REPORTING STANDARDS
`
`Reporting standards for endovascular aortic
`aneurysm repair
`
`Elliot L. Chaikof, MD, PhD, Jan D. Blankensteijn, MD, Peter L. Harris, MD, Geoffrey H. White, MD,
`Christopher K. Zarins, MD, Victor M. Bernhard, MD, Jon S. Matsumura, MD, James May, MD,
`Frank J. Veith, MD, Mark F. Fillinger, MD, Robert B. Rutherford, MD, and K. Craig Kent, MD, for the
`Ad Hoc Committee for Standardized Reporting Practices in Vascular Surgery of The Society for
`Vascular Surgery/ American Association for Vascular Surgery
`
`In 1997, an initial set of reporting standards for studies
`directed at the endovascular repair of the infrarenal aortic
`aneurysm was introduced by the Ad Hoc Committee for
`Standardized Reporting Practices in Vascular Surgery of
`the Society for Vascular Surgery/ International Society for
`Cardiovascular Surgery. 1 Since this initial report, the rapid
`evolution of the field has led to new concepts and insights
`into factors that define success and clinicopathologic fea(cid:173)
`tures that have a measurable impact on outcome. Thus, a
`revised set of recommended standards is presented to ac(cid:173)
`commodate these and other developments in this emerging
`discipline.
`
`CLASSIFICATION CRITERIA FOR ARTERIAL
`ANEURYSMS
`
`Recommendations regarding the definition and classi(cid:173)
`fication of arterial aneurysm have been previously pub(cid:173)
`lished.2 As noted in this initial document, precise nomen(cid:173)
`clature is desirable, but a risk of overclassification exists that
`may result in small patient subgroups that preclude mean(cid:173)
`ingful data analysis. Moreover, no classification of an eu(cid:173)
`rysms on the basis of a single factor has proved to be entirely
`satisfactory. Therefore, classification of aneurysm is recom(cid:173)
`mended with respect to a combination of fac tors, includ(cid:173)
`ing: ( 1) site; (2) etiology; and ( 3) clinicopathologic mani-
`
`Competition of interest: ELC has been paid a consu lting fee by and received
`clinical research funding from Guidant. MFF has received a speaking fee
`from Medtronic. JSM has been paid a consulting fee by and has received
`clinical research funding from Guidant, Medtronic, and WL Gore . He has
`also received research support from Boston Scientific. VMB is a consu ltant
`to and owns stock in Guidant. JM has been paid a consulting fee by
`Medtronic. CZ has been paid a consulting fee by and owns shares in
`Medtronic.
`Add itional material for this article may be found online at www.mosby.
`com/ jvs.
`Reprint requests: Elliot L. Chaikof, MD, PhD, 1639 Pierce Dr, Rm 5 105,
`Emory University, Atlanta, GA 30322 (e-mail: echaiko@cmory.edu ).
`J Vase Surg 2002;35: I 048 -60.
`Co pyright © 2002 by The Society for Vascular Surgery and The American
`Association for Vascular Surgery.
`0741 -5214/ 2002/ $35.00 + 0 24/1/123763
`doi:10. 1067 / mva.2002.123763
`1048
`
`fes tations. In any one specific report, selection of only one
`of these factors as the basis for classification may be appro(cid:173)
`priate.
`Anatomic classification. All reports shou ld classify
`aortic aneurys ms on the basis of site and extent of disease
`(Table I ). Respective definitions and recommended classi(cid:173)
`fication schemes for nonaortic arterial aneurysms have been
`described in detail elsewhere. 2
`Etiologic classification. It is recommended that re(cid:173)
`ports identify arterial aneurysm etio logy. Specifically,
`distinction sh ould be made between degenerative ( ar(cid:173)
`teriosclerotic ), anastomotic ,
`infectious,
`inflammatory
`(noninfectious), traumatic, and congenital aneu rysms. Al(cid:173)
`thou gh dissection may be associated with any of a number
`of underlying causes, such as Marfan's syndrome or hyper(cid:173)
`tension, it represents a distinct pathologic entity that
`shou ld be noted accordingly in clinical reports.
`Clinical classification. Aneurysms shou ld be catego(cid:173)
`rized by clinical presentation as asymptomatic or symptom(cid:173)
`atic. Specific symptoms and their time course should be
`documented, including those related to compression or
`erosio n into neighboring strucn1 res, thrombosis, or embo(cid:173)
`lization. Rupnire shou ld distinguish between a free rup(cid:173)
`n1re, fistulization into an adjacent organ, and a contained
`rupture. In this regard, central hemodynamic status sho uld
`be reported, including blood pressure and response to
`initial resuscitation (stable, unstable, cardiac arrest ).
`
`OUTCOME CRITERIA AND DEFINITIONS
`
`The motivation for aneurysm treatment is elimination
`of the risk of rupture and death. Therefore, by definition,
`the primary oittcorne criteria for endovascu lar aneurysm
`repair include the prevention of: ( 1) aneurysm rupture; (2)
`death from aneurysm rupture; and (3 ) aneurysm-related
`death that may result from primary or seconda1y treatment.
`Nonetheless, the mere presence of a device does not nec(cid:173)
`essarily prohibit anem·ysm rupnire and death. Thus, surro(cid:173)
`gate markers that suggest a continuing or increasing risk of
`rupture, such as aneu1ysm enlargement or endoleak, al(cid:173)
`though designated as secondary outcome criteria, play a
`
`TMT 2118
`Medtronic v. TMT
`IPR2021-01532
`
`

`

`JO URNAL OF VASCULAR SURGERY
`Volume 35, Number 5
`
`Chaikof et al 1049
`
`Table I. Classification of ao rtic and iliac aneurysms by
`anatomic site
`
`Aorta
`Ascending thoracic
`Arch
`Descending thoracic
`Thoracoabdominal (types I to IV)
`Abdominal
`Suprarenal
`Splanchnic and renal arterial involvement
`Only renal artery involvement
`Juxatrenal or pararenal (no norm al aorta between upper
`extent of aneurysm and renal arteries )
`Infrarenal
`
`Iliac
`Common
`External
`Internal
`
`critical role in the overall assessment of the effecti ve ness o f
`endovascular treatment strategies. All told, significant value
`exists in repo rting an accepted and unifying measure of
`clinical success that combines the most significant of pri(cid:173)
`mary and secondary o utcome criteria, which reflect the
`goals of th is treatment method. The fo llowing section
`provides recommendations for the definition and reporting
`of clinical success and associated outcome criteria that
`impact on the clinical effectiveness of endovascu lar repair.
`
`Definition of success
`
`Defining the success of endovascu lar ane urysm repair
`remains dependent o n a consideration of both clinical and
`radiographic criteria withi n the context of a histo ric stan(cid:173)
`dard established by open surgical repair in which the aneu(cid:173)
`rysmal segment is treated wi th in situ prosthetic graft
`replacement. In this regard , a similar result can only be
`accomplished with an endograft, if complete exclusion of
`the aneurysm from the ci rculatory system is ac hieved. Some
`investigators, howeve r, have suggested that the presence of
`a persistent type II endoleak after endovascular repair may
`no t be an absolute predictor of late aneurysm expansion
`and rupture. 3 Moreover, it has also been argued that a
`measured reductio n in rupture risk may be possible even in
`the presence of a type II endoleak. In the absence o f
`definitive data, the relative predictive value for this and
`other outcome criteria with respect to the risk o f aneurysm
`rupture remains incompletely defined. Thus, future inves(cid:173)
`tigations may lead to further refinement of the recom(cid:173)
`mended definitions of success presented herei n .
`Definition of technical success. Technical mccess re(cid:173)
`lates to periprocedural events that occur from the initiation
`of the procedure and extend through the first 24-hour
`postoperative period . Primary technical success is defined on
`an intent-to-treat basis and requires the successful intro(cid:173)
`ductio n and deployment of the device in the absence of
`surgical conversion or mortality, type I or III endoleaks, or
`graft limb obstruction. A technical success thus implies the
`following quali fying details:
`
`1. Successfu l access to the arterial system using a remo te
`site (ie, the femoral , external iliac, common iliac, or
`brachiocephalic arteries with o r without use of a tempo(cid:173)
`rary o r permanent prosthetic conduit to access these
`arteries);
`2. successfu l deployment of the endoluminal graft with
`secure proximal and distal fixation;
`3. absence o f either a type I or III endoleak;
`4 . patent endo luminal graft witho ut significant twist,
`kinks, o r obstruction (> 30% luminal stenosis o r a pres(cid:173)
`sure gradient > 10 mm H g ) by intraoperative measure(cid:173)
`ments.
`
`Primary technical success can include the use of addi(cid:173)
`tional modular co mponents, stents, or angioplasty, and
`adjunctive surgical procedures. H owever, if un planned en(cid:173)
`dovascular o r surgical procedures are necessitated, the
`terms assisted primary or secondary technical sitccess, respec(cid:173)
`tively, should be used.
`Secondary endpoints should be reported , such as pro(cid:173)
`cedure time , blood loss, blood transfusion , flu oroscopy
`time, contrast load , recove1-y time, range and average num(cid:173)
`ber of days in an intensive care unit, and hospital length of
`stay. H owever, these parameters do not enter into the
`consideration of technical success rates.
`Definition of clinical success. Clinical success should
`be reported o n an intent-to-treat basis and requires success(cid:173)
`fu l deployment of the endovascular device at the intended
`location without death as a result of aneurysm-related
`treatment, type I or III endoleak, graft infection or throm(cid:173)
`bosis, aneurysm expansion (diameter ~ 5 111111 , o r vol ume
`~ 5%), aneu1-ysm rupture, or conversion to o pen repair.
`Mo reover, the presence of graft dilatatio n of 20% o r mo re
`by diameter, graft migration, or a failure of device integrity
`classifies a case as a clinical failure. Clinical success can be
`claimed for those cases with a type II endoleak only in the
`absence o f aneurys m expansion . As long as the significance
`of a type II endoleak and its implication as a marker for late
`clinical failure remains an area of active investigation, it is
`recommended that reports clearly indicate the proportion
`o f patients classified as clinical successes that harbor a type
`II end oleak.
`The presentation of clinically meaningful success rates
`mandates that the data are statistically valid for the time
`period in questio n. Specifically, the standard deviation of
`life table or Kaplan -Meier estimates should not exceed 10%.
`The fo llowing temporaI characterization of clinical success
`is o ffered with this consideratio n in mind. Initial or 30-day
`clinical success encompasses 30-day d ata. Short-term clinical
`success includes o utcome measures reported within a 30-day
`to 6-month time frame. Mid-term clinical mccess refers to
`all outcome measures that are statistically significant up to 5
`yea rs after endograft implantatio n. Long-term clinical suc(cid:173)
`cess includes all outcome measures that are statistically
`significant beyond 5 years.
`Primary clinical success is clinical success withou t the
`need for an additional or secondary surgical or endovascu(cid:173)
`lar procedure . Assisted primary clinical success is clinical
`
`

`

`1050 Chailwf et al
`
`JOURNAL OF VASCULAR SURGERY
`May 2002
`
`success achieved with the use of an additional or secondary
`endovascular procedure . Secondary clinical success is clinical
`success obtained with the use of an additional o r seconda1y
`surgical procedure ( eg, the performance of a femoral(cid:173)
`femoral bypass for treatment of a unilateral limb occlusion
`of a bifurcated endograft). Conversely, clinical failure in(cid:173)
`cludes a failure to deploy the endovascular device at the
`intended location, the presence of a type I or III endoleak,
`graft thrombosis or infection , aneurysm expansion ( diam(cid:173)
`eter 2':: 5 mm , or volume 2':: 5%), aneu1ysm rupture, co nver(cid:173)
`sion to open repair, or death as a result of aneurysm rupture
`or anemysm-related treatment. Moreover, the presence of
`graft dilatatio n of 20% or more by diameter, graft migra(cid:173)
`tion, or a failure of device integrity classifies a case as a
`clinicaI fai lure. Aneurysm rupture should be reported as
`either a procedure-related ane11rysm ntpture (ie, perforation
`of the aneutysm during the course of the implantation
`procedure) or as a late aneurysm rupture that follows device
`deployment.
`Longitudinal reporting of clinical outcome mea(cid:173)
`sures. The following parameters are recommended for
`inclusion in any comprehensive report of endovascu lar
`aneurysm repair: survival; rupture-free survival; prevalence
`o f aneurysm rupture, death from aneutysm rupture, and
`aneurysm-related death ; freedom from aneurysm expan(cid:173)
`sion; freed o m from type I and III endoleaks; prevalence o f
`type II endoleak; prevalence o f seconda1y endo leak; en(cid:173)
`dograft patency; and technical and clinical success rates.
`Life tables or Kaplan-Meier curves should be calcu lated
`for presentation of survival and rupture-free survival , main(cid:173)
`tenance of clinical success, risk of aneurysm-related death,
`as well as for freedom from aneurysm expansion and type I
`or III endoleaks. Endograft patency should be reported in
`life table o r Kaplan-Meier format as primary, assisted pri(cid:173)
`mary, or seconda1y depending o n the use of additionaI
`endovascular or surgical procedures.
`Changes in renal function may occur as a result of
`suprarenal fixation of an endograft. Therefore, for those
`patients in whom an endograft extends above the renal
`arte1y orifices, a separate analysis sho uld consider reporting
`parameters relevant to long-term renal function , including
`renal artery patency, incidence of segmental renal infarcts,
`estimated glomerular filtration rates, hypertension , and
`dialysis-free survival.
`Comparing the clinical success of endovascular and
`open surgical repair. Investigations that compare open
`surgical and endovascular repair should report primaty
`ou tcome criteria for both treatment groups, as previously
`defined. H owever, the use o f a related although distinct
`definition of clinical success is necessary for patients treated
`with open surgety . Primary technical mccess for open mr;gi(cid:173)
`cal repair should be reported on an intent-to-treat basis and
`should require replacement or bypass of the aneurysmal
`segment with a prosthetic graft in the absence of mortality
`or graft thrombosis either during surgery or during the
`initial 24-hour postoperative period. If an unplanned sur(cid:173)
`gical procedure is necessitated, such as a splenectomy or
`
`reexploration for bleeding, the term secondary technical
`mccess should be used.
`The definition o f clinical success for open surgical repair
`includes the absence of death as the result of aneurysm(cid:173)
`related treatment, graft infection or thrombosis, failure o f
`device integrity, including graft dilatation 20% or more by
`diameter, and paraanastomotic aneurysm
`forma tion.
`Should open repair consist of aneutysm exclusion and
`bypass grafting, aneurysm expansion ( diameter 2':: 5 mm , or
`volume 2':: 5%) o r rupture would classify a case as a clinical
`failure. Definitions o f initial (30-day), short-term, mid(cid:173)
`term, and long -term clinical mccess and of primary, assisted
`primary, and secondary clinical sitccess otherwise remain
`unchanged , as do recommendations for longitudinal re(cid:173)
`porting o f clinical data. Other significant o utcome vari(cid:173)
`ables, such as device integrity, qua.lity of life, and cost
`effectiveness can be compared with guidelines outlined
`below. Likewise, grading schemes for reporting complica(cid:173)
`tions and their severity, although primarily focused in this
`report on endovascular treatment, can be adapted with
`little modification for open repair. Finally, in comparison of
`t\¥0 or more patient populations treated with open surgety
`and endovascular approaches, adjusting for case severity
`mix , particularl y with respect to comorbid medical condi(cid:173)
`tions, can be performed with schemes described else(cid:173)
`where.4
`
`Aortic and iliac artery remodeling
`
`Because anemysm size is the predominant factor deter(cid:173)
`mining risk of aneutysm rupture,5 changes in ane urysm
`dimension have been used as a surrogate marker for clinical
`efficacy after endovascular repair. Other mo rphologic
`changes, including progressive iliac angulation and aortic
`neck enlargement, may occur in response to either aneu(cid:173)
`rysm exclusio n or associated degenerative changes in adja(cid:173)
`cent segments, respectively. These late changes in aortoiliac
`anatomy have no t been used in the definition of clinical
`success after endovascu lar aneurysm repair.
`onetheless,
`they may have a significant impact on late device perfor(cid:173)
`mance and treatment durability. Thu s, characterizing mor(cid:173)
`p ho logic respo nses to endovascular intervention is useful
`both in defining clinical outcome and in serving to suggest
`mechanisms and engineering remedies in the event of a
`device failure.
`Changes in dimension of the aortic aneurysm. In
`endovascular repair, the aneurysm sac is left intact and, as a
`consequence, this feature plays an important role in out(cid:173)
`co me assessment. Principally, changes in tl1e dimension of
`tl1e residual sac assist in defining the success or failure o f
`anemysm exclusion. Specifically, clinical correlation sug(cid:173)
`gests that aneurysm growth after endovascular repair is an
`indicator of inco mplete aneurysm exclusion, co ntinued risk
`of aneurysm rupture, and a presumed treatment failure. 6 In
`addition, seco ndaty changes in the configuration of the
`aortoiliac segment, in response to a reductio n in aneurysm
`size, may jeopardize the integrity or function of the endo(cid:173)
`vascular graft.
`
`

`

`JO URNAL OF VASCULAR SURGERY
`Volume 35, Number 5
`
`Chaikofetal 1051
`
`Because variations in size occur in three dimensio ns,
`bo th sac volume and diameter are relevant parameters for
`defining changes in aneurysm size. It is no tewo rthy, how(cid:173)
`ever, that relatively small diameter shifts of 1 to 2 mm, that
`may otherwise be difficult to accurately measure with con(cid:173)
`ventional imaging techniq ues, may be correlated with a
`significant change in aneurysm volume. 7 -9
`Methodology for measuring changes in aneurysm di(cid:173)
`ameter, volume, and length. Aneurys m size should be
`expressed as either maximum diameter o r volume. M odal(cid:173)
`ity, method , and definitions sho uld be clearly described ,
`and comparisons sho uld o nl y be made bet\veen identical
`sources. Preferably, maximum aneurysm diameter should
`be measured perpendicular to the flow line of the vessel
`with three-dimensional reconstructed computed to mo(cid:173)
`graphic (CT) scan images. Because the aneurysm cross
`sectio n often appears elliptic o n axial images, the mino r axis
`of the ellipse (smaller diameter) is generally a closer approx(cid:173)
`imatio n of true maximum aneurysm diameter. 10 The in(cid:173)
`traobserver and interobserver variability of diameter mea(cid:173)
`surements o btained from CT scan images range beween 2
`and 5 mm or 5% and 15%. 11 •12 Therefore, a diameter change
`of 5 mm or more is considered significant. Blinding of
`observe rs and data sets are reco mmended to minimize bias.
`Total aneurysm volume is defined as the volume within
`the native aortic wall, which extends from the level of the
`most caudal renal artery to a reprod ucible distal landmark,
`such as the aortic or iliac bifurcation. Lumenal aneurysm
`volitme is defined as the volume circumscribed by the
`endograft, while nonlumenal aneurysm volume is com(cid:173)
`prised o f thrombu s and, if present, endoleak. End ograft
`dilatation may be associated with an increase in lumenal
`volume , while reduction in aneurysm size is principally
`related to a decrease in nonlumenal volume . The term
`complete aneurysm resolzttion should be used if the nonlu(cid:173)
`menal aneurysm volume is less than 10% of the o riginal
`no nlumenal volume noted after endograft implantation.
`The intrao bserver and interobserver variability for volume
`measurements have ranged bet\Veen 3% and 5%. Therefore,
`a volume change of 5% or more is considered significant. 8
`A red uction in aortic length has been noted in associa(cid:173)
`tion with a reduction in aneu rysm size and may be associ(cid:173)
`ated with predisposition for endograft buckling, kinking,
`and component dislocation. 13 Aortic length is recom(cid:173)
`mended to be measured from the most caudal renal artery
`to a reproducible distal endpoint, such as the aortic bifur(cid:173)
`cation, along the flow line axis, as depicted in three-dimen(cid:173)
`sional reconstructed spiral CT scan images. 14 • 15 Of note, in
`the absence of three-dimenstional reconstruction , aneu (cid:173)
`rysm length is underestimated as determined from axial CT
`scan images. 16 - 18 Moreover, limitations also exist for angiog(cid:173)
`raphy performed with a calibrated marker catheter. 14• 19
`Changes in dimension of the aortic neck. Lo ng(cid:173)
`term aneurysm exclusion and device stabilization is depen (cid:173)
`dent o n the maintenance of an effective attachment, con(cid:173)
`nection , o r seal bet\veen the endograft and the host aorta.
`Therefore, dilatation of the aorta at the site or sites in(cid:173)
`tend ed for primary endograft fixation may lead to treat-
`
`ment fa ilure either with device migratio n or via the occur(cid:173)
`rence of a new endoleak with aneurysm expansion.20-23
`Methodology for measuring changes in neck dimen(cid:173)
`sion. Bo th diameter and cross-sectio nal area of the aortic
`neck at the sites intended for sea ling o r graft attachment are
`reportable parameters, and measurement perpendicular to
`the flow line is recommended. If the aortic segment does
`not run perpendicular to the plane of measurement, the
`smallest diameter (minor axis o f the elliptical cross sectio n )
`is an appropriate approximate of the true neck diameter. 20
`Notably, the outer perimeter of the aortic neck wall should be
`used as the reference point for all measurements. Modality,
`method, and definitions should be clearly described , and
`comparisons should only be made bet\veen identical sources.
`Changes in dimension of the iliac arteries. Progres(cid:173)
`sive angulation of the aortoiliac segment leading to d istor(cid:173)
`tions o f the endograft can accompany reductio ns in aneu(cid:173)
`rysm size. T his may result in endograft disruption or limb
`occlusion. Likewise, progressive dilatation of an iliac artery
`may also contribute to device instability o r loss of effective
`anem-ysm exclusion. Accordingly, alteratio ns in iliac arte1-y
`angulation and size after endovascular grafting and the
`conseq uent responses of the prosthesis may be important
`determinants of o utcome.
`Methodology for measiwing changes in iliac artery
`diameter and tortuosity. T hree- dimensional image analy(cid:173)
`sis with appropriate anatomic referencing is recommended
`for accurate determinatio n of iliac angulatio n and tortu os(cid:173)
`ity. A second option is the presentation of measures derived
`from o blique, anteroposterior, and lateral plain abdominal
`x-rays that provide a simpler, albeit more approximate,
`approach. Admittedly, if the latter approac h is used , varia(cid:173)
`tions in position, angulatio n, and equipment may have a
`significant impact on the reproducibility of derived mea(cid:173)
`su res. Definitions and categorizatio n of iliac tortuosity and
`angulation have been detailed elsewhere_ 4
`Methodology for reporting sequential changes in
`aortoiliac morphology. After placement of an endovas(cid:173)
`cular prosthesis, the local biomechanical and hemod ynamic
`environments of the native aorta are altered. As a conse(cid:173)
`quence, dynamic changes in aortoiliac mo rpho logy are
`observed over time. T he rate , magnitude, and directio n of
`this response will be influenced by properties inherent to
`the chosen device and host aorta. Defining, with assured(cid:173)
`ness, the o nset and durability of an unqu alified treatment
`success wi ll require capturing the time course and the
`direction of these associated mo rpho logic changes.
`R eporting ntorphologic changes that occur during a
`study period. C hanges in anem-ysm dimension shou ld be
`repo rted in terms o f prevalence (%) with the number of
`available subjects at each time period clearly stated. In this
`regard , the investigator should report the following o ut(cid:173)
`come measures: ( 1) reduction in abdominal aortic aneu(cid:173)
`rysm (AAA) diameter or volume by 5 mm or mo re o r 5% or
`more, respectively; (2 ) enlargement in AAA diameter or
`volume by 5 mm or more or 5% or mo re, respectively; and
`(3) absence of significant change either in AAA diameter or
`volume. In characterizing the absence of aneurysm expan-
`
`

`

`1052 Chailwf et al
`
`JO URNAL OF VASCULAR SURGERY
`May 2002
`
`Table II. C lassification of endoleak
`
`Type
`
`Canse of perigraft floiv
`
`II
`
`III
`
`a) Inadequate seal at proximal end of endograft
`b ) Inadequate seal at distal end of endograft
`c) Inadequate seal at iliac occluder plug
`Flow from visceral vessel (lumbar, IMA, accessory
`renal, hypogastric ) without attachment site
`connection
`a) Flow from module disconnection
`b ) Flow from fabric disruption
`Minor ( < 2 mm )
`Major (;:;, 2 mm )
`Flow from porous fabric ( < 30 days after graft
`placement)
`Endoleak of Flow visuali zed but source unidentified
`undefined
`origin
`
`rv
`
`sion at any time during a study period, life table or Kaplan(cid:173)
`Meier analysis is an approp riate form for data presentation,
`as in freedom from aneurysm expansion.
`Long-term sn1dies are anticipated to provide data for
`two additional outcome measures ofincreasing interest: ( 1)
`complete aneu1ysm resolution; and (2) the rate of change
`of aneu1y sm diameter or volume, with a distinction be(cid:173)
`tween population subgroups that show either an enlarge(cid:173)
`ment or reduction in aneurysm size. Measurable changes in
`aneu1ysm dimensions have been reported immediately after
`endograft deployment . Therefore, changes in aneu rysm
`size should be referenced to those measurements obtained
`from the first set of postoperative images. Likewise, similar
`methodologic approaches can be adopted for presenting
`the changes of other morphologic features, such as aortic
`neck diameter, aortic length, or iliac angu lation .
`Endoleak and endotension
`
`Endoleaks, including their detection , potential clinical
`significance, and treatment, remain an active area of inves(cid:173)
`tigation. However, although it is now evident that an
`endoleak may resolve spontaneously, a proportion of those
`that do persist have been associated wi th late aneu1ysm
`rupture. 24-27 Thus, endo leak remains an important mea(cid:173)
`sure of clinical outcome.
`Definition of endoleak. Endoleak is defined by the
`persistence of blood flow outside the lumen of the endolu (cid:173)
`minal graft but within the aneu1ysm sac, as determined by
`an imaging study.28·29 An endoleak is evidence of incom(cid:173)
`plete exclusion of the ane u1ysm from the circu lation and
`may be the result of an incomplete seal between the en(cid:173)
`dograft and the wall of the blood vessel, an inadequate
`connection between components of a modular prosthesis,
`fabric defects or porosity, or retrograde blood flow from
`patent aortic side branches.29•30 Although intrasac pressure
`may approach systemic arterial pressure in the presence of
`an endoleak, some type II endoleaks have been associated
`with shrinking aneu1ysms and intrasac pressures that are
`substantially less than systemic values.31
`Classification of endoleak. An endoleak can be clas(cid:173)
`sified according to time of occurrence relative to the aper-
`
`ative procedure and site of origin. 32 An endoleak first
`observed during the perioperative ( :::=::30 days ) period is
`defined as a prirnary endoleak, and initial detection there(cid:173)
`after is termed a secondary endoleak.29 The reappearance of
`an end oleak either after spontaneous resolution or after an
`intervention that was considered successful is defined as a
`recttrrent endoleali. Further categorization requires precise
`information regarding the course of blood flow into the
`aneu1ysm sac (Table II ). A Type I endoleali is indicative o f a
`persistent perigraft channel of blood flow caused by inade(cid:173)
`quate or ineffective seal at either the proximal or distal graft
`ends or attac hment zones. Use of subscripts a orb indicate
`a type I endoleak originating at the proximal or distal ends
`of the endograft, respectively. In the case of an aortoiliac
`prosthesis, a type I endoleak may also refer to blood flow
`around an iliac occluder plug and should be indicated with
`use of subscript c. A type II endoleali is attributed to retro(cid:173)
`grade flow from lumbar arte ries, the inferior mesenteric
`artery (IMA), or other collateral vessels. Origin and ou tflow
`sources of a type II endoleak should be specified , such as
`lumbar-lumbar,
`lumbar-IMA, accesory
`renal-lumbar/
`IMA, hypogastric- lumbar/ lMA, or undefined. It should
`be emphasized that any connection of flow to proximal or
`distal graft ends or attachment zones, even in the presence
`ofretrograde flow from a lumbar or IMA vessel, classifies an
`endo leak as type I. An endoleak caused by fabric tears o r
`disruption , component disconnection, o r graft disintegra(cid:173)
`tion is classified as a type III endoleali. Distinction between
`modular disconnection and fa bric tear can be indicated
`through the use of subscripts a and b, respectively. A type
`III b endoleak can be further stratified with respect to the
`extent of fabric disruption as inajor (2'.:2 mm ) or rninor ( < 2
`mm ). An example of the latter is late "microleaks," which
`can develop at the site of suture attachment of a fabric to
`supporting elements. Blood flow through an intact but
`o therwise porous fabric, observed during the first 30 days
`after graft implantation , is termed a type I V endoleak. 32 This
`designation is not applicable to fab ric-related endoleaks
`observed after the first 30-day period. If an endoleak is
`visualized on imaging studies but the precise source cannot
`be determined , the endoleak is categorized as an endoleak of
`undefined origin.
`Endotension. It is now appreciated that an AAA may
`continue to enlarge after endovascular repair, even in the
`absence of a detectable endoleak, and that this enlargement
`may lead to aneurysm rupture. 31 ·33-36 It has been proposed
`that this condition be classified as endoleak-related. 32 Ex(cid:173)
`planations for persistent or recurrent pressurization of an
`anemysm sac include blood flow that is below the sensitiv(cid:173)
`ity limits fo r detection with current imaging technolo(cid:173)
`gy32•34•37 or pressure transmission thro ugh thrombus38,39
`or endograft fabric. On physical examination , the aneurysm
`may be pulsatile and intrasac measurements may reveal
`pressures that approach or are equal to systemic values .
`At this time, endotension remains an investigative term
`for which some disagreement persists regarding its precise
`definition, appropriate usage, and current applicability. En(cid:173)
`dotension has been va riabl y used in the literature to refer to
`
`

`

`JOURNAL OF VASCULAR SURGERY
`Volume 35, Number 5
`
`Chaikof et al 1053
`
`aneurysm enlargement in the presence or absence of an
`endoleak and to persistent or recurrent pressurization of an
`aneurysm sac, as determined with direct intrasac measure(cid:173)
`ment. 36 It is anticipated that future research may lead to the
`development of a surrogate marker for sac pressurization ,
`allowing more precise identification and definition of this
`phenomenon. H owever, in the interim, the majority view
`of this committee is that the term endotension be confined
`to instances of aneurysm enlargement after endovascular
`repair in the absence of a detectable endoleak. As such,
`endotension should not be classified as an endoleak of
`undefined origin .
`Methodology for measuring the presence, source,
`magnitude, and physiologic significance of endoleak
`and endotension. Although conflicting studies exists,
`most recent investigations suggest that the sensitivity of
`contrast-enhanced CT scan imaging is superior to that of
`other noninvasive techniques, such as duplex ultrasonogra(cid:173)
`phy, for endoleak detection.40 Nonetheless, CT scan imag(cid:173)
`ing may fa il to identify an endoleak if delayed images are not
`obtained after infusion of contrast medium.41 Further(cid