key: cord-305205-ilxxkm0f authors: Cochennec, Frédéric; Kobeiter, Hicham; Gohel, Manj S.; Majewski, Marek; Marzelle, Jean; Desgranges, Pascal; Allaire, Eric; Becquemin, Jean Pierre title: Impact of intraoperative adverse events during branched and fenestrated aortic stent grafting on postoperative outcome date: 2014-09-30 journal: Journal of Vascular Surgery DOI: 10.1016/j.jvs.2014.02.065 sha: doc_id: 305205 cord_uid: ilxxkm0f Objective Fenestrated and branched endovascular devices are increasingly used for complex aortic diseases, and despite the challenging nature of these procedures, early experiences from pioneering centers have been encouraging. The objectives of this retrospective study were to report our experience of intraoperative adverse events (IOAEs) during fenestrated and branched stent grafting and to analyze the impact on clinical outcomes. Methods Consecutive patients treated with fenestrated and branched stent grafting in a tertiary vascular center between February 2006 and October 2013 were evaluated. A prospectively maintained computerized database was scrutinized and updated retrospectively. Intraoperative angiograms were reviewed to identify IOAEs, and adverse events were categorized into three types: target vessel cannulation, positioning of graft components, and intraoperative access. Clinical consequences of IOAEs were analyzed to ascertain whether they were responsible for death or moderate to severe postoperative complications. Results During the study period, 113 consecutive elective patients underwent fenestrated or branched stent grafting. Indications for treatment were asymptomatic complex abdominal aortic aneurysms (CAAAs, n = 89) and thoracoabdominal aortic aneurysms (TAAAs, n = 24). Stent grafts included fenestrated (n = 79) and branched (n = 17) Cook stent grafts (Cook Medical, Bloomington, Ind), Ventana (Endologix, Irvine, Calif) stent grafts (n = 9), and fenestrated Anaconda (Vascutek Terumo, Scotland, UK) stent grafts (n = 8). In-hospital mortality rates for the CAAA and TAAA groups were 6.7% (6 of 89) and 12.5% (3 of 24), respectively. Twenty-eight moderate to severe complications occurred in 21 patients (18.6%). Spinal cord ischemia was recorded in six patients, three of which resolved completely. A total of 37 IOAEs were recorded in 34 (30.1%) patients (22 CAAAs and 12 TAAAs). Of 37 IOAEs, 15 (40.5%) resulted in no clinical consequence in 15 patients; 17 (45.9%) were responsible for moderate to severe complications in 16 patients, and five (13.5%) led to death in four patients. The composite end point death/nonfatal moderate to severe complication occurred more frequently in patients with IOAEs compared with patients without IOAEs (20 of 34 vs 12 of 79; P < .0001). Conclusions In this contemporary series, IOAEs were relatively frequent during branched or fenestrated stenting procedures and were often responsible for significant complications. Impact of intraoperative adverse events during branched and fenestrated aortic stent grafting on postoperative outcome Fenestrated and branched endovascular aneurysm repairs (FEVAR and BEVAR) have become an attractive alternative to open repair for complex abdominal aortic aneurysms (CAAAs) and thoracoabdominal aortic aneurysms (TAAAs). In many countries, these complex procedures are still under evaluation and generally available only in tertiary centers. In France, fenestrated and branched Cook devices have been approved for reimbursement from the national health care system. However, intraoperative difficulties and complications are not rare. 1 Safe target vessel cannulation and stenting is a concern, particularly in the presence of stenotic ostial lesions and small or angulated target vessels. Malpositioning of stent graft components can also have devastating consequences. As delivery devices are larger than in standard infrarenal endovascular aneurysm repair (EVAR) and the procedure duration is generally longer, patients are more prone to access complications. The real incidence of those intraoperative adverse events (IOAEs) and their impact on the postoperative course are poorly documented. In this retrospective study, we report the incidence of IOAEs during fenestrated or branched stent grafting and analyze to what extent these adverse events may influence early postoperative outcomes. Study setting. Consecutive patients undergoing FEVAR or BEVAR between February 2006 and October 2013 in a tertiary vascular unit (Henri Mondor Hospital, Créteil) were included. Patients were treated for CAAAs and TAAAs. CAAAs included short-necked infrarenal, juxtarenal, pararenal, and suprarenal abdominal aortic aneurysms, considered unsuitable for conventional EVAR. TAAAs were classified according to the Crawford classification. 2 In our institution, all patients with CAAAs and TAAAs are considered for open, hybrid, or endovascular repair in a multidisciplinary meeting including vascular surgeons, interventional radiologists, and anesthesiologists. Demographic, anatomic, intraoperative, and postoperative data were recorded by means of a prospectively collected database. Preoperative assessment and device sizing. All patients underwent a high-resolution computed tomography scan preoperatively and before discharge. Procedure planning and device sizing were performed with a dedicated three-dimensional vascular imaging workstation (Aquarius WS; TeraRecon Inc, Mateo, Calif) with centerline luminal reconstructions. The aneurysm morphology was assessed by a vascular surgeon (M.M.) and an interventional radiologist (H.K.), both with considerable experience with EVAR. Device designs proposed by the implanting physicians were systematically reviewed and approved by the planning center of the corresponding device manufacturer. Details of procedures. Procedures were performed in an angiography suite (Philips FD20; Philips Healthcare, Cleveland, Ohio) in a sterile environment. An experienced proctor physician was present during the procedure for the first five Cook fenestrated cases, the first two Cook branched cases, and the first fenestrated Anaconda and Ventana cases. Eight physician-modified fenestrated stent grafts were excluded. For each device, the implantation techniques have been described previously. [3] [4] [5] [6] [7] [8] [9] Control angiograms were obtained once each target vessel was cannulated with a long sheath, after deployment of bridging covered stents in each target vessel, and at the end of the procedure. Each control angiogram was saved and images were stored in a database. Technical problems and subsequent IOAEs were also recorded in the database. Definitions. IOAEs were defined as any intraoperative complication or technical problem occurring during stent graft implantation that required additional and unexpected endovascular manipulations. IOAEs were classified in three distinct types: Type 1: Problems with target vessel cannulation; Type 2: Malpositioning of one of the following graft components: bridging stents, bifurcated component, or iliac extensions; and Type 3: Difficulty with intraoperative access. Complications were defined according to the Society for Vascular Surgery criteria. 10 Only moderate and severe complications were reported in the current series. Study protocol. A radiologist (H.K.) and a vascular surgeon (F.C.) reviewed intraoperative and postoperative data recorded in the prospectively maintained FEVAR or BEVAR database and all intraoperative angiograms for each patient to identify IOAEs. Both clinicians also participated in the majority of the procedures. To evaluate the clinical consequences of IOAEs for postoperative outcome, the composite end point Table II . The mean (standard deviation) number of target vessels per patient was 2.8 6 0.8. Clinical outcomes and IOAEs. The overall inhospital mortality was 8.0% (9 of 113 patients). Inhospital mortality rates for the CAAA and TAAA groups were 6.7% (6 of 89) and 12.5% (3 of 24), respectively (Table III) . A total of 28 moderate to severe complications occurred in 21 (18.6%) patients (Table IV) ; 17 complications were related to IOAEs, and 11 occurred without any IOAE. In patients with TAAAs, complete paraplegia occurred in three cases (12.5%). Two other patients presented with paraparesis that resolved completely after spinal fluid drainage. In the CAAA group, one patient with a juxtarenal aneurysm who underwent fenestrated stent grafting with three fenestrations presented with hypoesthesia of both limbs that resolved completely after spinal fluid drainage. A total of 37 IOAEs were recorded in 34 (30.1%) patients ( (Table V) . The incidence of IOAEs was not significantly different in patients who underwent a Cook fenestrated stent graft compared with patients who underwent an Anaconda or a Ventana fenestrated stent graft. The incidence of IOAE did not change over time (Fig) . Type 1 IOAEs (problems with target vessel cannulation) occurred in 22 of 113 cases (19.4%), resulting in death (n ¼ 3) or moderate to severe complication (n ¼ 10) in 13 patients (59.1%). They led to target vessel loss in five cases (22.7%). Details of type 1 IOAEs, intraoperative management, and related outcomes are given in Table VI. Type 2 IOAEs (malpositioning of one of the graft components) occurred in 5 of 113 cases (4.4%), resulting in death or complication in two patients (Table VII) . Type 3 IOAEs (related to access site problems) occurred in 10 of 113 patients (8.8%), leading to early postoperative death in one patient and moderate to severe complications in another six patients (Table VIII) . The utility of fenestrated and branched devices for the treatment of CAAAs and TAAAs has gained widespread acceptance, with several large series confirming satisfactory early and midterm results. 5, [11] [12] [13] [14] [15] In a recent systematic review of juxtarenal abdominal aortic aneurysms treated by FEVAR, 16 368 FEVAR cases from eight cohort studies were evaluated. The reported 30-day mortality was 1.4%, and the incidence of permanent renal dialysis was 1.4%. Data from national registries and from high-volume centers have provided similar results, with 30-day mortality rates varying from 2% to 4%. 13, 14, 17 In expert hands, endovascular repair of TAAA with FEVAR or BEVAR has been associated with encouraging short-term results, with 30-day mortality ranging from 5% to 12% and spinal cord ischemia from 3% to 17%. 4, 12, [18] [19] [20] With 30-day mortality rates of 6.7% for juxtarenal patients and 12.5% for TAAA patients, our results do not compare favorably with previously published data. Several reasons might explain this observation. In contrast with pioneering series, in which one physician performed the majority of the cases, two vascular surgeons in our institution performed fenestrated and branched procedures as the first operator, although a highly experienced interventional radiologist (H.K.) was present during almost all procedures. This, combined with the fact that three different types of device were used, might have contributed to IOAEs in some patients. However, more than a simple learning curve of technical skills, one could argue that our results are mainly due to suboptimal patient selection. On review of the mortalities in this series, all patients had challenging aneurysm morphology or significant comorbidity (American Society of Anesthesiologists class 4). Four patients had "shaggy" aortas with floating thrombus, resulting in fatal embolic complications. As FEVAR and BEVAR are being increasingly used and disseminated, the results of our initial experience confirm that they remain complex procedures that need to be centralized in highvolume centers. They also raise the question of whether encouraging results of initial series can be reproduced in the "real world," knowing these series came from a very few pioneering expert centers and included highly selected patients. Type 1 IOAEs (problems with target vessel cannulation) were the most frequent in our experience (n ¼ 22). Target vessel occlusions and dissections occurred in 11 cases and were mainly due to target vessel injury or thrombus formation in the long 7F sheaths. Even if most of them could be rescued with additional bailout endovascular maneuvers and without permanent damaging consequences for the patient (Table VI) , they are considered avoidable. Thrombus formation in the long 7F sheaths may also have been avoided by more frequent flushing. Cannulation failure was relatively rare (n ¼ 4) but had devastating consequences for the patient, leading directly or indirectly to severe complications or death (Table VI) . This was mainly due to sizing errors, malpositioning of the fenestrated component, or difficult anatomy and occurred predominantly at the beginning of our experience. Type 2 IOAEs (malpositioning of bridging stents, bifurcated component, or iliac extensions) are also considered avoidable technical errors. In our series, they occurred in five cases. In one patient with a type II TAAA, a sac perfusion branch had to be covered as it was inadvertently located at an overlap zone. The patient subsequently died of meningoencephalitis after spinal drain placement for paraplegia, although the use of sac perfusion branches to reduce paraplegia risk remains controversial. With the exception of one endovascular reintervention, all remaining IOAEs relating to graft malpositioning were managed successfully without harm to the patient. FEVAR and BEVAR procedures are frequently long, require large-diameter introducer sheaths, and are prone to access vessel complications. In this series, intraoperative access site problems (type 3 IOAEs) occurred in 10 patients (8.8%), leading to moderate or severe complications in seven cases. In our current practice, we use a percutaneous approach for most standard EVAR cases. Because of the increased risk of access complications during FEVAR and BEVAR procedures, we still favor surgical cutdown of femoral arteries. Caution should be taken in interpreting the results of this series as the definition of CAAA is broad, including short neck, juxtarenal, pararenal, and suprarenal aneurysms. Consequently, there was considerable heterogeneity in complexity of aneurysm morphology and difficulty of procedures, which varied from stents with two fenestrations to challenging branched stent grafting for type II TAAAs in high-risk patients. Despite these limits, this study suggests that in routine practice, technical difficulties and IOAEs during FEVAR and BEVAR procedures are not rare, particularly with difficult aneurysm morphology. It is likely that continuous improvements in endovascular and imaging technologies will improve the safety of complex endovascular aortic procedures. Development of lower profile stent grafts and bridging stents, with better visibility and repositionability, will probably play a key role in reducing IOAEs. New tools for improved catheter navigation, such as robotic navigation, 21 may also facilitate target vessel cannulation. Although few data exist, fusion of images may be useful to reduce doses of contrast material and radiation. 22 In our series, IOAEs during branched and fenestrated stent grafting were frequent, occurring in 25% of patients with CAAAs and 50% of patients with TAAAs. As branched and fenestrated devices are being increasingly used and disseminated in vascular centers, additional studies are needed to determine if the encouraging results from pioneering expert centers can be reproduced in "real-life" practice. Incidence and management of complications after branched and fenestrated endografting Thoracoabdominal aortic aneurysms: preoperative and intraoperative factors determining immediate and long-term results of operations in 605 patients Repair of thoracoabdominal aortic aneurysms with fenestrated and branched endovascular stent grafts Endovascular repair of thoracoabdominal aortic aneurysms A prospective analysis of fenestrated endovascular grafting: intermediateterm outcomes Fenestrated endovascular repair for juxtarenal aortic aneurysm Outcomes of fenestrated endografts in the treatment of abdominal aortic aneurysm in Western Australia Initial experience with a new fenestrated stent graft Ventana fenestrated stent-graft system for endovascular repair of juxtarenal aortic aneurysms Reporting standards for endovascular aortic aneurysm repair Branched endografts for thoracoabdominal aneurysms Contemporary analysis of descending thoracic and thoracoabdominal aneurysm repair: a comparison of endovascular and open techniques Intermediate results of a United States multicenter trial of fenestrated endograft repair for juxtarenal abdominal aortic aneurysms An analysis of the French multicentre experience of fenestrated aortic endografts: medium-term outcomes Thoracoabdominal aortic aneurysm branched repair Modern treatment of juxtarenal abdominal aortic aneurysms with fenestrated endografting and open repairda systematic review British Society for Endovascular Therapy and the Global Collaborators on Advanced Stent-Graft Techniques for Aneurysm Repair (GLOB-ALSTAR) Registry. Early results of fenestrated endovascular repair of juxtarenal aortic aneurysms in the United Kingdom Endovascular treatment of thoracoabdominal aortic aneurysms Endovascular treatment of thoracoabdominal aortic aneurysms Present and future of branched stent grafts in thoraco-abdominal aortic aneurysm repair: a single-centre experience The role of robotic endovascular catheters in fenestrated stent grafting Intraoperative C-arm cone-beam computed tomography in fenestrated/branched aortic endografting Debate: whether abdominal aortic aneurysm surgery should be centralized at higher-volume centers A shaggy aorta is associated with mesenteric embolisation in patients undergoing fenestrated endografts to treat paravisceral aortic aneurysms Fenestrated endovascular repair for juxtarenal aortic pathology The report by Dr Cochennec highlights some very significant issues related to the practice and dissemination of complex aortic procedures.The authors present 113 cases performed during a period of 7 years, which equates to fewer than 20 cases per annum divided among four operators using three endovascular devices. Previous studies have shown a link between surgeon as well as center volumes and outcomes. 1 Complex aneurysms represent a small proportion of all aneurysms treated, and it is difficult, even in large vascular centers, to achieve volumes large enough to reap the benefits of truly high volumes. In contrast to "simpler" aortic procedures, treatment of complex aneurysms is not only a question of individual operator experience and skill, but it places significant demands on the supporting structure of the center.Preoperative planning of complex cases is paramount for success. Detailed anatomic analysis as well as extensive specific device knowledge is needed. The learning curve is significantly longer than for infrarenal endovascular aneurysm repair. This indicates that it is better to use a single device and to learn it well rather than to stray between different devices, trying to take advantage of individual stent graft benefits.Certain morphologic features need to be specifically addressed. (1) A "shaggy" aorta is a risk factor for embolic events and poor outcome. 2 (2) Poor access can lead to difficulties in orienting the device, leading to target vessel loss, resulting in end-organ dysfunction. Very long procedures can lead to compartment syndrome as well as a systemic reperfusion injury causing severe physiologic instability and multiorgan failure in the immediate postoperative phase. 3 Patients with complex aneurysmal disease often carry significant comorbidities and generally have a much lower tolerance for perioperative and postoperative complications, which is clearly seen in the current study with significant morbidity and mortality. This furthermore emphasizes the need for appropriate patient selection, workup, and optimization before surgery as well as extensive involvement of anesthetic and intensive care expertise in the preoperative, perioperative, and postoperative period. This can most likely be achieved only in very select high-volume centers dedicated to treatment of complex aortic disease.I thank the authors for their honest reporting and for highlighting the issues in dealing with complex endovascular aortic procedures.