Prediction of Progression of Coronary Artery Disease and Clinical Outcomes Using Vascular Profiling...

U Coskun Michail I Papafaklis and Charles L FeldmanSatoru Otsuji Fuminobu Yoshimachi Junko Honye Dawn Harwood Martha Reitman Ahmet

Atsushi Hirohata Toshiyuki Matsumura Seiji Yamazaki Hiroyoshi Yokoi Shinji TanakaTomohiro Kawasaki Akihiko Takahashi Takaaki Katsuki Sunao Nakamura Atsuo Namiki

Peter H Stone Shigeru Saito Saeko Takahashi Yasuhiro Makita Shigeru NakamuraPREDICTION Study

Vascular Profiling of Endothelial Shear Stress and Arterial Plaque Characteristics The Prediction of Progression of Coronary Artery Disease and Clinical Outcomes Using

Print ISSN 0009-7322 Online ISSN 1524-4539 Copyright copy 2012 American Heart Association Inc All rights reserved

is published by the American Heart Association 7272 Greenville Avenue Dallas TX 75231Circulation doi 101161CIRCULATIONAHA1120964382012126172-181 originally published online June 21 2012Circulation

httpcircahajournalsorgcontent1262172World Wide Web at

The online version of this article along with updated information and services is located on the

httpcircahajournalsorgcontentsuppl20120620CIRCULATIONAHA112096438DC1htmlData Supplement (unedited) at

httpcircahajournalsorgsubscriptions

is online at Circulation Information about subscribing to Subscriptions

httpwwwlwwcomreprints Information about reprints can be found online at Reprints

document Permissions and Rights Question and Answer this process is available in the

click Request Permissions in the middle column of the Web page under Services Further information aboutOffice Once the online version of the published article for which permission is being requested is located

can be obtained via RightsLink a service of the Copyright Clearance Center not the EditorialCirculationin Requests for permissions to reproduce figures tables or portions of articles originally publishedPermissions

by guest on July 9 2012httpcircahajournalsorgDownloaded from

Coronary Heart Disease

Prediction of Progression of Coronary Artery Disease andClinical Outcomes Using Vascular Profiling of Endothelial

Shear Stress and Arterial Plaque CharacteristicsThe PREDICTION Study

Peter H Stone MD Shigeru Saito MD Saeko Takahashi MD Yasuhiro Makita MD Shigeru Nakamura MDTomohiro Kawasaki MD Akihiko Takahashi MD Takaaki Katsuki MD Sunao Nakamura MD

Atsuo Namiki MD Atsushi Hirohata MD Toshiyuki Matsumura MD Seiji Yamazaki MDHiroyoshi Yokoi MD Shinji Tanaka MD Satoru Otsuji MD Fuminobu Yoshimachi MD Junko Honye MD

Dawn Harwood PhD Martha Reitman MD Ahmet U Coskun PhD Michail I Papafaklis MD PhDCharles L Feldman ScD for the PREDICTION Investigators

BackgroundmdashAtherosclerotic plaques progress in a highly individual manner The purposes of the Prediction ofProgression of Coronary Artery Disease and Clinical Outcome Using Vascular Profiling of Shear Stress and WallMorphology (PREDICTION) Study were to determine the role of local hemodynamic and vascular characteristics incoronary plaque progression and to relate plaque changes to clinical events

Methods and ResultsmdashVascular profiling using coronary angiography and intravascular ultrasound was used to reconstructeach artery and calculate endothelial shear stress and plaqueremodeling characteristics in vivo Three-vessel vascular profiling(27 arteries per patient) was performed at baseline in 506 patients with an acute coronary syndrome treated with apercutaneous coronary intervention and in a subset of 374 (74) consecutive patients 6 to 10 months later to assess plaquenatural history Each reconstructed artery was divided into sequential 3-mm segments for serial analysis One-year clinicalfollow-up was completed in 992 Symptomatic clinical events were infrequent only 1 (02) cardiac death 4 (08)patients with new acute coronary syndrome in nonstented segments and 15 (30) patients hospitalized for stable anginaIncrease in plaque area (primary end point) was predicted by baseline large plaque burden decrease in lumen area (secondaryend point) was independently predicted by baseline large plaque burden and low endothelial shear stress Large plaque sizeand low endothelial shear stress independently predicted the exploratory end points of increased plaque burden and worseningof clinically relevant luminal obstructions treated with a percutaneous coronary intervention at follow-up The combinationof independent baseline predictors had a 41 positive and 92 negative predictive value to predict progression of anobstruction treated with a percutaneous coronary intervention

ConclusionsmdashLarge plaque burden and low local endothelial shear stress provide independent and additive prediction toidentify plaques that develop progressive enlargement and lumen narrowing

Clinical Trial RegistrationmdashURL httpwwwclinicaltrialsgov Unique Identifier NCT01316159 (Circulation 2012126172-181)

Key Words atherosclerosis endothelium natural history shear stress

Atherosclerosis is a systemic disease with focal and eccentricmanifestations1 In a patient with coronary artery disease

(CAD) and systemic risk factors each coronary lesion pro-gresses regresses or remains quiescent in an independentmanner2 indicating that local vascular factors must be a majordeterminant responsible for the behavior of individual plaques

Editorial see p 161Clinical Perspective on p 181

The vascular endothelium is in a unique and pivotalposition to respond to the extremely dynamic forces acting onthe vessel wall because of the complex 3-dimensional (3D)

Received January 27 2012 accepted May 16 2012From the Brigham amp Womenrsquos Hospital Harvard Medical School Boston MA (PHS ST MR MIP CLF) Shonan Kamakura General Hospital Kamakura

(SS ST JH) Hakodate Municipal Hospital Hakodate (YM) Kyoto Katsura Hospital Kyoto (SN) Shin Koga Hospital Kurume (TK) Sakurakai TakahashiHospital Kobe (AT) Jichi Medical University Hospital Shimotsuke (TK) New Tokyo Hospital Matsudo (SN) Kanto Rosai Hospital Kawasaki (AN)Sakakibara Heart Institute of Okayama Okayama (AH) Kumamoto Rosai Hospital Yatsushiro (TM) Sapporo Higashi Tokushukai Hospital Sapporo (SY)Kokura Memorial Hospital Kitakyushu (HY) Shonan Atsugi Hospital Atsugi (ST) Higashi Takarazuka Satoh Hospital Takarazuka (SO) AomoriPrefectural Central Hospital Aomori Japan (FY) REGISTRAT-MAPI Lexington KY (DH) Northeastern University Boston MA (AUC)

Guest Editor for this article was Steven E Nissen MDThe online-only Data Supplement is available with this article at httpcircahajournalsorglookupsuppldoi101161CIRCULATIONAHA

112096438-DC1Correspondence to Peter H Stone MD Cardiovascular Division Brigham amp Womenrsquos Hospital 75 Francis St Boston MA 02115 E-mail

pstonepartnersorgcopy 2012 American Heart Association Inc

Circulation is available at httpcircahajournalsorg DOI 101161CIRCULATIONAHA112096438

172 by guest on July 9 2012httpcircahajournalsorgDownloaded from

geometry of the artery associated with its natural curvatureand the acquired presence of focal atherosclerotic obstruc-tions Mechanical forces in general and fluid shear stress inparticular elicit a large number of humoral metabolic andstructural responses in endothelial cells3ndash5 The response of anumber of genes sensitive to local low endothelial shearstress (ESS) leads to creation of a raised plaque subsequenthemodynamic forces created by the enlarging plaque maylead to a cycle of progressive atherogenesis367 which mayculminate in plaque rupture that may become clinicallymanifest as obstructive thrombus68 or rapid progression of afixed obstruction9

Identification of an early coronary atherosclerotic plaquelikely to acquire high-risk characteristics and precipitate a newcoronary event may allow for development of preemptivestrategies to avert adverse events The recent Providing RegionalObservations to Study Predictors of Events in the Coronary Tree(PROSPECT) Study10 demonstrated that coronary lesions re-sponsible for new cardiac events in patients following a percu-taneous coronary intervention (PCI) for an acute coronarysyndrome (ACS) were associated with large plaque burden asmall lumen area thin-cap fibroatheroma (TCFA) morphologyas assessed by intravascular ultrasound (IVUS) and minimalobstruction by angiography This natural history study waslimited however by acquisition of vascular characteristics atonly a single point in time and pathobiological mechanismspotentially responsible for ongoing and progressive plaqueinstability such as local ESS were not investigated

The purposes of the Prediction of Progression of CoronaryArtery Disease and Clinical Outcome Using Vascular Profilingof Shear Stress and Wall Morphology (PREDICTION) Studywere to identify the detailed coronary hemodynamic and plaquecharacteristics in high-risk patients following a PCI for an ACSto follow the serial natural history of the plaques over a 6- to10-month period in a large number of consecutive patients andto investigate whether high-risk lesions likely to rupture andcause an ACS or rapid progression of an obstruction could beidentified early in its natural history

MethodsStudy DesignThe PREDICTION Study was designed as both an anatomic naturalhistory study and an event-attribution study Patients with an ACSundergoing PCI for a culprit lesion were enrolled and underwentcoronary vascular profiling (VP) as described below at the time ofthe index catheterization procedure A large subset of consecutiveunselected patients underwent follow-up VP after 6 to 10 months toassess anatomic natural history and antecedent vascular characteris-tics responsible for the natural history All patients had clinicalfollow-up at 1 year The study was performed in Japanese clinicalsites because Japanese patients routinely undergo follow-up cathe-terization and IVUS 6 months after successful PCI for an ACS andthis clinical practice facilitated performance of a large natural historystudy

Inclusion criteria included age 18 years presentation with anACS CAD with at least 1 coronary segment requiring PCI accordingto usual clinical indications and at least 1 vessel suitable for IVUSnot planned for PCI Exclusion criteria included heart failure NewYork Heart Association class IIIIV unstable clinical status leftmain disease or 3-vessel disease significant coronary calcificationprecluding IVUS evaluation renal failure such that additional

contrast material would be contraindicated clinically significantvalvular disease and life expectancy 12 months

Study MethodsDetailed ESS and plaquearterial wall characteristics were identifiedwith use of VP methods combining intracoronary IVUS and biplanecoronary angiography to represent the artery accurately in 3D spaceand to measure flow as previously described1112 and presented indetail in the online-only Data Supplement In brief the 3D anatomyof the artery was reconstructed from digitized radiofrequency IVUSsignals and two planes of coronary angiography The arterial lumenand outer vessel wall were reconstructed from digitized and seg-mented end-diastolic IVUS frames A structured grid utilizing abody-fitted coordinate system was used to represent the lumenvolume Coronary blood flow was calculated directly from the timerequired for the previously calculated true 3D volume of bloodcontained within the arterial section to be displaced by radio-opaquematerial during a contrast injection The detailed intravascular flowcharacteristics were obtained by solving the transport equationsgoverning the conservation of mass and momentum (PhoenicsCHAM England)11 The ESS at the luminal surface of the artery wascalculated as the product of viscosity (calculated from the measuredhematocrit) and the gradient of blood velocity at the wall The 3Dgeometry of the outer vessel wall (area within the external elasticmembrane [EEM]) was recreated in a manner similar to thatdescribed for the lumen geometry The 3D geometry of the plaque(plaque plus media thickness) was taken as the difference betweenthe outer vessel wall and the lumen13 The processes of dataacquisition and analysis are highly reproducible12

Study AssessmentsWe divided the entire reconstructed artery into consecutive 3-mmsegments starting at the ostium We chose 3-mm segments becausethis length was methodologically reliable and would also accuratelyreflect the local hemodynamic and plaque characteristics and theheterogeneous and highly focal changes occurring within the plaqueover time as well Within each 3D segment we assessed local ESSplaque characteristics and vascular remodeling14 The same coro-nary artery segments were evaluated at the time of the follow-upcatheterization 6 to 10 months later to determine the change invascular characteristics Assurance that the arterial segments to bemeasured were identical at the initial and the follow-up procedureswas accomplished by using fixed anatomic landmarks primarilymultiple arterial branches as reference fiducial points

Coronary artery natural history outcomes were analyzed by utilizinga variety of different approaches to represent different magnitudes ofCAD progression (1) Each 3-mm segment was evaluated to determinethe effect of the baseline local ESS and plaque characteristics (ascontinuous variables) on the change in vascular outcome variables (ascontinuous variables) in the same segment (2) We analyzed the naturalhistory of obstructions at baseline because arterial areas with a lumennarrowing at baseline may be particularly likely to exhibit worseningobstruction in follow-up We analyzed those arterial areas of 5 sequen-tial 3-mm segments at baseline within which a discrete luminalnarrowing (ldquothroatrdquo) was present in the middle surrounded by two3-mm segments with progressive narrowing on either side of the throatEach two 3-mm segments proximal (upstream) to the throat had toexhibit progressive decrease in lumen area 01 mm2 in comparisonwith the preceding 3-mm segments and each 3-mm segment distal(downstream) to the throat had to exhibit progressive increase in lumenarea 01 mm2 in comparison with the preceding 3-mm segmentsThese baseline lumen narrowings were evaluated for change in obstruc-tion at follow-up and the baseline ESS was investigated to determine itsrole in the outcome (3) We identified those patients who underwent aPCI during the follow-up because of new clinical symptoms or substan-tial worsening of the luminal obstruction and we investigated theantecedent vascular characteristics responsible for the clinically relevantchange in lumen obstruction

The PREDICTION Study is an observational natural historystudy The primary end point when the study was designed in 2006was change in plaque area but we also identified a variety of

Stone et al Prediction of CAD Progression in Humans 173


secondary exploratory end points We include plaque burden as anend point to reflect a combined end point of both plaque and EEMareas

Statistical AnalysisCategorical variables are presented as counts and percentages contin-uous variables are summarized as meanSEM or median and interquar-tile range as appropriate Several statistical methods were used to correctfor systematic error (nonindependence of observations) introduced bythe clustering of multiple arterial 3-mm segments or lumen narrowingswithin patients First to investigate the association of continuousresponse variables (eg change in plaque burden) with categoricalvariables (eg baseline ESS category) mixed-effects ANOVA with thepatient and artery designated as random effects was used Probabilityvalues were adjusted for multiple comparisons with the use of theScheffe method Second to investigate the relationship between contin-uous response variables and continuous predictors (eg ESS magnitude)linear mixed modeling was used Third to investigate the association ofbinary anatomic outcomes (eg progressive decrease in lumen areaworsening luminal narrowing treated with PCI) with baseline variablesmixed-effects logistic regression was implemented Baseline variablesassociated with anatomic outcomes on univariable analysis at P level01 were considered for entry in the respective multivariable modelsand final selection of independent predictors was performed with abackward-stepping algorithm (criterion for retention P01) In logisticregression modeling cut points for dichotomizing continuous anatomicvariables were selected by receiver operator characteristic analysiswhich best predicted each outcome (receiver operator characteristiccriterion max[sensitivityspecificity]) All statistical tests were2-tailed and an -level of 005 was used to determine statisticalsignificance All analyses were performed with Stata 100 (StataCorpLP College Station TX) and SPSS 170 (SPSS Inc Chicago IL)

The study was approved by the institutional review board at eachhospital and each patient gave written informed consent

ResultsEnrollment was initiated on April 7 2007 and the last subjectfollow-up visit was on October 18 2010 Five hundred sixsubjects were enrolled at 17 clinical sites in Japan The indexACS event included an ST-elevation myocardial infarction in291 patients (575) nonndashST-elevation myocardial infarc-tion in 60 patients (119) and unstable angina in 155patients (306) One-year follow-up was completed in 502(992) patients Two patients (04) were lost to follow-upand 2 patients (04) withdrew consent to participate

Baseline VP data were obtained in all enrolled patients andwere analyzable for 496 (980) patients a total of 1341arteries were analyzable (27 arteriespatient) Follow-up VPdata for serial anatomic natural history analyses were ob-tained in 374 (74) consecutive patients Baseline andfollow-up VP data were analyzable for 329 (880) of the374 patients including VP data from 824 pairs of arteries

The reconstructed coronary arteries (mean length per artery476044 mm) were divided into consecutive 3-mm-lengthsegments (19 875 3-mm segments 13 788 3-mm segments innative areas) Each segment was characterized by local predom-inant ESS value (defined as the minimum averaged ESS valueover a 90 deg arc in each 3-mm segment) plaque and lumen areaplaque burden (plaque areaEEM area) and remodeling patternFor analysis of serial anatomic changes each arterial segment atbaseline was compared with the identical segment at follow-up(8137 3-mm segments in native areas with available baselineand follow-up VP data)

This report focuses on the natural history and outcomes ofthe native (nonstented) coronary segments

Patient DemographicsThe patient characteristics are presented in Table 1 A majorityof patients had substantial coronary risk factors but only 59(117) had a history of previous CAD Blood pressure atbaseline was well controlled and fasting lipids were modestlyelevated Cardiac risk assessed by C-reactive protein was lowAt hospital discharge most patients were on dual-antiplatelettherapy and routine vasculoprotective therapies (Table 2)

Safety of VP ProceduresEight hundred ninety-five VP procedures were performedThere were 5 complications (06) coronary artery dissec-tion in 3 and transient neurological deficit in 2 patients Therewere no long-term consequences from these complicationsand hospital stay was not prolonged

Table 1 Baseline Demographics and Clinical Characteristics

CharacteristicAll Patients

(n506)

Sex

Male n () 404 (798)

Female n () 102 (202)

Age y 65 (58ndash72)

Risk factors

Hypertension n () 321 (634)

Dyslipidemia

LDL 100 mgdL n () 374 (739)

HDL 40 mgdL n () 196 (387)

Cigarette Smoking (within past 2 y) n () 249 (492)

Diabetes mellitus n () 179 (354)

Insulin dependent n () 15 (30)

Family history of premature CAD n () 32 (63)

Previous CAD history n () 59 (117)

Stable angina n () 27 (53)

Myocardial infarction n () 35 (69)

PCI n () 48 (95)

CABG n () 0 (00)

Physical examination

Heart rate beatsmin 70 (63ndash77)

Systolic blood pressure mm Hg 126 (111ndash140)

Diastolic blood pressure mm Hg 72 (64ndash80)

Laboratory data

Hematocrit 39 (35ndash42)

LDL mgdL 109 (88ndash135)

HDL mgdL 43 (35ndash52)

Total cholesterol mgdL 180 (152ndash215)

Triglycerides mgdL 111 (75ndash155)

C-reactive protein mgdL 04 (01ndash11)

Continuous variable data are presented as median (interquartile range) LDLindicates low-density lipoprotein HDL high-density lipoprotein CAD coronaryartery disease PCI percutaneous coronary intervention and CABG coronaryartery bypass graft

174 Circulation July 10 2012


Analysis of Anatomic Natural History Outcomes

Effect of Baseline Vascular Characteristics on VascularOutcomes at Follow-Up in Each 3-mm Segment

Primary End Point Change in Plaque AreaPlaque area increase at follow-up after adjustment for thebaseline plaque area was independently associated withbaseline large plaque burden lumen area and excessiveexpansive remodeling (Table 3 and online-only Data Supple-ment Table I) Using baseline ESS as a categorical variablebased on terciles low ESS was not associated with a changein plaque area (Figure 1A)

Secondary End Point Change in Lumen AreaLumen area decrease after adjustment for baseline lumenarea was independently associated with baseline large plaque

burden low ESS excessive expansive remodeling and distallocation of the segment (Table 3) Baseline low ESS as acategorical variable was associated with a significant de-crease in lumen area (Figure 1B)

Exploratory End Point Change in Plaque BurdenPlaque burden increase after adjustment for baseline plaqueburden was independently associated with baseline low ESSexcessive expansive remodeling large lumen area and moreproximal location of the segment (Table 3) With use of ESSas a categorical variable baseline low ESS was associatedwith a significant increase in plaque burden (Figure 1C)

Natural History of Baseline Discrete Luminal NarrowingsBased on Local Vascular CharacteristicsThe majority of lumen obstructions (53) at follow-uporiginated from areas with preexisting mild lumen narrowingat baseline In these baseline narrowings ESS was primarilymoderate proximal to the throat high just proximal to and atthe throat and low distal to the throat (Figure 2)

Low ESS distal to the throat was significantly associatedwith different magnitudes of worsening of luminal area obstruc-tion at follow-up (Figure 3)

Natural History of Clinically Relevant LuminalObstructions Treated With a PCI at Follow-UpPCI was performed in follow-up in 59 native lesions in 53patients (105) either for development of a new clinicalevent or because of identification of a significantly worseningobstruction in a patient who underwent routine follow-upcoronary angiography (Tables 4 to 6 online-only Data

Table 2 Medications at Hospital Discharge

Medical Therapy All Patients (n506)

Statin n () 355 (702)

Other lipid-lowering medications n () 30 (59)

Acetylsalicylic acid n () 451 (891)

-blocker n () 169 (334)

Calcium channel blocker n () 130 (257)

Long-acting nitrate n () 98 (194)

ACE inhibitorangiotensin receptor blocker n () 300 (593)

Ticlopidine n () 146 (289)

Clopidogrel n () 347 (686)

ACE indicates angiotensin-converting enzyme

Table 3 Independent Baseline Predictors of the Change (Follow-Up MeasurementBaseline Measurement) inAnatomic Outcomes in 3-mm Segments (n8137)

Outcome Baseline Independent Predictor (95 CI) P

Change in plaquearea (mm2)

ESS (per 1-Pa decrease) 002 (006 to 002) 032

Lumen area (per 1-mm2 increase) 006 (004 to 008) 0001

Plaque area (per 1-mm2 increase) 016 (018 to 013) 0001

Plaque burden (per 10 increase) 011 (004 to 019) 0003

Remodeling pattern excessive expansive vscompensatoryconstrictive

009 (0001 to 0176) 0047

Change in lumenarea (mm2)




Longitudinal arterial location (per 10-mm increase) 015 (020 to 010) 0001


025 (039 to 012) 0001

Change in plaqueburden ()






074 (032 to 115) 0001

Variables entered in the multivariable model (A) change in plaque area ESS lumen area plaque area plaque burden andremodeling pattern (B) change in lumen area ESS lumen area plaque burden longitudinal arterial location and remodeling pattern(C) change in plaque burden ESS lumen area plaque area plaque burden longitudinal arterial location and remodeling pattern ESSindicates endothelial shear stress

The results for change in plaque area are presented without removing the primary variable of interest (ie ESS) which was notsignificant from the backward stepwise process of the multivariable analysis



Supplement Table II and Figures 4 and 5) New clinicalevents consisted of a new ACS in 13 patients (26) whichoccurred in a native portion of the coronary artery in only 4patients (08) and worsening stable angina in 15 patients(3) which occurred in a native segment in 10 patients (2)Thirty-nine patients (78) had a PCI performed in the absence

of symptoms because a worsening native luminal obstructionwas observed at the time of the routine follow-up cardiaccatheterization (luminal diameter stenosis measured by quan-titative coronary angiography changed from 42416 atbaseline to 58615 at follow-up P0001)

Independent vascular predictors of baseline substantialluminal narrowings (ie minimal lumen area 6 mm2 whichcorresponded approximately to the lower tercile of lumenarea) treated with PCI for lesion progression or occurrence ofsymptoms during follow-up were large plaque burden andlow ESS (Tables 5 and 6 Figure 4) There was no significantrelationship between plaque burden at the throat and ESSdistal to the throat

The positive predictive value of baseline vascular character-istics to identify a lesion that progressed clinically at follow-uprose from 22 if only large plaque burden was present to 41if both large plaque burden and low ESS were present (Figure 5)Negative predictive value remained high if both large plaqueburden and low ESS were absent One hundred ten patients(223) had at least 1 substantial luminal narrowing with largeplaque burden at baseline 50 patients (100) had a substantialluminal narrowing with low ESS at baseline and 31 patients(62) had a substantial luminal narrowing with both largeplaque burden and low ESS

There was no significant meaningful association of clini-cal characteristics at baseline and subsequent vascular change(online-only Data Supplement Tables III and IV)

Analysis of Clinical OutcomesClinical events were infrequent during the 1-year followup(Table 7) There were a total of 7 deaths (14) and only 1(02) cardiac death An ACS occurred in 13 patients (264 patients [08] with the culprit segment in a native vesselarea) Hospitalization for worsening stable angina occurred in15 patients (30 7 patients [08] with the culprit segmentin a native vessel area 3 patients [06] with both a nativeand a stented segment treated with revascularization)

There were too few patients with symptomatic clinical eventsto make meaningful statistical inferences between baselinevascular characteristics and symptomatic clinical outcomeevents

DiscussionThe PREDICTION Study is unique in that it presents the largestand most comprehensive serial anatomic natural history study ofcoronary atherosclerosis ever performed and used innovativemethodologies to investigate potential pathophysiological mech-anisms responsible for CAD progression The goal of thesenatural history investigations in consecutive patients with high-risk CAD was to identify the early plaque and arterial wallcharacteristics that precede the subsequent progression of lesionsleading to acute plaque rupture or accelerated luminal obstruc-tion Our multi-tiered natural history analyses indicate thatplaque burden is the most powerful predictor of plaque progres-sion and luminal obstruction and that low ESS provides substan-tial additive independent prognostication The combination ofthe independent baseline predictors of plaque burden and lowESS had a 41 positive predictive value to predict clinicallyrelevant obstruction progression treated with PCI but the com-

Figure 1 Effect of baseline ESS on vascular outcomes at fol-lowup in coronary artery segments A Change in plaque areaB Change in lumen area C Change in plaque burden Cutpoints for the 3 ESS categories were derived from the terciles ofthe frequency distribution in 3-mm segments Probability valuesrefer to the univariable analysis and are corrected for the clus-tering of arteries and segments within patients and for multiplecomparisons Error bars represent SEM ESS indicates endothe-lial shear stress



bination of these vascular characteristics at baseline was infre-quent (6) Symptomatic clinical events were uncommon inthese patients whose disease was well controlled following anACS and lack of statistical power precluded analyses to inves-tigate the role of vascular characteristics at baseline to predictnew symptomatic clinical events

The changes in coronary plaque and arterial remodelingobserved in the PREDICTION Study and the prominent roleof local ESS in the atherogenic processes associated withthose changes in coronary anatomy are remarkably similar tothe observations in the atherosclerotic pig model2715 and

small pilot studies in humans1617 In the diabetic hypercho-lesterolemic pig low ESS was an independent predictor ofthe development of high-risk TCFAs and the magnitude ofthe atherogenic phenotype was inversely related to the mag-nitude of local ESS71518 The magnitude of progression ofcoronary lesions we observed in humans in the PREDICTIONStudy was not as marked as the progression observed in thepig model but the pigs were rendered diabetic and intenselyhypercholesterolemic (mean serum cholesterol 61128 mgdL)and the duration of follow-up was relatively prolonged715 Thecentral proatherogenic role of low ESS in humans is underscored

Figure 2 Baseline ESS patterns along the courseof a coronary artery obstruction The 3 ESS cate-gories (low 1 Pa moderate 1ndash17 Pa high17 Pa) in the bar graph were derived from theterciles of the ESS frequency distribution in 3-mmsegments NC indicates necrotic core ESS endo-thelial shear stress

Figure 3 Effect of baseline ESS on magnitude of worsening severity of luminal obstruction at follow-up (408 narrowings in 241patients) A Decrease in lumen area 24 mm2 (20th percentile n86) B Decrease in lumen area 36 mm2 (10th percentile n40)C Decrease in lumen area 47 mm2 (5th percentile n20) The 3 ESS categories (low 1 Pa moderate 1ndash17 Pa high 17 Pa)were derived from the terciles of the baseline ESS frequency distribution in 3-mm segments Probability values are for testing the inde-pendence between ESS category (3 groups) and the binary outcome ESS indicates endothelial shear stress



by the observation in the PREDICTION Study that baselinelow ESS was an independent predictor responsible for plaqueprogression and worsening luminal obstruction both in theroutine natural history of CAD and in the development ofclinically relevant lesions treated with PCI This relationshipbetween local low ESS and atherosclerosis is particularlyimpressive because these coronary patients were followedclinically over a brief 12-month follow-up period (and assess-ment of anatomic natural history was performed only 6ndash10months after the ACS) they were aggressively managed withvasculoprotective agents and their cardiovascular risk waslow

Baseline large plaque burden and local low ESS wereindependently associated at follow-up with an increase inplaque burden and decrease in lumen area and clinicalworsening of luminal obstruction treated with a PCI as wellLocal low ESS environment may lead to plaque developmentprogression and formation of rupture-prone TCFA2715

which initially may not obstruct the lumen but may ruptureand precipitate either a new ACS or rapid progression of afixed plaque Plaque progression may have occurred in theseregions as a result of intraplaque hemorrhage due to subclin-ical plaque rupture or disruption of intraplaque vasa vasorumassociated with consequent fibrosis and worsening luminal

obstruction91920 Plaque progression with a decrease in lu-men and EEM area may also result from a more directphenotypic expression of fibrosis and scarring

We did not observe that baseline high ESS was associatedwith plaque progression or progressive luminal obstructionOther investigators have suggested that high ESS may pro-mote transformation of a plaque to a high-risk phenotype2122

Small studies have observed that plaque rupture in thecoronary and carotid circulations may be related to areas ofhigh ESS2324

We observed very few ACS events in our patients despitethe presence of substantial traditional risk factors and a recentACS This relatively benign outcome may be due to theeffective vasculoprotective treatments provided to these pa-tients or may reflect the relatively low genetic cardiovascularrisk of the Japanese population in comparison with a Westernpopulation25

Although we did not observe an effect of systemic riskfactors on coronary anatomic outcomes local factors are onlyimportant in the context of the presence of systemic riskfactors Recent studies suggest that high-risk plaque charac-teristics in the carotid artery may be associated with adversevascular outcomes in other vascular territories26

Although low ESS was an independent predictor of plaqueburden increase in follow-up we found that low ESS did notindependently predict plaque area increase The majority ofpatients in the PREDICTION Study were statin naive beforetheir index coronary event and were administered statins onlyafter their presentation with an ACS There may have beenglobal regression of atherosclerotic plaque area from theinitiation of statins during the follow-up period27 (Figure 1A)and an additional effect of plaque morphology change orEEM shrinkage from local low ESS and associated pathobi-ology as well manifesting as an associated increase inplaque burden and decrease in lumen area (Figure 1C)

The observations in the PREDICTION Study are comple-mentary to the recent observations in the PROSPECT Studyand provide new mechanistic insights that may elucidate theunderlying pathobiology of plaque progression and the de-velopment of clinical events In PROSPECT 697 patientswith an ACS underwent 3-vessel coronary angiography andIVUS imaging after PCI10 Multivariable analysis indicatedthat nonculprit lesions associated with subsequent majoradverse cardiovascular events were characterized at baselineby large plaque burden (70) small minimal luminal area(40 mm2) and appearance of a TCFA by radiofrequencyIVUS The hazard ratio for major adverse cardiovascularevents at 3 years was high (1105) if all 3 lesion character-

Table 4 Coronary Lesions Treated With PCI During Follow-Up

PCI Outcomes During Follow-Up Patients (n502)TreatedLesions

New acute coronary syndrome 13 (26) 13

Native area 4 (08) 4

Worsening stable angina 15 (30) 20


Absence of symptoms butsubstantial obstruction progressiondagger

102 (203) 122


Data on patient classification are not mutually exclusive PCI indicatespercutaneous coronary intervention 3D 3-dimensional

A total of 59 native areas and 96 previously stented areas (in-stentrestenosis or thrombosis) were treated with PCI at follow-up

daggerSubstantial obstruction progression defined as (1) lumen area decrease18 mm2 or 20 by 3D-modelndashbased measurements in cases withavailable baseline and follow-up vascular profiling data and (2) lumen diameterdecrease 20 by angiography in cases without available follow-up vascularprofiling data

Table 5 Baseline Anatomic Predictors (Continuous Data) ofPCI for Clinically Relevant Baseline Luminal Narrowings(n250) Univariable Analysis

Baseline Predictor Odds Ratio (95 CI) P

Plaque burden at throat (per 10increase)

261 (163ndash416) 0001

ESS distal to throat (per 1-Pa decrease) 160 (097ndash263) 0067

Plaque area at throat (per 1-mm2 increase) 126 (113ndash140) 0001

Lumen area at throat (per 1-mm2 increase) 088 (056ndash139) 059

Longitudinal arterial location at throat (per10-mm increase)

078 (059ndash104) 0090

PCI indicates percutaneous coronary intervention ESS endothelial shearstress

Table 6 Baseline Anatomic Predictors (Continuous Data) ofPCI for Clinically Relevant Baseline Luminal Narrowings(n250) Multivariable Analysis

Baseline Independent Predictor Odds Ratio (95 CI) P

Plaque burden at throat (per 10 increase) 279 (169ndash462) 0001


Variables entered in the multivariable model ESS distal to the throat plaquearea plaque burden and longitudinal arterial location at the throat PCIindicates percutaneous coronary intervention ESS endothelial shear stress



istics were present at baseline but only 42 of patientsmanifested these criteria Assessment of local ESS was notperformed in PROSPECT As in PROSPECT only 6 ofPREDICTION patients manifested the highest risk lesionprofile of large plaque burden and low ESS but these patientshad a 41 incidence of PCI for clinically relevant obstruc-tions We speculate that the small proportion (33) ofplaques with baseline large plaque burden that progressed tocause cardiac events in PROSPECT were those plaquesexposed to a low ESS environment which stimulated pro-gressive inflammation and atherogenesis culminating in anabrupt luminal change

Although plaque characterization and estimation of TCFAby radiofrequency IVUS analyses were not available in thePREDICTION Study large plaque burden was an indepen-dent predictor of an adverse outcome in both PREDICTIONand PROSPECT In PREDICTION there was no relationshipbetween plaque burden and low ESS The variable of localESS uniquely available in the PREDICTION Study pro-vided substantial independent incremental predictive insightwith the combined presence of low ESS and large plaqueburden having a positive predictive value of 41 to identifylesions at baseline likely to require PCI in the next year

LimitationsThe study is limited by the small number of clinical eventsthat occurred in the 1-year follow-up Many of the clinical

events were also primarily asymptomatic and consisted ofsignificant lumen progression treated with PCI There werefew patients with hard ischemic end points There were only14 (3) symptomatic clinical events (related to native seg-ments) in the 1-year follow-up Furthermore a number of theculprit lesions in native areas responsible for clinical eventswere either proximal or distal to the IVUS acquisitionpullback and thus not available for detailed VP investigation(proximal n4 distal n7 left main coronary artery n3VP data not available or usable n4) We are also limited bynot having radiofrequency IVUS characterization of plaqueconstituents available although the accuracy of the histolog-ical correlation of such characterization has recently beencalled into question2829

Our predictive analyses use both continuous and categor-ical baseline variables because the continuous variables areessential to prove the pathobiological relationship betweenbaseline vascular characteristics and outcomes but the cate-gorical variables derived either by nonbiased terciles or cutpoints determined by receiver operator characteristic analy-ses provide a way to compare the quantitative relationshipswe observed to the relationships in similar studies performedby investigators who use categorical baseline variables221ndash23

We emphasize that this study is a natural history study and ourconclusions are primarily exploratory All of our analyses havebeen prespecified but some of the specific threshold values of

Figure 4 Independent predictors of PCI for clinically relevant baseline luminal narrowings Prediction of PCI for baseline luminal nar-rowings with small minimal lumen area (6 mm2 n250) Variables entered in the multivariable model ESS plaque area plaque bur-den and longitudinal arterial location Cut points for continuous baseline variables were selected by ROC analysis which best predictedPCI occurrence low ESS (098Pa) large plaque area (730 mm2) large plaque burden (580) and proximal arterial location(254 mm) Error bars are 95 confidence intervals for odds ratios PCI indicates percutaneous coronary intervention ESS endotheli-al shear stress and ROC receiver operator characteristic

Figure 5 Incidence of PCI for clinically relevantbaseline luminal narrowings Incidence of PCI(n31) for baseline luminal narrowings with smallminimal lumen area (6 mm2 n250) accordingto the presence of independent predictors largeplaque burden (n131) low ESS (n53) and theircombination (n32) Probability values are forcomparing the incidence of PCI in baseline luminalnarrowings with versus without the predictor(s)Data on prevalence are for one or more suchbaseline luminal narrowings per patient PCI indi-cates percutaneous coronary intervention ESSendothelial shear stress



ESS plaque areaburden and lumen narrowings were deter-mined post hoc empirically within the PREDICTION data setIn addition all patients were post-PCI for an ACS and we donot know how the findings translate to asymptomatic orstable patients

In summary the PREDICTION Study demonstrated thatroutine 3-vessel IVUS can be safely performed in high-riskpatients following successful PCI for an ACS and thatprogression of plaque burden and luminal obstruction can bepredicted on the basis of the presence of substantial plaqueand low ESS at baseline Most clinical events in follow-up inthis population consisted of either asymptomatic or symptom-atic progressive luminal narrowing with stable symptomsACS events were very infrequent It remains to be confirmedwhether early identification of high-risk lesions is enhancedby determination of local ESS and whether such insights willbe clinically useful to guide patient management

AcknowledgmentsWe thank Gail MacCallum Michelle Lucier Emily Bomba-Rienzoand Nicholas Cefalo for technical evaluation of IVUSangiographicimages Drs Hiroyuki Daida and Mitsuyasu Terashima for earlyparticipation in the study and Dr Kazuhiro Sase for his invaluablehelp with the initial study design and implementation

Sources of FundingThis investigator-initiated trial was supported by Boston ScientificCompany We acknowledge the support of the George D BehrakisFellowship and the Hellenic Heart Foundation

DisclosuresThe investigators had sole direct access to the primary data andperformed all data analyses

References1 Asakura T Karino T Flow patterns and spatial distribution of athero-

sclerotic lesions in human coronary arteries Circ Res 1990661045ndash1066

2 Koskinas KC Feldman CL Chatzizisis YS Coskun AU Jonas MMaynard C Baker AB Papafaklis MI Edelman ER Stone PH Naturalhistory of experimental coronary atherosclerosis and vascular remodelingin relation to endothelial shear stress a serial in vivo intravascularultrasound study Circulation 20101212092ndash2101

3 Malek AM Alper SL Izumo S Hemodynamic shear stress and its role inatherosclerosis JAMA 19992822035ndash2042

4 Gimbrone MA Jr Topper JN Nagel T Anderson KR Garcia-Cardena GEndothelial dysfunction hemodynamic forces and atherogenesis Ann NY Acad Sci 2000902230ndash239

5 Chatzizisis YS Coskun AU Jonas M Edelman ER Feldman CL StonePH Role of endothelial shear stress in the natural history of coronaryatherosclerosis and vascular remodeling molecular cellular and vascularbehavior J Am Coll Cardiol 2007492379ndash2393

6 Libby P Inflammation in atherosclerosis Nature 2002420868ndash8747 Chatzizisis YS Jonas M Coskun AU Beigel R Stone BV Maynard C

Gerrity RG Daley W Rogers C Edelman ER Feldman CL Stone PHPrediction of the localization of high-risk coronary atheroscleroticplaques on the basis of low endothelial shear stress an intravascularultrasound and histopathology natural history study Circulation 2008117993ndash1002

8 Virmani R Kolodgie FD Burke AP Farb A Schwartz SM Lessons fromsudden coronary death a comprehensive morphological classificationscheme for atherosclerotic lesions Arterioscler Thromb Vasc Biol 2000201262ndash1275

9 Kolodgie FD Gold HK Burke AP Fowler DR Kruth HS Weber DKFarb A Guerrero LJ Hayase M Kutys R Narula J Finn AV Virmani RIntraplaque hemorrhage and progression of coronary atheroma N EnglJ Med 20033492316ndash2325

10 Stone GW Maehara A Lansky AJ de Bruyne B Cristea E Mintz GSMehran R McPherson J Farhat N Marso SP Parise H Templin BWhite R Zhang Z Serruys PW A prospective natural-history study ofcoronary atherosclerosis N Engl J Med 2011364226ndash235

11 Feldman CL Ilegbusi OJ Hu Z Nesto R Waxman S Stone PH Deter-mination of in vivo velocity and endothelial shear stress patterns withphasic flow in human coronary arteries a methodology to predict pro-gression of coronary atherosclerosis Am Heart J 2002143931ndash939

12 Coskun AU Yeghiazarians Y Kinlay S Clark ME Ilegbusi OJ WahleA Sonka M Popma JJ Kuntz RE Feldman CL Stone PH Reproduc-ibility of coronary lumen plaque and vessel wall reconstruction and ofendothelial shear stress measurements in-vivo in humans Catheter Car-diovasc Interv 20036067ndash78

13 Mintz GS Nissen SE Anderson WD Bailey SR Erbel R Fitzgerald PJPinto FJ Rosenfield K Siegel RJ Tuzcu EM Yock PG AmericanCollege of Cardiology Clinical Expert Consensus Document on Standardsfor Acquisition Measurement and Reporting of Intravascular UltrasoundStudies (IVUS) A report of the American College of Cardiology TaskForce on Clinical Expert Consensus Documents J Am Coll Cardiol2001371478ndash1492

14 Feldman CL Coskun AU Yeghiazarians Y Kinlay S Wahle AOlszewski ME Rossen JD Sonka M Popma JJ Orav J Kuntz RE StonePH Remodeling characteristics of minimally diseased coronary arteriesare consistent along the length of the artery Am J Cardiol 20069713ndash16

15 Chatzizisis YS Baker AB Sukhova GK Koskinas KC Papafaklis MIBeigel R Jonas M Coskun AU Stone BV Maynard C Shi GP Libby PFeldman CL Edelman ER Stone PH Augmented expression and activityof extracellular matrix-degrading enzymes in regions of low endothelialshear stress colocalize with coronary atheromata with thin fibrous caps inpigs Circulation 2011123621ndash630

16 Stone PH Coskun AU Kinlay S Clark ME Sonka M Wahle A IlegbusiOJ Yeghiazarians Y Popma JJ Orav J Kuntz RE Feldman CL Effectof endothelial shear stress on the progression of coronary artery diseasevascular remodeling and in-stent restenosis in humans in-vivo 6-monthfollow-up study Circulation 2003108438ndash444

17 Stone PH Coskun AU Kinlay S Popma JJ Sonka M Wahle A Yeghi-azarians Y Maynard C Kuntz RE Feldman CL Regions of low endo-thelial shear stress are the sites where coronary plaque progresses andvascular remodelling occurs in humans an in-vivo serial study EurHeart J 200728705ndash710

18 Koskinas K Sukhova GK Baker AB Chatzizisis YS Papafaklis MICoskun AU Jonas M Shi G-P Libby P Edelman ER Stone PHFeldman CL Coronary thin-capped atheromata exhibit increasedexpression of interstitial collagenases in regions of persistently low en-dothelial shear stress a serial in-vivo natural history study in pigsCirculation 2010122A19592

19 Virmani R Kolodgie FD Burke AP Finn AV Gold HK Tulenko TNWrenn SP Narula J Atherosclerotic plaque progression and vulnerabilityto rupture angiogenesis as a source of intraplaque hemorrhage Arte-rioscler Thromb Vasc Biol 2005252054ndash2061

20 Takaya N Yuan C Chu B Saam T Polissar NL Jarvik GP Isaac CMcDonough J Natiello C Small R Ferguson MS Hatsukami TSPresence of intraplaque hemorrhage stimulates progression of carotidatherosclerotic plaques a high-resolution magnetic resonance imagingstudy Circulation 20051112768ndash2775

Table 7 Clinical Outcomes After 1-Year Follow-Up

Outcome n ()

Follow-up complete 502 (992)

Total death 7 (14)

Cardiac death 1 (02)

Noncardiac death 6 (12)

Acute coronary syndrome 13 (26)

Unstable angina 11 (22)

NonndashST-elevation MI 1 (02)

ST-elevation MI 1 (02)

Hospitalization for stable angina 15 (30)

MI indicates myocardial infarctionData classification is not mutually exclusive



21 Samady H Eshtehardi P McDaniel MC Suo J Dhawan SS Maynard CTimmins LH Quyyumi AA Giddens DP Coronary artery wall shearstress is associated with progression and transformation of atheroscleroticplaque and arterial remodeling in patients with coronary artery diseaseCirculation 2011124779ndash788

22 Gijsen FJ Wentzel JJ Thury A Mastik F Schaar JA Schuurbiers JCSlager CJ van der Giessen WJ de Feyter PJ van der Steen AF SerruysPW Strain distribution over plaques in human coronary arteries relates toshear stress Am J Physiol Heart Circ Physiol 2008295H1608ndashH1614

23 Fukumoto Y Hiro T Fujii T Hashimoto G Fujimura T Yamada JOkamura T Matsuzaki M Localized elevation of shear stress is related tocoronary plaque rupture a 3-dimensional intravascular ultrasound studywith in-vivo color mapping of shear stress distribution J Am CollCardiol 200851645ndash650

24 Tang D Teng Z Canton G Yang C Ferguson M Huang X Zheng JWoodard PK Yuan C Sites of rupture in human atherosclerotic carotidplaques are associated with high structural stresses an in-vivo MRI-based3D fluid-structure interaction study Stroke 2009403258ndash3263

25 Ueshima H Sekikawa A Miura K Turin TC Takashima N Kita YWatanabe M Kadota A Okuda N Kadowaki T Nakamura Y Okamura

T Cardiovascular disease and risk factors in Asia a selected reviewCirculation 20081182702ndash2709

26 Hellings WE Peeters W Moll FL Piers SR van Setten J Van der SpekPJ de Vries JP Seldenrijk KA De Bruin PC Vink A Velema E deKleijn DP Pasterkamp G Composition of carotid atherosclerotic plaqueis associated with cardiovascular outcome a prognostic study Circu-lation 20101211941ndash1950

27 Nicholls SJ Ballantyne CM Barter PJ Chapman MJ Erbel RM LibbyP Raichlen JS Uno K Borgman M Wolski K Nissen SE Effect of twointensive statin regimens on progression of coronary disease N EnglJ Med 3652078ndash2087

28 Thim T Hagensen MK Wallace-Bradley D Granada JF Kaluza GLDrouet L Paaske WP Botker HE Falk E Unreliable assessment ofnecrotic core by virtual histology intravascular ultrasound in porcinecoronary artery disease Circ Cardiovasc Imaging 20103384ndash391

29 Shin ES Garcia-Garcia HM Ligthart JM Witberg K Schultz C van derSteen AF Serruys PW In-vivo findings of tissue characteristics usingiMap IVUS and virtual histology IVUS EuroIntervention 201161017ndash1019

CLINICAL PERSPECTIVECoronary plaques progresses in a highly individual manner Identification of early stages of high-risk plaque may enabledevelopment of preemptive strategies to alter the natural history of high-risk plaque and avert adverse cardiac events Thepurposes of the PREDICTION Study were to determine the role of local vascular characteristics on coronary plaqueprogression in humans and to relate plaque changes to clinical events Five hundred six patients with an acute coronarysyndrome treated with a percutaneous coronary intervention were enrolled Vascular profiling using coronary angiographyand intravascular ultrasound was used to reconstruct the artery and calculate endothelial shear stress and plaqueremodelingcharacteristics in vivo Each reconstructed artery was divided into sequential 3-mm segments for serial analysisThree-vessel vascular profiling was performed at baseline and in a subset of 374 patients 6 to 10 months later to assessplaque natural history Clinical follow-up was performed at 1 year Symptomatic clinical events were infrequent Increasein plaque area at follow-up was predicted by baseline large plaque burden decrease in lumen area was independentlypredicted by baseline large plaque burden and low endothelial shear stress Large plaque size and low endothelial shearstress independently predicted increased plaque burden and worsening of luminal obstructions treated with a percutaneouscoronary intervention at follow-up Large plaque burden and low local endothelial shear stress provide independent andadditive prediction to identify plaques that develop progressive enlargement and lumen narrowing These observations mayjustify prospective randomized trials to identify early stages of high-risk plaque and investigate the value of localpreemptive interventions



Coronary Heart Disease

Prediction of Progression of Coronary Artery Disease andClinical Outcomes Using Vascular Profiling of Endothelial

Shear Stress and Arterial Plaque CharacteristicsThe PREDICTION Study

Peter H Stone MD Shigeru Saito MD Saeko Takahashi MD Yasuhiro Makita MD Shigeru Nakamura MDTomohiro Kawasaki MD Akihiko Takahashi MD Takaaki Katsuki MD Sunao Nakamura MD

Atsuo Namiki MD Atsushi Hirohata MD Toshiyuki Matsumura MD Seiji Yamazaki MDHiroyoshi Yokoi MD Shinji Tanaka MD Satoru Otsuji MD Fuminobu Yoshimachi MD Junko Honye MD

Dawn Harwood PhD Martha Reitman MD Ahmet U Coskun PhD Michail I Papafaklis MD PhDCharles L Feldman ScD for the PREDICTION Investigators

BackgroundmdashAtherosclerotic plaques progress in a highly individual manner The purposes of the Prediction ofProgression of Coronary Artery Disease and Clinical Outcome Using Vascular Profiling of Shear Stress and WallMorphology (PREDICTION) Study were to determine the role of local hemodynamic and vascular characteristics incoronary plaque progression and to relate plaque changes to clinical events

Methods and ResultsmdashVascular profiling using coronary angiography and intravascular ultrasound was used to reconstructeach artery and calculate endothelial shear stress and plaqueremodeling characteristics in vivo Three-vessel vascular profiling(27 arteries per patient) was performed at baseline in 506 patients with an acute coronary syndrome treated with apercutaneous coronary intervention and in a subset of 374 (74) consecutive patients 6 to 10 months later to assess plaquenatural history Each reconstructed artery was divided into sequential 3-mm segments for serial analysis One-year clinicalfollow-up was completed in 992 Symptomatic clinical events were infrequent only 1 (02) cardiac death 4 (08)patients with new acute coronary syndrome in nonstented segments and 15 (30) patients hospitalized for stable anginaIncrease in plaque area (primary end point) was predicted by baseline large plaque burden decrease in lumen area (secondaryend point) was independently predicted by baseline large plaque burden and low endothelial shear stress Large plaque sizeand low endothelial shear stress independently predicted the exploratory end points of increased plaque burden and worseningof clinically relevant luminal obstructions treated with a percutaneous coronary intervention at follow-up The combinationof independent baseline predictors had a 41 positive and 92 negative predictive value to predict progression of anobstruction treated with a percutaneous coronary intervention

ConclusionsmdashLarge plaque burden and low local endothelial shear stress provide independent and additive prediction toidentify plaques that develop progressive enlargement and lumen narrowing

Clinical Trial RegistrationmdashURL httpwwwclinicaltrialsgov Unique Identifier NCT01316159 (Circulation 2012126172-181)

Key Words atherosclerosis endothelium natural history shear stress

Atherosclerosis is a systemic disease with focal and eccentricmanifestations1 In a patient with coronary artery disease

(CAD) and systemic risk factors each coronary lesion pro-gresses regresses or remains quiescent in an independentmanner2 indicating that local vascular factors must be a majordeterminant responsible for the behavior of individual plaques

Editorial see p 161Clinical Perspective on p 181

The vascular endothelium is in a unique and pivotalposition to respond to the extremely dynamic forces acting onthe vessel wall because of the complex 3-dimensional (3D)

Received January 27 2012 accepted May 16 2012From the Brigham amp Womenrsquos Hospital Harvard Medical School Boston MA (PHS ST MR MIP CLF) Shonan Kamakura General Hospital Kamakura

(SS ST JH) Hakodate Municipal Hospital Hakodate (YM) Kyoto Katsura Hospital Kyoto (SN) Shin Koga Hospital Kurume (TK) Sakurakai TakahashiHospital Kobe (AT) Jichi Medical University Hospital Shimotsuke (TK) New Tokyo Hospital Matsudo (SN) Kanto Rosai Hospital Kawasaki (AN)Sakakibara Heart Institute of Okayama Okayama (AH) Kumamoto Rosai Hospital Yatsushiro (TM) Sapporo Higashi Tokushukai Hospital Sapporo (SY)Kokura Memorial Hospital Kitakyushu (HY) Shonan Atsugi Hospital Atsugi (ST) Higashi Takarazuka Satoh Hospital Takarazuka (SO) AomoriPrefectural Central Hospital Aomori Japan (FY) REGISTRAT-MAPI Lexington KY (DH) Northeastern University Boston MA (AUC)

Guest Editor for this article was Steven E Nissen MDThe online-only Data Supplement is available with this article at httpcircahajournalsorglookupsuppldoi101161CIRCULATIONAHA

112096438-DC1Correspondence to Peter H Stone MD Cardiovascular Division Brigham amp Womenrsquos Hospital 75 Francis St Boston MA 02115 E-mail

pstonepartnersorgcopy 2012 American Heart Association Inc

Circulation is available at httpcircahajournalsorg DOI 101161CIRCULATIONAHA112096438

172 by guest on July 9 2012httpcircahajournalsorgDownloaded from
























(n506)

Sex

Male n () 404 (798)

Female n () 102 (202)


Risk factors


Dyslipidemia

LDL 100 mgdL n () 374 (739)

HDL 40 mgdL n () 196 (387)








PCI n () 48 (95)

CABG n () 0 (00)





Laboratory data





















Statin n () 355 (702)



-blocker n () 169 (334)















009 (0001 to 0176) 0047







025 (039 to 012) 0001







074 (032 to 115) 0001








































102 (203) 122








261 (163ndash416) 0001





078 (059ndash104) 0090














































Outcome n ()


Total death 7 (14)















































(n506)

Sex

Male n () 404 (798)

Female n () 102 (202)


Risk factors


Dyslipidemia

LDL 100 mgdL n () 374 (739)

HDL 40 mgdL n () 196 (387)








PCI n () 48 (95)

CABG n () 0 (00)





Laboratory data





















Statin n () 355 (702)



-blocker n () 169 (334)















009 (0001 to 0176) 0047







025 (039 to 012) 0001







074 (032 to 115) 0001








































102 (203) 122








261 (163ndash416) 0001





078 (059ndash104) 0090














































Outcome n ()


Total death 7 (14)



































Statin n () 355 (702)



-blocker n () 169 (334)















009 (0001 to 0176) 0047







025 (039 to 012) 0001







074 (032 to 115) 0001








































102 (203) 122








261 (163ndash416) 0001





078 (059ndash104) 0090














































Outcome n ()


Total death 7 (14)



























































102 (203) 122








261 (163ndash416) 0001





078 (059ndash104) 0090














































Outcome n ()


Total death 7 (14)





























































Outcome n ()


Total death 7 (14)
























Prediction of Progression of Coronary Artery Disease and Clinical Outcomes Using Vascular Profiling...

Documents

Transcript of Prediction of Progression of Coronary Artery Disease and Clinical Outcomes Using Vascular Profiling...