Clinical Experience with a CustomizedFenestrated Endograft for Juxtarenal AbdominalAortic Aneurysm RepairRenan Uflacker, MD, Jacob D. Robison, MD, Claudio Schonholz, MD, and Krassi Ivancev, MD

Despite the numerous stent-graft devices available, unsuitable anatomy is still the greatest exclusion criterion forendovascular abdominal aortic aneurysm (AAA) repair (EVAR). The present report describes an on-site preproceduralcustomization of a conventional Zenith stent-graft device just before the endovascular procedure that includes thecreation of fenestrations and scallops as necessary for the patient’s anatomy. Three patients with difficult anatomy inwhom conventional AAA repair posed a high degree of risk were treated with customization of the stent-graft deviceto fit disparate renal arteries. A single fenestration for the left renal artery was made in two cases, and a single scallopwas made in the other case to accommodate the superior mesenteric artery. Gold beads were used to mark the locationof the fenestration and scallop. The three cases were successfully performed without perceptible endoleaks in thefollow-up period, which ranged from 4 to 14 months. No procedure-related complications were detected; however,pneumonia developed in one patient 3 weeks after EVAR. The initial results with this technique are encouraging, andthe role of EVAR can be significantly increased with the use of this customization technique when the intervention-alist does not have access to the commercially available devices or when the waiting time is too prolonged toaccommodate the patient’s clinical situation.

J Vasc Interv Radiol 2006; 17:1935–1942

Abbreviations: AAA � abdominal aortic aneurysm, EVAR � endovascular abdominal aortic aneurysm repair, SMA � superior mesenteric artery

THERE has been a rapid evolution inthe technology used for endovascularabdominal aortic aneurysm (AAA) re-pair (EVAR) since the inception of thetechnique by Parodi and colleagues in1991 (1). Despite the numerous de-vices available, unsuitable anatomy isstill the greatest exclusion criterion forEVAR. Inadequate proximal neck anat-omy is the most frequent rejection factordespite recent developments in fixation

systems. The presence of a short “land-ing zone” in the proximal neck requiresaccurate delivery. In addition, angula-tion and mural thrombus may even fur-ther limit the success of the procedure.Because of these difficulties, a fenes-trated device was developed by Park etal in 1996 (2) and refined by Faruqi et alin 1999 (3) in an attempt to increase thesealing area. Further development in-cluded a modular system, which is theplatform for the currently available fe-nestrated devices used for EVAR (4–6),in addition to reinforced fenestratedbranched grafts and directional branchesfor more complex aneurysms (7). Fenes-trated devices are currently commer-cially available and have been used inAustralia and in Europe, but the devicesare in short supply, and the waitingtime can reach several weeks. The fenes-trated endograft is manufactured in alengthy process based on detailed mea-surements of the patient’s aorta and re-view of imaging studies, and the device

is eventually made for that specific pa-tient. Sterilization time and long-dis-tance shipping lengthen the process. Inaddition, fenestrated endografts are notreadily available in the United Statesoutside of clinical trials. As a result ofthese limitations, there is an increasingunderserved population of patientswith unsuitable anatomy for the regulardevices. Particularly at risk are symp-tomatic patients who may not be able towait for the manufactured device andpatients with juxtarenal aneurysmsseeking endovascular treatment in areasof the world where the device is notavailable.

The present report describes a use-ful technique of on-site preproceduralcustomization of the conventional Ze-nith device (Cook, Bloomington, IN)with creation of fenestrations and scal-lops as necessary for the patient justbefore the endovascular procedure,according to the technique developedby Ivancev (personal communication).

From the Departments of Interventional Radiology(R.U., C.S.) and Vascular Surgery (J.D.R.), MedicalUniversity of South Carolina, 169 Ashley Avenue,Charleston, South Carolina 29425; and EndovascularCenter (K.I.), University of Lund, Malmo, Sweden.Received April 27, 2006; revision requested August27; final revision received September 3; and ac-cepted September 20. Address correspondence toR.U.; E-mail: uflacker@musc.edu

None of the authors have identified a conflict ofinterest.

© SIR, 2006

DOI: 10.1097/01.RVI.0000248828.92093.1A

Brief Reports

1935

MATERIALS AND METHODS

Three male patients (age, 71–80years) with AAAs larger than 5.5 cmthat were anatomically unsuitable fortraditional infrarenal endovasculargrafts, and who were deemed physio-logically to be at high risk for openrepair, were selected for endovasculartreatment with the preprocedural cus-tomized fenestrations made in a con-ventional Zenith device. In two pa-tients, a single lateral fenestration wasconstructed to accommodate an ex-cessively low renal artery on the leftside, and in one case an anterior scal-lop was open to accommodate the su-perior mesenteric ostium in a patientwith chronic renal failure in whomthe small renal arteries were electedto be covered by the stent-graft. In-stitutional review board review waswaived for this retrospective study.

Preoperative Assessment

Preoperative assessment of aorticmorphology was done by high-resolu-tion multislice helical computed tomo-graphic (CT) angiography and abdom-inal aortography with a sizing pigtailcatheter. Careful measurement of theproximal neck diameter was per-formed. Location of the renal arteriesin the circumference of the proximalneck was recorded in a clock-dial con-figuration. The distance of the ostiumof the renal arteries from the end of theshort neck was carefully measured, aswere the distances between the levelsof the two renal arteries and the rela-tionship with the superior mesentericartery (SMA). The distance from thesuperior mesenteric ostium to thehighest renal artery was also recorded.

In the third case, which was a renalartery fenestration, a bare stent wasplaced within the ostium of the leftrenal artery (the target of the fenestra-tion) to facilitate the location of theartery during the treatment procedure.The presence of this stent, placed dur-ing the diagnostic study as a reference,facilitated positioning, catheterization,and deployment of the stent-graft. Pa-tients with excessive tortuosity of theiliac arteries and/or an angle of more30° in the proximal neck were ex-cluded.

Device Preparation

Under sterile conditions in the op-erating procedure room, a commer-cially available regular Zenith en-dograft (Cook, Bloomington, IN) wascustomized for the endovascular pro-cedure. The first two covered stents ofthe device were deployed by pullingthe external sheath of the delivery sys-tem (Fig 1a). Then, the safety triggerwire was partially pulled, and the barefixation stent was deployed from in-side the cap by advancing the innermetal tubing connected to the cover-ing cap (Fig 1b). Fluoroscopy of thedeployed stents was performed toidentify the radiopaque markers (Fig1c, d). According to the measurementsperformed on the patient’s CT angio-gram, the Dacron graft was markedwith a sterile marker pen (Fig 1e). Thefenestration was carefully drawn inthe proper position along the circum-ference of the graft, coinciding withthe clock dial position previously de-fined, and at the proper distance fromthe top edge of the graft (Fig 1f). Next,with low-power ophthalmic cautery(Medtronic Xomed, Jacksonville, FL)

the fenestration or scallop was openedin the fabric of the graft. Dacron isflammable, so a low temperatureshould be used in cautery and greatcare is necessary to avoid excessive oruncontrolled burning of the materialby reducing contact time of the cau-tery with the fabric (Fig 1g). The diam-eter of the fenestration should be atleast 1 mm smaller than the diameterof the ostium of the renal artery, thetarget for fenestration. The reason forthis sizing is that the stent has to besnug against the edge of the Dacronfabric fenestration to be flared againstthe internal surface of the graft, rivet-ing the graft against the aortic walland fixing the fabric around the stentand into the orifice of the visceral ar-tery. The excess fabric has been dem-onstrated in bench tests to actuallywedge within the ostium of the renalartery, bridging the wall of the graft tothe inner surface of the proximal renalartery (R. Uflacker, MD, and C. Schon-holz, MD, unpublished data, March2005) close to the ostium and therebyreducing the chances of endoleak.

The fenestration or scallop was po-sitioned within the space of the stentstruts. When the fenestration wasopen at the proper size and location,gold marker beads (Accellent, Wil-mington, MA) were sutured with 6–0monofilament suture (Ethicon, Piscat-away, NJ) in the four quadrants of thefenestration (Fig 1h). Fluoroscopy ofthe fenestrated device was performedto identify the new radiopaque mark-ers (Fig 1i). The bare stent was recap-tured with a long 2–0 nylon suture(Ethicon) intertwined within the strutsof the device and pulled together (Fig1j). At this point, the suture was passedwithin the lumen of the delivery system

™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™3Figure 1. Sequence of events shows the construction of the fenestration in the commercial graft. (a) Zenith stent-graft device is partiallydeployed with the two top covered stents expanded. Note the safety wire already pulled from the hole in the cap (arrow). (b) Graft afterdeployment of the bare stent at the top. Note the black restraining cap, which was pushed upwards to liberate the bare stent with thehooks (arrow). (c) Radiograph of the segment with the two covered stents partially expanded to identify the radiopaque markers. (d)Radiograph of the deployed bare stents with the barbs, in which the radiopaque markers are readily visible. (e) The distance from theedge of the graft is measured. (f) Ink marks are made in the area of the fenestration. (g) Ophthalmic cautery is used to open thefenestration. (h) When the fenestration is finished, the gold markers are sutured to the four quadrants of the fenestration orifice (arrow).(i) A radiograph of the graft with the fenestration and the markers (arrow) is taken for identification. (j) The bare stent is recapturedwith a long 2–0 suture intertwined within the struts of the bare stent and pulled together. The nylon suture is passed within the lumenof the delivery system cap and exteriorized through the side hole used to introduce the safety wire. When the suture is in position, thebare stent is gently directed inside the restraining cap while the metal core tubing is pulled back to the original position. (k) The safetywire is repositioned through the side hole to retain the bare stent in place (arrow). (l) Umbilical tape is passed around each of the coveredstents, and the external sheath of the delivery system is carefully advanced over the stent-graft without torsion of the elements. (m)When the external sheath is back to the original position and it is confirmed that no damage has been inflicted to the sheath, the deviceis ready for use.

1936 • Customized Fenestrated Endograft for Juxtarenal AAA Repair December 2006 JVIR

Uflacker et al • 1937Volume 17 Number 12

Figure 2. Sequence of events during the AAA repair with the modified stent-graft. (a) Maximum-intensity projection reconstruction ofthe proximal neck of the aneurysm shows the disparate renal arteries and presence of disease in the wall of the neck below the right renalartery. (b) Aortogram shows the extent of the aneurysm and the relationship of the disparate renal arteries. (c) Intermittent injections ofcontrast material (10 mL at a rate of 20 mL/sec) used to identify the renal arteries during positioning of the device before partial de-ployment. (d) After the device was deployed and the gold markers were positioned in the orientation of the renal arteries accordingto the clock-dial configuration (arrow), a C2 Cobra catheter is used to catheterize the renal artery through the graft fenestration, keep-ing the bare stent restrained within the delivery system cap (curved arrow). Selective catheterization of the renal artery was achieved ina conventional manner through the fenestration and a guide wire was advanced deeply into the targeted renal artery. (e) Catheter-ization of the left renal artery was followed by introduction of a guiding catheter to allow stent advancement and deployment. (continues)

1938 • Customized Fenestrated Endograft for Juxtarenal AAA Repair December 2006 JVIR

Figure 2. (continued). (f,g) When the guiding catheter is advanced and stable within the renal artery,the bare stent at the top of the graft is released by pulling the safety wire and moving the deliverysystem cap upward. A balloon-expandable stent 8 mm in diameter and 17 mm in length was deployedwith two thirds within the artery and one third extending within the aorta/graft. (h) The aortic portionof the stent is flared with the tip of a 10-mm balloon. (i) Subsequently, with a large compliant balloon,the flare is completed (arrow). (j) The contralateral iliac extension of the endograft is then deployed, thecompliant balloon is used for fixation of the proximal bare stent, and the ipsilateral iliac leg is deployedas usual, followed by placement of an adequate iliac extension. (k) Final aortogram shows the stent-graft in place with patent renal arteries and no endoleak.

Uflacker et al • 1939Volume 17 Number 12

cap and exteriorized through the sidehole used to introduce the safety wire.When the suture was in position, thebare stent was gently directed inside therestraining cap while the metal core tub-ing was pulled back to the original po-sition (Fig 1k). The safety wire was re-positioned through the side hole toretain the bare stent in place. One um-bilical tape was passed around each ofthe covered stents, and the externalsheath of the delivery system was care-fully advanced over the stent-graft with-out torsion (Fig 1l). When the externalsheath was back to the original positionand it was assured that no damage hadbeen inflicted to the sheath, the devicewas ready for use (Fig 1m).

Procedure

After anticoagulation with heparin,with bilateral femoral cutdown, a stiff

Lunderquist wire (Cook) was used toplace the device in the aorta at thelevel of the renal arteries. From thecontralateral side, a pigtail catheterwith multiple side holes (Cook) waspositioned at the level of the renal ar-teries for aortography (Fig 2a) and in-termittent injections of contrast mate-rial (10 mL at a rate of 20 mL/sec) toidentify the renal arteries. With thebare stent remaining restrained, thefirst two covered stents of the en-dograft were deployed, and the goldmarkers were positioned in the previ-ously defined orientation of the renalarteries according to the clock-dialconfiguration (Fig 2b). Injections ofcontrast material were used to prop-erly identify the renal arteries. Mini-mal rotational and forward/backwardadjustments were necessary to prop-erly orient the fenestrations with therespective renal artery ostium or the

scallop to the SMA orifice. With thebare stent still restrained within thedelivery system cap, the endograftwas deployed until the contralateralshort leg was liberated within the an-eurysm sac. Access to the lumen of themain body of the endograft was ob-tained in the conventional mannerwith selective catheterization of theshort contralateral leg. Selective cath-eterization of the renal artery wasachieved in a conventional mannerthrough the fenestration, and a guidewire was advanced deep into the tar-geted renal artery (Fig 2c), followed byan 8-F Veripath guiding catheter(Guidant, St. Paul, MN) to allow theParamount stent (ev3, Plymouth, MN)to be advanced and deployed (Fig 2d).If a contralateral renal artery needs tobe catheterized, the same procedurewould be followed on the oppositeside. When the guiding catheter is ad-vanced and stable within the renal ar-tery, the bare stent of the endograft isreleased by pulling the safety wireand moving the delivery system capupward. A balloon-expandable stent17–28 mm in length with adequate di-ameter (7–9 mm) was deployed, withtwo thirds within the artery and onethird extending within the aorta/graft(Fig 2e). The aortic portion of the stentwas flared with the tip of a 10-mmangioplasty balloon (Fig 2f–h) andsubsequently with a large compliantCoda balloon (Cook) (Fig 2i). Thesame process may be used for the con-tralateral renal artery. If a scallop isused for the SMA, as in our case, stentplacement may not be necessary. Thecontralateral iliac extension of the en-dograft was then deployed and thecompliant balloon was used for fixa-tion of the proximal bare stent (Fig 2j),and the ipsilateral iliac leg was de-ployed as usual, followed by theplacement of an adequate iliac exten-sion (Fig 2k). Follow-up was the stan-dard for those patients, which in-cluded CT examinations at 1, 6, and 12months (Fig 3).

RESULTS

Three patients were treated withon-site preprocedural customized fen-estration. All procedures were suc-cessful. Two patients required unilat-eral left renal fenestration with singlestent placement, and one patient re-quired a scallop for the SMA without

Figure 3. CT angiographic follow-up at 6 months shows AAA exclusion and patentrenal arteries without endoleaks. The SMA was cut to improve visualization of the leftrenal artery.

1940 • Customized Fenestrated Endograft for Juxtarenal AAA Repair December 2006 JVIR

stent implantation. Two single fenes-trations were constructed on the leftside. Only one scallop for the SMAwas necessary and was constructed inthe anterior aspect of the graft. Theproximal necks were all more than 10mm in length below the right upperrenal artery but less than 3 mm belowthe lower left renal artery. The neckdiameters ranged from 24 to 26 mm.Angulation was not more than 20°.Follow-up times ranged from 4 to 14months. There were no endoleaks inthis group of patients after EVAR withthe fenestrated device or scallop. Noobstruction of branches was identified.There was no AAA growth, but onlyone aneurysm sac shrank within the1-year follow-up period. There was atransient creatinine increase in one pa-tient, with improvement within 1week with hydration. No dialysis wasnecessary. The two stent-implanted re-nal arteries incorporated within thefenestrated prosthesis were patent atfollow-up. The SMA scallop is alsopatent. There were no deaths or casesof major morbidity in the group, al-though the patient with chronic renalfailure experienced pneumonia 3 weeksafter the procedure and required hospi-tal admission.

DISCUSSION

Fenestrated stent-grafts were de-veloped to extend the sealing zone ofthe proximal neck in patients withdifficult anatomy that would haveprecluded endovascular repair. Thefenestrations are designed to accom-modate any juxtarenal aortic aneu-rysms, particularly renal arteries withshort necks. Large fenestration in thegraft fabric may be a potential cause ofmassive endoleak, and creative solu-tions have had to be developed toovercome that problem, such as flaredstents to rivet the graft against theaortic wall or branch. However, it isstill necessary to have a short segmentof proximal neck available for propersealing in the infrarenal position. Fourmajor requirements are necessary tothe performance of fenestrated graftprocedures: careful patient selection,exact measurement and device con-struction, proper technical skills withendovascular grafting, and visceralvessel catheterization and stent im-plantation skills (6). In addition, it isnecessary to understand the design,

the function of the device, and itsmultiple parts, as well as to mastertechniques such as disassembling andassembling the device. The on-sitecustomization described herein usesthe same basic steps and deploymenttechniques developed for the manu-factured stent-grafts except for thetethered tube graft used for adjust-ment of the fenestration position, aspreviously described in the literature(5,6), which adds an additional com-ponent to the commercially availableendograft set and a third trigger wireto the delivery set. The absence of thetethered tube graft in fact limits theuse of the customized device to pa-tients with a relatively straight aortaand iliac arteries. It is difficult to rotateand adjust a fully deployed body ofthe main device compared with themanufacturer’s tethered tube graft.However, we believe that the mostdisparate renal arteries and nearbySMA can be properly accommodatedwith the customized device, providedthe neck angle is acceptable. Excessiveangulation may generate increasedproximal endoleak risks, as describedby Greenberg et al (8). Not all balloon-expandable stents can be used to sealthe renal artery fenestration. Flaring ofthe intragraft portion of the stent isnecessary to rivet the graft fabricagainst the aortic wall or even dragsome of the fabric into the renal arteryitself. Some stents have larger cellsizes and flare better than those withsmall-cell stents. At this time we theprefer the double-strut balloon-ex-pandable stent (ev3) or the Herculinklow-profile stent (Guidant, MenloPark, CA).

The technique described herein is avariation of the technique describedby Anderson et al (4) and Stanley et al(5) and currently used by the industryto manufacture fenestrated endografts,with the main difference being that wedo not “fold and zip” to restrain ex-pansion of the body of the tethered tubegraft as described (4,8,9). The stent-graft customization was done on siteimmediately before the procedure andnot in advance, following a planningprocedure described in the literature (4).Although the fenestrated endograftsare becoming increasingly accepted andnecessary, this customization techniqueis not a replacement for the manufac-tured devices; however, it may help inthe treatment of many patients who are

not treatable otherwise. Many problemsmay be generated if the customization isnot properly designed. If the alignmentis inaccurate or catheterization of thevisceral artery is impossible, the proce-dure fails, and conversion to open sur-gery is probably the only option becausethe Zenith endograft device is not re-trievable. In the event of failure to alignthe fenestration with the visceral vessel,one option may be to occlude the renalartery and place an aortic cuff to closethe fenestration and proceed with a reg-ular stent-graft placement, losing thecovered kidney artery.

Our initial results with this cus-tomization technique are encouraging,and the role of EVAR may be signifi-cantly increased with the use of thistechnique when the interventionalistdoes not have access to the commer-cially available devices or the waitingtime is too prolonged to accommodatethe patient’s clinical situation. How-ever, we can hardly suggest that thistechnique could or should be used as aroutine because several questions re-lated to reliability, durability, and out-comes have not been properly ad-dressed.

References1. Parodi JC, Palmaz JC, Barone HD.

Transfemoral intraluminal graft implan-tation for abdominal aortic aneurysms.Ann Vasc Surg 1991; 5:491–499.

2. Park JH, Chung JW, Choo IW, et al.Fenestrated stent-grafts for preservingvisceral arterial branches in the treat-ment of abdominal aortic aneurysms:preliminary experience. J Vasc IntervRadiol 1996; 7:819–823.

3. Faruqi RM, Chuter TA, Reilly LM, et al.Endovascular repair of abdominal aorticaneurysm using a pararenal fenestratedstent-graft. J Endovasc Surg 1999; 6:354–358.

4. Anderson JL, Berce M, Hartley DE.Endoluminal aortic grafting with renaland superior mesenteric artery incorpo-ration by graft fenestration. J EndovascTher 2001; 8:2–15.

5. Stanley BM, Semmens JB, Lawrence-Brown MM, et al. Fenestration in en-dovascular grafts for aortic aneurysmrepair: new horizons for preservingblood flow in branch vessels. J EndovascTher 2001; 8:16–24.

6. Verhoeven ELG, Prins TR, Tielliu IFJ,et al. Treatment of short-necked in-frarenal aortic aneurysms with fenes-trated stent-grafts: short-term results.Eur J Vasc Endovasc Surg 2004; 27:477–483.

Uflacker et al • 1941Volume 17 Number 12

7. Kaviani A, Greenberg R. Current sta-tus of branched stent-graft technologyin treatment of thoracoabdominal an-eurysms. Sem Vasc Surg 2006; 19:60–65.

8. Greenberg RK, Haulon S, Lyden SP, etal. Endovascular management of jux-tarenal aneurysms with fenestrated en-dovascular grafting. J Vasc Surg 2004;39:279–284.

9. Greenberg RK, Haulon S, O’Neill S, etal. Primary endovascular repair of jux-tarenal aneurysms with fenestrated en-dovascular grafting. Eur J EndovascSurg 2004; 27:484–491.

1942 • Customized Fenestrated Endograft for Juxtarenal AAA Repair December 2006 JVIR