Eur J Vasc Endovasc Surg 16, 419-426 (1998)

The Role of Infrarenal Aortic Side Branches in the Pathogenesis of Endoleaks after Endovascular Aneurysm Repair

I. A. M. J. Broeders, J. D. Blankensteijn* and B. C. Eikelboom

Department of Vascular Surgery, Universzty Hospital Utrecht, The Netherlands

Aim: to investigate the relatzon between the number of preoperatzve patent szde branches and the presence or absence of postoperatwe endoleaks, and to study the fate of patent branches after operation. Patients and Methods: tlurty consecutzve patwnts were included Czne mode viewing of axzaI CT angiography zmages was apphed to detect mfrarenal aortic szde branches The posltzon of side branches relatwe to the renal arterzes, branch patency and run-off pathways were studied. Results: a total of 160 patent side branches were found. All patzents had two or more patent sMe branches. A patent mferzor mesenteric artery was found m 22/30 patients (73%) Postoperatwe CT scans revealed major endoleaks m five patzents (16%) and minor endoleaks zn eight (27%). There was no szgni~cant dzfference m the number of preope~atzve patent szde branches m patzents with a completely thrombosed aneurysm sac (five; range 2-8) compared to patzents with postoperatzve endoleaks (szx; range 3-9, p = 0.12) Backbleedmg from patent szde branches as the sole cause of endoleak was seen zn one patient only (3 3%) Conclusion: postoperatzve endoleaks are not related to the number of preoperatwe patent s~de branches. In patzents wzthout endoleaks, contrast enhancement of szde branches was repeatedly seen m the wczmty of the aneurysm wail. Although close follow-up of these branches is warranted, they dzd not affect the outcome of endovascular a neurysm repazr.

Key words: Endovascular aneurysm repazr, Patent branches; Endoleaks; Houncefield index.

Introduction identified as important determinants of endoleak out- come, there might be a role for pre-emptive coll em-

Incomplete exclusion of an aneurysm of the lnfrarenal bolisation of these arteries. We therefore studied the aorta after endovascular aneurysm repair may result role of patent side branches in the pathogenesis of in continuing expansion of the aneurysm sac. 1-4 If endoleaks. The aim of the s tudy was to correlate the blood flow outside the endovascular graft but inside number of preoperat ive patent infrarenal aortic side the excluded aortic segment can be demonstrated, this branches with the presence or absence of postoperat ive situation is described as an endoleak. 5 endoleaks, and to s tudy the fate of patent branches

Patent side branches of the diseased infrarenal aorta after operation. have been identified as part of endoleak circuits ~'~ These m a y serve as run-off channels for leaks ori-

ginatmg from areas of incomplete seal of the stents of Patients and Methods the endograft to the aortic wall On the other hand,

persistent backflow from patent lumbar and inferior Thirty patients with pre- and postoperat ive CT scans mesenterlc arteries into the aneurysm sac has been available for measurementswere inc luded . All patients repor ted despite a well posi t ioned graft. 7's The exact were treated by a first or second generation endo- role of patent side branches in the pathogenesls of vascular technologies endograft, as part of an FDA endoleaks is currently unknown, supervised international trial. A tube endograf t was

If patent infrarenal aortic side branches could be inserted in 13 patients, a bifurcated endograft in 17.

Twenty-six patients were male. The median age was 68 (51-86), the median preoperat ive aneurysm size was 58.4 m m (36.5-74.5).

* Please address all correspondence to J D Blankensteljn, De- partment of Vascular Surgery G04 232, Umverslty Hospital Utrecht, Spiral CT angiography was used for the detection of Heldelberglaan 100, 3584 CX Utrecht, The Netherlands mfrarenal aortic side branches. Patients were scanned

1078-5884/98/110419 +08 $12.00/0 © 1998 WB Saunders Company Ltd

420 I . A . M . J . Broeders et aL

Fig. 1. Tracmg of small cahbre aortic side branches usmg cme-mode vmwmg of axial CTA images.

preoperatively, on the third postoperat ive day, at 6 was documented on the preoperat ive and the post- months and every 12 months after the procedure. In operative scans. Side branches were classified as case of postoperat ive endoleaks, an additional scan lumbar, sacral, inferior mesenterlc (IMA), accessory was made at 3 months follow-up. All CT investigations renal or other. Only those side branches with an ost ium were per formed according to a protocol designed for covered by the endograft were evaluated. In patients abdominal aortic aneurysm, using a Philips EV-AP CT with a tube endograft , this meant that sacral arteries scan (Philips Medical Systems; Best, The Netherlands). originating at the level of the aortic bifurcation were According to this protocol, scanning started at the excluded. upper limit of the twelfth thoracic vertebra, which was The analysis of side branches was based on three the p resumed level of the coeliac trunk. At least 50 parameters. First, the level of the ost ium of the side continuous rotations of i s each were made. The col- branches was registered relative to the level of the left l imation was set at 5 m m and the table speed at 5 m m renal artery, using the vertical body axis as a scale. per second, resulting in a pitch of 1. The length of the The levels of patent branches on preoperat ive scans scanned volume was 25 cm at the minimal number of were used as a tool to trace both patent and throm- 50 rotations, stretching from the level of the coeliac bosed branches on postoperat ive and follow-up scans. t runk to the level of the external iliac arteries. Intra- Second, the ana tomy of side-branches was docu- venous contrast was given at a rate of 1-2 ml per s, mented. The ost ium and path of the arteries, and the starting injection 30 s ahead of scanning. Out of the level of the first branch bifurcation were registered acquired volume of CT data, axial images with an using the aorta and the vertebrae as anatomic land- imaginary thickness of 5 m m were reconstructed every marks. This documenta t ion was used to investigate 2 mm. This resulted in a dataset of at least 123 over- the extent of side branch thrombosis at postoperat ive lapping axial reconstructions. The images were trans- and follow-up investigations. Third, side branches ferred to an EasyVision Workstation (Philips Medical were classified into four groups: patent, partially Systems; Best, The Netherlands). thrombosed, completely thrombosed and un-

Cine mode viewing was applied to s tudy aortic side detectable. Patency was defined by detection of con- branches. This apphcat lon of the workstat ion allows trast enhancement extending from the aortic or visualization of axial CT images in a movie-like fashion endoleak lumen into the side branches. In case a side (Fig. 1). 9 The number of side branches for each patient branch was classified as partially thrombosed, contact

Eur J Vasc Endovasc Surg Vol 16, November 1998

Side Branches and Endoleaks 421

between the branch and the aortic lumen or endoleak analysis of the relation between the median number was no longer present, but contrast enhancement was of preoperat ive patent side branches and the presence seen m the vicinity of side branch bifurcations, of postoperat ive endoleak. A Chi-squared test was

In order to suppor t the differentiation between pat- applied to investigate the relation between a patent ency or thrombosis of side branches in a quantitative inferior mesenteric artery (IMA) and postoperat ive manner, the maximal houncefield (HF) index of side endoleaks. branches was measured and compared to the mean HF index of the aortic thrombus. HF indices were measured using the computer calibrated HF definition tool of the Easy Vision CT/MRI program. This tool Results provides a graphic display of HF indices along a hand d rawn straight hne of variable length. The mean HF Patent aortic szde branches on preoperative CT images index along this line is displayed, together with the minimal and maximal value. The maximal HF index A total of 160 patent side branches were found on the of side branches was found by drawing the line of preoperat ive CT images (Table 2). At least two patent measurement in a 90 degree angle over the most clearly branches could be identified in each patient. The me-

dian number of patent branches per patient was six enhanced part of the side branch (Fig. 2). The mean thrombus HF index was defined by drawing a line of (range 2-9). The second left and right lumbar arteries 2-3 cm in the thrombus at the level displaying the appeared to be patent most frequently, being identified most extensive amount of aortic thrombus mass. The in 24 patients each (80%). The fourth left lumbar artery HF ratio was defined by dividing the HF index of the was thrombosed most often. This branch was patent side branch by the HF index of the thrombus, in 13 out of 30 patients (43%). The IMA was patent in

22 patients (73%), the sacral artery in 19 (63%). How- Reference data on these type of measurements are not available. Therefore, the HF ratios of all branches ever, in five patients the ostium of the sacral ar tery

was outside the aortic segment covered by a tube on pre- and postoperat ive scans that were defined as completely patent or fully thrombosed by visual endograft and therefore in these particular cases not analysis were hsted (Table 1). The mean HF ratio of included in the figures of the study. In four patients, patent branches was 3.7 (range 1.1-9 2), compared to one accessory renal artery was found and m one a mean HF ratio of 0.9 of thrombosed branches (range patient two accessory renal arteries.

-0.6-1.6). After evaluating the data listed in Table 1, a HF ratio of 1.5 was chosen as cut-off level; a ratio

up to 1.5 was considered to be support ive for side Endoleaks branch thrombosis, a ratio higher than 1.5 for side

branch patency. This HF ratio classification was cor- Major endoleaks were encountered in five patients related to visual analysis in all branches studied. Re- (16%) and minor endoleaks in eight (27%) on the investigation was per formed in case of discrepancy, immediate postoperat ive CT scans (Fig. 3). Patent resulting in a change of classification in three cases, aortic side branches circuits could be identified in all

Endoleak was defined as the presence of contrast cases of endoleak. In cases of major endoleak, the area enhancement outside the endograft but inside the wall of incomplete seal of the stent to the aortic wall could of the aortic aneurysm. 5 Patients were divided in two be identified at the level of the proximal neck in three groups, one with and one wi thout postoperat ive en- patients, at the distal neck in one and in the right doleaks. Endoleaks were categorised in two groups, common iliac artery in one. All five endoleaks were

Major endoleaks were those due to an incomplete due to technical failures; a mismatch in limb size of seal of one of the endograf t stents to the aortic wall. the endograft and common iliac artery (n = 1), proximal These endoleaks presented as large areas of contrast dep loyment to distal f rom the renal arteries (n =3), enhancement in the partially thrombosed aneurysm, and entanglement of two hooks of a distal aortic stent

Minor endoleaks were solely encountered in patients (n = 1). The endoleak circuit of minor endoleaks was with a correctly posi t ioned endograft. These endoleaks found to be in contact with the at tachment area of the presented as streaks of contrast enhancement in the proximal stent in three patients and to the at tachment thrombosed aneurysm sac or as small contrast en- area of the left or right common iliac arteries in four. hanced areas m the vicinity of the ost ium of aortic In one case of a minor endoleak involving the left side branches, fourth lumbar and the sacral artery, there was no

The Mann-Whi tney U-Test was used for statistical indication of contact with the aortic or iliac lumen.

Eur J Vasc Endovasc Surg Vol 16, November 1998

422 I . A . M . J . Broeders et aL

Pro~l 1 e

I ~li,'irnum -?.fl

M~Odma~ 22~

~'~md~u'a devl~on 6,0

Profile ~ .

Prot*J le H

/ 1

f MIti~UIYl ~BN,O ~laglmum ~,~

Idraan -27 ~3

L~mg~ EE ~ [mnl]

Fig. 2. M e a s u r e m e n t t echmque of the m e a n houncehe ld ratzo of the aortic t h r o m b u s mass and the m a x i m u m houncef ie ld ratio of aortm side branches

Table 1. O v e r v i e w of houncef ie ld ratios; patent versus thrombosed infrarenal aortic s ide branches.

Houncehe ld ratio 1-0 0-1 1-1 5 1 5-2 2-3 3-5 5-10

N u m b e r of pa ten t b ranches 0 0 1 12 64 84 40 N u m b e r of th rombosed branches 1 18 24 1 0 0 0

In order to depict the relation between preoperative had a completely excluded aneurysm sac after the patent side branches and postoperative endoleaks, side procedure (p = 0.12). When considering minor en- branch patency was listed separately for the endoleak doleaks only, a median number of 6.5 (four to eight) group and the no endoleak group (Table 2). A me&an patent branches was found (p=0.10). A patent IMA of six (three to nine) patent branches was found on was found in 11/13 cases in the endoleak group (85%) the preoperatxve scan in patients that developed an and in 11/17 cases in the no endoleak group (65%). endoleak compared to five (two to eight) in those that The positive predictive value for the presence of post-

Eur J Vasc Endovasc Surg Vol 16, N o v e m b e r 1998

Side Branches and Endoleaks 423

Table 2. Number and percentage of preoperative patent side branches on pre- and postoperative CT images in the endoleak and no endoleak group.

Preoperatlve CT Postoperative CT Endoleak group No endoleak group Endoleak group

(n = 13) (n = 17) (n = 13)

Total 76 84 41 Second left lumbar 9 (70%) 15 (88%) 3 (23%) Second right lumbar 11 (85%) 13 (76%) 4 (31%) Third left lumbar 8 (62%) 11 (65%) 4 (31%) Third right lumbar 8 (62%) 12 (71%) 3 (23%) Fourth left lumbar 7 (54%) 6 (35%) 4 (31%) Fourth right lumbar 8 (62%) 6 (35%) 6 (46%) Sacral 10 (77%) 4 (24%) 8 (62%) IMA 11 (85%) 11 (65%) 7 (54%) Accessory renal 2 (15%) 4 (24%) 1 (8%) Other 2 (8%) 1 (6%) 1 (8%)

A B Fig. 3. Examples of endoleak (A) Major endoleak due to incomplete seahng of the distal stent to the inferior aneurysm neck (B) Minor endoleak revolving the fourth lumbar arteries.

o p e r a t i v e e n d o l e a k in case of a p a t e n t I M A w a s 50%; c i rcui t s on p o s t - o p e r a t i v e CT i m a g e s (54%; Table 2), the p o s i t i v e p r e d i c t i v e v a l u e for the absence of p o s t - the r e m a i n i n g 35 b r a n c h e s occ luded . A n IMA, a sacra l o p e r a t i v e e n d o l e a k in case of a t h r o m b o s e d I M A w a s or a l u m b a r a r t e r y as a s ing le run -o f f c h a n n e l for the 75%. A m e d i a n n u m b e r of fou r l u m b a r a r t e r i e s w a s e n d o l e a k w a s seen m th ree cases (23%). f o u n d in the e n d o l e a k g r o u p , c o m p a r e d to th ree in the no e n d o l e a k g r o u p (p = 0.10).

The fate of preoperatzve patent side branches in patients without endoleak

Involvement of szde branches m endoleak clrcuits All p a t e n t s ide b r a n c h e s in the g r o u p of 17 p a t i e n t s

Seven ty - s ix of the 160 p a t e n t s ide b r a n c h e s w e r e f o u n d w i t h a fu l ly t h r o m b o s e d a n e u r y s m sac o c c l u d e d to xn the g r o u p of 13 p a t i e n t s t ha t d e v e l o p e d an e n d o l e a k s o m e ex ten t b y l o s ing the i r con tac t w i t h the aor t ic a f te r e n d o v a s c u l a r a n e u r y s m r e p a m F o r t y - o n e of these l u m e n . The p o s t o p e r a t i v e fate of these b r a n c h e s v a r i e d 76 b r a n c h e s c o u l d be i den t i f i ed as p a r t of e n d o l e a k f r o m c o m p l e t e occ lus ion to p a r t i a l p a t e n c y d u e to

Eur J Vasc Endovasc Surg Vol 16, November 1998

4 2 4 I . A . M . J . Broeders et al.

Table 3. Fate of preoperative patent aortic side branches in the no endoleak group.

Inferior mesentenc Lumbar and Accessory arterms sacral arteries renal arteries (n=l) (n=68) (n=5)

Complete thrombosis or no 1 22 4 longer detectable

Patent up to outer aneurysm 7 - - - - wall

Partial thrombosis between - - 10 1 aneurysm and first side branch bifurcation

Patient at level of hrst side 3 10 - - branch bifurcation

Left to right collateral pathways - - 26 - - at backside aneurysm

A B Fig. 4. The fate ot side branches after successful endovascular aneurysm repair (A) Lumbar collateral clrcmts at the backside of the aneurysm (B) Patency of the IMA up to the aneurysm wall

r e t rograde filling (Table 3). In seven ou t of 10 pat ients on backb leed ing f r o m aortic side b ranches du r ing the wi th a pa ten t IMA, re t rograde filling was in this a r te ry latter p rocedure . In contras t to o p e n a n e u r y s m repair, obse rved u p to outer hmi t of the a n e u r y s m sac (Fig. backf low f r o m aortic side b ranches canno t be seen 4). L u m b a r pa t ency due to collateral p a t h w a y s f r o m before inser t ion of the vascular prosthesis . Howeve r , left to r ight l umbar arteries at the dorsa l side of the pa ten t side b ranches are s u p p o s e d to close spon- a n e u r y s m was seen in 9 / 1 7 pat ients w i t h o u t endo leaks t aneous ly due to cut-off of inflow, and obs t ruc t ion of (53%) A total of 26 l u m b a r arteries were r evo lved m ou t f low in case of backbleeding. This hypo thes i s is those collateral circuits (Fig. 4) s u p p o r t e d b y research on the fate of the exc luded

a n e u r y s m sac after aortic hga t ion and bypas s su rge ry for a n e u r y s m s of the infrarenal aorta. Resnikof et al. r epor t ed on 831 pat ients ope ra t ed on a 10-year per iod. 1°

Discussion In 17 patients , aortic side b ranches were l igated du r ing the p r o c e d u r e because of incomple te exclus ion of the

One of the differences b e t w e e n t r an s abdomina l and a n e u r y s m sac, assessed b y pressure measu remen t . endovascu l a r a n e u r y s m repair is the lack of control Persis tent f low m the a n e u r y s m sac w a s d i agnosed in

Eur J Vasc Endovasc Surg Vol 16, November 1998

Side Branches and Endoleaks 425

another 17 pat ients dur ing follow-up. Fourteen branches or collateral lumbar circuits, close fol low-up pat ients were reoperated, demons t ra t ing the IMA or is warranted. This was demons t ra ted by one case of lumbar arteries as the cause of endoleaks in six. In- a persistent endoleak, in which an initially par t ly complete exclusion of the aneu rys m sac due to patent th rombosed IMA became par t of the endoleak circuit aortic side branches was therefore demons t ra ted in dur ing follow-up. 23/813 (2.8%) of all pat ients in this group, including In conclusion, pa tent infrarenal aortic side branches endoleaks that might have d i sappeared in the early p lay an impor tan t role as run-off channels in pat ients pos topera t ive phase. Blumenberg et al. per fo rmed a wi th persistent endoleaks due to incomplete sealing prospect ive s tudy on 100 consecutive pat ients t reated of the endograft . Persistent backbleeding f rom side by exclusion of the aneurysm, using a stapling device, branches in patients wi th an adequate ly posi t ioned c o m b m e d wi th an aortobl-iliac bypass graft, tI Per- endograf t is encountered infrequently. Arguments for sistent f low in the aneu rysm sac was encountered in p re -empt ive coil embolisat ion could not be found 5% of the patients, p robab ly f rom lumbar arteries in based on the results of this study. all cases.

Various authors have repor ted on persistent en- doleaks due to re t rograde perfus lon by aortic side branches following successful p lacement of an aortic References endograft . Persistent endoleaks due to lumbar or IMA backb leedmg only were repor ted in 0-9.3% of the cases 1 MALINA M, IVANCEV K, CHUTER TA, LINDt{ M, LANNE T, LIND-

BLAD B, BRUNKWALL J, RISBERG B Changing aneurysmal mor- in seven studies that included more than 30 pat ients phology after endovascular graftmg relation to leakage or (median 2%)J '8't2-16 In all cases of lumbar backbleeding, persmtent perfuslon J Endovasc Surg 1997; 4 23-30 collateral pa thw ays involving l leolumbar branches be- 2 MATSUMURA IS, PEARCE WH, MCCARTHY WJ, YAO JS Reduction

m aortic aneurysm size early results after endovascular graft tween the hypogas tnc and the lumbar arteries could placement For the EVT Investigators ] Vasc Surg 1997; 25 be demonst ra ted . Coil embolisat ion of these pa thways 113-123 was successful in 90% of all at tempts. 3 BROEDERS IAMJ, BLANKENSTEIJN JD, GVAKHARIA A, MAY J, BELL

P, SWEDENBORG J, COLLIN J, EIKELBOOM B. The efficacy of trans- A relation be tween the n u m b e r of pa tent side femoral endovascular aneurysm management a study on

branches on preopera t ive CT images and the presence changes of the abdominal aorta at mid-term follow-up Eur J of pos topera t ive endoleaks could not be demons t ra ted Vasc Endovasc Surg 1997, 14 84-90

4 MaY J. The ms and outs of excluded abdommal aortic aneurysm in our study; contrast enhancement indicating flow decreasing diameters and dllatmg necks ] Endovasc Stag 1997, was found in a relatively high percentage of all side 4 31-32 branches in bo th the endoleak and the no endoleak 5 WHITE GH, YU W, MAY J, CHAUFOUR X, STEPHEN MS Endoleak

as a complication of endolummal grafting of abdominal aortic group. These results do not suppor t the hypothes is aneurysms classificatlon, mcidence, dlagnosls, and management that mfrarenal aortic side branches are the direct cause J Endovasc Surg 1997, 4 152-168 of failure of endovascular aneu rysm repair. Al though 6 BEEBE HG, BERNHARD VM, PARODI JC, WHITE GH Leaks after

endovascular therapy for aneurysm detection and classification the n u m b e r of pa tent side branches is not predict ive for J Endovasc Surg 1996, 3 445448 pos topera t ive endoleak, they are p robab ly an essential 7 KHILNANI NM, SOS TA, TROST DW, WINCHESTER PA, JAGUST par t of the endoleak circuits in patients wi th an in- MB, MITCHELL RS, DALE MD Embohzatlon of backbleedmg

lumbar arteries filling an aortic aneurysm sac after endovascular complete inclusion of the aneu rys m by the endograft , stent-graft placement J Vasc Inte, v Radzol 1996, 7 813-817 Go lzanan et al. demons t ra ted that the aneurysms of 8 IVANCEV K, CHUTER T, LINDH M, LINDBLAT B, BRUNKWALL J, these pat ients m a y be excluded complete ly after coil RISBERG B Optlons for treatment of persmtent aneurysm per-

fusion after endovascular repair World J Surg 1996, 20 673-678 embolisat ion of the involved side branches. I3 The ira- 9 BROEDERS IAMJ, BLANKENSTEIJN JD, OLgEE M, MALI WT, pact of these procedures on the size of the aneu rysm EIKELBOOM BC Preoperative sizing of grafts for transfemoral sac will have to be fol lowed wi th care, as the th rombus endovascular aneurysm management- a prospective comparative

study of spiral CT angmgraphy, artermgraphy and conventional m a y still be subjected to arterial pressure. Preferably, CT scanning J Endovasc Surg 1997, 4 252-261 this type of endoleak should be treated by secondary 10 RESNIKOFF M, DARLING RC, CHANG BB, LLOYD WE, PATY PS, sealing of the area of incomplete stent to wall contact, L~ATHER RP, SHAH DM Fate of the excluded abdommal aortic

aneurysm sac long-term follow-up of 831 patients J Vasc Surg for example by inserting a second endograf t or an 1996, 24 851-855 extension cuff. 11 BLUMENBERG RiM, SKUDDER PA, JR, GELTAND ML, BOWERS CA,

In patients wi th a complete ly excluded aneu rysm BARTON EA Retroperltoneal non-resective staple exclusion of abdommal aortic aneurysms chmcal outcome and fate of the

sac, contrast enhancement of lumbar and inferior mes- excluded abdommal aortic aneurysms J Vasc Sung 1995, 21: enteric arteries was frequent ly seen in close vicinity 623-634 of the aneu rys m wall. Al though the outcome of en- 12 MIALHEC, AMICABILEC, BEcQUEMINJP Endovascular treatment

of mfrarenal abdommal aneurysms by the Stentor system pre- dovascular ane u rys m repair in patients wi thout en- hmmary results of 79 cases Stentor Retrospective Study Group doleaks was not affected by part ial ly th rombosed side l Vasc Surg 1997, 26 199-209

Eur J Vasc Endovasc Surg Vol 16, November 1998

426 I . A . M . J . Broeders et aL

13 GOLZARIAN J, STRUYVEN J, ABADA HT, WERY D, DUSSAUSSOIS L, 16 WHITE GH, Yu W, MAY J, WAUGH R, CHAUEOUR X, HARRIS IP, MADANI A, ~FERREIRA J, DEREUME JP Endovascular aortic stent- STEPHEN MS. Three-year experience wlth the Whlte-Yu Endo- grafts transcatheter embohzatlon of persistent pengraft leaks vascular GAD Graft for translummal repair of aortic and lhac Ra&ology 1997, 202 731-734 aneurysms J Endovasc Surg 1997, 4 124-136

14 BLUM U, VOSHAGE G, LAMMER J, BEYERSDORF F, TOLLNER D, 17 VAN SCHIE G, SIEUNARINE K/ HOLT M, LAURENCE-BROWN M, KRETSCHMER G, SPILLNER G, POLTERAUER Pr NAGEL G, HOL- HARTLEY D, GOODMAN MA1 PRENDERGAST FJ, KHANGURE M ZENBEiNTetalEndolummalstent-graftsformfrarenalabdomlnai Successful embohzatlon of persistent endoleak from a aortlcaneurysms [see comments] N EnglJMed1997,336.13-20 patent inferior mesenterlc artery J Endovasc Surg 1997, 4

15 MOORE WS, RUTHERFORD RB Transfemoral endovascular repair 312-315 of abdominal aortic aneurysm results of the North American EVT phase 1 trial EVT Investigators J Vasc Surg 1996, 23 543-553. Accepted 5 May 1998

Eur J Vasc Endovasc Surg VoI 16, November 1998