242
`
`J ENDOVASC SURG
`1997;4:242-251
`
`
`
`Abdominal Aortic Aneurysms: Experience With
`the lvancev-Malm6 Endovascular System for
`Aortomonoiliac Stent-Grafts
`
`Krassi lvancev, MD; Martin Malina, MD*; Bengt Lindblad, MD*;
`Timothy A.M. Chuter, MD*; Jan Brunkwall, MD*; Mats Lindh, MD;
`Ulf Nyman, MD; and Bo Risberg, MD*
`Departments of Radiology and *Vascular Surgery, Malm6 University Hospital,
`Lund University, Malmé6, Sweden
`
`@
`
`a
`Purpose: To describe a
`component-based aortomonoiliac stent-graft system and the first
`clinical results achieved with this device in endovascular abdominal aortic aneurysm (AAA)
`repair.
`Methods: From November 1993 to October 1996, 45 patients aged 60 to 86 years underwent
`endoluminal exclusion of true AAAs (median diameter 60 mm) involving the common iliac
`right and 15 mm feft) using
`arteries (median diameter 16 mm
`unilimb stent-grafts deployed
`with the lvancev-Malm6 system.
`Results: Six immediate conversions occurred in the beginning of the series due to endo-
`grafts that were too short. Complications, including 2 inadvertent renal artery occlusions,
`7 kinked grafts, 6 iliac artery dissections, and 3 perioccluder leaks, were
`features
`prominent
`in the first 15 patients. Five patients died in the postoperative period, four of whom were
`one 3 weeks after
`nonsurgical candidates. There werefive significant stent-graft migrations:
`to
`surgery due to mechanical injury of the proximal stent and four after 1 year owing
`continuous dilation of a wide proximal neck, stent-graft placementin a conical, thrombus-
`lined proximal neck, and two instances of proximal extension separation from the main
`graft. Translumbar aneurysm perfusion required embolization in 3 patients.
`Conclusions: Despite early complications associated with a
`learning curve, exclusion of
`large AAAs using unilimb stent-grafts is feasible. Strict inclusion criteria are necessary in
`order to improve mortality among nonsurgical candidates and minimize the risk for late
`migration.
`
`J Endovasc Surg 1997;4:242-251
`
`Key words: endovascular grafts, endograft, aortoiliac, migration, endoleak
`
`Sd
`
`e
`
`Endovascular stent-grafts for abdominal aortic
`on suitable im-
`aneurysm (AAA) exclusion rely
`plantation sites. While the
`proximal neck,i.e.,
`the nondilated aorta between the renal arteries
`and the aneurysm, is considered a safe im-
`
`and
`reprints: Krassi lvancev,
`Address for correspondence
`MD, Departmentof Radiology, Malm6 University Hospital,
`S-205 02 Malm6, Sweden. Fax: 46-40-96-99-77.
`
`plantation site within certain limits for all
`the distal fixation point varies ac-
`stent-grafts,
`to the graft configuration. This has an
`cording
`important effect on the utility of any given sys-
`tem, because a
`patient’s candidacy for endo-
`on the suita-
`vascular exclusion often depends
`bility of the distal implantation sites.
`Since it is rare to find more than a
`tiny ring
`aor-
`of nondilated aorta abovethe bifurcation,’
`are of limited use. Aortomono-
`toaortic grafts
`
`MEDTRONIC 1123
`
`

`

`J ENDOVASC SURG
`1997;4:242-251
`
`ABDOMINAL AORTIC ANEURYSM
`IVANCEVET AL.
`
`243
`
`iliac (straight, tapered) and bifurcated grafts,
`on the other hand, have a much wider poten-
`true for the
`tial application. This is particularly
`monoiliac version,2-® becauseits larger choice
`stent-graft sizes can accommodate a
`of distal
`broader range of distal
`implantation sites.
`Moreover, an aortomonoiliac approach is
`con-
`technically simple in terms of stent-graft
`struction and delivery.
`The lvancev-Malm6 system of endovascular
`aneurysm repair incorporates these advan-
`tages of the monoiliac
`endograft design. The
`was assembled from an
`system
`esophageal
`ma-
`stent
`delivery system, conventional graft
`terial, and modified venous Gianturco Z-
`our
`stents. We report
`3-year experience with
`to AAAs
`this unilimb endograft approach
`deemed unsuitable for conventional tube en-
`dografts owing to
`aneurysmal involvementin
`the common iliac arteries (CIAs).
`
`METHODS
`
`Stent-Graft Construction
`
`The fabric of the stent-graft is a standard
`available woven
`commercially
`polyester
`(Cooley Verisoft, Meadox/Boston Scientific,
`Oakland, NJ, USA). The crimps in the material
`are ironed out so that the graft
`can be com-
`pressed into the limited space available inside
`the delivery system. The graft size is tailored
`to each patient according to the preoperative
`and all grafts
`are
`tapered to
`imaging studies,
`match the diameters of the aorta
`proximally
`and the CIA distally. Thus, the graft is made
`from twodifferent sizes of cylindrical graft seg-
`ments, which are
`spatulated and sutured end-
`a CV-6 Gore-Tex suture (WL Gore
`to-end using
`and Associates, Flagstaff, AZ, USA).
`Both proximal and distal graft orifices are
`supported by modified venous Gianturco Z-
`stents (William Cook Europe, Bjaeverskov,
`Denmark). The proximal stent, with a
`resting
`diameter of 45 mm, consists of twenty 25-mm-
`on each
`long struts that meet to form 10 angles
`serve as attachment
`end. These junctions
`points for the graft, with four sets each of cau-
`dally oriented barbs and cranially oriented
`hooks. The distal stent, which has a
`resting di-
`ameter of 25 mm, has no barbs or hooks.
`
`Delivery System Design
`was
`The Ivancev-Malmé6 delivery system
`adapted for aortic aneurysm repair from an
`stent
`delivery system (William
`esophageal
`Cook Europe). Its 20F inner diameter (24F outer
`diameter) accepts stent-grafts up to 34 mm
`wide. The Teflon sheath and dilator are distin-
`special fittings. A cone at the
`guished by two
`outer end of the sheath facilitates compression
`as it is squeezed into the
`of the stent-graft
`sheath. The cone locks into a
`corresponding
`on the dilator. The pusher has a
`long
`fitting
`shaft that is attached to a
`25-cm-long metal
`cannula with a smooth bulbous tip
`at its up-
`stream end. A central lumen runs
`throughout
`from the bulbous tip to the end of the shaft.
`
`Iliac Artery Occluder
`A variety of CIA occluders were evaluated,
`all of which were some
`type of closed-ended
`a fabric bag and an un-
`stent-graft comprising
`barbed Gianturco Z-stent. In some cases, the
`was closed at both ends, like a
`graft
`pillow,
`and in others, it was closed at
`one
`end,
`only
`like a
`case. The diameters varied ac-
`pillow
`cording to the diameter of the iliac artery,
`ranging from 15 to 25 mm. After trying
`a vari-
`ety of systems for occluder insertion, a 16F
`(inner diameter) introducer was selected. The
`valve is crossed with a
`loading capsule of the
`same diameter containing the occluder. The
`latter is then pushed in place with the blunt
`end of the dilator.
`
`Stent-Graft Loading
`The stent-graft and the occluder are steri-
`lized separately; the delivery systems are pur-
`on
`chased sterile. Loading is performed
`side
`table at the time of operation. The stent-graft is
`mounted over the metal cannula of the pusher
`with the proximal end of the stent-graft next
`to the bulbous tip. A single, unknotted length
`of multifilament suture is passed through op-
`at the proximal end ofthe proxi-
`posite angles
`mal stent. Both ends of this suture are tied to
`a
`guidewire and pulled through the central
`lumen of the pusher, from the bulbous tip of
`the metal cannula at the inner end to the blunt
`end of the shaft at the outer end. The two ends
`are
`of the suture
`clampedat the outer end
`
`loop
`
`

`

`244
`
`ABDOMINAL AORTIC ANEURYSM
`IVANCEV ETAL.
`
`J ENDOVASC SURG
`1997;4:242-251
`
`under a small amountof tension. In this way,
`the suture
`to the
`loop attaches the stent-graft
`so that when the
`inner end of the pusher,
`pusher is advanced through the sheath, the
`stent-graft is towed along behind it. Currently,
`the stent-graft is preloaded into a
`capsule iden-
`tical to the 20F delivery sheath with the cone
`at the outer end.
`
`Procedure
`
`Patients were treated in an
`room
`operating
`equipped with a mobile C-arm fluoroscopy
`unit (Siremobil 2000, Siemens, Erlangen, Ger-
`many) with digital subtraction angiography
`(DSA), road mapping, and hard-copy output
`capabilities. General or
`regional anesthesia
`was used according
`to
`patient preference and
`the anesthesiologist’s recommendation. With
`the patient prepared and draped in standard
`a
`fashion
`transabdominal
`for
`approach
`(should conversion be necessary), both com-
`mon femoral arteries (CFAs) were
`exposed
`surgically. Using the Seldinger technique, the
`and a stiff guidewire
`CFAs were
`punctured
`(Amplatz Extra Stiff, Cook, Bloomington, IN,
`USA) was
`a catheter on the
`passed through
`side selected for delivery system insertion. To
`facilitate serial angiography during endograft
`a 5F straight-tipped, multiple
`deployment,
`sidehole angiographic catheter was inserted
`a sheath introduced in the contralat-
`through
`eral CFA and advanced until the catheter tip
`rested just abovethe renal arteries.
`were
`Five to ten thousand units of heparin
`administered intravenously before the appli-
`cation of the arterial occlusive clamps and
`snares and repeated when necessary. In most
`cases, the sheath and dilator of the delivery
`were inserted through
`a
`system
`longitudinal
`at the guide-
`femoral arteriotomy performed
`wire entrance site and then advanced into the
`(Fig. 1A) until the end of the sheath was
`aorta
`abovethe level of the renal arteries. External
`compression and retroperitoneal exposure
`were sometimes used to
`straighten tortuous
`iliac arteries and facilitate introduction of the
`access
`delivery system, but direct iliac artery
`was
`rarely required.
`were re-
`The dilator and the stiff guidewire
`moved from the sheath, and Fogarty clamps
`were
`applied to the sheath at the femoral level
`
`to prevent brisk
`hemorrhage. The pusher and
`were introduced into the sheath
`the stent-graft
`by either squeezing them through the cone or,
`into a
`alternatively, placing the stent-graft
`loading capsule to facilitate introduction into
`was then used
`the primary sheath. The pusher
`to advancethe stent-graft up the sheath with
`the aid of the suture attached to the proximal
`stent.
`wereinside
`Oncethe stent-graft and pusher
`were removed,
`the sheath, the Fogarty clamps
`hemostasis being provided by the bulk of the
`stent-graft and the dilator. When the proximal
`stent reached the level of the aneurysm neck,
`an
`was taken to confirm the posi-
`angiogram
`tion of the renal arteries. This was
`particularly
`important when the neck was short or tor-
`tuous. If the aortic neck was
`angled sharply
`anterior, the C-arm had to be rotated into a
`more cranial-caudal axis. Contrast was
`in-
`jected through the
`contralateral
`femoral
`sheath to assist in locating the bifurcation,
`which served as a landmark to
`guide deploy-
`ment of the distal stent and thus assure correct
`length of the stent-graft.
`Oncein satisfactory position, the proximal
`stent was held in place by the pusher
`as the
`sheath was
`slowly withdrawn, deploying the
`self-expanding Z-stent as it emerged from the
`were often re-
`sheath (Fig. 1B). Angiograms
`peated immediately before full deployment to
`allow last-minute adjustmentsof the stent po-
`sition.
`was
`The stent-graft
`released from the
`pusher by removing the suture line, and a
`was introduced through the pusher
`guidewire
`into the proximal aorta before delivery system
`removal (Fig. 1C) to facilitate subsequent in-
`strumentation of the stent-graft.
`The angio-
`graphic catheter from the contralateral femo-
`was
`ral
`down
`into
`the
`artery
`pulled
`aneurysmal sac.
`To reinforce the distal segment of the stent-
`a Wallstent (Schneider AG, Europe, Bu-
`graft,
`lach, Switzerland) of appropriate diameter
`(usually 16 mm) was
`deployed from the ta-
`pered portion of the main stent-graft body into
`the limb all the way to the distal Gianturco
`stent.
`
`was
`performed to
`Completion angiography
`look for perigraft leak and to assess
`stent-graft
`position and patency. Subsequent adjunctive
`
`

`

`J ENDOVASC SURG
`1997;4:242-251
`
`ABDOMINAL AORTIC ANEURYSM
`IVANCEVETAL.
`
`245
`
`Guidewire —_
`
`Dilator
`
`fl
`If
`|“
`
`ye
`
`|
`
`|
`
`_—Pusher
`
`7
`‘sy
`\
`
`~
`
`Suture Loop
`Npre ximal stent
`
`
`
`f
`Ss
`
`Suture Loop
`
`/
`
`-
`
`Shaft of Pusher
`
`
`
`~Teflon Loading Cone
`
`:
`
`——
`
`(a rail
`
`we
`
`em
`
`\, —Distal Stent
`
`Shaft of Pusher
`
`
`
`po
`Pet
`
`Cc
`
`¢
`
`—
`
`Graft
`
`_~ Occluder
`
`aN
`
`
`
`oe
`
`~
`Femorofemoral Bypass
`
`D
`
`Figure 1@ (A) Insertion of the introducer sheath and its dilator over a stiff guidewire. Note
`the Teflon cone at the outer end of the sheath. (B) Deployment of the proximal stent, guided
`by serial angiography. (C) Removal of the delivery system. (D) Completion of the procedure
`with insertion of a contralateral common iliac artery occluder and femorofemoral bypass.
`
`

`

`246
`
`ABDOMINAL AORTIC ANEURYSM
`IVANCEV ETAL.
`
`J ENDOVASC SURG
`1997;4:242-251
`
`was used primarily for measurementof stent-
`graft length and assessment of the size and
`ofthe iliac arteries, which werefre-
`angulation
`quently too calcified and/or tortuous to be as-
`sessed by CT alone.
`
`Patient Selection
`
`During the 3-year period ending October
`1996, 45 patients (mean age 73 years, range
`60 to 86) were selected for and gave informed
`consent to treatment of their true AAAs using
`the Ivancev-Malm6 system. Approximately
`were
`half of the patients
`symptomatic of is-
`chemic heart disease and/or hypertension
`(Table 1); 32% had congestive heart failure.
`The median AAA diameter was 60 mm; me-
`was 20 mm, as measured by
`dian neck length
`spiral CT. Other relevant dimensions are
`given
`in Table 1.
`
`Follow-up
`DSA and intra-arterial contrast-enhanced
`spiral CT were
`performed at 1 and 12 months
`and annually thereafter. Spiral CT enhanced
`was re-
`with intravenous contrast
`injection
`peated at 3, 6, and 18 months after stent-graft
`placement. Any contrast outside the lumen of
`was
`as
`the graft
`persistent peri-
`interpreted
`or
`endoleak,? which was inves-
`graft perfusion
`tigated further by angiography. The treatment
`
`0T
`
`ABLE 1
`Risk Factors and Arterial Dimensions* in 45
`Patients With Atherosclerotic AAAs Selected for
`Endoluminal Repair
`
`Comorbidities
`53%
`Hypertension
`48%
`Angina
`Congestive heart failure
`32%
`Previous myocardial infarction
`45%
`Maximum Minimum Median
`
`Arterial Dimensions
`60
`39
`AAAdiameter
`95
`72
`Neck length
`20
`9
`24
`14
`33
`Neck diameter
`0
`42
`16
`Right CIA diameter
`65
`Left CIA diameter
`15
`9
`> YF
`*
`Measuredin millimeters as shortest distance on
`spiral
`=
`abdominal aortic aneu-
`computed tomography. AAA
`= common iliac artery.
`rysm; CIA
`
`maneuvers
`on the angio-
`depended mainly
`graphic findings. Angiographic assessment
`was sometimes supplemented by intravascu-
`no clinical de-
`lar ultrasound (IVUS), although
`cisions have yet been based on the IVUS find-
`ings.
`The angiographic catheter and introducer in
`the contralateral CFA were
`exchanged for a
`stiff guidewire. A longitudinal femoral arte-
`over the entrance site of the guidewire
`riotomy
`was used to introduce the delivery system for
`in the CIA.
`the iliac occluder and position it
`With the dilator removed, the occluder was
`pushed from a
`loading capsule into the outer
`end of the sheath. From there, it
`was extruded
`into the CIA to
`expand under the action of the
`Gianturco Z-stent. Contrast was
`injected
`through the sheath to assess the complete-
`ness of CIA occlusion and the patency of the
`external to internal iliac artery pathway. A con-
`ventional femorofemoral bypass (Fig. 1D) was
`performed between the femoral arteriotomies
`to
`complete the procedure.
`
`Protocol
`
`a
`In November 1993, we
`hospital-ap-
`began
`to evaluate the
`proved investigational protocol
`safety and efficacy of endovascular AAA repair
`as described
`using the lvancev-Malm6 system
`above. To test the functional limits of this de-
`vice, the inclusion criteria were liberal. Early
`were
`required to be
`in the series, the patients
`candidatesfor surgery; however,this criterion
`was
`was relaxed as
`gained, and
`experience
`eight nonsurgical candidates were
`accepted
`was
`for treatment in this series. The protocol
`to
`identify exclusion criteria pro-
`designed
`spectively.
`
`Preoperative Evaluation
`were evaluated with contrast-
`All patients
`enhanced spiral computerized tomography
`(CT) using 5-mm collimation, 5-mm/s table
`and 3-mm_
`axial
`reconstructions.
`speed,
`were used to assess
`Shaded surface displays
`the aortic shape, and aortic dimensions were
`measured from multiplanar reconstructions.
`DSAwith a calibrated intra-aortic catheter was
`often performed concurrently by delivering
`the contrast medium intra-arterially during the
`spiral CT to minimize the contrast load. DSA
`
`

`

`J ENDOVASC SURG
`1997;4:242-251
`
`ABDOMINAL AORTIC ANEURYSM
`IVANCEV ET AL.
`
`247
`
`of persistent perigraft perfusion varied accord-
`ing to the cause.®
`
`Definitions
`Successful exclusion was defined as the ab-
`sence of endoleak (contrast outside the graft)
`on
`or decreased
`follow-up CT and unchanged
`aneurysm diameter (Fig. 2). The use of the
`term
`“morbidity” warrants elucidation be-
`cause some
`“‘complications” of endovascular
`no
`morbidity in the traditional
`repair produce
`sense. The most extreme
`example of this is
`a
`conversion to open repair, which represents
`total failure of endovascular aneurysm exclu-
`sion but may lead to an
`entirely uncomplicated
`recovery. Other examples include kinking and
`perigraft leak, which may also be asymptom-
`atic. As a
`result, these are listed as causes of
`morbidity only if they led to additional open
`surgical procedures, such as
`placementof ad-
`large Wallstents stent or
`ditional
`stent-grafts;
`were deemed complications if
`however, they
`were
`diagnosed and corrected percutane-
`they
`ously in the angiographic suite.
`
`RESULTS
`were done on an emer-
`Three procedures
`gency basis becauseof a contained rupture of
`
`the aneurysm; the remaining patients under-
`went elective surgery. Adjunctive
`maneuvers
`were
`performed during the initial stent-graft
`insertion to correct endoleaks or
`as-
`problems
`sociated with delivery; these were not consid-
`ered complications.® The ancillary procedures
`included:
`reorientation of the proximal
`(1)
`stent with valvulotomy balloons to correct
`proximal perigraft leak (n =10); (2) insertion
`of a second
`stent-graft either above or below
`=
`the main stent-graft to correct proximal (n
`=
`5) or distal (n
`3) perigraft leak; and (3) retro-
`peritoneal exposureofthe iliac arteries to as-
`sist in delivery system insertion or to exclude
`=
`flow to CIA aneurysms (n
`5).
`were
`Endograft procedures
`successfully
`completed in 37 patients (82%); there were 8
`conversions (18%) to
`surgical repair. However,
`5 of the successfully treated patients (11%)
`died from myocardial
`infarction within 30
`a 71% successrate. All the deaths
`days, giving
`werein the high-risk nonsurgical candidates.
`The outcome data were divided chronologi-
`cally into three groups comprising 15 patients
`each to better appreciate the effects of cumula-
`tive experience. Procedural
`time and blood
`loss were reduced (Table 2). The complication
`rate also improved (Table 3), as did mortality
`and the rate of successful exclusions (Table 4).
`
` Figure 2 @ Digital subtraction angiography demonstrating
`
`an aortomonoiliac stent-graft 1 year
`after placement. Note the slight kink at the transition to the small limb. (The right kidney had
`been removed becauseof carcinoma.)
`
`

`

`248
`
`ABDOMINAL AORTIC ANEURYSM
`IVANCEVET AL.
`
`J ENDOVASC SURG
`1997;4:242-251
`
`YP
`TABLE 2
`Mean Duration and Estimated Blood Loss for the
`Three 15-Patient Groups Undergoing
`Endovascular Aneurysm Exclusion
`Group 1
`
`Group 2
`
`Group 3
`
`220
`250
`244
`Duration (min)
`500
`550
`725
`Blood loss (mL)
`°°
`
`Among the complications, two renal arteries
`were
`inadvertently occluded by the endograft.
`In one case
`a stent-
`early in our
`experience,
`graft migrated proximally during removal of
`In another instance, uncer-
`the suture line.
`tainty regarding the position of the renal arter-
`ies in a short angulated proximal
`neck led to
`over the renal ori-
`proximal stent deployment
`fice. This hazard has been avoided in subse-
`quent cases
`by performing serial angiography
`stent placement to
`identify renal artery
`during
`position.
`Five of the six cases of external iliac artery
`dissection were on the contralateral side from
`were
`the stent-graft. They
`presumably caused
`by the insertion of the large (20F) occluder de-
`livery system used early in the study. Asmaller
`was
`16F delivery system
`developed and has
`since markedly reduced the incidence of dis-
`section.
`There werefive cases of proximal stent mi-
`gration that resulted in failure to exclude the
`oc-
`aneurysm. One case of early migration
`
`¢—__-——
`TABLE3
`Complications and Morbidity in Patients
`Undergoing Endovascular Aneurysm Exclusion
`
`Group Group Group
`1
`2
`3
`
`Total
`
`8 (18%)
`Conversion
`1
`6
`0
`5
`Immediate
`1
`2
`0
`1
`Late
`6 (13%)
`0
`3
`3
`lliac artery dissection
`Morbidity (return to OR)
`6 (13%)
`0
`3
`3
`1
`1
`2 (4%)
`OQ
`Renal artery occlusion
`0
`1
`4
`5(11%)
`Migration
`5
`7 (15%)
`0
`2
`Kinking
`2 (4%)
`0
`0
`2
`Late proximal leak
`1
`2
`3 (7%)
`0
`Perioccluder leak
`4 (9%)
`1
`2
`1
`Collateral perfusion
` —_
`
`
`
`¢
`
`¢
`
`TABLE 4
`Outcome by 15-Patient Group for Endovascular
`Aneurysm Exclusion
` Group 1 Group2
`Group 3
`Total
`8 (53%)
`32 (71%)
`Success
`11(73%)
`13(87%)
`2 (13%)
`8 (18%)
`6 (40%)
`0
`Conversion
`1 (7%)
`2 (13%)
`5 (11%)
`2 (13%)
`Mortality
`oo
`
`curred following open anastomosis of the dis-
`tal end of a too-short stent-graft
`to the proxi-
`mal CIA during which the proximal neck and
`stent were
`clamped. Three weeks
`proximal
`later, when the graft thrombosed and this pa-
`tient was converted to an open repair, the
`stent was found to have fractured
`proximal
`and migrated.
`The other four cases of stent migration
`oc-
`curred 1 year after stent-graft placement. Con-
`was
`version to open repair
`required in two
`cases. The causefor migration in the first was
`implantation of the proximal stent in a short,
`thrombus-lined conical neck. In the second pa-
`tient, the wide (32.6 mm) proximal neck dilated
`further to 38 mm over the courseof the year,
`resulting in migration of the stent-graft (Fig.
`3). For this reason, the upper limit of proximal
`neck diameter for secure
`placement of stent-
`grafts is now 28 mm.
`two casesof proximal
`stent
`In the remaining
`migration, the neck was < 10 mm in length,
`and hemostatic seal was achieved initially by
`a
`single polyester-covered stent above
`placing
`the main stent-graft. One year later, the main
`stent and
`stent-graft separated from the single
`migrated. Conversion was not
`required in
`a
`either case; instead,
`stent-graft extension
`was inserted proximally.
`Three of the four cases of persistent collat-
`eral perfusion by patentiliolumbar arteries re-
`an-
`quired coil embolization due to increasing
`eurysmal diameter. In the remaining patient,
`the aneurysmal size has remained unchanged
`for 2.5 years without intervention.
`were de-
`Seven cases of stent-graft kinking
`tected during follow-up in our first 20 patients.
`was overcome
`a
`This problem
`by inserting
`to provide longi-
`Wallstent into the endograft
`tudinal as well as circumferential support to
`the device.
`
`

`

`J ENDOVASC SURG
`1997;4:242-251
`
`ABDOMINAL AORTIC ANEURYSM
`IVANCEV ET AL.
`
`249
`
`an
`endograft’s proximal! stent
`multiplanar reformation format showing
`2 months after placement at the level of the renal arteries in a
`large inflammatory aortic
`aneurysm. (B) One year later, the stent-graft has migrated into the aneurysm because of pro-
`gressive dilatation of the proximal neck.
`
` Figure 3 @ (A) CT in a
`
`vers, including retroperitoneal iliac artery
`current
`to treat
`is
`posure. Our
`policy
`nonsurgical candidates only if there is a
`high
`likelihood of success based on
`preoperative
`
`ex-
`
`DISCUSSION
`was
`One of our
`original goals in this protocol
`to determine the functional limits of the sys-
`tem and thus define the necessary anatomic
`as a basis for
`substrate for endovascular repair
`we
`future exclusion criteria. To do this,
`initially
`took a very aggressive approach to patient in-
`clusion, as reflected in the challenging arterial
`anatomy in these patients (Table 1). Under
`wetreated only those pa-
`these circumstances,
`tients who were able to withstand a
`possible
`open repair. After 3 years, we have as yet not
`identified any absolute exclusion criteria ex-
`cept the absence of a suitable proximal neck.
`we have amassed
`list of factors
`However,
`more difficult
`that make endovascular repair
`and dangerous (Table 5).
`As these anatomic predictors of success be-
`welifted our ban on
`came clear,
`surgical candi-
`dacy and included patients with poor cardio-
`reserve.
`4 of the 5
`pulmonary
`Unfortunately,
`deaths were among
`these 8 nonsurgical candi-
`dates, a 50% mortality rate (compared
`to 3%
`among the 37 surgical candidates). All these
`patients had undergone challenging endovas-
`cular procedures requiring additional maneu-
`
`imaging.
`Ahigh rate of immediate conversion to open
`was a
`ex-
`prominentfeature of our
`early
`repair
`——___—
`
`9
`
`+
`
`TABLE 5
`Relative Contraindications to Endovascular AAA
`Repair Identified With the lvancev-Malm6 System
`for Aortomonoiliac Stent-Grafts
`
`1. Short neck (< 10 mm)
`2. Wide neck (> 30 mm)
`3. Angulated neck (> 75° to the long axis of the aneu-
`rysm)
`4. Conical neck (widening at a rate > 3 mm/10 mm from
`proximal to distal)
`5. Thrombus-lined neck
`6. Severeiliac artery tortuosity (angulation
`point)
`7. Severeiliac artery stenoses
`lliac artery dilatation (> 25 mm) at or distal to the im-
`8.
`plantation site
`9. Short commom iliac artery (< 30 mm
`long)
`Iliac arteries fixed to the retroperitoneum either due
`10.
`to previous surgery or to inflammatory reaction
`TS
`
`> 120° at any
`
`

`

`250
`
`ABDOMINAL AORTIC ANEURYSM
`IVANCEV ET AL.
`
`J ENDOVASC SURG
`1997;4:242-251
`
`perience. The main cause was inaccurate siz-
`ing of the graft
`length. Determining graft
`length remains difficult in tortuous aortas,par-
`ticularly those with short distal implantation
`sites. However, since instituting the current
`for preoperative assessment and
`protocol
`maneuversto aid device delivery, only
`using
`one case among the subsequent 40 patients
`treated failed due to
`improper graft sizing. This
`involved an emergency case in which a con-
`tained AAA rupture went to complete rupture
`during distal elongation of the main stent-
`graft.
`The rapid progression up the
`“learning
`curve” reflected increasing skill with the de-
`vice, improved patient selection, and the de-
`velopmentof a
`maneu-
`variety of adjunctive
`vers. These techniques enabled us to treat
`problems, such as
`perigraft leak, kinking, and
`stent-graft malposition, without resorting to
`open surgery.® The five postoperative deaths
`in this series included four patients in whom
`the iliac arteries were
`exposed, but the retro-
`incision was
`only part of a
`peritoneal
`long,
`is difficult to say
`It
`complicated procedure.
`whether a
`planned, uncomplicated iliac artery
`exposure would have been so morbid.
`lliac artery dissection has been reported
`even in those
`only rarely in other series,2->
`with large delivery systems. External iliac ar-
`or
`no
`tery dissection often produces
`signs
`were per-
`symptoms, and unless angiography
`formed routinely,
`the diagnosis might be
`missed. Indeed, the significance of these le-
`sions remains unclear. However, minimizing
`the profile of the occluder delivery system has
`reduced the incidence of this sequela.
`None of the 45 patients in this series were
`candidates for aortoaortic or aortobi-iliac
`stent-grafts. All required implantation distally
`in the ClIAs, which were wider than 12 to 15
`mm either bilaterally (majority of the cases) or
`unilaterally. Such wide ClAs preclude the use
`of currently known bifurcated stent-grafts, in-
`cluding the Chuter device, which we have used
`in another 15
`patients.”
`The major benefit of a monoiliac stent-graft
`one suitable CIA implanta-
`is the need for only
`tion site. Such versatility is particularly impor-
`tant in patients with large aneurysms, which
`have iliac artery involvement. Un-
`frequently
`fortunately, the need for femorofemoral by-
`
`pass and contralateral CIA occlusion repre-
`sents potential drawbacks of this technique.
`The patency rates of femorofemoral bypass
`are
`generally inferior to thoseof direct arterial
`reconstruction; however, these data are from
`arterial occlusive disease patients and may not
`be applicable in this setting. Overall, the com-
`bination of aortomonoiliac stent-graft implan-
`tation and direct surgical treatmentofiliac dis-
`ease opens the opportunity of endovascular
`to a wider range of candidates.?? Using
`repair
`this approach, approximately 70% of the AAA
`patients admitted to our
`facility during the last
`24 months have had endovascular repair.
`on
`The three most important late findings
`were
`changesin the size of the in-
`follow-up
`frarenal aortic aneurysm, migration of the
`stent-grafts, and collateral perfusion of the an-
`eurysmal sac. The postoperative diminution in
`aneurysm diameter seems to be dependent
`on
`complete exclusion of the aneurysm sac from
`the circulation, as
`reported by Mayet al.° In
`contrast, the mean diameter of the aneurysm
`increased postoperatively in cases with peri-
`graft leak shown on CT or
`angiography. We
`a more detailed description of
`have reported
`changes of aneurysmal size as related to the
`ofthe proxi-
`degree of exclusion and changes
`mal aneurysm neck with time.’° However, one
`aspect of the decline of aneurysmal size that
`deserves a note here is kinking of the stent-
`graft.
`no
`or
`Although kinking often produces
`signs
`symptoms until late in the patient’s course,
`a technical failure of the ini-
`this event is really
`tial procedure. We believe that with the de-
`crease of aneurysmal diameter there is also
`remodeling and shortening of the infrarenal
`aorta
`causing kinking of redundant unsup-
`a common
`experience in
`ported stent-grafts,
`our first 15-patient group. The prosthesis used
`in the current version of the lvancev-Malm6
`twostents, one at each end.
`system has only
`Wehave been able to eliminate the problem
`a
`of kinking by routinely inserting
`Wallstent,
`which provides the prosthesis with stent sup-
`port throughout the length of its narrow distal
`segment whereit is most at risk for clinically
`significant kinking. It would probably be better
`if a
`long central stent(s) were
`incorporated in
`the prosthesis. This enhancement would pre-
`
`

`

`J ENDOVASC SURG
`1997;4:242-251
`
`ABDOMINAL AORTIC ANEURYSM
`IVANCEVET AL.
`
`251
`
`vent
`potentially dangerous kinks and also min-
`imize the risk for migration.
`Most of the other causes for migration iden-
`tified so far in our
`are related to a
`experience
`deficient proximal neck. Wide (> 30 mm)
`necks, for example, appear to
`undergo contin-
`uous
`dilatation, while short (< 10 mm) necks
`offer no
`satisfactory fixation site. Conical,
`thrombus-lined necks are likewise inappro-
`priate because the thrombus is unstable, and
`over time. These pa-
`endoleak may develop
`tients should be excluded from endoluminal
`treatment.
`Lack of mural thrombus in the aneurysmal
`sac On
`preoperative contrast-enhanced CT has
`been associated with translumbar aneurysm
`perfusion in our
`experience. Although emboli-
`zation can
`successfully halt resultant aneurys-
`may
`mal enlargement, such patients
`be better
`excluded from endoluminal repair or, when
`possible, treated preoperatively with emboli-
`zation. Unfortunately,
`collateral perfusion
`through lumbar arteries is not amenable to
`technical improvementsin the system.
`As a final note on
`technique, oversizing the
`stent-grafts by at least 4 mm relative to the
`neck is a
`diameter of the proximal
`good way of
`enhancing the important proximal stent-graft
`fixation. Also, if proximal elongation is used,
`we
`try to
`overlap at least one stent
`length (25
`mm) between the stent-grafts in order to
`achieve secure anastomosis. In patients with
`status and a
`an-
`poor cardiopulmonary
`large
`eurysm with a
`we
`very short proximal neck,
`recommend that the uncovered portion of the
`proximal stent be placed abovethe level of the
`a secure anchor. As
`renal arteries to provide
`an
`a
`importantaside, after damaging
`proximal
`stent with a vascular clamp,
`we now feel that
`any stent that has been clamped
`cannotbe re-
`lied upon as a means of attachment between
`the graft and the aorta.
`In conclusion, the two main advantages of
`the lvancev-Malm6 system areits simplicity
`and the ready availability of the necessary
`components. Thereis very
`little in this system
`to go wrong. Thatis not to say that the system
`never
`fails, but we have found its limitations
`to be predictable consequencesofindividual
`
`arterial anatomy. Some patients simply lack
`the anatomic substrate for successful endo-
`we knowthat the
`vascular repair. For example,
`Gianturco Z-stent used in this system will not
`secure hemostatic attachment in a
`produce
`short angulated, conical, thrombus-lined neck
`that is wider than 30 mm. However,it is doubt-
`ful whether any of the current systems could
`be applied in such a case.
`
`6.
`
`REFERENCES
`1. Chuter TAM, Green RM, Ouriel K, et al.
`In-
`frarenal aortic aneurysm morphology: Impli-
`cations for transfemoral repair. J Vasc Surg
`1994;20:44-50.
`2. Parodi JC. Endovascular repair of abdominal
`aortic aneurysms and other arterial lesions. J
`Vasc Surg 1995;21:549-557.
`3. May J, White G, Waugh R, et al. Treatment of
`a
`complex abdominal aortic aneurysms by
`combination of endoluminal and extraluminal
`aortofemoral grafts. J Vasc Surg 1994;19:
`924-933.
`4.