REVIEW ARTICLESRichard P. Cambria, MD, Section Editor

FromDC

TheBiNexND

AuthAddRep4Q

Thetom

0741Cophttp

242

Open repair versus fenestrated endovascularaneurysm repair of juxtarenal aneurysmsRohini Rao, BSc, Tristan R. A. Lane, MRCS, Ian J. Franklin, FRCS(Gen Surg), andAlun H. Davies, DM, FRCS, London, United Kingdom

Background: Open repair is the gold standard management for juxtarenal aneurysms. Fenestrated endovascular aneurysmrepair (FEVAR) is indicated for high-risk patients. The long-term outcomes of FEVAR are largely unknown, and there isno Level I comparative evidence. This systematic review and meta-analysis of case series compares elective juxtarenalaneurysm surgery by open repair and FEVAR.Methods: A systematic literature search was conducted for all published studies on elective repair of juxtarenal aneurysms byFEVAR and open repair. TheMEDLINE, EMBASE, and Cochrane databases were searched from 1947 to April 2013. Theexclusion criteria were case series of <10 patients or ruptured aneurysms. The primary outcomeswere perioperativemortalityand postoperative renal insufficiency. The secondary outcomes were secondary reinterventions and long-term survival.Results: We identified 35 case series with data on 2326 patients. Perioperative mortality was 4.1% in open repair andFEVAR case series (odds ratio for open repair with FEVAR, 1.059; 95% confidence interval, 0.642-1.747; P [ .822).Postoperative renal insufficiency was not significantly different (odds ratio for open repair with FEVAR, 1.136; 95%confidence interval, 0.754-1.713; P [ .542). FEVAR patients had higher rates of secondary reintervention, renalimpairment during follow-up, and a lower long-term survival compared with open repair patients.Conclusions: FEVAR and open repair have similar short-term outcomes but have diverging long-term outcomes that maybe secondary to the selection bias of FEVAR being offered to high-risk patients. FEVAR is a favorable option in high-riskpatients, and open repair remains viable as the gold standard. (J Vasc Surg 2015;61:242-55.)

Juxtarenal abdominal aortic aneurysms are defined asaneurysms with an infrarenal short proximal neck of<15 mm that require suprarenal cross-clamping duringopen repair.1 Open surgical repair of juxtarenal aneurysmshas been evolving and improving during the last 50 yearsby addressing the issue of renal ischemia from suprarenalcross-clamping through renal protection methods such ascold perfusion and renal artery bypass.2 It has achieved

the Academic Section of Vascular Surgery, Division of Surgery,epartment of Surgery & Cancer, Imperial College London, Charingross Hospital.research was supported by the National Institute for Health Researchomedical Research Centre based at Imperial College Healthcareational Health Service Trust and Imperial College London. The viewspressed are those of the authors and not necessarily those of theational Health Service, National Institute for Health Research, or theepartment of Health.or conflict of interest: none.itional material for this article may be found online at www.jvascsurg.org.rint requests: Rohini Rao, BSc, 30 St. James Rd, Flat 45, London SE16J, UK (e-mail: rohinirao91@gmail.com).editors and reviewers of this article have no relevant financial relationshipsdisclose per the JVS policy that requires reviewers to decline review of anyanuscript for which they may have a conflict of interest.-5214yright � 2015 by the Society for Vascular Surgery.://dx.doi.org/10.1016/j.jvs.2014.08.068

low mortality rates of between 0.8% and 5.1%,3-5 withestablished long-term graft durability.2

However, this procedure is still associated with bloodloss and prolonged in-hospital recovery as well as cardiacand pulmonary events, anastomotic bleeding, limb and renalischemia, and cerebrovascular events.6 Furthermore, withan aging population and an increasing prevalence of comor-bidities that serve as contraindications for open repair, clini-cians and patients are seeking to increase the availability anduse of minimally invasive endovascular procedures.

Endovascular aneurysm repair (EVAR) is in widespreaduse for infrarenal aneurysms and involves placing a fabric-covered metal stent graft, under fluoroscopic guidance,intraluminally into the abdominal aorta, usually throughthe femoral arteries. The endograft often has multiple com-ponents that are assembled in vivo, and the procedure usu-ally results in a bifurcated “inverted Y” graft inserteddistally in both the common iliac arteries and fixed proxi-mally just above the neck of the aneurysm to achieve a sta-ble graft seal between the nonaneurysmal aorta and theendograft.7 However, juxtarenal aneurysms are short-necked and thus do not have a sufficient length of infrare-nal aorta to achieve a good graft seal with minimal risks ofendoleak or graft migration, and hence are not suitable forEVAR. Furthermore, there is a risk that the endograftcould occlude the renal arteries.8

Fig 1. Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) diagram.

JOURNAL OF VASCULAR SURGERYVolume 61, Number 1 Rao et al 243

Fenestrated EVAR (FEVAR) describes a custom-madeendograft intended for short-necked aneurysms where theproximal seal lies above the renal or visceral arteries toachieve a sufficient attachment length and graft fixationand alignment.9 These endografts incorporate fenestra-tions, or holes in the sides of the endograft, to maintainarterial flow to the aortic branches covered by the stentgraft. These openings are carefully aligned with the renaland visceral artery ostia in vivo, thereby allowing the prox-imal sealing zone of the graft to extend to a region ofnormal aorta located above the renal arteries withoutoccluding the branching arteries. FEVAR is currentlylargely indicated for patients with juxtarenal aneurysmsthat are unsuitable for open repair due to their high surgicalrisk.10 Many centers advocate FEVAR as a viable alterna-tive to open repair of juxtarenal aneurysms.9

Studies comparing FEVAR with open repair of juxtare-nal aneurysms have found a significantly lower 30-day mor-tality rate for FEVAR compared with open repair.11,12

Whether FEVAR or open repair is superior in the long-term is still largely unknown. Because the procedure is

technically challenging and requires advanced skills inendovascular manipulation, it is limited to high-volumespecialist centers.1

However, because the apparent greater short-termbenefits of endovascular surgery cannot be denied, thereis a high level of uncertainty in the vascular communityregarding whether FEVAR or open repair is the moreeffective and safe management option for the narrow selec-tion criteria of patients with juxtarenal aneurysms.

This systematic review aims to compare the short-termand long-term outcomes in published case series during thepast 20 years of juxtarenal aneurysms treated by open repairand FEVAR. These results can determine whether there isa need to gain randomized controlled trial evidence ofopen repair vs FEVAR to answer the question of which isthe most effective management strategy to improve patientoutcomes and wellbeing.

METHODS

Search strategy. Two independent observers per-formed an Internet-based systematic literature search of

Table I. Patient demographics of all open repair case series included

StudyYear

published CountryPatients,

No.Mean

age, years

Meananeurysm

diameter, cmMales,No. (%)

Preoperativemean serumcreatinine,a

mmol/L

Preoperativerenal dysfunctionrate, No. (% of

patients)

Dubois19 2013 Canada 169 72.4 6.8 128 (75.7) 102.0 12 (7)Donas10 2012 Germany 31 71.2 6.0 27 (87) 97.2 2 (6)Tsai2 2012 USA 199 74.0 142 (71) 39 (20)Totsugawa20 2010 Japan 26b

Speziale21 2010 Italy 92 71.5 5.3 86 (94) 122.9 36 (39)Marrocco-Trischitta22 2009 Italy 32 70.4 5.9 28 (88) 100.8Chisci16 2009 Italy/Sweden 61 67.8 6.3 55 (89) 6 (10)Mukherjee23 2008 USA 56 69.0 Median, 5.7 45 (80)Knott3 2008 USA 126 73.6 6.3 98 (78) 22 (17)Yeung24 2008 Netherlands 23 70.0 6.7 18 (78) 146.7 12 (52)c

Pearce25 2007 USA 134 71.0 5.9 99 (74) 106.1 36 (27)Back26 2005 USA 78d

Ryan27 2004 USA 58Kudo28 2004 Japan 18Bicknell29 2003 UK 44 Median, 70.0 37 (84) Median, 113Shortell30 2003 USA 112 72.1 6.2 87 (78) 14 (12)Sarac5 2002 USA 138 71.7 6.4 92 (67) 27 (20)Ayari31 2001 France 53 Median, 69.0 Median, 6.0 51 (96) 21 (40)Giulini32 2000 Italy 56 Median, 67.0 Median, 6.0 51 (91) 4 (7)Allen33 1993 USA 31e

Poulias34 1992 Greece 38 Median, 66.0 38 (100) 6 (16)Pooled proportion Total 1575 70.9 6.1 1082 (82) 110.7 237 (19)

aPreoperative mean serum creatinine values were converted from mg/dL to mmol/L where required.bExcluding four cases of ruptured aneurysms.cPreoperative serum creatinine >1.25 mg/dL.dExcluding other visceral aneurysms.eExcluding suprarenal and ruptured aneurysms.

JOURNAL OF VASCULAR SURGERY244 Rao et al January 2015

medical databases, with the assistance of the Imperial Col-lege London Library staff. The medical databases searchedwere OVID MEDLINE In-Process & Other Non-Indexedcitations, and OVID MEDLINE from 1946 to April 7,2013. EMBASE classic and EMBASE were searched from1947 to April 11, 2013. The Cochrane Library was searchedon April 30, 2013. A combination of the following searchterms was used: “juxtarenal aneurysm,” “open repair,”“fenestrated endograft,” and “fenestrated graft.”

Also conducted was a search of PubMed using thefollowing Medical Subject Heading terms: AbdominalAortic Aneurysms, Vascular Surgical Procedure, BloodVessel Prostheses, Vascular Prosthesis Implantation, Endo-vascular Procedure, Stent and Endoleaks.

Hand searching of relevant past reviews and confer-ences was also done. Where information on the outcomesof elective juxtarenal repair was not available exclusively,or in the case of conference presentations, with no paperspublished, authors were contacted for unpublished data.Two authors conducted the literature searches indepen-dently using the same search terms with the same databasesand then discussed together which studies fit the inclusioncriteria to produce the final list of studies.

Inclusion criteria. Cohort studies or case series con-taining at least 10 patients, published after 1990 in English,were included. Any studies describing patients treated forelective open repair of juxtarenal aneurysms were consid-ered. The term “juxtarenal aneurysm” was sufficient to

warrant inclusion, without the requirement of a specificanatomic definition. No exclusion criteria were based onthe operative technique of open repair (eg, certain renalprotection measures).

Studies describing elective FEVAR of juxtarenal aneu-rysms using all currently licensed stent grafts were included.The literature search resulted in the inclusion of studies thatused the Zenith Fenestrated AAA Endovascular Graft(CookMedical, Bloomington, Ind). This graft incorporatestwo fenestrations for the renal artery ostia and a scallop witha covered stent for the superior mesenteric artery. TheFEVAR patients included in the review were all treatedwith purely fenestrated devices without branched/fenes-trated device combinations.

Exclusion criteria. Studies of complex aneurysms(pararenal and type IV thoracoabdominal) and rupturedaneurysms were excluded, unless the author could provideunpublished data for elective juxtarenal aneurysms only.Patients treated with endografts other than custom-madefenestrated stent grafts, such as hybrid repairs, branchedendografts, chimney grafts, snorkel, and endowedge tech-niques, were excluded. Registry data were not included dueto duplicate publication.

Outcomes. The primary outcomes were perioperativemortality (30-day and in-hospital mortality), postoperativerenal insufficiency (based on the study definition of renalinsufficiency), postoperative permanent dialysis, and renalfailure development during follow-up (persistent renal

Table II. Patient demographics of all fenestrated endovascular aneurysm repair (FEVAR) case series included

StudyYear

published CountryPatients,

No.Mean

age, years

Meananeurysm

diameter, cmMales,No. (%)

Preoperativecreatininemean,a

mmoll/L

Preoperativerenal dysfunctionrate, No. (% of

patients)

Donas10 2012 Germany 29 73.7 6.5 29 (100) 106.1 5 (17)Constantinou35,b 2012 UK 64 53 (83) 19 (30)Metcalfe36,c 2012 UK 24 70.5 6.5 23 (96) 99Coscas37 2012 France 38 74 5.8 34 (89) 3 (8)Verhoeven38 2010 Netherlands/Germany 100 72.6 Median, 6.0 87 (87) 46 (46)Mastracci39 2010 USA 150 75.1 6.6 127 (85) 107.3 33 (22)Chisci16 2009 Italy/Sweden 52 71.5 6 42 (81) 12 (23)Beck40 2009 Netherlands 18 72.4 16 (89) 1 (6)Greenberg41 2009 USA/Japan 30 75 6.1 24 (80) 2 (7)Kristmundsson42 2009 Sweden 54 Median, 72.0 Median, 6.0 46 (85) 24 (44)Scurr43 2007 UK 45 Median, 73.0 Median, 6.8 41 (91) 6 (13)Semmens44 2006 Australia 58 75.5 51 (88)Halak45 2006 Australia 17Haddad46 2005 USA 72 75 6.2 97.2 23 (32)Pooled proportion Total 751 73.2 6 573 (86) 103.3 129 (21)

aPreoperative mean creatinine values were converted from mg/dL to mmol/L where required.bIncluding patients with temporary axillobifemoral bypass grafting.cUnpublished data.

Table III. Comorbidity rates in open repair andfenestrated endovascular aneurysm repair (FEVAR)studies

Preoperative comorbidity Open repair FEVAR Pa

Mean creatinine, mmol/L 105.8 100.9 <.0001Rate of

Renal insufficiency, % 19 21.4 <.0001Cardiac dysfunction, % 43.3 52.2 .03Respiratory insufficiency, % 28.9 33.7 .01Diabetes, % 11.8 17.4 .001Hypertension, % 78.8 76.8 .02

aP values were calculated by the Fisher Z test for significance.

JOURNAL OF VASCULAR SURGERYVolume 61, Number 1 Rao et al 245

failure at discharge or progression to renal failure overfollow-up). The secondary outcomes included majorpostoperative complications, secondary reintervention ratesduring follow-up, and long-term survival (Kaplan-Meierdata). Independent outcomes for FEVAR included targetvessel patency and endoleak rates.

Heterogeneity. The open repair and FEVAR cohortswere assessed for statistical heterogeneity in perioperativemortality using the c2 Cochran Q statistic to determinethe validity of performing a meta-analysis of outcomes.

Statistical analysis. The rates of each outcome (eg,perioperative mortality, postoperative renal insufficiency)were calculated for each study and compared among studiesby a meta-analysis of proportions with a random effectsmodel. The pooled event rate in the open repair studies wascompared with the pooled event rate in the FEVAR studiesto calculate an odds ratio comparing the odds of an outcomeoccurring in open repair studies vs FEVAR studies. Forcontinuous variables (eg, patient age, mean preoperativeserum creatinine), where studies had provided a medianvalue and range, the mean and standard deviation were

estimated using a formula by Hozo et al13 to pool the re-sults. All data analysis was performed on ComprehensiveMeta-analysis 2 software (Biostat, Englewood, NJ).

For studies where Kaplan-Meier data were available (sur-vival curves and numbers at risk), the number of deaths atdifferent time intervals in each individual study was calcu-lated using the Parmar technique for summary statisticextraction.14 The curves were then reproduced on asingle graph. For the purpose of testing the difference inlong-term survival between open repair and FEVAR, thenumber of deaths and numbers at risk in all open repairand all FEVAR studies were pooled. These data werethen used to produce a hypothetical Kaplan-Meier graphwith two survival curves (open repair vs FEVAR) usingSPSS software (IBM Corp, Armonk, NY). A log-rank(Mantel-Cox) test for a significant difference in survivalbetween the two groups was calculated.15

RESULTS

We identified 33 relevant studies from 1296 articles.The inclusion and exclusion criteria are outlined in thePreferred Reporting Items for Systematic Reviews andMeta-Analyses (PRISMA) diagram (Fig 1).

From these 33 studies, 31 were case series and twowere cohort studies.10,16 The two cohort studies were splitinto four independent case series (two open repair and twoFEVAR) for the purpose of comparison and meta-analysiswith the other case series. Thus, a total of 35 case serieswere included in the final analysis: 21 case series of openrepair and 14 case series of FEVAR.

Interobserver agreement

The k statistic for interobserver agreement for inclusionor exclusion into the study was 0.47, which shows moder-ate agreement between the two authors.17

Fig 2. Forest plot shows the mortality rate per study with 95% confidence intervals (CIs; horizontal lines), and pooledmortality rate (represented by a diamond) for all open repair (OR) and all fenestrated endovascular aneurysm repair(FEVAR) case series combined. The odds ratio favors FEVAR but is not significant.

JOURNAL OF VASCULAR SURGERY246 Rao et al January 2015

Study quality assessment

The included studies were assessed using a validated qual-ity of observational studies checklist (Appendix, onlineonly).18 The checklist was amended with certain parametersrelevant to this study, and nonrelevant items were removed.The items of the checklist are binary data, with 0 representingabsence of the information and 1 representing presence of theinformation in the study. The total score acts as a comparatorof the quality of the included studies. Themaximum scorewas13 for open repair studies and 15 for FEVAR studies.

Publication bias

The Egger test showed evidence of publication biaswith P < .0001, bias ¼ �1.32, and 95% confidence interval(CI) �1.98 to �0.66 (Supplementary Fig 1, online only).

Patient characteristics

A total of 2326 patients were included, of whom 1575underwent open repair and 751 underwent FEVAR. Pa-tient characteristics are outlined in Tables I and II. In 16open repair studies and 12 FEVAR studies, men made up82% and 86% of the study population, respectively, whichwas not a significant difference (P ¼ .05). The mean age

was 71 years in the 15 open repair studies and 73 yearsin the 12 FEVAR studies. The standard difference inmean age was 4.986 years (95% CI, 4.802-5.170; P <.0001) suggesting that FEVAR patients were significantlyolder than open repair patients. The mean aneurysm diam-eter was 6.1 cm in 11 open repair studies and 6.0 cm ineight FEVAR studies.

Comorbidities

FEVAR patients had significantly higher rates ofpreoperative renal insufficiency, cardiac dysfunction, res-piratory insufficiency, and diabetes (Table III). Howev-er, open repair patients had a significantly higherpreoperative mean serum creatinine levels and rates ofhypertension.

Confounding factors

Various confounding factors, such as the year of studypublication (Supplementary Fig 2, online only), the coun-try of publication (Supplementary Fig 3, online only), theseries size (Supplementary Fig 4, online only), and the pre-operative serum creatinine (Supplementary Fig 5, onlineonly) may have influenced outcomes significantly.

Fig 3. Forest plot shows the postoperative renal insufficiency rate per study with 95% confidence intervals(CIs, horizontal lines), and pooled rate (represented by a diamond) for all open repair (OR) and all fenestratedendovascular aneurysm repair (FEVAR) case series combined. The odds ratio favors FEVAR but is not significant.

JOURNAL OF VASCULAR SURGERYVolume 61, Number 1 Rao et al 247

Outcomes

Perioperative mortality. The pooled perioperativemortality was 0.041 (95% CI, 0.031-0.053) or 4.1% in21 case series of open repair and 0.041 (95% CI 0.028-0.059) or 4.1% in 14 case series of FEVAR (Fig 2). Theodds ratio for open repair with FEVAR was 1.059 (95% CI,0.642-1.747; P ¼ .822) and not statistically significant.The Cochran Q test showed low heterogeneity betweenstudies (Q ¼ 28.8; P ¼ .752).

Postoperative renal insufficiency. Postoperative renalinsufficiency rates were reported in 20 open repair case se-ries and in five FEVAR studies (Fig 3). The definitionsvaried between studies, but most involved the patient’sserum creatinine rising acutely by w44 mmol/L (0.5 mg/dL) or a certain percentage (20%-50%) above baseline, oran absolute rise above 133 to 354 mmol/L (1.5-4 mg/dL)before discharge. The pooled rates were 0.139 (95% CI,0.101-0.188) or 13.9% for open repair and 0.114 (95% CI,0.041-0.276) or 11.4% for FEVAR. The odds ratio foropen repair with FEVAR was 1.136 (95% CI, 0.754-1.713;P ¼ .542) and not statistically significant.

Postoperative permanent dialysis. The rates of newpermanent hemodialysis (not resolved before discharge)were reported in 15 open repair studies and in seven

FEVAR studies (Fig 4). The pooled rates were 0.028(95% CI, 0.017-0.044) or 2.8% for open repair and 0.019(95% CI, 0.008-0.044) or 1.9% for FEVAR. The odds ratiofor open repair with FEVAR was 1.660 (95% CI, 0.627-4.397; P ¼ .308) and not statistically significant.

Renal failure during follow-up. Persistent renalinsufficiency at discharge or development of renal failureduring follow-up was reported in 11 of 21 open repairstudies and in six of 14 FEVAR studies (Fig 5). The pooledevent rate was 0.077 (95% CI, 0.050-0.118) or 7.7% inopen repair and was 0.197 (95% CI, 0.147-0.258) or19.7% in FEVAR. The odds ratio for open repair withFEVAR was 0.375 (95% CI, 0.263-0.536; P < .0001),which was significant in favor of open repair, indicating thatmore FEVAR patients developed renal failure duringfollow-up than open repair patients.

Secondary reintervention rate. Nine of 21 openrepair studies and nine of 14 FEVAR studies reported therequirement for graft-related secondary reinterventionsduring follow-up (between 0 and 120 months.) Most openrepair reinterventions were for bleeding during surgery(Table IV), whereas those for FEVAR were for endoleak orvessel occlusion/stenosis. The pooled event rate was 0.049(95% CI, 0.029-0.082) or 4.9% for open repair and 0.127(95% CI 0.090-0.117) or 12.7% for FEVAR (Fig 6). The

Fig 4. Forest plot shows the postoperative permanent hemodialysis rate per study with 95% confidence intervals (CIs,horizontal lines), and pooled rate (represented by a diamond) for all open repair (OR) and all fenestrated endovascularaneurysm repair (FEVAR) case series combined. The odds ratio favors FEVAR but is not significant.

Fig 5. Forest plots show the rate of renal failure over follow-up per study, with 95% confidence intervals (CIs, horizontallines), and pooled rate (represented by a diamond) for all open repair and fenestrated endovascular aneurysm repair(FEVAR) case series combined. The odds ratio favors open repair and is significant.

JOURNAL OF VASCULAR SURGERY248 Rao et al January 2015

Table IV. Secondary reinterventions during follow-up in open repair and fenestrated endovascular aneurysm repair(FEVAR) studies

Study YearDetails of complication requiring

secondary reinterventionTime of

reinterventionMorbidity fromreintervention

Meanfollow- up

Open repairDonas10 2012 d Acute left popliteal artery

occlusion and rest pain treatedby popliteal vein bypass grafting

Perioperative Uneventful 24 months

Tsai2 2012 d 5 remote descending thoracicaneurysms: 4 treated byTEVAR and 1 by open thoracicaortic aneurysm repair

d Infection of left aortofemoralbypass treated by graft resectionand right femoropoplitealbypass

1 monthpostoperative

Uneventful 56 months

Speziale21 2010 d Bleeding treated by secondlaparotomy

d Acute lower limb ischemiatreated by Fogarty catheterembolectomy

Perioperative 54 months

Chisci16 2009 d Fogarty embolectomy þ iliacstent

d 3 incisional hernia repairsd Fogarty embolectomy þrecanalization

d Major amputation

19.5 months

Sarac5 2002 d Bleeding treated by secondlaparotomy

d Lower extremity embolic eventstreated

Allen33 1993 d 2 cases of bleeding treated bysecond laparotomy

Poulias34 1992 d Bleeding treated by secondlaparotomy

Perioperative

FEVARDonas10 2012 d Symptomatic left renal artery

occlusion treated by openthrombectomy of artery andDacrona ileorenal bypassgrafting

d Symptomatic superiormesentericartery restenosis treated byadditional stent placement

6 months Uneventful

Metcalfe36 2012 d Type II endoleak treated byrenal stent dilation

d Washout of groin wound withsartorius flap reconstruction

d Iliac limb occlusion treated byfemorofemoral crossover graft

d Kinked iliac limb treated by self-expanding stent implantation

4-6 months,12 months

13 months

Verhoeven38 2010 d Suspected mesenteric ischemiatreated by laparotomy, bypassright renal artery and stent graftpull down by laparotomy

d Iliac stenosis treated by stentingd Late renal artery occlusionin 2 patients treated bycatheterization

d Persistent inferior mesentericartery thrombosis treated bycoil embolization

d Persistent lumbar arteriestreated by laparotomy

d Iliac limb occlusion treated bycross-over bypass

Perioperative,30 days,2 months-18months

d Failed catheterizations ofsevere renal artery stenosisat 2 and 18 monthsresulting in renal arteryocclusion

d Failed renal arterycatheterization at13 months

d Cross-over bypassdisconnect between tubeand bifurcation treated bythoracic stent graft at20 months

24 months

Beck40 2009 None 23 months

(Continued on next page)

JOURNAL OF VASCULAR SURGERYVolume 61, Number 1 Rao et al 249

Table IV. Continued.

Study YearDetails of complication requiring

secondary reinterventionTime of

reinterventionMorbidity fromreintervention

Meanfollow- up

Greenberg41 2009 d 3 patients treated by angioplastyat 6, 12, and 24 months forin-stent stenosis of right renalartery; 1 patient treated forbilateral renal artery in-stentstenosis at 6 months.

d Left renal artery occlusiontreated by left iliac-renal bypass

6, 12, 24 months Uneventful 24 months

Chisci16 2009 d Fogarty embolectomy þfibrinolysis

d Major amputationd Fogarty embolectomy þfibrinolysis

d Stent graft þ coil embolizationof superficial femoral artery

d Renal artery stent graftd Superior mesenteric artery stent

19.5 months

Kristmundsson42 2009 d Distal type I endoleak treatedby Giant Palmazb stent in rightiliac artery

d Type II endoleak treated by coilembolization of inferiormesenteric artery and glueembolization of lumbar arteries

d Renal artery stenoses treated bypercutaneous transluminalangioplasty at 12 months.

Perioperative,1-18 months

25 months

Scurr43 2007 d Distraction of graft componentstreated by bridging cuffaddition at 28 months

d 2 iliac limb extensions at 1 and24 months

d Progressive stenotic diseasetreated by iliac angioplasty at6 months

d Stenosis within stented superiormesenteric artery treated byangioplasty at 15 days

Perioperative,1 month,28 months

d Patient with distraction ofgraft components laterdied of pancreatitis

d Patient with superiormesenteric artery stenosisdeveloped critical stenosisat 6 months, which wassuccessfully treated byiliac-superior mesentericartery bypass.

24 months

Semmens44 2006 d Unresolved type I endoleaktreated by Giant Palmaz Stent

d Persistent type II endoleakstreated by coil embolizationand laparotomy

d Distal component graftmigration downwards treatedby bridge reinforcement

d Distal extension to rightexternal iliac artery treated bycoil occlusion

d Debridement and resuture ofwound

Perioperative,16.8 months

d 2 patients died #30 daysof secondary procedure

d 1 patient died aftersecondary procedure(performed 2 years afterprimary procedure)

16.8 months

TEVAR, Thoracic aortic aneurysm repair.aCordis, Miami Lakes, Fla.bDuPont, Wilmington, Del.

JOURNAL OF VASCULAR SURGERY250 Rao et al January 2015

odds ratio for open repair with FEVAR was 0.334 (95% CI,0.212-0.525; P < .0001), which was significant in favorof open repair, suggesting that FEVAR patients have toundergo a greater number of repeat procedures duringfollow-up.

Major complications. Open repair studies had asignificantly higher postoperative major complication rate

compared with FEVAR (25.0% vs 15.7%; P ¼ .001). Theseincluded cardiac events (myocardial infarction, arrhyth-mias), pulmonary infections, ischemic colitis, pulmonaryembolism, and wound complications.

Long-term survival. Only two open repair studiesand five FEVAR studies provided Kaplan-Meier curveswith numbers at risk. The survival data for open repair

Fig 6. Forest plot shows the secondary intervention rate per study with 95% confidence intervals (CIs, horizontal lines),and pooled rate (represented by a diamond) for all open repair (OR) and all fenestrated endovascular aneurysm repair(FEVAR) case series combined. The odds ratio favors open repair and is significant.

Fig 7. Hypothetical estimated Kaplan-Meier survival curves(pooled) of all patients combined from two open repair studies(Ayari,31 2001 and Knott3 2008) and five fenestrated endovascularaneurysm repair (FEVAR) studies (Beck,40 2009; Kristmundsson,42

2009; Mastracci,39 2010; Verhoeven,38 2010; and Metcalfe,36

2012).

JOURNAL OF VASCULAR SURGERYVolume 61, Number 1 Rao et al 251

and FEVAR studies were pooled to produce two hy-pothetical Kaplan-Meier curves that were compared by alog-rank (Mantel-Cox) test for significance (Fig 7).FEVAR had a significantly lower long-term survival thanopen repair (P ¼ .09) with 93%, 74%, and 55% survivalat 1, 3 and 5 years, respectively; whereas open repairhad 89%, 80%, and 73% survival at 1, 3, and 5 years,respectively.

Outcomes for FEVAR

Target vessel patency. Target vessel patency is animportant measure of long-term graft resilience and likeli-hood of survival. The case series reported in Table V had apooled vessel patency of 98.2% at completion angiography.In the long-term (5 years), the studies reported a 95.8%patency during follow-up.

Endoleak. The rates of early types I and III endoleakwere 5.8% and 2.6%, respectively, and rates of late endo-leaks were 2.2% and 3.2% (Table VI).

DISCUSSION

This systematic review demonstrates that open repairand FEVAR of juxtarenal aneurysms have an equivalentsafety profile in perioperative mortality and early postoper-ative renal impairment. FEVAR has been shown to portraythe benefits of minimally invasive surgery with lower ratesof infection and bleeding. In addition, this review hashighlighted low rates of endoleak and high rates of targetvessel preservation after FEVAR, which are a testamentto excellent technical application, and it provides a useful

option for clinicians whose patients are unfit for openrepair.

The FEVAR cohort had an increased progression torenal impairment during follow-up, higher secondary

Table V. Target vessel patency in fenestratedendovascular aneurysm repair (FEVAR) case series atcompletion angiography and during follow-up

Study Year

Target vessel patency

At completionangiography, No. (%)

During follow-up,No. (%)

Donas10 2012 43/44Constantinou35 2012 159/161Metcalfe36,a 2012 42/42Verhoeven38 2010 272/275 263/275Beck40 2009 56/56 53/56Greenberg41 2009 77/77Kristmundsson42 2009 131/134 129/134Scurr43 2007 82/82 79/82Halak45 2006 35/35Semmens44 2006 105/116Haddad46 2005 141/142Total target

vessel patency1143/1164 (98.2) 524/547 (95.8)

aUnpublished data.

JOURNAL OF VASCULAR SURGERY252 Rao et al January 2015

reinterventions, and a lower long-term survival. This maybe attributable to higher rates of preoperative comorbidity;however, definitive evidence of this would require a ran-domized controlled trial or cohort study with matched pa-tients and long-term follow-up of 5 to 10 years, which maybe unfeasible. FEVAR presents other factors that maycontribute to this progressive renal damage, includingcontrast load from lifelong computed tomography (CT)surveillance.47 Because FEVAR showed excellent targetvessel patency rates in this study, target vessel occlusion isunlikely to be a cause of the increased rate of renal failurein the long-term.

The FEVAR patients were on average 5 years olderthan those undergoing open repair, with higher rates ofpreoperative renal impairment, diabetes, cardiovascular dis-ease, and respiratory insufficiency; therefore, the worsenedoutcomes for mortality and development of renal insuffi-ciency may be expected. There is, however, significantclinical and methodologic heterogeneity among thestudies, and this may confound the interpretation of theseoutcomes.

Varied definitions of preoperative and postoperativerenal impairment were used and not always based on an ab-solute value of serum creatinine. This inconsistency mayexplain the discrepancy found that open repair patientshad a higher mean preoperative serum creatinine thanFEVAR patients; however, the pooled preoperative renalinsufficiency rate (based on study definition) was higheramong FEVAR patients. Consistent reporting amongstudies and the use of standardized assessment of renalimpairment in the future is needed to avoid confounding re-sults. Other confounding factors that may have an influenceon improved outcomes are highlighted by meta-regressionanalysis (Supplementary Figs 2-5), and include more recentstudies, studies that took place in countries such as theUnited States, Japan, Italy, or the Netherlands, increasingstudy size, and lower preoperative serum creatinine.

The perioperative mortality rates in open repair andFEVAR studies were exemplary and equivalent (4.1%),despite a higher rate of postoperative major complicationswith open repair (P ¼ .001) and reinterventions due tobleeding and embolic events. Jongkind et al48 found a2.9% rate for pooled perioperative mortality after electiveopen juxtarenal repair, and Nordon et al11 reported 3.6%in their systematic review. These low mortality rates maybe attributable to improvements in renal and visceral pro-tective strategies during open repair (cold perfusion andmannitol).12

The improved 5-year survival rates seen with openrepair (73% vs 55%) may be secondary to selection bias,with higher-risk patients offered FEVAR; alternatively,with secondary reintervention rates of FEVAR significantlyhigher than in open repair, such reintervention may ac-count for the decreased long-term survival. Most of thereinterventions for FEVAR occurred between 6 and12 months postoperatively, whereas those required inopen repair were in the short-term (#30 days), whichputs into question the durability of FEVAR (Table IV).

Whereas most patients with secondary reinterventionsin the open repair group had uneventful recoveries, somecases of reintervention from FEVAR led to significantmorbidity and mortality. These cases were due to distrac-tion of graft components, endoleaks, graft kinking, andrenal artery occlusion leading to failed catheterizationsand death #30 days of reintervention. This would beconsistent with the increased risk of aneurysm rupturefrom endoleaks, graft migration, and kinking seen inEVAR.49 However, considering the excellent target vesselpatency rates and low endoleak rates seen in this review,as well as the literature findings of very low graft migrationrates of between 1.5% and 4.2%,50,51 the effect of graftdurability on survival appears complex.

With respect to the development of renal failure in thelong-term, although limited by heterogenous reporting ofrenal outcomes, this study indicates that FEVAR patientsdo not risk significantly higher rates of immediate postop-erative renal dysfunction due to procedurally induced athe-roemboli. The higher rates of progression to renalimpairment are likely secondary to worse baseline renalfunction. However, our findings lend weight to the sugges-tion that as FEVAR patients undergo CT angiography sur-veillance at short intervals, contrast nephropathy is also alikely significant contributor to the development of long-term renal impairment.47 This should encourage the devel-opment and practice of ultrasound follow-up for FEVARpatients as for EVAR.52 Greenberg et al53 hypothesizethat delayed renal impairment in the FEVAR group maybe due to contrast-induced nephropathy, delayed deposi-tion of emboli triggered during the procedure, and renaldamage from suprarenal stent fixation. This may alsoexplain the increased progression to renal impairment dur-ing follow-up with FEVAR in this study.

The main limitation of this study is the reporting biasof the studies, which showed a lack of information on theoperative experience of the surgeon or the volume of

Table VI. Endoleak rates in fenestrated endovascular aneurysm repair (FEVAR) case series at completion angiographywithin the first 6 months after surgery and after 6 months

Study Year

Endoleaks at completion angiography Early endoleaks (#6 months) Late endoleaks (>6 months)

Type I,No. (%)

Type II,No. (%)

Type III,No. (%)

Type I,No. (%)

Type II,No. (%)

Type III,No. (%)

Type I,No. (%)

Type II,No. (%)

Type III,No. (%)

Donas10 2012 1 (3.4)Coscas37 2012 2 (5.3) 2 (5.3) 1 (2.6) 6 (15.8) 7 (18.4) 1 (2.6) 1 (2.6) 4 (10.5) 1 (2.6)Metcalfe36 2012 3 (12.5) 3 (12.5) 1 (4.2) 2 (8.3) 1 (4.2)Greenberg41 2009 8 (26.7) 6 (20.0) 4 (13.3)Kristmundsson42 2009 3 (5.6) 13 (24.1) 1 (1.9) 3 (5.6)Chisci16 2009 2 (3.8) 1 (1.9) 13 (25.0)Scurr43 2007 4 (8.9)Semmens44 2006 4 (6.9) 2 (3.4)Halak45 2006 0 (0.0)Total endoleak rate, % 5.8 12.9 2.6 15.8 14.4 2.6 2.2 13.1 3.2

JOURNAL OF VASCULAR SURGERYVolume 61, Number 1 Rao et al 253

outcomes during follow-up, which would have enabled amore robust comparison of open repair and FEVAR.Also lacking is follow-up and long-term outcomes informa-tion and target vessel patency for open repair studies. Therewas a lack of follow-up data on the rates of graft thrombosisand stenosis in the open repair studies, which preventedanalysis and comparison with long-term graft durabilitydata in the FEVAR group. Furthermore, FEVAR hadmore variable outcomes than open repair, which may beattributable to the heterogeneity of this complex procedureinvolving intricate and significant endovascular manipula-tion, or the relative youth of the procedure that confoundsthe validity of outcomes. The systematic literature searchdid not include potentially valuable sources of data dueto duplicate publications.54

From the study quality assessment, it was clear that low-scoring studies failed to provide clear inclusion and exclu-sion criteria, did not report fully the statistical parametersrequired for a meta-analysis (eg, standard deviation), anddid not report long-term outcomes. The heterogenous def-initions of renal insufficiency and juxtarenal aneurysmsmadethe accurate and valid assimilation of the data difficult.

In the future, surgeons may wish to focus on continuingto improve endovascular techniques to improve the long-term durability and reduce the rate of reinterventions andtheir consequent morbidity. All stent grafts described inthis study were Cook Zenith endografts. These stents arebespoke for each patient, taking 4 to 6 weeks to manufac-ture, with high expense. Multiple other devices are nowavailable, and off-the-shelf fenestrated grafts are being eval-uated for juxtarenal aneurysms, which should reduce costsfor FEVAR treatments and might alter the balance ofpreferred treatments.55

A randomized controlled trial comparing open repairwith FEVAR would provide Level 1 evidence to guide clini-cians in the future, based on this study’s findings that openrepair appears to be safe in appropriate patients and durable,with good long-term outcomes, whereas FEVAR has showna similar efficacy to open repair and appears to be a safe andvalid option for high-risk patients who are unsuitable for

open repair. A comparative study would require consider-able numbers to investigate further the small, nonsignificantdifferences in procedural safety seen in this analysis.

CONCLUSIONS

Overall, open repair and FEVAR can both be usedsafely and with low postoperative morbidity and mortalityin the short-term. The concerns of suprarenal cross-clamping and periprocedural atheroemboli causing highrates of postoperative renal impairment in both open repairand FEVAR, respectively, are not supported by the evi-dence, with both offering similar and low rates of postop-erative renal impairment. However, FEVAR has a higherrate of progression to renal impairment during follow-up,a greater rate of secondary reinterventions, and a lowerlong-term survival, which may be partly attributed toFEVAR currently being performed on high-risk patientsbut may also be due to perioperative contrast use andrecurrent contrast CT surveillance. FEVAR has shownexcellent outcomes for high-risk patients, and continuedimprovements in technique and devices to lower therequirement for secondary reinterventions may, in thefuture, encourage a FEVAR first strategy in anatomicallyfeasible patients.

We thank the members of the Academic Section ofVascular Surgery of Imperial College London for theiradvice and help and the Imperial College Library Staff fortheir support with the systematic literature search.

AUTHOR CONTRIBUTIONS

Conception and design: TL, AD, IFAnalysis and interpretation: RR, TL, ADData collection: RR, TLWriting the article: RR, TLCritical revision of the article: RR, TL, IF, ADFinal approval of the article: RR, TL, IF, ADStatistical analysis: RR, TLObtained funding: TL, ADOverall responsibility: RR

JOURNAL OF VASCULAR SURGERY254 Rao et al January 2015

REFERENCES

1. Health Quality Ontario. Fenestrated endovascular grafts for the repairof juxtarenal aortic aneurysms: an evidence-based analysis. Ont HealthTechnol Assess Ser 2009;9:1-51.

2. Tsai S, Conrad MF, Patel VI, Kwolek CJ, LaMuraglia GM,Brewster DC, et al. Durability of open repair of juxtarenal abdominalaortic aneurysms. J Vasc Surg 2012;56:2-7.

3. Knott AW, Kalra M, Duncan AA, Reed NR, Bower TC, Hoskin TL,et al. Open repair of juxtarenal aortic aneurysms (JAA) remains a safeoption in the era of fenestrated endografts. J Vasc Surg 2008;47:695-701.

4. West CA, Noel AA, Bower TC, Cherry KJ, Gloviczki P, Sullivan TM,et al. Factors affecting outcomes of open surgical repair of pararenalaortic aneurysms: a 10-year experience. J Vasc Surg 2006;43:921-7;discussion: 927-8.

5. Sarac TP, Clair DG, Hertzer NR, Greenberg RK, Krajewski LP,O’Hara PJ, et al. Contemporary results of juxtarenal aneurysm repair.J Vasc Surg 2002;36:1104-11.

6. Arko FR, Hill BB, Olcott C, Harris EJ, Fogarty TJ, Zarins CK.Endovascular repair reduces early and late morbidity compared to opensurgery for abdominal aortic aneurysm. J Endovasc Ther 2002;9:711-8.

7. Greenhalgh RM, Powell JT. Endovascular repair of abdominal aorticaneurysm. N Engl J Med 2008;358:494-501.

8. Bruen KJ, Feezor RJ, Daniels MJ, Beck AW, Lee WA. Endovascularchimney technique versus open repair of juxtarenal and suprarenalaneurysms. J Vasc Surg 2011;53:895-904; discussion: 904-5.

9. Monahan TS, Schneider DB. Fenestrated and branched stent grafts forrepair of complex aortic aneurysms. Semin Vasc Surg 2009;22:132-9.

10. Donas KP, Eisenack M, Panuccio G, Austermann M, Osada N,Torsello G. The role of open and endovascular treatment with fenes-trated and chimney endografts for patients with juxtarenal aortic an-eurysms. J Vasc Surg 2012;56:285-90.

11. Nordon IM, Hinchliffe RJ, Holt PJ, Loftus IM, Thompson MM.Modern treatment of juxtarenal abdominal aortic aneurysms withfenestrated endografting and open repairda systematic review. Eur JVasc Endovasc Surg 2009;38:35-41.

12. Tsilimparis N, Perez S, Dayama A, Ricotta JJ. Endovascular repair withfenestrated-branched stent grafts improves 30-day outcomes for com-plex aortic aneurysms compared with open repair. Ann Vasc Surg2013;27:267-73.

13. Hozo SP, Djulbegovic B, Hozo I. Estimating the mean and variancefrom the median, range, and the size of a sample. BMC Med ResMethodol 2005;5:13.

14. Parmar MK, Torri V, Stewart L. Extracting summary statistics toperform meta-analyses of the published literature for survival endpoints.Stat Med 1998;17:2815-34.

15. Bland JM, Altman DG. The logrank test. BMJ 2004;328:1073.16. Chisci E, Kristmundsson T, de Donato G, Resch T, Setacci F,

Sonesson B, et al. The AAA with a challenging neck: outcome of openversus endovascular repair with standard and fenestrated stent-grafts.J Endovasc Ther 2009;16:137-46.

17. Viera AJ, Garrett JM. Understanding interobserver agreement: thekappa statistic. Fam Med 2005;37:360-3.

18. Downs SH, Black N. The feasibility of creating a checklist for theassessment of the methodological quality both of randomised and non-randomised studies of health care interventions. J Epidemiol Com-munity Health 1998;52:377-84.

19. Dubois L, Durant C, Harrington DM, Forbes TL, Derose G,Harris JR. Technical factors are strongest predictors of postoperativerenal dysfunction after open transperitoneal juxtarenal abdominal aorticaneurysm repair. J Vasc Surg 2013;57:648-54.

20. Totsugawa T, Yoshitaka H, Kuinose M, Ishida A, Tsushima Y,Chikazawa G. Prosthetic sleeve reinforcing for juxtarenal abdominalaortic aneurysm: experience of 30 cases. Ann Thorac Cardiovasc Surg2010;16:380-4.

21. Speziale F, Ruggiero M, Sbarigia E, Marino M, Menna D. Factorsinfluencing outcome after open surgical repair of juxtarenal abdominalaortic aneurysms. Vascular 2010;18:141-6.

22. Marrocco-Trischitta MM, Melissano G, Kahlberg A, Vezzoli G,Calori G, Chiesa R. The impact of aortic clamping site on glomerularfiltration rate after juxtarenal aneurysm repair. Ann Vasc Surg 2009;23:770-7.

23. Mukherjee D, Becker TE. An update on the “fast-track” abdominalaortic aneurysm repair. Int J Angiol 2008;17:93-7.

24. Yeung KK, Jongkind V, Coveliers HM, Tangelder GJ, Wisselink W.Routine continuous cold perfusion of the kidneys during elective jux-tarenal aortic aneurysm repair. Eur J Vasc Endovasc Surg 2008;35:446-51.

25. Pearce JD, Edwards MS, Stafford JM, Deonanan JK, Davis RP,Corriere MA, et al. Open repair of aortic aneurysms involving the renalvessels. Ann Vasc Surg 2007;21:676-86.

26. Back MR, Bandyk M, Bradner M, Cuthbertson D, Johnson BL,Shames ML, et al. Critical analysis of outcome determinants affectingrepair of intact aneurysms involving the visceral aorta. Ann Vasc Surg2005;19:648-56.

27. Ryan SV, Calligaro KD, McAffee-Bennett S, Doerr KJ, Chang J,Dougherty MJ. Management of juxtarenal aortic aneurysms andocclusive disease with preferential suprarenal clamping via a midlinetransperitoneal incision: technique and results. Vasc Endovascular Surg2004;38:417-22.

28. Kudo FA, Nishibe T, Miyazaki K, Murashita T, Yasuda K, Ando M,et al. Postoperative renal function after elective abdominal aorticaneurysm repair requiring suprarenal aortic cross-clamping. Surg Today2004;34:1010-3.

29. Bicknell CD, Cowan AR, Kerle MI, Mansfield AO, Cheshire NJW,Wolfe JH. Renal dysfunction and prolonged visceral ischaemia increasemortality rate after suprarenal aneurysm repair. Br J Surg 2003;90:1142-6.

30. Shortell CK, Johansson M, Green RM, Illig KA. Optimal operativestrategies in repair of juxtarenal abdominal aortic aneurysms. Ann VascSurg 2003;17:60-5.

31. Ayari R, Paraskevas N, Rosset E, Ede B, Branchereau A. Juxtarenalaneurysm. Comparative study with infrarenal abdominal aortic aneu-rysm and proposition of a new classification. Eur J Vasc Endovasc Surg2001;22:169-74.

32. Giulini SM, Bonardelli S, Portolani N, Giovanetti M, Galvani G,Maffeis R, et al. Suprarenal aortic cross-clamping in elective abdominalaortic aneurysm surgery. Eur J Vasc Endovasc Surg 2000;20:286-9.

33. Allen BT, Anderson CB, Rubin BG, Flye MW, Baumann DS,Sicard GA. Preservation of renal function in juxtarenal and suprarenalabdominal aortic aneurysm repair. J Vasc Surg 1993;17:948-58; dis-cussion: 958-9.

34. Poulias GE, Doundoulakis N, Skoutas B, Prombonas E, Haddad H,Papaioannou K, et al. Juxtarenal abdominal aneurysmectomy.J Cardiovasc Surg (Torino) 1992;33:324-30.

35. Constantinou J, Giannopoulos A, Cross J, Morgan-Rowe L, Agu O,Ivancev K. Temporary axillobifemoral bypass during fenestrated aorticaneurysm repair. J Vasc Surg 2012;56:1544-8.

36. Metcalfe MJ, Holt PJ, Hinchliffe RJ, Morgan R, Loftus IM,Thompson MM. Fenestrated endovascular aneurysm repair: graftcomplexity does not predict outcome. J Endovasc Ther 2012;19:528-35.

37. Coscas R, Becquemin JP, Majewski M, Mayer J, Marzelle J, Allaire E,et al. Management of perioperative endoleaks during endovasculartreatment of juxta-renal aneurysms. Ann Vasc Surg 2012;26:175-84.

38. Verhoeven EL, Vourliotakis G, Bos WT, Tielliu IF, Zeebregts CJ,Prins TR, et al. Fenestrated stent grafting for short-necked and jux-tarenal abdominal aortic aneurysm: an 8-year single-centre experience.Eur J Vasc Endovasc Surg 2010;39:529-36.

39. Mastracci TM, Greenberg RK, Hernandez AV, Morales C. Defininghigh risk in endovascular aneurysm repair. J Vasc Surg 2010;51:1088-95.e1.

40. Beck AW, Bos WT, Vourliotakis G, Zeebregts CJ, Tielliu IF,Verhoeven EL. Fenestrated and branched endograft repair of juxtarenalaneurysms after previous open aortic reconstruction. J Vasc Surg2009;49:1387-94.

41. Greenberg RK, Sternbergh WC, Makaroun M, Ohki T, Chuter T,Bharadwaj P, et al. Intermediate results of a United States multicenter

JOURNAL OF VASCULAR SURGERYVolume 61, Number 1 Rao et al 255

trial of fenestrated endograft repair for juxtarenal abdominal aorticaneurysms. J Vasc Surg 2009;50:730-7.e1.

42. Kristmundsson T, Sonesson B, Malina M, Björses K, Dias N, Resch T.Fenestrated endovascular repair for juxtarenal aortic pathology. J VascSurg 2009;49:568-74; discussion: 574-5.

43. Scurr JR, Brennan JA, Gilling-Smith GL, Harris PL, Vallabhaneni SR,McWilliams RG. Fenestrated endovascular repair for juxtarenal aorticaneurysm. Br J Surg 2008;95:326-32.

44. Semmens JB, Lawrence-Brown MMD, Hartley DE, Allen YB,Green R, Nadkarni S. Outcomes of fenestrated endografts in thetreatment of abdominal aortic aneurysm in Western Australia (1997-2004). J Endovasc Ther 2006;13:320-9.

45. Halak M, Goodman MA, Baker SR. The fate of target visceral vesselsafter fenestrated endovascular aortic repairdgeneral considerations andmid-term results. Eur J Vasc Endovasc Surg 2006;32:124-8.

46. Haddad F, Greenberg RK, Walker E, Nally J, O’Neill S, Kolin G, et al.Fenestrated endovascular grafting: the renal side of the story. J VascSurg 2005;41:181-90.

47. Muhs BE, Verhoeven EL, Zeebregts CJ, Tielliu IF, Prins TR,Verhagen HJ, et al. Mid-term results of endovascular aneurysm repairwith branched and fenestrated endografts. J Vasc Surg 2006;44:9-15.

48. Jongkind V, Yeung KK, Akkersdijk GJ, Heidsieck D, Reitsma JB,Tangelder GJ, et al. Juxtarenal aortic aneurysm repair. J Vasc Surg2010;52:760-7.

49. Harris PL, Vallabhaneni SR, Desgranges P, Becquemin JP, vanMarrewijk C, Laheij RJ. Incidence and risk factors of late rupture,conversion, and death after endovascular repair of infrarenal aortic an-eurysms: The EUROSTAR experience. J Vasc Surg 2000;32:739-49.

50. Greenberg RK, Lawrence-Brown M, Bhandari G, Hartley D, Stelter W,Umscheid T, et al. An update of the Zenith endovascular graft for

abdominal aortic aneurysms: initial implantation and mid-term follow-up data. J Vasc Surg 2001;33:157-64.

51. Stanley BM, Semmens JB, Mai Q, Goodman MA, Hartley DE,Wilkinson C, et al. Evaluation of patient selection guidelines forendoluminal AAA repair with the Zenith Stent-Graft: the Australasianexperience. J Endovasc Ther 2001;8:457-64.

52. Wolf YG, Johnson BL, Hill BB, Rubin GD, Fogarty TJ, Zarins CK.Duplex ultrasound scanning versus computed tomographic angiog-raphy for postoperative evaluation of endovascular abdominal aorticaneurysm repair. J Vasc Surg 2000;32:1142-8.

53. Greenberg RK, Chuter TAM, Lawrence-Brown M, Haulon S, Nolte L.Analysis of renal function after aneurysm repair with a device usingsuprarenal fixation (Zenith AAA Endovascular Graft) in contrast toopen surgical repair. J Vasc Surg 2004;39:1219-28.

54. British Society for Endovascular Therapy and the Global Collabora-tors on Advanced Stent-Graft Techniques for Aneurysm Repair(GLOBALSTAR) Registry. Early results of fenestrated endovascularrepair of juxtarenal aortic aneurysms in the United Kingdom. Circu-lation 2012;125:2707-15.

55. Sobocinski J, d’Utra G, O’Brien N, Midulla M, Maurel B, Guillou M,et al. Off-the-shelf fenestrated endografts: a realistic option for morethan 70% of patients with juxtarenal aneurysms. J Endovasc Ther2012;19:165-72.

Submitted May 7, 2014; accepted Aug 11, 2014.

Additional material for this article may be found onlineat www.jvascsurg.org.

Supplementary Fig 1 (online only). Egger funnel plot displayspublication bias. The funnel plot shows the perioperative mortalityrate on a logarithmic scale against the standard error for each studyin the meta-analysis. The vertical line with the diamond shows thepooled estimate of the logit perioperative mortality rate in all of thestudies, and the diagonal lines show the 95% confidence interval(CI) for a given standard error.

Supplementary Fig 2 (online only). Meta-regression shows the relationship between the year of study publicationand the logit perioperative mortality rate of the studies.

JOURNAL OF VASCULAR SURGERY255.e1 Rao et al January 2015

Supplementary Fig 3 (online only). Chart outlines the periop-erative mortality rates based on country of study publication. Thevarious symbols, including triangles, crosses, and squares, representindividual studies. FEVAR, Fenestrated endovascular aneurysmrepair.

Supplementary Fig 4 (online only). Meta-regression shows the relationship between study size and the logitperioperative mortality rate of the studies. FEVAR, Fenestrated endovascular aneurysm repair.

JOURNAL OF VASCULAR SURGERYVolume 61, Number 1 Rao et al 255.e2

Supplementary Fig 5 (online only). Meta-regression shows the relationship between mean preoperative creatinine ofthe patients included in the studies and the logit perioperative mortality rate of the studies.

JOURNAL OF VASCULAR SURGERY255.e3 Rao et al January 2015

Appendix (online only). Study quality assessment of 21 open repair case seriesa

Author Year Study typeHypothesis/

aimMain

outcomes

Inclusion/exclusion

criteria clearlydescribed

Interventionsclearly

described

Clear definitions

Juxtarenalaneurysm

Renalfailure

Pearce25 2007 Case series 1 1 1 1 1 1Knott3 2008 Case series 1 1 1 1 1 1Ayari31 2001 Case series 1 1 0 1 1 1Dubois19 2013 Case series 1 1 1 1 1 1Tsai2 2012 Case series 1 1 1 1 0 1Speziale21 2010 Case series 1 1 1 1 0 1Chisci16 2009 Case control 1 1 1 1 0 1Back26 2005 Case series 1 0 1 1 1 1Shortell30 2003 Case series 1 1 1 1 1 1Sarac5 2002 Case series 1 1 1 1 1 1Allen33 1993 Case series 1 0 1 1 1 1Donas10 2012 Case control 1 1 1 1 1 1Totsugawa20 2010 Case series 1 1 1 1 1 0Marrocco-Trischitta22 2009 Case series 1 1 0 1 1 1Bicknell29 2003 Case series 1 1 1 1 1 1Kudo28 2004 Case series 1 0 1 1 1 1Giulini32 2000 Case series 1 0 0 1 0 1Yeung24 2008 Case series 1 0 1 1 0 1Mukherjee23 2008 Case series 0 0 0 1 0 0Ryan27 2004 Case series 0 0 1 1 0 1Poulias34 1992 Case series 0 0 0 1 1 0

Table adapted from Downs and Black18 (1998).aA score of 1 indicates the presence of a parameter, and 0 indicates absence of the parameter. The maximum score for each open repair study is 13 if all theparameters of the checklist scale are present.

Appendix (online only). Study quality assessment of 14 fenestrated endovascular aneurysm repair (FEVAR) case seriesa

Author Year Study typeHypothesis/

aimMain

outcomes

Inclusion/exclusioncriteria clearly

described

Interventionsclearly

described

Cleardefinitions

Mainfindingsdescribed

Juxtarenalaneurysm

Renalfailure

Verhoeven38 2010 Case series 1 1 1 1 1 1 1Metcalfe36 2012 Case series 1 1 1 1 0 1 1Greenberg41 2009 Case series 1 1 1 1 0 1 1Kristmundsson42 2009 Case series 1 1 1 1 0 1 1Scurr43 2008 Case series 1 1 0 1 1 1 1Beck40 2009 Case series 1 0 1 1 0 0 1Semmens44 2006 Case series 1 1 1 1 0 1 1Mastracci39 2010 Case series 1 1 1 1 0 0 1Coscas37 2012 Case series 1 1 0 1 0 1 1Constantinou35 2012 Case series 1 1 0 1 0 0 1Halak45 2006 Case series 1 1 0 1 1 0 1

Table adapted from Downs and Black18 (1998).a1 indicates the presence of a parameter, and 0 indicates absence of the parameter. The maximum score for each FEVAR study is 15 if all the parameters of thechecklist scale are present.

JOURNAL OF VASCULAR SURGERYVolume 61, Number 1 Rao et al 255.e4

Appendix (online only). Continued.

Mainfindingsdescribed

Range of values foroutcomes outlined(eg standard error)

All importantadverse events

reported

Survivalanalysisconducted

Appropriatestatistical

tests

Valid andreliable outcomemeasures used

Losses tofollow uptaken intoaccount

Totalscore (/13)

1 1 1 1 1 1 1 131 1 1 1 1 1 0 121 1 1 1 1 1 1 121 1 1 0 1 1 0 111 1 1 1 1 1 0 111 1 1 1 1 1 0 111 1 1 1 1 1 0 111 1 1 1 1 1 0 111 1 1 0 1 1 0 111 1 1 0 1 1 0 111 1 1 0 1 1 1 111 1 0 0 1 1 0 101 1 1 0 1 1 0 101 1 1 0 1 1 0 101 1 0 0 1 1 0 101 1 0 0 1 1 0 91 1 1 0 1 1 0 81 0 0 0 1 1 0 71 1 1 0 1 1 0 61 0 0 0 0 1 0 51 0 1 0 0 1 0 5

Appendix (online only). Continued.

Range of values foroutcomes outlined(eg standard error)

All importantadverse events

reported

Survivalanalysisconducted

Appropriatestatistical

tests

Valid andreliableoutcomes

Loss to followup taken

into account

Endoleak andreintervention rates reported

Totalscore(/15)

Endoleakrate

Reinterventionrate

1 1 1 1 1 1 1 1 151 1 1 1 1 1 1 1 141 1 1 1 1 1 1 1 141 0 1 1 1 1 1 1 131 1 0 1 1 0 1 1 121 1 1 1 1 1 0 1 111 0 0 1 1 0 1 1 111 0 1 1 1 1 0 0 101 0 1 1 1 0 0 0 91 1 0 1 1 0 0 0 81 0 0 0 1 0 1 0 8

JOURNAL OF VASCULAR SURGERY255.e5 Rao et al January 2015