Received: 13 September 2002Revised: 27 December 2002Accepted: 10 February 2003Published online: 12 April 2003© Springer-Verlag 2003

Abstract The aim of this study wasto evaluate incidence, potential riskfactors and effects on stent-graft mi-gration of proximal neck dilatationafter endoluminal repair of abdomi-nal aortic aneurysm (EVAR), and therole of ultrasound (US) in detectingneck enlargement. From November1998 to October 2001, 90 patientsunderwent EVAR. On follow-up, USand CT angiography (CTA) wereperformed, and diameters of the su-prarenal and infrarenal aortic neckswere monitored. Incidence of signifi-cant neck enlargement (≥2.5 mm)and distal stent-graft migration(>10 mm) was calculated. Severalfactors were evaluated as predictiveof neck enlargement. Ultrasound andCTA measurements were compared.The US and CTA examinations wereavailable in 68, 39, and 11 patients at 1, 2, and 3 years follow-up (meanfollow-up 15 months). Incidence of significant neck dilatation was21.8% at the infrarenal level (13, 33,

and 36% at 1, 2, and 3 years follow-up) and 13.8% at the suprarenal level(9, 18, and 27% at 1, 2, and 3 yearsfollow-up). Significant stent-graftmigration occurred in 14 of 87 pa-tients (16%) and was associated withneck dilatation in 8 (2 suprarenal and6 infrarenal). No risk factors wereidentified. Ultrasound was less accu-rate than CT in measuring neck di-ameter, in particular at the suprarenallevel. Proximal aortic neck enlarge-ment occurs in up to 30% of patientsafter EVAR and represents the mainrisk factor for stent-graft migration.The risk of infrarenal neck dilatationis higher at 2 years follow-up,whereas the suprarenal neck enlargeslater. Ultrasound is not useful inmonitoring neck diameter.

Keywords Abdominal aortic aneurysm · Endoluminal repair ·Neck enlargement · Suprarenalneck · Infrarenal neck · Stent migration

Eur Radiol (2003) 13:1962–1971DOI 10.1007/s00330-003-1859-y VA S C U L A R - I N T E RV E N T I O N A L

Vinicio NapoliSavino G. SardellaIrene BargelliniPasquale PetruzziRoberto CioniClaudio VignaliMauro FerrariCarlo Bartolozzi

Evaluation of the proximal aortic neck enlargement following endovascular repair of abdominal aortic aneurysm: 3-years experience

Introduction

It is now well known that changes in proximal aorticneck diameter represent a critical point in defining thesuccess of the endoluminal repair of abdominal aorticaneurysm (EVAR). Significant modification in the proxi-mal aortic neck has been demonstrated after surgicaltreatment of abdominal aortic aneurysm (AAA) [1, 2];however, evidence of dilatation at the level of the supra-and infrarenal aortic neck is still controversial [3, 4, 5,6], as well as its pathogenesis and etiology. In a recent

study, Badran et al. concluded that the proximal neck di-lates in one-third of patients, mostly in the first 2 yearsafter EVAR [7]. A strict follow-up is therefore requiredto evaluate modification of the neck, since this event candetermine distal stent-graft migration and type-I endo-leaks, which represent a risk factor for aortic rupture.

The purpose of our prospective study was to evaluatethe morphological changes of the proximal aortic neck,both at the supra- and infrarenal levels, after EVAR. Inparticular, we evaluated: (a) incidence of suprarenal andinfrarenal neck dilatation; (b) annual incidence of neck

V. Napoli (✉) · I. Bargellini · P. PetruzziR. Cioni · C. Vignali · C. BartolozziDivision of Diagnostic and InterventionalRadiology, Department of Oncology,Transplants and Advanced Technologies in Medicine,University of Pisa,Via Roma 67, 56126 Pisa, Italye-mail: irenebargellini@hotmail.comTel.: +39-050-992509Fax: +39-050-551461

S. G. Sardella · M. FerrariDivision of Vascular Surgery,Cisanello Hospital,Pisa, Italy

dilatation (1, 2, and 3 years after treatment); (c) presenceof predictive factors of neck enlargement; (d) incidenceof distal stent-graft migration related to neck dilatation;and (e) relation between US and CT in the measurementsof proximal aortic neck diameters (maximum, minimum,and biaxial) in patients with neck enlargement.

Materials and methods

From November 1998 to October 2001, 90 patients (88 men and 2 women; mean age±SD: 69.9±7 years, age range 63–77 years)with infrarenal abdominal aortic aneurysm underwent EVAR. Theprocedure was performed in the surgical suite, by interventionalradiologists and vascular surgeons, after appropriate patient selec-tion on the basis of clinical, anatomical, and anesthesiological fac-tors. Patients were classified as ASA II (n=21, 23%), III (n=68,76%), and IV (n=1, 1%).

Preprocedurally, all patients underwent CT Angiography(CTA; HiSpeed and LightSpeed Plus, GE Medical Systems, Milwaukee, Wis.) for the evaluation of the abdominal aortic aneu-rysm; the mean diameter±SD of the aneurysm was 49.4±11.4 mm(range 40–80 mm).

A bifurcated stent graft was implanted in 83 patients, whereasin 7 cases a straight endoprosthesis was required. Several types of prosthesis were used: 44 Aneurx (Medtronic, Minneapolis,Minn.), 2 Vanguard (Min Tec, Freeport, Bahamas), 12 Talent(World Medical Manufacturing, Sunrise, Fla.), 9 Endologix (En-dologix, Irvine, Calif.), 1 White-Yu Sis (Baxter, Deerfield, Ill.), 13 Excluder (Gore, Flagstaff, Ariz.), and 9 Zenith (Cook, Bloom-ington, Ind.).

On follow-up, 3 patients died of unrelated causes; therefore, 87 patients were strictly followed-up by means of CTA (at 7 days,6 and 12 months, and annually thereafter) and color-coded DuplexUS (AU5; Esaote Biomedica, Genova, Italy) (at 1 and 3 months,and associated with CTA thereafter). Digital subtraction angiogra-phy (DSA; Multistar, Siemens, Erlangen, Germany) was per-formed only in case of suspected complications.

CTA protocol and post-processing

Computed tomographic angiography was performed from the celi-ac artery to the common femoral arteries both before and after in-travenous contrast administration (Visipaque 320, Nycomed, Oslo,Norway) at the dose of 120 cc with a flow-rate of 3 cc/s. Acquisi-tion parameters for spiral CT were 3-mm thickness, 1-mm recon-struction spacing, and variable pitch.

In the past year a multidetector spiral CT (LightSpeed Plus,GE Medical Systems, Milwaukee, Wis.) had become available,and CTA was performed with the following parameters: High-Speed modality, gantry rotation 0.5–0.6, table speed 7.5 mm/rot,spacing 2.5 mm, and interval 1.2 mm.

Scan delay ranged between 20 and 40 s, according to patientcirculation time determined by an automated bolus time test(SmartPrep, GE Medical Systems). The venous phase was ac-quired with the same parameters 80 s after contrast injection.

Images were processed with a dedicated software package onan independent workstation (Advantage Windows 3.1 and 4.1; GEMedical Systems) by a single experienced operator, to generatemultiplanar reformations (MPRs), maximum intensity projections(MIPs), and volume renderings (VRs). In particular, the antero-posterior and latero-lateral diameters were measured in the trueaxial plane (perpendicular to the main axis of the aorta) at the ori-gin of the superior mesenteric artery (suprarenal neck) and imme-diately below the origin of the lowest renal artery, where at leasthalf of the circumference of the stent graft is visualized (infrarenal

neck). Diameters were measured wall to wall including parietalthrombosis, in two perpendicular planes; minimum, maximum,and biaxial (mean of minimum and maximum) diameters of theproximal aortic neck (PAD) were obtained.

The variation of PAD was calculated in millimeters as the dif-ference (∆PAD) between the diameter calculated at 7 days (refer-ence value) and diameters at 1, 2, and 3 years follow-up.

Enlargement of the proximal aortic neck was considered sig-nificant when a variation of at least 10% of the mean value of thereference diameter was observed.

The mean±SD suprarenal PAD was 23.9±3 mm, whereas theinfrarenal PAD was 23.3±3 mm. The suprarenal neck was <21 mmin 6 of 87 (7%) patients, between 21 and 25 mm in 61 of 87(70%), and >25 mm in 20 of 87 (23%) cases. The infrarenal neckwas <21 mm in 9 of 87 (10%) patients, between 21 and 25 mm in62 of 87 (70%), and >25 mm in 16 of 87 (20%) cases. On the ba-sis of these data a cutoff value of 2.5 mm was calculated to identi-fy a significant variation of ∆PAD.

Position of the stent graft was evaluated on MPRs measuringthe distance between the origin of the lowest renal artery and theproximal end of the stent graft. Stent-graft migration was definedas any observed variation of this distance, and was considered sig-nificant when ≥10 mm.

Duplex US

Color-coded Duplex US (AU5, Esaote Biomedica, Genova, Italy)was performed by a single operator in a blinded manner, with anabdominal phase array (2.5–3.5 MHz). The supra- and infrarenalportions of the aorta were visualized with the patient in a lateraldecubitus through a translumbar approach, obtaining a longitudi-nal scan of the aorta, thereby visualizing the entire aorta and theorigin of the ipsilateral renal artery. Maximum, minimum, and bi-axial PAD were calculated, including the aortic walls, 1 cm proxi-mally and distally to the renal artery. Measurements were repeatedthree times at each level and obtained bilaterally. Variations ofPAD on follow-up were calculated in millimeters using as refer-ence values those obtained 1 month after treatment.

Statistical analysis

Statistical analysis was performed with a JMP 4.0.0 StatisticsMade Visual software package (SAS Institute, Cary, N.C.).

Due to the asymmetric distribution of the measurements ob-tained at duplex US and CTA, continuous numerical data weregiven as mean±SD, median value, and interquartile range (IQRbetween 25 and 75° quartiles); means were compared using theStudent’s t test; medians were compared using the Wilcoxon rank-sum test.

Categorical data were analyzed with the chi-square test and theFisher exact test. When not otherwise specified, the statistical sig-nificance of a test was defined by p<0.05.

Coefficient of variation (CV%), expressed by the formula(Standard Deviation/Mean × 100), was calculated to evaluate in-tra-observer variability, both at CTA and US.

The relation between proximal aortic neck dilatation and po-tential risk factors was evaluated with the use of scatter plots forcontinuous data (such as reference PAD) and chi-square test forcategorical data such as type of prosthesis, suprarenal fixation,presence of hooks, presence of endoleak, aneurysm dilatation, andhypertension. These factors were also analyzed in relation to distalstent-graft migration.

1963

1964

Results

The mean (±SD) proximal diameter of the implantedstent grafts was 27 mm (±3 mm), ranging from 22 to34 mm. All stent grafts were dilated after implantation atthe level of the proximal and distal necks and at the gate,to ensure fixation of the prosthesis.

The US and CTA examinations were available in 68 of87 patients at 1 year (78%), 39 of 87 patients at 2 years(45%), and 11 patients at 3 years (13%) follow-up.

The global incidence (number of new cases per year)of significant neck dilatation (∆PAD ≥2.5 mm) was 21.8%at the infrarenal level and 13.8% at the suprarenal level.

Variation of the proximal aortic neck on follow-upobserved at CTA is extensively reported in Table 1. A

significant increase in the incidence of aortic neck en-largement (p=0.001 by chi-square analysis) was ob-served at 2 years follow-up (33% of patients vs 13% at1 year; Figs. 1, 2).

Significant dilatation of the suprarenal aortic neck oc-curred in a lower number of patients and later on follow-up (Fig. 3).

Median, IQR, and mean±SD values of ∆PAD at thesuprarenal and infrarenal levels are summarized in Ta-ble 2. The negative value represents the enlargement ofthe neck (group D, dilatation), whereas a positive valuerepresents no variation or reduction of the neck (groupND, no dilatation).

The median value of the infrarenal PAD at the 7-daycontrol was 23 mm (IQR 21–24 mm). It progressively

Table 1 Variation of the proximal aortic neck at the infra- and suprarenal levels: computed tomographic angiography (CTA) results onfollow-up. ND no dilatation, D dilatation, NC/Y new cases per year

CTA study control 1 year (n=68) 2 years (n=39) 3 years (n=11) Cumulative incidence

N % N % N % N %

InfrarenalND 49 72 15 38 5 45 – –D 19 28 24 62 6 55 – –∆PAD ≥2.5 mm 9 13 13 33 4 36 – –NC/Y – – 8 20 2 18 19 21.8

SuprarenalND 44 65 20 51 6 55 – –D 24 35 19 49 5 45 – –∆PAD ≥2.5 mm 6 9 7 18 3 27 – –NC/Y – – 4 10 2 18 12 13.8

∆PAD difference in proximal aortic neck diameter calculated between the reference study (after 7 days) and the examinations at 1, 2,and 3 years follow-up

Table 2 Infrarenal and suprarenal aortic neck diameters (in millimeters) on follow-up: mean, standard deviavion, median, and quartilevalues

CT study control 1 year (n=68) 2 years (n=39) 3 years (n=11)

ND D ∆PAD ≥2.5 mm ND D ∆PAD ≥2.5 mm ND D ∆PAD ≥2.5 mm

InfrarenalMean 1.01 −2.9 −4.7 1.4 −2.6 −3.6 0.6 −4 −5.2SD 1.3 2 1.2 1.3 1.4 1.07 0.5 2.1 0.9Median 0 −2* −5 1 −2.7* −3 1 −4.5** −5.575% 0 −5 −6 0 −3 −4.5 0 −6 −625% 2 −1 −3.5 2 −2 −3 1 −1.7 −4.2

SuprarenalMean 0.7 −2 −4 0.9 −2.1 −3.8 0.8 −3.5 −4.8SD 0.8 1.4 1.3 1.1 1.4 0.7 0.7 2.5 2.4Median 0.7 −2*** −3.7 1 −2* −4 1 −3**** −475% 0 −2.4 −5.2 0 −3.5 −4 0 −5.7 −7.525% 1.4 −1 −2.9 1 −1 −3 1.2 −1.5 −3

∆PAD difference in proximal aortic neck diameter calculated between the reference study (after 7 days) and the examinations at 1, 2,and 3 years follow-up*p<0.0001; **p<0.006; ***p<0.00001; ****p=0.05

1965

increased in group D, with a statistically significant dif-ference compared with group ND, at 1, 2, and 3 yearsfollow-up (Wilcoxon rank-sum test; Fig. 4a–c). A statis-tically significant difference of ∆PAD (p<0.01) at 1, 2,and 3 years follow-up was appreciated only in patientswith a significant (≥2.5 mm) neck dilatation (Fig. 4d).

The median value of the suprarenal PAD at the 7-dayCTA examination was 23 mm (IQR 21–25 mm). It pro-

gressively and significantly increased in group D with astatistically significance difference relative to group ND,on follow-up. Nevertheless, no statistically significantdifference was appreciated between the 3 years of fol-low-up in patients with a significant (≥2.5 mm) neck di-latation.

No relation was demonstrated between significantneck enlargement (whether infra- or suprarenal) and theconsidered risk factors: (a) proximal aortic neck diame-ter at 7 days (≤21, 21–25, ≥25 mm); (b) stent-graft over-sizing (10–20, 20–30, >30%); (c) infrarenal neck length,angulation, and morphology (cylindric, conic, conic-re-verse, tortuous), at 7 day follow-up; (d) detection of en-doleaks; (e) aneurysm dilatation; (f) supra- or infrarenalstent-graft fixation; (g) stent-graft fixation modality; (h)type and model of the prosthesis; and (i) presence of hy-pertension.

The scatter plot analysis did not identify a significantrelation between infrarenal PAD calculated at the 7-daycontrol and ∆PAD at 1, 2, and 3 years follow-up (R2=0.05,

Fig. 1a–e Patient with infrarenal aortic neck dilatation and distalstent-graft migration. In this patient, follow-up CTA performeda immediately after treatment and b 1 year later demonstrated asignificant infrarenal aortic neck dilatation and distal stent-graftmigration. Infrarenal (1) and suprarenal (2) diameters were as-sessed also by means of c duplex US performed with a left trans-lumbar approach. Patency of the renal artery is seen. LRA left re-nal artery. d Computed tomographic angiography and e duplexUS, performed after deployment of a proximal extender cuff, con-firmed correct positioning of the cuff just below the origin of therenal arteries. Again, infrarenal (I) and suprarenal (II) aortic neckdiameters were assessed

1966

Fig. 1c Legend see page 1965

Table 3 Correlation between US and CT measurements of theproximal aortic neck diameters (in millimeters) in patients with asignificant (≥2.5 mm) neck dilatation. All follow-up period is con-

sidered globally. IQR interquartile range, AD aortic diameter,CV% coefficient of variation

Study US CTAD

Minimum Maximum Biaxial Minimum Maximum Biaxial

InfrarenalMedian 21.6 21.9 21.6 23 23 23IQR 20–23 20–24 20–23 22–28 22–28 22–27Mean 22.2 22.4 22.3 24.4 24.4 24.4SD 2 2 2 1.2 1.2 1.2R2 +0.32 +0.28 +0.31 – – –p value <0.0001 <0.0001 <0.0001 – – –CV% 9 8.9 8.9 4.9 4.9 4.9

SuprarenalMedian 24 24 24 24 24 24IQR 21.4–25.6 22–25.2 22–25.5 22–26 22–26 22–26Mean 23.6 23.8 23.7 24.1 24.2 24.2SD 2.5 2.5 2.5 1.6 1.6 1.4R2 0.0002 0.005 0.0004 – – –p value 0.9 0.6 0.8 – – –CV% 10.5 10.5 10.5 6.6 6.6 5.7

1967

Fig. 1d, e Legend see page 1965

Fig. 2 Frequency of dilatation of the infrarenal aortic neck on fol-low-up. The infrarenal neck dilates progressively over time reach-ing the highest incidence 2–3 years after treatment (33–36% of pa-tients treated)

p=0.05 at 1 year; R2=0.04, p=0.2 at 2 years; R2=0.11,p=0.4 at 3 years follow-up). A significant relation wasdemonstrated between stent-graft oversizing (87% of stentgrafts were oversized ≥20%, median 26%, IQR 24–28%)and PAD at 7 days (R2=+0.23, p<0.0001), but no relationwas demonstrated between oversizing and infrarenal∆PAD on follow-up (R2=0.006, p=0.5 at 1 year; R2=0.01,p=0.4 at 2 years; R2=0.22, p=0.1 at 3 years follow-up).

No statistically significant difference was observedbetween patients with supra- and infrarenal significantneck dilatation (55, 54, and 0% at 1, 2, and 3 years fol-low-up, respectively) and patients with significant infra-renal neck enlargement alone; therefore, these eventsseem to be unrelated.

A good correlation was observed between US and CTmeasurements of the minimum, maximum, and biaxial

1968

Fig. 3 Frequency of dilatation of the suprarenal aortic neck on fol-low-up. The suprarenal aortic neck dilates in a lower number of pa-tients and later on follow-up compared with the infrarenal neck. Thehighest incidence is reached 3 years after treatment (27% of patients)

Fig. 4a–d Median values of the infrarenal ∆proximal aortic neck(PAD): comparison between patients with aortic neck dilatation(D) and patients without dilatation (ND). A statistically significantdifference is observed in median ∆PAD between groups D and ND

at a 1, b 2, and c 3 years follow-up (Wilcoxon rank-sum test). Inthe group of patients with d dilatation ≥2.5 mm, a statistically sig-nificant difference is observed in median values of ∆PAD at 1, 2,and 3 years (p<0.01)

PAD at the infrarenal level (R2=0.26, p<0.0001;R2=0.24, p<0.0001; R2=0.27, p<0.0001, respectively)and at the suprarenal level (R2=0.22, p<0.0001; R2=0.18,p<0.0001; R2=0.22, p<0.0001, respectively). In particu-lar, the relation was satisfactory in patients with signifi-cant infrarenal neck dilatation, whereas no correspon-dence was demonstrated in patients with suprarenal neckenlargement (Table 3).

Distal stent-graft migration was observed in 17 of 87patients (19.5%). It was significant (≥10 mm) in 14 of 87patients (16%) and occurred in 5 of 78 patients (6.4%)6 months after the procedure, with 3 of 68 (4.4%) newcases at 1 year follow-up, 7 of 39 (18%) at 2 years, and 2of 11 (18.1%) new cases at 3 years follow-up. Data onpatients with stent-graft migration are extensively report-ed in Table 4.

The incidence of distal migration associated with sig-nificant (≥2.5 mm) neck dilatation was 2 of 17 patients

1969

(12%) for the suprarenal level and 6 of 17 patients(35.3%) for the infrarenal level. Only 1 case of distal mi-gration (5.8%) was associated with a significant increaseof the aneurysmal sac. Endoleaks were detected in 5 pa-tients (29.4%). Type-I and type-IV endoleaks were treat-ed by deployment of an extender cuff at the proximalend and into one branch, respectively. Type-II endoleakswere treated only when associated with increasing sacdiameter (n=1). Treatment was ineffective and surgicalconversion was required.

Thirteen patients with distal stent-graft migrationwere not treated and no further migration was observedon follow-up (median 12 mm, IQR 15–10 mm; medianof variation between two examinations 0 mm, IQR2–0 mm). In these cases the proximal aortic neck wassufficiently long to guarantee the sealing of the sac de-spite graft migration; therefore, no endoleak occurredand the aneurysm sac did not enlarge.

No significant relation was observed between migra-tion and potential risk factors such as (a) type and modelof the prosthesis, (b) detection of endoleak, (c) stent-graft oversizing >20%, (d) presence of proximal fixinghooks, (e) suprarenal graft fixation, and (f) hypertension.

Discussion

Proximal aortic neck dilatation after EVAR occurs in ap-proximately one-third of patients after EVAR [7], andour data are in agreement with these findings. In particu-lar, we demonstrated that proximal aortic neck enlarge-ment can occur with different incidence and modalitiesat the suprarenal and infrarenal levels.

Published reports on this phenomenon are rare, con-troversial, and non-comparable due to important differ-ences in follow-up time, number of patients studied, dataprocessing, and measurement modalities.

According to our experience, the measurement of theminimum, maximum, and biaxial wall-to-wall diameterson two perpendicular planes represents an easy and re-producible method for the evaluation of neck changes, inparticular when the CT true axial planes are obtained.Measuring the area of the proximal neck seems to over-estimate neck variation [8].

The suprarenal and infrarenal levels have alreadybeen defined [3, 8, 9, 10]. The infrarenal neck dilatationoccurs in 21–23% of patients 2–3 years after treatment[8, 9], whereas the suprarenal neck enlargement seems tobe less frequent [3, 5, 11, 12]. Nevertheless, neck en-largement is still controversial. Walker et al. demonstrat-ed no significant changes in the proximal aortic neck di-ameters after EVAR, but this study included a low num-ber of patients followed-up for over 1 year [6]. Accord-ing to Mahnken et al. the risk of neck enlargement andproximal type-I endoleak is very low [13].

In our series aortic neck enlargement occurred in 22%of patients at the infrarenal level and 14% at the suprare-nal level.

Defining measurement modalities and reference val-ues represents a key point in evaluating neck changes onfollow-up. The reported annual degree of neck enlarge-ment is extremely variable: from <1 mm/year [10, 14], to1–2 mm/year [3, 15, 16] and over 2 mm/year [4, 8, 17].There is an error of approximately 1–2 mm in the mea-surement procedure [3, 7, 18], and early neck dilatationis related to stent-graft oversizing [4, 9]. These variables

Table 4 Patients with distal stent-graft migration on follow-up. The sign − indicates an increase of the diameter; the sign + defines a decrease of the diameter. AAA abdominal aortic aneurysm

Patient Time of Migration Changes in Changes in Changes in Presence Follow-upno. migration (mm) suprarenal infrarenal AAA sac of endoleak

(months) neck diameter neck diameter diameter (mm) (mm) (mm)

1 24 12 −2 –1 +10 No Observation2 24 10 –2 –2 +2.5 No Observation3 24 12 0 0 +3 No Observation4 24 11 0 –1 +6 No Observation5 12 6 0 –3 +12.5 Type I Cuff6a 12 8 0 –5 +13 Type I Cuff7 6 10 0 –3 +17.5 No Observation8 24 10 –1 0 +1.5 No Observation9 24 10 0 –5 +3.5 No Observation

10 36 13.6 –2 –4 +21.5 Type II Observation11 6 10 0 –2 –20 Type II Conversion12 36 10 –2 0 +2.5 No Observation13 24 10 –4 –1.5 +2.5 No Observation14 6 9 –4 –1 +10 No Observation15 12 10 0 0 +9.5 No Observation16 6 10 0 –4 +13.5 No Observation17 6 13.5 0 0 –4.5 Type IV Cuff

a See Fig. 1

determine the need for the definition of a cutoff value toestablish the statistical significance of the dilatation.Some authors have used a value of 2–2.5 mm [3, 8]. Inour series a cutoff value of 2.5 mm and a CV% <10%were used.

It is possible to describe the dilatation process overtime calculating the incidence of new cases per year ofaortic neck enlargement. In a series of 84 patients, Reschdemonstrated that the infrarenal enlargement occurs inapproximately 23% of patients, with 2-mm increase at18 months follow-up; in approximately 68% of these pa-tients the neck dilates progressively [17]. According toMakaroun and Deaton, of 314 treated patients, dilatationoccurs in 13, 21, and 19% of patients at 1, 2, and 3 yearsfollow-up, respectively, with the highest peak during thesecond year [8]. Badran et al. demonstrated a higher in-cidence of neck dilatation during the second year of fol-low-up (up to 33% in a series of 73 patients) [7]. Simi-larly, in our series the incidence of significant infrarenalneck dilatation was 13, 33, and 36% at 1, 2, and 3 yearsfollow-up, respectively. Considering the incidence ofnew cases per year (13, 20, and 18%; not significant),the enlargement seems to start during the first year offollow-up reaching its highest peak in the second yearand remaining stable in the third year. Accordingly, thevariation of the infrarenal aortic neck, expressed in milli-meters, progressively increases during the first and sec-ond years, stabilizing later. In fact, the median value of∆PAD at 24 months differs significantly from the valueat 1 and 3 years follow-up (p=0.01, Wilcoxon rank-sumtest).

The suprarenal aortic neck dilates less frequently andlater on follow-up than the infrarenal neck. Incidence ofsuprarenal neck dilatation progressively increases (9, 18,27% at 1, 2, and 3 years, respectively), reaching themaximum peak at 3 years follow-up. A longer follow-upis still required to further investigate this event.

In patients with significant neck dilatation, the varia-tion of suprarenal ∆PAD does not show a significant dif-ference over time (median values of −3.7 mm at 1 year,−4 mm at 2 years, and −4 mm at 3 years; not significant).

In our experience no relation was observed betweensuprarenal and infrarenal neck dilatation, possibly due tothe different anatomy of these districts [12, 19].

Up to now, no risk factors of neck dilatation havebeen identified [3, 8, 10, 12]. These data are confirmedby our study.

In our experience the dilatation seems to occur morefrequently in patients with larger necks at discharge [7],which is not in accordance with Makaroun and Deaton’sdata [8]. Nevertheless, most patients in our seriesshowed neck diameters at discharge >21 mm and under-went significant (>20%) stent-graft oversizing. Thesefactors could be responsible for the difference observedbetween our study and Makaroun and Deaton’s [8] anal-ysis.

According to our data, US examination is not able toprecisely evaluate the proximal aortic neck on follow-up,despite the translumbar bilateral approach and the use ofseveral planes to visualize the renal arteries. This is par-ticularly true for the suprarenal level. In fact, a satisfac-tory correspondence between CT and US measurementswas observed only at the infrarenal level, in particular inpatients with neck dilatation. Besides, US inter- and in-traobserver variability is well known, and CT seems tobe superior to US both in the evaluation of aneurysmmeasurements and in the detection of small endoleaks[20]. Recent studies have demonstrated a good correla-tion between CT and MR imaging in providing the rele-vant information in the follow-up after EVAR [20]. In-deed, the role of MR imaging in treated patients has yetto be defined. It is able to provide good aneurysm mea-surements, and it seems to be more sensitive than CT inthe detection of type-II endoleaks, with fewer metal arti-facts [21, 22, 23]. Nevertheless, it is more expensive, notwidely available, contraindicated in some patients, andsusceptibility artifacts occur depending on the type ofstent graft implanted [20]; therefore, CT remains themost adequate and easily available imaging modality forthe evaluation of PAD and for the detection of complica-tions after EVAR.

Neck dilatation represents the main cause of distalstent-graft migration [24, 25]. In our series significantstent-graft migration occurred in 16% of patients, with apeak at 24 and 36 months follow-up. Approximately35% of patients with migration showed infrarenal aorticneck dilatation. Among these patients, only 2 casesshowed a proximal type-I endoleak and only 4 patientsrequired treatment. The length of the proximal neck rep-resents a key factor in determining the presence of an en-doleak and, therefore, the need for further intervention:in those patients with stent-graft migration and a suffi-ciently long proximal neck, the sealing of the sac can beensured despite the migration.

The main weak point of our study is represented bythe large number of different types of endografts im-planted. This makes our statistical analysis less consis-tent, in particular for the detection of risk factors. For ex-ample, the use of stent grafts with suprarenal fixation,with or without side branches, could determine a signifi-cant reduction in the incidence of proximal neck enlarge-ment and stent migration [26].

Conclusion

Our study demonstrates that the proximal aortic neck canenlarge after EVAR. This enlargement represents themain risk factor for stent-graft migration and loss of sta-bility, leading eventually to the development of an endo-leak and the increase of the sac pressure, with a higherrisk of aneurysm rupture.

1970

1971

Infrarenal neck enlargement can occur early in fol-low-up because of the stent-graft oversizing, but thehighest incidence is observed in the second year of fol-low-up. The suprarenal neck dilates later; therefore, along and accurate follow-up is always required afterEVAR for the early identification and treatment of thiscomplication.

No risk factors were identified, and the supra- andinfrarenal neck dilatation seem to be unrelated events.A relation could be hypothesized between neck dilata-

tion and progression of the aneurysmal disease, as wellas neck dilatation and stent-graft radial force impressedon the aortic walls [3, 10]. Differences between supra-renal and infrarenal neck enlargement could be ex-plained by the different anatomical features of thesesites [12, 19].

Finally, US does not represent a useful tool in moni-toring neck modifications, whereas it seems to be ex-tremely important in the detection and characterizationof endoleaks.

References

1. Illig KA, Green RM, Ouriel K et al.(1997) Fate of the proximal aortic cuff:implications for endovascular aneu-rysm repair. J Vasc Surg 26:492–501

2. Lipski DA, Ernst CB (1998) Naturalhistory of the residual infrarenal aortaafter infrarenal abdominal aortic aneu-rysm repair. J Vasc Surg 27:805–812

3. Sonesson B, Malina M, Ivancev K et al. (1998) Dilatation of the infrarenalaneurysm neck after endovascular ex-clusion of abdominal aortic aneurysm.J Endovasc Surg 5:195–200

4. May J, White GH, Yu W et al. (1996)Prospective study of anatomo-patho-logical changes in abdominal aortic an-eurysms following endoluminal repair:Is the aneurysmal process reversed?Eur J Vasc Endovasc Surg 12:11–17

5. Broeders IA, Blankensteijn JD, Gvakharia A et al. (1997) The efficacyof transfemoral endovascular aneurysmmanagement: a study on size changesof the abdominal aorta during mid-termfollow-up. Eur J Vasc Endovasc Surg14:84–90

6. Walker SR, Macierewicz J, ElmarasyNM et al. (1999) A prospective studyto assess changes in proximal aorticneck dimensions after endovascular repair of abdominal aortic aneurysm. J Vasc Surg 29:625–630

7. Badran MF, Gould DA, Raza I et al.(2002) Aneurysm neck diameter afterendovascular repair of abdominal aortic aneurysm. J Vasc Interv Radiol13:887–892

8. Makaroun MS, Deaton DH (2001) Is proximal aortic neck dilatation afterendovascular aneurysm exclusion acause for concern? J Vasc Surg33:S39–S45

9. Prinssen M, Wever JJ, Mali WPThM et al. (2001) Concerns for the durabilityof the proximal abdominal aortic aneu-rysm endograft fixation from a 2-yearand 3-year longitudinal computed tomography angiography study. J VascSurg 33:S64–S69

10. Matsumura JS, Chaikof EL (1998)Continued expansion of aortic necksafter endovascular repair of abdominalaortic aneurysm. J Vasc Surg28:422–431

11. Matsumura JS, Pearce WH, McCarthyWJ et al. (1997) Reduction in aorticaneurysm size: early results after endo-vascular graft placement. J Vasc Surg25:113–123

12. Wever JJ, de Nie AJ, Blankensteijn JDet al. (2000) Dilatation of the proximalneck of infrarenal aortic aneurysm afterendovascular AAA repair. Eur J VascEndovasc Surg 19:197–201

13. Mahnken AH, Chalabi K, SchurmannK et al. (2000) Changes in the regionand the proximal aneurysm neck after endovascular repair of infrarenalaortic aneurysms. Rofo Fortschr GebRontgenstr Neuen Bildgeb Verfahr172:842–846

14. Thompson MM, Boyle JR, Hartshorn Tet al. (1998) Comparison of computedtomography and duplex imaging in assessing aortic morphology followingendovascular aneurysm repair. Br J Surg 85:346–350

15. Sonesson B, Resch T, Lanne T et al.(1998) The fate of the infrarenal aorticneck after open aneurysm surgery. J Vasc Surg 28:889–894

16. Malina M, Ivancev K, Chuter TAM,Lindh M et al. (1997) Changing aneu-rysmal morphology after endovasculargrafting: relation to leakage or persis-tent perfusion. J Endovasc Surg4:23–30

17. Resch T, Ivancev K, Brunkwall J et al.(2000) Midterm changes in aortic an-eurysm morphology after endovascularrepair. J Endovasc Ther 7:279–285

18. Lederle FA, Wilson SE, Johnson GR et al. (1995) Variability in measure-ment of abdominal aortic aneurysms.Abdominal aortic aneurysm detectionand management veterans administra-tion cooperative study group. J VascSurg 21:945–952

19. Wolinsky H (1970) Comparison of medial growth of human thoracic andabdominal aortas. Circ Res 27:531–538

20. Golzarian J, Struyven J (2001) Imagingof complications after endoluminaltreatment of abdominal aortic aneurysms. Eur Radiol 11:2244–2251

21. Haulon S, Lions C, McFadden EP et al.(2001) Prospective evaluation of magnetic resonance imaging after endovascular treatment of infrarenalaortic aneurysms. Eur J Vasc EndovascSurg 22:62–69

22. Engellau L, Larsson EM, AlbrechtssonU et al. (1998) Magnetic resonance imaging and MR angiography in endoluminally treated abdominal aorticaneurysms. Eur J Vasc Endovasc Surg15:212–219

23. Kramer SC, Gorich J, Pamler R, Aschoff AJ, Wisianowski C, BrambsHJ (2002) The contribution of MRI to the detection of endovascular aneurysm repair. Rofo Fortschr Geb Rontgenstr Neuen Bildgeb Verfahr174:1285–1288

24. Resch T, Ivancev K, Brunkwall J et al.(1999) Distal migration of stent-graftsafter endovascular repair of abdominalaortic aneurysms. J Vasc Interv Radiol10:257–264

25. Cao P, Verzini F, Zannetti S et al.(2002) Device migration after endolu-minal abdominal aortic aneurysm re-pair: analysis of 113 cases with a mini-mum follow-up period of 2 years. J Vasc Surg 35:229–235

26. Uflacker R, Robison J (2001) Endovas-cular treatment of abdominal aortic aneurysms: a review. Eur Radiol11:739–753