera G, Hoch JR. A quantitative assessment of the impact of intercostal artery reimplantation on paralysis risk in thoracoabdominal aortic aneurysm repair. Ann Surg 2008;248:529-40. Coselli JS, LeMaire SA, Koksoy C, Schmittling ZC, Curling PE. Cerebrospinal fluid drainage reduces paraplegia after thoracoabdominal aortic aneurysm repair: results of a randomized clinical trial. J Vasc Surg 2002;35:631-9. Gasper WJ, Reilly LM, Rapp JH, Grenon M, Hiramoto JS, Sobel JD, et al. Assessing the anatomic applicability of the multibranched endovascular repair of thoracoabdominal aortic aneurysm technique. J Vasc Surg 2013;57:1553-9. Bischoff MS, Scheumann J, Brenner RM, Ladage D, Bodian CA, Kleinman G, et al. Staged approach prevents spinal cord injury in hybrid surgical-endovascular thoracoabdominal aortic aneurysm repair: an experimental model. Ann Thorac Surg 2011;92:138-46. Chang CK, Chuter TA, Niemann CU, Shlipak MG, Cohen MJ, Reilly LM, et al. Systemic inflammation, coagulopathy, and acute renal insufficiency following endovascular thoracoabdominal aortic aneurysm repair. J Vasc Surg 2009;49:1140-6.

Submitted Mar 18, 2015; accepted May 15, 2015.