REVIEW ARTICLESRichard P. Cambria, MD, Section Editor

FromRanG

AuthRepDUN

Thetom

0741Cophttp

Current state of experimental imaging modalitiesfor risk assessment of abdominal aortic aneurysmRuben V. C. Buijs, BSc,a Tineke P. Willems, MD, PhD,b René A. Tio, MD, PhD,c

Hendrikus H. Boersma, PharmD, PhD,d,e Ignace F. J. Tielliu, MD, PhD,a

Riemer H. J. A. Slart, MD, PhD,d and Clark J. Zeebregts, MD, PhD,a Groningen, The Netherlands

Background: Abdominal aortic aneurysm (AAA) is a major cause of death in developed countries. Patients often lackclinical symptoms, most acute AAA patients do not survive rupture, and subsequent surgical repair has a significantpostoperative mortality. Diagnostics for AAAs are currently centered on aneurysm diameter, but recent studies claim thismethod to be insufficiently accurate. More accurate diagnostic criteria need to be indentified to minimize the amount ofunnecessary interventions and to provide earlier diagnosis of rupture-prone AAAs.Methods: A literature study using the MEDLINE database followed by manual cross-referencing provided original studiesconcerning AAA diagnostics.Results: The currently validated imaging modalities such as ultrasound, computed tomography, and magnetic resonanceimaging allow AAA research to develop in several directions. Some studies investigate whether clinically visible entitieslike thrombus, calcification, and vascular anatomy could be implemented directly into clinical practice through use ofultrasound or computed tomography. Experimental studies on intravascular ultrasound, positron emission tomography-computed tomography, ultrasound particle image velocimetry and superparamagnetic particles in magnetic resonanceimaging propose new methodologies to benefit AAA research. Other studies focus on available technology towardinflammation, metabolism, and the effects of hemodynamics on vascular integrity.Conclusions: Contradictory outcomes, low availability of experimental imaging modalities, and an often small populationsize hamper research in this field. Introducing new techniques and biomarkers in current or experimental modalitiesmay prove to be the next step in the development of new diagnostic criteria for the risk assessment of AAA rupture. Untilthen, the AAA diameter remains the gold standard as a clinical risk factor. (J Vasc Surg 2013;57:851-9.)

Abdominal aortic aneurysm (AAA) currently is a signif-icant cause of sudden death in developed countries. Its inci-dence quickly follows atherosclerosis and hypertension incardiovascular mortality with over 15,000 AAA-relateddeaths in the United States and 8000 deaths in the UnitedKingdom yearly.1 Since an AAA most frequently occurs inthe former or current smoking elderly population, it issuspected that its incidence will increase rapidly.

the Departments of Surgery, Division of Vascular Surgery,a

adiology,b Cardiology,c Nuclear Medicine and Molecular Imaging,d

d Hospital and Clinical Pharmacy,e University Medical Centerroningen, University of Groningen.or conflict of interest: none.rint requests: Clark J. Zeebregts, MD, PhD, Department of Surgery,ivision of Vascular Surgery, University Medical Center Groningen,niversity of Groningen, P.O. Box 30 001, 9700 RB Groningen, Theetherlands (e-mail: czeebregts@hotmail.com).editors and reviewers of this article have no relevant financial relationshipsdisclose per the JVS policy that requires reviewers to decline review of anyanuscript for which they may have a competition of interest.-5214/$36.00yright � 2013 by the Society for Vascular Surgery.://dx.doi.org/10.1016/j.jvs.2012.10.097

Up to now, measurement of aneurysm diameter is theclinically approved tool for the diagnosis and follow-up ofAAAs. Small aneurysms (diameter <5.5 cm in men,<5.0 cm in women) are followed up by routine monitoringof growth. Should the aneurysm growmore than 1 cm/y orover 5.5 cm (men; 5.0 cm in women), surgical interventionis indicated. The results of these population-based studies,however, are just partly helpful when extrapolated to theindividual level. Recent studies suggested this method ofdiameter assessment only is insufficient.2-4 Subjects withaneurysm diameters well under 5.0 cm can also rupture,whereas large aneurysms (>10 cm diameter) may alsoremain intact until death from other causes. In the searchfor new diagnostic methods, many different approacheshave been explored but have yet to be validated. This reviewassesses current options and promising new imaging possi-bilities in AAA rupture risk diagnostics.

METHODS

Both the UpToDate and MEDLINE/PubMed data-bases were searched for the following terms: “abdominalaortic aneurysm” and “AAA” as heading and “diagnosis,”

851

JOURNAL OF VASCULAR SURGERY852 Buijs et al March 2013

“computed tomography,” “CT,” “ultrasound,” “Doppler,”“magnetic resonance,” “MR,” “imaging,” “angiography,”“PET,” “PET-CT,” or “rupture” as keywords. Furthermanual cross-referencing provided the remaining literatureneeded. No limitations were set for either languages ortime periods.

Clinical risk assessment

AAA remains quiescent in most of the cases. Eventu-ally, AAA rupture classically presents with lower backpain, tenderness of the abdomen, and a pulsating abdom-inal mass that is painful on palpation. A hypovolemic shockcan eventually occur. Although a less common entity,symptomatic nonruptured AAAs present with one ormore previously stated symptoms without any form ofrupture. Often, asymptomatic patients are diagnosed inci-dentally after receiving ultrasound (US), computed tomog-raphy (CT), or magnetic resonance imaging (MRI) forother indications. A few clinically available patient charac-teristics are known as significant risk factors. A history ofsmoking and the amount of cigarettes consumed aredirectly dose- and time period-dependent factors to thedevelopment of an AAA.5 Age, male sex, and a positivefamily history for AAA also increase the risk for developingan AAA.1

Imaging modalities

Ultrasound-based techniques. US is currently thegold standard inmonitoring growth in aneurysms. Being rela-tively inexpensive, easy to use with low interobserver vari-ability, and no radiation burden, US is considered theperfect AAA screening modality.

In a retrospective cohort study, Lindholt et al6 providednew insights in the effects of calcification on AAA develop-ment. Through the application of US, the authors foundsignificantly slower expansion rates (Wilcoxon rank-sum;P < .001) of AAAs in men with small aneurysms(<30 mm) containing calcification in more than 50% ofthe total AAA wall circumference in comparison to menwith calcification in less than 50% of the wall circumference.In spite of these findings, mortality was similar in bothgroups (hazard ratio, 0.89; P¼ .604). AAA-related hospitaladmissions were only significantly lower in >50% calcifiedAAAwalls with univariate analysis and not as an independentrisk factor for hospital admission when tested in a multi-variate model. Therefore, although calcification is notprotective against AAA symptoms and death, this studysuggests itmight be protective against aneurysm expansion.6

Some researchers specifically use the duplex andDoppler modality of US. Whereas duplex research is mainlyfocusing on discovering postsurgical endoleaks and moni-toring AAA growth,7 Long et al8 discovered new insightsin presurgical risk assessment using Doppler imaging.They studied 56 patients using tissue Doppler imagingaiming to evaluate AAA behavior at different levels ofcompliance and diameter. A higher maximal meansegmental dilatation and other raised compliance markerswere found as AAA diameter increased. The authors

speculated this to be the consequence of elastin andcollagen degradation and, thus, a marker for rupturetendency. On the other hand, Long et al8 did not includea control group with nonaneurysmatic abdominal aortas.Also, the relationship with AAA outcome was left out,therefore, they were not in the position to align theirresults with firm conclusions about rupture risk.

AAA development is not just bound by morphologicvariables. Researchers have found links between hemody-namic variables and aneurysm growth and rupture. Liuet al9 constructed a new velocimetry technique, echoparticle image velocimetry (PIV). They combined US withmicrobubbles acting as flow tracers in several cardiovascularmodels, including an AAA model. The authors managed toaccurately measure several different complex flow patterns,such as vorticity, stagnation, and recirculation.9 Recently,Zhang et al10 performed a preliminary in vivo study usingthis technique in five human carotid arteries. Optical PIVis currently the gold standard for wall shear rate (WSR)and wall shear stress (WSS) measurement. Phase-contrastMRI also acts comparable to echo PIV in measuring WSRand WSS. Compared with the results of optical PIV andphase-contrast MRI, echo PIV measurements showedhighly resembling error bars. This suggests that echo PIVis a valid method for calculating WSS and WSR. Develop-ment of this technique could lead to clinical reproducibilityas a complement to US and may provide clinicians withhemodynamic information or prove useful in assessingWSS and its effects on aneurysm growth and rupture.

In the search for a more accurate diameter measure-ment, Van Essen et al11 studied 22 AAAs with intravascularultrasound (IVUS) as part of preoperative assessment.IVUS seemed not only to perform as well as CT angiog-raphy (CTA), it also seemed “the most accurate way todetermine the morphology of vascular structures (ie,calcium, thrombus). Previous to this research, Whiteet al12 already stated in their review article that IVUS isthe ultimate tool for AAA assessment. In spite of this,IVUS remains an invasive procedure that is far less availablethan US, CTA, or MRI. Its use would be an addition to thecurrent standard instead of an alternative, and whether theadditional costs outweigh the potential benefits is stillunclear. It, therefore, is unlikely that, in its current state,IVUS will have a great impact on AAA diagnostics.

CTA. CTA is currently the first choice in determiningthe specific AAA anatomy for preoperative assessment.13

This choice revolves solely around its ability to renderhigh-resolution images in acute and nonacute circum-stances. US may perform well in screening, yet somespecific morphologic details are only presented by CTA.Exact size, width, and length of the aneurysm are variablesthat are vital in preparing AAA surgery.

Although aneurysm diameter and growth are the majorrisk factors for rupture, other morphologic details are beingconsidered to be valuable in estimating the chances of AAArupture. Not only the anatomic aspects of the vessel itselfbut also their association to other tissues and vertebraeare likely to attribute to risk for AAA rupture. In a CTA-

Fig 1. Computational wall stress and rupture risk assessment in an abdominal aortic aneurysm (AAA). A three-dimensional model is recreated from raw computed tomography (CT) data. Wall stress is measured quantitativelyand translated to a color gradient. Blue, green, and red portray low, intermediate, and high wall stress respectively.14

(Reprinted with permission from Elsevier.)

JOURNAL OF VASCULAR SURGERYVolume 57, Number 3 Buijs et al 853

based computational study, Fillinger et al14 found an asso-ciation between peak AAA wall stress and risk for rupture(Fig 1). In a later study, Fillinger et al15 failed to finda significant correlation between thrombus size and AAArupture in a noncomputational retrospective CTA study.Speelman et al16 opposed these results when their findingssuggested a link between intraluminal thrombi and theirinfluence on wall stress. Using computational interpreta-tions of CTA scans, they discovered that an increase inthrombus size would increase the AAA growth rate butwould also be associated with lower wall stress. Consid-ering these ambivalent findings, the effects of intraluminalthrombus remain controversial.

AAA is known for its rigorous change in vessel anatomy.The aneurysmal sac grows in an unpredictable fashionwith curving and sloping against the high intravascularpressure. These structural changes, and concomitant throm-bosis, calcification, and atherosclerosis, affect the dynamicsof passing blood. However, these pathologic and incidentalentities are not the only cause for damaging blood flowpatterns. In-born tortuosity of the vessel and otherwisephysiologically formed anatomic structures might coinci-dentally result in damaging hemodynamics. For example,as the abdominal aorta follows the spinal curvature, theposterior wall is subjected to a higher hemodynamic stresscompared with the anterior wall. In the risk assessment ofAAA, a hydrodynamic engineering perspective has alwaysbeen the mainstay. The diameter criterion is based on thelaw of Laplace, although it is argued that the diameter playsonly a partial role in the biomechanics of AAAs. Its influ-ence on rupture may be trivial in comparison to many otherinfluences on the vessel stress and strength ratio.4

Therefore, the focus has shifted toward computationalAAA models for the evaluation of the different effects ofvessel anatomy. Doyle et al17 showed that asymmetricAAAs with localized wall thickness variation portray highermechanical stresses and an increase in AAA rupture risk.They also theorized how the risk of rupture is connectedto aneurysm diameter asymmetry. This might consequentlyinfluence the amount of wall stress on the posteriorabdominal aortic wall. The asymmetry variation was calcu-lated as the difference between the major and minor axes atthe maximum width. The authors showed that increasedasymmetry leads to increased posterior wall stress, implyingfuture rupture. In eight and nine out of 15 patients, diam-eter and diameter asymmetry, respectively, were found tobe significantly influential (P < .05) on posterior wallstress. Doyle et al17 claimed that diameter asymmetry ison par with current risk assessment using aneurysmdiameter.

In an earlier retrospective cohort study by Fillingeret al,15 diameter asymmetry was calculated in CTA scans oftwo different AAA risk groups, elective vs ruptured. A signif-icantly greater diameter asymmetry (P ¼ .03; OR, 3.2) wasfound with patients receiving acute aneurysm surgery.Besides this, Fillinger et al15 found minimized tortuosityof the aorta to be as influential a factor as smoking andsex were on future rupture (P ¼ .01; OR, 3.3).

A number of promising software packages is beingdeveloped for the computation analysis of available CTand CTA images. Blood flow, pressure distribution, shearstress, and the interaction of curvature, diameter asymme-try, and intraluminal structures on any of these factors arepart of a host of influences on the mechanics of wall

Fig 2. A male patient with impending rupture of an abdominal aortic aneurysm (AAA). Bleeding into the intraluminalthrombus is portrayed in unenhanced (A) and contrast-enhanced (B) axial computed tomography (CT) images asa crescentic form with hyperattenuation (arrow in A).40 (Reprinted with kind permission of D. Rakita, Department ofRadiology, Division of Body Imaging, Long Island Jewish Medical Center.)

JOURNAL OF VASCULAR SURGERY854 Buijs et al March 2013

deterioration and rupture proneness. We expect that thisbiomechanics-based perspective will be of great importancefor future research.

A very different entity found in CT imaging is theflowing of contrast into the thrombus combined withtransformation of the lumen. This might be the cause ofa sign called hyperattenuating crescents (Fig 2). Histopath-ologic testing showed this phenomenon to be hemorrhageinto the thrombus and, consequently, the aneurysm wall.Mehard et al18 and Consalves et al19 described a sign ofimpending rupture characterized by high-attenuatingcrescents in the AAA wall. Consalves et al19 claimed thecrescent sign to be of prognostic value for AAA rupture.In the study by Mehard et al,18 the specificity of high-attenuating crescents vs aneurysm complication was 93%.Hyperattentuation was defined by Siegel et al20 as a signalbeing higher than contrast visibility of the psoas muscle inenhanced scans or a signal higher than that of patent lumenin unenhanced scans. In their study, 21% of patients withruptured aneurysms showed hyperattenuating crescents,while no patients with intact aneurysms showed this sign.Roy et al21 most recently investigated signs of bleeding inthe intraluminal thrombus and rupture site in AAA patients.Although the crescentic form was found significantly morein the ruptured group rather than the stable aneurysm group(38% vs 14%; P ¼ .02), localized areas of hyperattenuationwere found in both groups without significant variation.Therewas,however, a significantlyhigher (P¼ .02) thrombustotal attenuation in the ruptured group. Roy et al21 compro-misingly stated that “whether these findings also predict AAArupture, remains to be established.”

Positron emission tomography-computed tomog-raphy (PET-CT). Up until recently, only anatomicallyfocused imaging modalities were used in risk assessmentof AAAs. The hybridization of PET-imaging with CT,provides added value over the separate use of PET and CTalone. PET enables functional imaging of cellular activity inAAA tissue, whereas the addition of CT grants improvedanatomic localization and characterization.

The radiopharmaceutical tracer 18F-fluorodeoxyglu-cose (FDG) is designed to image high-glucose-using cells.

FDG accumulates in inflammatory sites because of its richmacrophage colonization. AAAs have also been proved toattract inflammatory cells and cytokines like matrix metal-loproteinases (MMPs).22 Mycotic thrombus in AAAs isa more severe infection of the AAA, often because ofbacteremia or otherwise circulating bacteria. Mycotic aneu-rysms tend to rupture more easily and are known for a highmortality rate. Diagnosis of mycotic aneurysm is difficult,even on CT images. There are several suggestive signs,such as the presence of perivascular fluid or raised inflam-matory blood markers, but none are either sensitive orspecific enough in most of the cases. Recently, PET-CThas shown to be very valuable.23 PET-CT might also beable to provide information on short-term outcomes frompharmaceutical interventions on this infection (Fig 3).

In 2002, Sakalihasan et al24 attempted to link FDGuptake with AAAs using PET. After performing staticwhole-body PET on 26 patients, 10 showed heightenedactivity in the abdominal aorta. Their preliminary researchsuggested PET has the capacity to portray metabolic activitywithin the aneurysm wall. In 2008, through a highly signi-ficant (r ¼ 0.93; P < .0001) association between 18F-FDGuptake and histologically assessed macrophage-density,Reeps et al25 were able to confirm that vascular inflam-mation could be detected accurately using PET-CT. Insymptomatic patients, this activity proved to be significantlyhigher than in asymptomatic patients (P < .001). Duringthe same year, Sakalihasan et al24 presented preliminaryresults of an ongoing research. In a patient group of 26patients, they found that subjects with positive PET scans(n¼ 10) showed either AAA symptoms (n ¼ 5), expansion,leakage of blood, and fulminant rupture (n¼ 3). Two PET-positive patients remained asymptomatic. Patients withnegative PET images were treated after delay of severalmonths out of convenience for the patient, without experi-encing the adverse effects as seen in the PET-positivepatients. Consequently, the authors claimed that FDG-PET could provide an argument whether or not to justifyan intervention.26

In a longitudinal observational study, Kotze et al27 mostrecently aimed to explain the relationship between FDG

Fig 3. Abdominal aortic aneurysm (AAA) images from computed tomography (CT) (A and B) and 18F-fluorodeox-yglucose (18F-FDG) positron emission tomography (PET) (C and D). The arrows show how a region with high FDGuptake in the PET image coincides with a mural thrombus in the CT image.41 (Reprinted with permission from Elsevier.)

JOURNAL OF VASCULAR SURGERYVolume 57, Number 3 Buijs et al 855

uptake and future growth rate of AAAs. Combined with USimaging, the expansion was measured after 1 year and asso-ciated with whole-vessel standardized uptake value. Aninverse correlation of –0.501 (P¼ .011) was found betweenwhole-vessel standardized uptake value and aneurysmalgrowth. Two major study limitations were identified,however. First, the follow-up lasted until 12 months.Second, the follow-up was performed using US, whichhas an error margin that can be larger than small AAAsexpand in a year. Nonetheless, the authors conclude theirpublication implying that less metabolically active aneu-rysms are more likely to grow and perhaps subsequentlyrupture. These contradicting conclusions on the value ofFDG undermine the results presented by both authorsand might set back the research in this field.

However, FDG is not the sole imaging agent inPET-CT imaging of AAAs. Nahrendorf et al28 investigateda modified dextran-coated iron oxide nanoparticle thatparticularly binds to macrophages. In a murine aneurysmmodel using ApoE -/- mice treated with angiotensin II,PET-CT showed significantly higher signals from theAAA model (2.46 6 0.48, standard uptake value) than inwild-type littermates (0.82 6 0.05; P < .05). Flow cytom-etry, immunohistologic analysis, and scintillation countingall portrayed how the nanoparticles migrated mainly tomacrophages and monocytes within the AAA wall. Severalother PET-CT reagents have been found in variousmedical fields such as interleukin-2 and choline-specificpharmaceuticals that could be applied in the field of inflam-mation PET diagnostics. It may only be a matter of timebefore these are introduced to AAA diagnostics. PET-CTis starting to prove its worth over CTA and US. Yet, thestudies performed all had a low number of patients in their

cohorts, and some authors questioned the reproducibilityof their study. These results should, therefore, be regardedas preliminary.

MRI/magnetic resonance angiogram (MRA).Although magnetic resonance is widely accepted as animaging tool, its place within AAA diagnostics has notbeen established. In spite of this, multiple groups havebeen looking into the value of MRI and MRA in AAArisk assessment. In 1995, Prince et al29 started assessingthe usefulness of gadolinium-enhanced MRA for thediagnosis of AAAs. It performed equally as well as CTA,without need of iodine-based contrast. This was a greatimprovement at the time, as renal complications attrib-utable to the high iodine concentrations in CTA contrastwere far more common.

In 2004, Kramer et al30 used gadolinium-pentetic acidin identifying atherosclerotic fibrous caps in AAAs. UsingT2-weighted MRI imaging, they managed to accuratelydelineate the fibrous cap and thrombus from the vesselwall (Fig 4). Although the study was focused on identi-fying vulnerable atherosclerotic plaques, it consequentlyprovided new insights on vulnerability of the AAA wall.Also, starting from an atherosclerosis-focused perspective,Sadat et al31 linked MRI to AAA extracellular matrix degra-dation. Ultrasmall superparamagnetic iron oxide (USPIO)particles are known for a strong interaction with macro-phages and leukocytes. By observing the amount ofUSPIO uptake in the cells, which translates to lower T2-and T2* signal intensity, Sadat et al31 theorized the phago-cytic activity in the AAA wall could be quantified. In theirstudy, T2- and T2* values in the AAA wall correlatedsignificantly (Spearman’s correlation coefficient ¼ .90;P < .001) after injection with USPIO. This propensity to

Fig 4. A T2-weighted magnetic resonance imaging (MRI) image (A) and T1-weighted MRI image after gadoliniumpentetic acid infusion (B) of an abdominal aortic aneurysm (AAA) with intraluminal thrombus shows three distinctlayers. These components were confirmed by histopathology (C). The fibrous cap contained high numbers (120 whiteblood cells/hpf) of polymorphonuclear leukocytes, as was seen using high power field microscopy (D).30 (Reprintedwith permission from Elsevier.)

JOURNAL OF VASCULAR SURGERY856 Buijs et al March 2013

USPIO uptake by the aneurysm wall suggests inflammationis abundant. Using this technology, it seems increasinglyfeasible to quantify inflammation and concomitantly toquantify stability of the AAA. Regrettably, there was nohistologic control for these findings, but as a feasibilitystudy, it provided encouraging results for larger cohortstudies.31

Nchimi et al,32 however, did manage to provide histo-logic backgrounds to their in vivo USPIO MRI study.Post-USPIO signal-to-noise ratios for thrombus tissue andmuscle tissue were significantly different (P ¼ .016), aswere the contrast-to-noise ratios for the luminalsublayer of the thrombus (P < .001) and deeper thrombus(P< .012).UsingUSPIO as a phagocytosis-specific imagingagent, accurate morphologic assessment of the thrombuscan be achieved by visualizing phagocytic activity corrobo-rating with immunohistochemical stainings. CD66b (poly-morphonuclears), CD68 (macrophages), and pro-MMP-9(extracellular matrix remodeling) were found in significantlevels (P ¼ .009, P ¼ .002, P ¼ .014, respectively) opposedto a decrease in signal intensity of the luminal sublayer of thethrombus. USPIO is known for a high rate of liver clearance.

It is, therefore, questionable whether the low signal intensitywas entirely due to high macrophage uptake or whether itwas partly due to the rapid clearance.32 Sadat et al31 alsostated that the USPIO agent used in their study is no longercommercially available. Therefore, repetition of this researchis highly improbable. Acknowledgement through otherstudies with larger cohorts, repeating, or enriching thisresearch is essential, as both authors state their researchwas bound by the limited amount of included patients.

Bio-optical imaging. Bio-optical imaging uses tech-niques such as chemiluminescence, bioluminescence, andnear-infrared fluorescence (NIRF). As of yet, bio-opticalimaging of AAAs has only been used in experimentalsettings. Luminescent probes react to a chosen substrate (eg,MMP, VEGF) by proteolytic cleavage of the substrate andconsequently emit fluorescence. Intensity of the signal is,therefore, directly correlated to the substrate concentrationin the imaged tissue. As inflammation seems likely to beinvolved in AAA growth and rupture, more and more indi-vidual factors are being discovered. Inflammation-inducedelastin degradation in the extracellular matrix has beenshown to be of influence on AAA development. The main

Fig 5. A female patient with a symptomatic nonrupturedabdominal aortic aneurysm (AAA). Calcification was found alongthe complete circumference of the vessel. The high-density signalis distinctly visible in spite of the contrast agent in the lumen.

JOURNAL OF VASCULAR SURGERYVolume 57, Number 3 Buijs et al 857

actors in this process seem to be MMP-2, MMP-9, andtheir counteractor tissue inhibitor of metalloproteinase-1(TIMP-1).33 Kaijzel et al34 proposed a new method ofAAA identification, using MMP-specific probes in fibulin-4reduced-expression allele mice. Fibulin-4 is relevant in theorganization of extracellular matrix structures and regardedan important factor for arterial integrity. Although mainlyaffecting the ascending and descending thoracic aorta, thisspecific knockdown perpetuates aortic aneurysm dilatationin more and less severe degrees for homozygous (R/R) andheterozygous (þ/R) knockdowns, respectively. Usinga NIRF imaging system, a dose-dependent rise in MMPfluorescence was shown in the þ/R (1.79-ns life-time) andR/R (2.02-ns life-time) mice vs the control littermates(<1-ns life-time). After whole mouse scanning, individualhearts and aortas were harvested and scanned for MMP.Control mouse aortic arch MMP signal intensity reached19.75 relative fluorescence units (rfu), whereas increased rfu(26.53 and 105.77) were measured in þ/R and R/R mice,respectively. Using histologic analysis, zymography, andfluorescence molecular tomography, similar results werefound. The authors suggested that NIRF imaging of MMPcould provide information on aneurysm development in thethoracic aorta.34

Neovascularization has also been proposed to be ofvalue in aneurysm pathology. Nonaneurysmal aortas andintact aneurysms demonstrate a lower degree of muralneovascularization than ruptured aneurysms.22 Vascularendothelial growth factor (VEGF) is a major proangiogenicmediator. Tedesco et al35 sought to associate VEGFreceptor (VEGFR) expression with aneurysm developmentin AAAs in an apolipoprotein E-deficient (Apo E -/-)murine AAA model. These murine models developedsuprarenal AAAs after infusion with angiotensin II (AngII mice). Selected mice were given either angiogenesisinhibitors (Exel 0862) or doxycycline (positive control),which is known to limit experimental AAA progression.36

Empty vehicle was given as a control to others and inacti-vated VEGF probes (scVEGF/in) acted as a control fornonreceptor-mediated reactions. NIRF imaging resultedin significantly higher VEGFR signal intensity for aneu-rysmal aortic segments in relation to nonaneurysmalsegments in Ang II mice and control mice and Ang IIscVEGF/in mice. This association was confirmed byfluorescence microscopy. Also, angiogenesis inhibitionseemed to significantly limit AAA growth in controls vsExel 0862 and doxycycline. Mural inflammation too corre-lated significantly in Exel 0862- and doxycylin-treated micevs control vehicle-treated mice. The authors concludedthat VEGFR has definitive experimental value as a param-eter for aneurysm progression. In the eventual case thatthese results may translate to human pathology of AAAs,this research may provide new diagnostic options in theform of bio-optical imaging.35

DISCUSSION

There seem to be many contenders for validation ofAAA rupture risk. Unfortunately, most of the preliminary

and introductory studies that have been performed lackscientific strength. Often, this was due to small or unbal-anced cohort population size. Studies focused on clinicalimplementation seem to be the heart of AAA research.Their main goal is to enhance the diagnostic capabilities ofalready existing imaging modalities. This clinically focusedresearch clusters known clinical entities, variables, andcomorbidities, and checks off against endpoints like aneu-rysm growth and rupture. Except for smoking, familialoccurrence, vessel tortuosity, and diameter asymmetry,most experimental variables were rejected as a risk factorfor AAA rupture. Calcification is a clinical entity that hasreceived little scientific attention (Fig 5). Just as thrombosis,the effect of calcification on the integrity of the abdominalaorta seems ambiguous. Calcification decreases elasticityand compliance of a vessel. It can, therefore be hypothesizedthat calcification adversely affects the vessel’s reactivity tostress and as a consequence increases rupture risk.37

Abdominal aortic calcification has also been associatedwith vascular morbidity and mortality.38 Contrary to thesestatements, using US, Lindholt et al6 showed an increaseof calcification would reduce the need for intervention,suggesting calcification has a protective role. Still, untilnow, little else other than computational calcificationmeasurement studies have been focused on the meaningof calcification. This lack of research on the subject istheorized to be due to several limitations in CTA scans ofAAAs. The Agatston score is currently regarded as thebest validated method in coronary calcification measure-ment. However, the Agatston score is inapplicable instandard AAA CTA scanning. Since CTA uses contrast fluidto identify the abdominal aorta, a background signal isproduced that floods all intravascular high-density particles.Noncontrast CT scans are of little value in routine vascularimaging, and it is questionable whether performing botha noncontrast and contrast CT is justifiable because thiswill double the radiation exposure. The lack of method-ology to investigate calcification in CTs needs to be over-come for this specific field of interest to be accurately

JOURNAL OF VASCULAR SURGERY858 Buijs et al March 2013

investigated. Dual-energy CT might provide an alternative.Dual-energy CT uses two different tube voltages, whichcould be set to both the absorption rates of calcium andsoft tissues like blood vessels. It, therefore, could bypassthe “contrast contamination.”This highly applicable updateto common CT imaging needs further investigation andvalidation, but its potential should be recognized. Resultsof these studies might be implemented inexpensively andbe highly feasible in medical practice.

Research in this field is limited to the boundaries of exist-ing modalities. Yet, experimental imaging modalities andsubstrates can go where existing modalities fall short. Forexample, the highly specific, radiation-independent imagingof human biomarkers for AAA risk holds great promise. Bio-optical imaging, however, is still limited to laboratorystudies, as both the modalities as the reagents are in manyways unsuitable to be used in humans. Further developmentthat focuses on creating supporting materials will improvepracticality and might propel this field to new heights.Experimental use of MRI and PET-CT seems to be themost promising new lead in AAA diagnostics. Their dualfunction of both anatomic and functional imaging of theabdominal aorta is a contribution and not an alternative tothe institutionalized US and CTA in AAA diagnostics. Ofvalue may be the next-generation PET-MRI camera forsimultaneous imaging of anatomy and function. The meritsof also being able to measure biochemical, inflammatory,metabolic activity, apoptosis, and angiogenesis are highlysignificant. Fluctuation in inflammation and metabolismmay provide a better instant understanding of pharmaceu-tical interactions with AAA or wall weakening on a muchshorter term. This is in stark contrast to US screening ofAAA growth, which is valuable only if repeated every 6months. Nevertheless, a study by Osman et al39 showedhow clinically significant findings such as cirrhosis, kidneylesions, and AAAs on the CT portion of PET-CT were notshown by PET or combined imaging. Although these find-ings are incidental, major diagnoses might be overlooked ifCTA interpretation moves to the background.

Fact remains that currently, aneurysm diameter is theonly criterion clinicians can rely on, even though this israpidly being considered to be less so. The amount ofunnecessary interventions for AAA repair and the degreeof risk taken when approaching a small aneurysm withwatchful waiting are equally unknown. We believe thatan expansion of the current paradigm surrounding theAAA diameter toward a broader interpretation of biome-chanical influences will be the next step in the predictionof AAA rupture. Randomized clinical trials with any ofthese methodologies or risk factors may still only be a futureprospect, but eventually their intellectual offspring mightprovide a broad yet accurate screening protocol for therisk assessment of the AAA.

AUTHOR CONTRIBUTIONS

Conception and design: RB, TW, CZAnalysis and interpretation: RB, TW, CZ

Data collection: RBWriting the article: RB, CZCritical revision of the article: TW, RT, HB, IT, RS, CZFinal approval of the article: RB, TW, RT, HB, IT, RS, CZStatistical analysis: Not applicableObtained funding: Not applicableOverall responsibility: CZ

REFERENCES

1. Sakalihasan N, Limet R, Defawe OD. Abdominal aortic aneurysm.Lancet 2005;365:1577-89.

2. Doyle BJ, Callanan A, Burke PE, Grace PA, Walsh MT, Vorp DA, et al.Vessel asymmetry as an additional tool in the assessment of abdominalaortic aneurysms. J Vasc Surg 2009;49:443-54.

3. Van de Geest JP, Di Martino ES, Bohra A, Makaroun MS, Vorp DA.A biomechanics-based rupture potential index for abdominal aorticaneurysm rupture risk assessment. Ann NY Acad Sci 2006;1085:11-21.

4. Vorp DA. Biomechanics of abdominal aortic aneurysm. J Biomech2007;40:1887-902.

5. Wilmink TB, Quick CR, Day NE. The association between cigarettesmoking and abdominal aortic aneurysms. J Vasc Surg 1999;30:1099-105.

6. Lindholt J. Aneurysmal wall calcification predicts natural history ofsmall abdominal aortic aneurysms. Atherosclerosis 2008;197:673-8.

7. Beeman BR, Murtha K, Doerr K, McAfee-Bennett S, Dougherty MJ,Calligaro KD. Duplex ultrasound factors predicting persistent type IIendoleak and increasing AAA sac diameter after EVAR. J Vasc Surg2010;52:1147-52.

8. Long A, Rouet L, Bissery A, Rossignol P, Mouradian D, Sapoval M.Compliance of abdominal aortic aneurysms evaluated by tissueDoppler imaging: correlation with aneurysm size. J Vasc Surg 2005;42:18-26.

9. Liu L, Zheng H, Williams L, Zhang F, Wang R, Hertzberg J, et al.Development of a custom-designed echo particle image velocimetrysystem for multi-component hemodynamic measurement: systemcharacterization and initial experimental results. Phys Med Biol2008;53:1397-412.

10. Zhang F, Lanning G, Mazzaro L, Barker AJ, Gates PE, Strain WD,et al. In vitro and preliminary in vivo validation of echo particle imagevelocimetry in carotid vascular imaging. Ultrasound Med Biol 2011;37:450-64.

11. Van Essen JA, Gussenhove EJ, Blankensteijn JD, Honkoop J, vanDijk LC, van Sambeek MR, et al. Three-dimensional intravascularultrasound assessment of abdominal aortic aneurysm necks. J EndovascTher 2000;7:380-8.

12. White RA, Donayre C, Kopchok G, Walot I, Wilson E, de Virgilio C.Intravascular ultrasound: the ultimate tool for abdominal aortic aneu-rysm assessment and endovascular graft delivery. J Endovasc Surg1997;4:45-55.

13. Isselbacher EM. Thoracic and abdominal aortic aneurysms. Circulation2005;111:816-28.

14. Fillinger MF, Marra SP, Raghavan ML, Kennedy FE. Prediction ofrupture risk in abdominal aortic aneurysm during observation: wallstress versus diameter. J Vasc Surg 2003;37:724-32.

15. Fillinger MF, Racusin J, Baker RK, Cronenwett JL, Teutelink A,Schermerhorn ML, et al. Anatomic characteristics of rupturedabdominal aortic aneurysm on conventional CT scans: implications forrupture risk. J Vasc Surg 2004;39:143-5.

16. Speelman L, Schurink GWH, Bosboom MH, Buth J, Breeuwer M, vande Vosse FN, et al. The mechanical role of thrombus on the growthrate of an abdominal aortic aneurysm. J Vasc Surg 2010;51:19-26.

17. Doyle BJ, Callanan A, Burke PE, Grace PA, Walsh MT, Vorp DA, et al.Vessel asymmetry as an additional diagnostic tool in the assessment ofabdominal aortic aneurysms. J Vasc Surg 2009;49:443-54.

18. Mehard WB, Heiken JP, Sicard GA. High-attenuating crescent inabdominal aortic aneurysm wall at CT: a sign of acute or impendingrupture. Radiology 1994;192:359-62.

JOURNAL OF VASCULAR SURGERYVolume 57, Number 3 Buijs et al 859

19. Gonsalves C. The hyperattenuating crescent sign. Radiology 1999;211:37-8.

20. Siegel CL, Cohan RH, Korobkin M, Alpern MB, Courneya DL,Leder RA. Abdominal aortic aneurysm morphology: CT features inpatients with ruptured and non-ruptured aneurysms. Am J Roentgenol1994;163:1123-9.

21. Roy J, Labruto F, Beckman MO, Danielson J, Johansson G,Swedenborg J. Bleeding into the intraluminal thrombus in abdominalaortic aneurysms is associated with rupture. J Vasc Surg 2008;48:1108-13.

22. Choke E, Thompson MM, Dawson J, Wilson WR, Sayed S, Loftus IM,et al. Abdominal aortic aneurysm rupture is associated with increasedmedial neovascularization and overexpression of proangiogenic cyto-kines. Arterioscler Thromb Vasc Biol 2006;4:129-49.

23. Davidson JM, Montilla-Soler JL, Broussard E, Wilson R, Cap A,Allen T. F-18 FDG PET-CT imaging of a mycotic aneurysm. Clin NuclMed 2005;30:483-7.

24. Sakalihasan N, Van Damme H, Gomez P, Rigo P, Lapiere CM,Nusgens B, et al. Positron emission tomography (PET) evaluation ofabdominal aortic aneurysm (AAA). Eur J Vasc Endovasc Surg 2002;23:431-6.

25. Reeps C, Essler M, Pelisek J, Seidl S, Eckstein HH, Krause BJ.Increased 18F-fluorodeoxyglucose uptake in abdominal aorticaneurysms in positron emission/computed tomography is associatedwith inflammation, aortic wall instability, and acute symptoms. J VascSurg 2008;48:417-23.

26. Sakalihasan N, Hustinx R, Gomez P, Nchimi A, Limet R. Can posi-tron emission tomography (PET) predict the risk of rupture ofabdominal aortic aneurysm (AAA)? In: Controversies and updates invascular surgery 2008. Torino, Italy: Ed. Minerva Medica; 2008. p.15-22.

27. Kotze CW, Groves AM, Menezes LJ, Harvey R, Endozo R, Kayani IA,et al. What is the relationship between 18F-FDG aortic aneurysmuptake on PET/CT and future growth rate? Eur J Nucl Med MolImaging 2011;38:1493-9.

28. Nahrendorf M, Keliher E, Marinelli B, Leushner F, Robbins CS,Gerszten RE, et al. Detection of macrophages in aortic aneurysms bynanoparticle positron emission tomography-computed tomography.Arterioscler Thromb Vasc Biol 2011;31:750-7.

29. Prince MR, Narasimham DL, Stanley JC, Wakefield TW, Messina LM,Zelenock GB, et al. Gadolinium-enhanced magnetic resonance angi-ography of abdominal aortic aneurysms. J Vasc Surg 1995;21:656-69.

30. Kramer CM, Cerilli LA, Hagspiel K, DiMaria JM, Epstein FH,Kern JA. Magnetic resonance imaging identifies the fibrous cap inatherosclerotic abdominal aortic aneurysm. Circulation 2004;109:1016-21.

31. Sadat U, Taviani V, Patterson AJ, Young VE, Graves MJ, Teng Z, et al.Ultrasmall superparamagnetic iron oxide-enhanced magnetic resonanceimaging of abdominal aortic aneurysms–a feasibility study. Eur J VascEndovasc Surg 2011;41:167-74.

32. Nchimi A, Defawe O, Brisbois TK, Defraigne JO, Magotteaux P,Massart B, et al. MR imaging of iron phagocytosis in intraluminalthrombi of abdominal aortic aneurysms in humans. Radiology2010;254:973-81.

33. Lemaître V, Soloway PD, D’Armiento J. Increased medial degradationwith pseudo-aneurysm formation in apolipoprotein E-knockout micedeficient in tissue inhibitor of metallopreoteinases-1. Circulation2003;107:333-8.

34. Kaijzel EL, van Heijningen PM, Wielopolski PA, Vermeij M,Koning GA, van Cappellen WA, et al. Multimodality imaging revealsa gradual increase in matrix metalloproteinase activity at aneurysmallesions in live fibulin-4 mice. Circulation Cardiovascular Imaging2010;3:567-77.

35. Tedesco MM, Terashima M, Blankenberg FG, Levashova Z, Spin JM,Backer MV, et al. Analysis of in situ and ex vivo vascular endothelialgrowth factor receptor expression during experimental aortic aneurysmprogression. Arterioscler Thromb Vasc Biol 2009;29:1452-7.

36. Thompson RW, Curci JA, Ennis TL, Mao D, Pagano MB, Pham CT.Pathophysiology of abdominal aortic aneurysms: insights from theelastase-induced model in mice with different genetic backgrounds.Ann N Y Acad Sci 2006;1085:59-73.

37. Speelman L, Bohra A, Bosboom EM, Schurink GW, van de Vosse FN,Makaroun MS, et al. Effects of wall calcifications in patient-specific wallstress analyses of abdominal aortic aneurysms. J Biomech Eng2007;129:105-9.

38. Wilson PW, Kauppila LI, O’Donnell CJ, Kiel DP, Hannan M,Polak JM, et al. Abdominal aortic calcific deposits are an importantpredictor of vascular morbidity and mortality. Circulation 2001;103:1529-34.

39. Osman MM, Cohade C, Fishman EK, Wahl RL. Clinically significantincidental findings on the unenhanced CT portion of PET/CT studies:frequency in 250 patients. J Nucl Med 2005;46:1352-5.

40. Rakita D, Newatia A, Hines JJ, Siegel DN, Friedman B. Spectrum ofCT findings in rupture and impending rupture of abdominal aorticaneurysms. 2007;27:497–507.

41. Ponce Herrera C, Borrego Dorado I, Ruiz Franco-Baux J, CabreraMoreno R. Imagen de un aneurisma de aorta abdominal con PET 18F-FDG. Rev Esp Med Nucl 2004;23:425-6.

Submitted Jun 3, 2012; accepted Oct 21, 2012.