Fenestrated and Chimney Technique for Juxtarenal Aortic Aneurysm: A Systematic Review and Pooled Data Analysis

Juxtarenal aortic aneurysms (JAA) account for approximately 15% of abdominal aortic aneurysms. Fenestrated endovascular aneurysm repair (FEVAR) and chimney endovascular aneurysm repair (CH-EVAR) are both effective methods to treat JAAs, but the comparative effectiveness of these treatment modalities is unclear. We searched the PubMed, Medline, Embase, and Cochrane databases to identify English language articles published between January 2005 and September 2013 on management of JAA with fenestrated and chimney techniques to conduct a systematic review to compare outcomes of patients with juxtarenal aortic aneurysm (JAA) treated with the two techniques. We compared nine F-EVAR cohort studies including 542 JAA patients and 8 CH-EVAR cohorts with 158 JAA patients regarding techniques success rates, 30-day mortality, late mortality, endoleak events and secondary intervention rates. The results of this systematic review indicate that both fenestrated and chimney techniques are attractive options for JAAs treatment with encouraging early and mid-term outcomes.

. The largest series (318 patients) was in the Globalstar study in F-EVAR group. Male patients predominated. In 8 CH/SN-EVAR studies and 9 F-EVAR studies, men comprised 82.9% and 85.7% of the study population, respectively. The mean age was 75 (59-88) years in the CH/SN-EVAR series and 74 (47-86) years in the F-EVAR series. The mean aneurysm diameter was 64.0 mm (47-112) in the F-EVAR group and 64.5 mm  in the CH/SN-EVAR group. The mean/median aneurysm neck length was 6.7 mm (0-14) in five F-EVAR studies and 2.3 mm (0-10) in six CH/SN-EVAR studies.
All 15 studies reported information on patient co-morbidities (Tables 1 and 2). Nine studies documented smoking status; patients with any history of smoking comprised 69.6% of the patients in the F-EVAR group and 66.1% of those in the CH/SN-EVAR group. Diabetes mellitus was reported in 16.9% and 16.2% of patients in 5 CH/SN-EVAR studies and 6 F-EVAR studies, respectively. The hypertension rates were high in each EVAR group (67% in the F-EVAR group and 87.8% in the CH/SN-EVAR group). Hyperlipidemia was not widely reported in the F-EVAR group. Cardiovascular disease rates were retrieved from all 15 studies. CAD, CHF, MI, arrhythmia, or any combination of these was noted in 33.1% of patients in the F-EVAR series and 44.6% of patients in the CH/ SN-EVAR series. In total, 36% and 29.7% of patients in the F-EVAR group and CH//SN-EVAR group, respectively, reported respiratory diseases. Here, 18.6% patients in the F-EVAR series had renal diseases, and 28% of the patients in the CH/SN-EVAR had renal diseases. Previous major abdominal surgery was reported in 4 studies of the CH/SN-EVAR group, which covered a total of 30 patients and 3 studies with 12 patients in the F-EVAR group. Regarding American Society of Anesthesiologists (ASA) grade, the factor was not widely reported, and groups were often combined, making it difficult to extrapolate exact numbers.
Postoperative. The length of the hospital stay was reported for the F-EVAR group in 7 studies (mean, 7 (1-100) days). This information was not widely reported for the CH/SN group. Only 3 articles provided this information. The average length of hospital stay was 4.4 (2-50) days, which is likely an underrepresentation. The patency rate was 95.9% in the F-EVAR group and 97% in the CH-EVAR group (Tables 7 and 8).
The 30-day mortality rate was 3.8% (6/158) in the CH/SN-EVAR group and 1.1% (6/542) in the F-EVAR group. Six deaths occurred after CH/SN-EVAR, including three cases of MOF (multiple organ failure) caused by bowel ischemia and three cardiac events with no intraoperative deaths. One patient in the fenestrated group died due to myocardial infarction (MI), one died from pneumonia, one died of multisystem organ failure, and one patient died of bowel ischemia related to AAA. The reasons for the last two deaths were not specifically listed but were not related to AAA.
The mean follow-up was 14.7 months in the CH/SN-EVAR group (range 0-46 months) and 12.8 months in the F-EVAR group (range 1-65 months). During follow-up, 15 (8.46%) deaths occurred in the CH/SN-EVAR group (range 1-26 months) In addition, 29 patients (5.4%) died in the fenestrated group, and the last death occurred in the 51th month postoperation. After the chimney/snorkel operation, 7 (7/15 46.6%) patients died of cardiac causes. Two patients died of AAA rupture, and 4 died for reasons that were not specified but not related to AAA 19 . The 29 deaths that occurred after the fenestrated operation, including 4 deaths caused by cardiac and respiratory factors, 3 deaths due to MOF, 3 deaths due to stroke, 1 death due to CKD, 1 death due to gastrointestinal bleeding, 1 death due to tumor, and 16 deaths with non-specific causes but not related to AAA. A total of 21 deaths (13.3%) occurred in the CH/SN group, and 35 deaths (6.4%) occurred in the F-EVAR group. Cardiac disease was the most common cause of postoperative death (reported death cause).
Secondary interventions included the resolution of endoleaks, target vessel occlusion or stenosis, limb occlusion or stenosis, and excess bleeding or hematoma. A total of 58 secondary intervention events occurred after fenestrated operation, 15 of which were noted in the CH/SN-EVAR group.
The comparison results were extracted from two groups of studies and provided in Table 5. We also performed a meta-analysis on two-arm studies (Fig. 2).
Endoleak events. The endoleaks were divided into groups according to leak site as described by Veith et al. 20 : type I, attachment site leak; type II, branch leak; type III, graft defect; and type IV, graft wall (fabric) porosity. Pooled analyses were performed by calculating the overall rates of events. All 15 articles reported endoleak   Post-procedural CT scans showed 12 type I (7.6%, 12/158), 16 type II (10.1%, 16/158) and 1 type III endoleaks in the CH/SN-EVAR group. Here, 6 patients (2 type I, 4 type II) required secondary intervention to resolve the endoleak. A total of 12 endoleaks resolved spontaneously during follow-up, ranging from 1 to 12 months. In addition, 7 type II endoleaks and one type I endoleak were under surveillance during follow-up. According to Schiro, two deaths were caused by aneurysm rupture related to a type I endoleak 19 .  Renal events. Here, 49 renal artery events were reported in F-EVAR group, including stenosis (26), occlusions (12), perforation/bleeding (7) and stent events (4). Of these, 5 events were resolved during the operation, and 32 events required secondary intervention. Four cases of renal artery occlusion were documented in the CH/ SN-EVAR group.
Of the 15 studies included in the review, 10 reported renal events. These events were defined as an increase in serum creatinine to > 2 mg/dl or by > 30% relative to baseline during the peri-operative period. Of the 542 relevant patients, 30 (5.5%) developed renal impairment or failure following F-EVAR. In addition, 16 suffered from postoperative renal impairment. This complication was temporary for 7 patients, and only 2 patients required temporary or persistent postoperative dialysis. In contrast, 31 patients (19.6%) in the chimney group developed this complication, and 1 patient required persistent dialysis.

Discussion
The present review compared the clinical outcomes of patients who underwent F-EVAR and the chimney/snorkel technique for treatment of juxtarenal aortic aneurysms. The fenestrated technique exhibited advantages compared with the chimney/snorkel technique with respect to 30-day mortality, late mortality and renal adverse events. However, patients in the CH/SN-EVAR group experienced shorter operative and fluoroscopy procedures, required lower contrast doses, and suffered less blood loss during the operation. The present study aimed to evaluate the safety and efficacy of fenestrated and chimney techniques for JAAA.
The fundamental goal of fenestrated and chimney/snorkel techniques is to extend the sealed area and maintain flow to a branch vessel with or without the use of a stent-graft. F-EVAR is an expensive procedure that is tailored specifically to each individual patient's anatomy. The design of each fenestrated device is very complicated and requires accurate calculations of the distances between the visceral vessels. This procedure can easily take 4 to 6 weeks or more in centers lacking staff experienced in this method, where measurements must be double-checked. This technique is costly, time-consuming and not suitable for urgent situations [21][22][23][24] . The chimney/snorkel technique is widely available and can be performed in smaller centers. The technique is less complex and can be performed with off-the-shelf endografts. The technique can be used to provide immediate treatment in acute cases 25 . The major concerns regarding chimney/snorkel are endoleaks and subsequent complications 25,26 .
With larger delivery systems, fenestrated grafts must use conduits to open the arteries to insert the transfer system. This procedure may increase the mean operative time and blood loss relative to chimney techniques (operative time: 261 min for F-EVAR vs 178 min for CH-EVAR; estimated blood loss: 534 ml for F-EVAR vs 332 ml for CH-EVAR). Fluoroscopy time and contrast dose were both slightly increased in the fenestrated series compared with the chimney series (64 min vs 54.6 min for fluoroscopy; 166 ml vs 146 ml for contrast dose). This finding is potentially attributed to the fact that the graft can be placed more accurately, and secondary procedures are often performed to verify that the new position is suitable.
Not surprisingly, 30-day mortality rates favor F-EVAR over CH/SN-EVAR (1% vs 3.8%) ( Table 9). The increased 30-day mortality rate of CH/SN-EVAR may be attributed to the inclusion of acute patients (acute or semi-acute) and patients with more challenging anatomical structures. Late mortality was 5.35% in the F-EVAR group and 9.5% in the CH/SN-EVAR group. The all-cause death rate was 6.46% (35 patients) in the F-EVAR series and 13.3% (21 patients) in CH/SN-EVAR series. One possible explanation for the relatively increased mortality in the CH group is postprocedural renal dysfunction, which is a strong indicator of poor long-term survival 27 . In the current review, 30 (5.5%) renal events (renal impairment or failure) were reported in the F-EVAR series; 21.5% (34/158) patients suffered from postprocedural renal impairment or failure. Age is also a well-known predictor of mortality after AAA repair.
Endoleak is the most common procedure-specific feature and complications of chimney/snorkel and fenestrated grafting. The postoperative rate of type I endoleak was 7.6% (12/158) in the CH/SN-EVAR group, which was increased compared with the F-EVAR group (3.7% (20/542)) in the current review, excluding nine endoleaks of F-EVAR and one endoleak of CH/SN-EVAR that was detected and treated intraoperatively. In contrast to  F-EVAR, chimney grafts were positioned along the outside of the main abdominal endograft and rely on the close conformation of the endograft and the aortic wall around the chimney stent. The gaps that formed between the grafts and the aortic wall can be imagined as small cylinders and conduits (CGs and main graft) within a larger cylinder (the aorta). The gaps may have increased the risk of type I endoleakage in the CH-EVAR group.
Oversizing was considered an effective method of narrowing the gaps. In this study series, 5 studies, including 2 F-EVAR studies, reported increases in the main stent size ranging from 10% to 30%. Lachat proposed an elliptical   model for the estimation of the appropriate aortic stent graft diameter. Generally, to facilitate the formation of a good seal, the graft should increase in size by 30%. Some authors also recommend that the endograft should be up to 40% oversized to minimize the effects of the chimney gaps 28 . The ideal amount of oversizing remains undetermined 29 . Recent in vitro data demonstrated that increasing oversizing significantly decreased the sizes of gap areas, but main endograft in-folding was also detected in most oversized stentgrafts 28 . Interestingly, 8 type I endoleaks disappeared during follow-up. We hypothesize that the longer the gutters, the more resistance to blood flow and the more likely the gutters will thrombose. However, no evidence is provided to support these hypotheses. A high secondary intervention rate was noted after F-EVAR and CH/SN-EVAR. The reintervention rate was approximately 10.7% in the F-EVAR group and 9.95% of in the CH/SN-EVAR group during follow-up. Persistent endoleakage, renal artery stenosis, occlusion, and bleeding all require secondary intervention to relieve these procedures.
One limitation of this study is that some studies did not report all relevant information (i.e., the aneurysm neck length, information regarding stents, fluoroscopy time, and blood loss are not widely reported). Second, case studies and technical reports were excluded. The small number of patients included was insufficient for analysis, and this limitation may have led to underestimation of the rate of post-procedural complications. A number of acute and semi-acute procedures were performed in the CH/SN-EVAR group, whereas fenestrated stents required 4 to 6 weeks to measure and manufacture. Publication bias must also be acknowledged. Nevertheless, this review describes the current state of experience with fenestrated and chimney/snorkel techniques and provides considerable insight into the potential indications, technical considerations, and complications associated with these procedures. Juxtarenal aneurysm has no standard classification system that was applied throughout the current published works on EVAR; however, each of these reports cases of JAA. JAA was defined as cases in which the cross-clamp could not be placed above the infrarenal area safely during open surgery. In studies of EVAR, the term JAA typically refers to normal inter-renal aortic aneurysms without renal artery involvement. There are two situations in which it is unclear whether the term JAA should be applies: 1) extension of the AAA immediately above the inter-renal aorta and 2) aneurismal involvement of renal artery origins with an otherwise normal inter-renal aorta 30 . True comparisons of F-EVAR and CH/SN-EVAR can be made only when study participants are anatomically homogeneous. In the endovascular era, any new classification of JRA should include the location and diameter of the aneurysm and the length and angulation of the aneurysm neck.

Conclusion
F-EVAR and CH-EVAR techniques are both effective treatment for JAAs patients. The fenestrated technique was considered the priority treatment for JAAs, whereas CH-EVAR is frequently performed in patients with more       Table 9. Preoperative patient demographics and main outcomes in F-EVAR and CH-EVAR cohorts.