Embed Size (px)
Citation preview
J A C C : C A R D I O V A S C U L A R I M A G I N G V O L . 9 , N O . 9 , 2 0 1 6
ª 2 0 1 6 B Y T H E AM E R I C A N C O L L E G E O F C A R D I O L O G Y F O UN DA T I O N I S S N 1 9 3 6 - 8 7 8 X / $ 3 6 . 0 0
P U B L I S H E D B Y E L S E V I E R h t t p : / / d x . d o i . o r g / 1 0 . 1 0 1 6 / j . j c m g . 2 0 1 6 . 0 5 . 0 0 5
STATE-OF-THE-ART PAPER
Coronary Artery Disease - Reportingand Data System (CAD-RADS)An Expert Consensus Document of SCCT,ACR and NASCI
Endorsed by the ACC
Ricardo C. Cury, MD, Suhny Abbara, MD, Stephan Achenbach, MD, Arthur Agatston, MD, Daniel S. Berman, MD,Matthew J. Budoff, MD, Karin E. Dill, MD, Jill E. Jacobs, MD, Christopher D. Maroules, MD, Geoffrey D. Rubin, MD,Frank J. Rybicki, MD, PHD, U. Joseph Schoepf, MD, Leslee J. Shaw, PHD, Arthur E. Stillman, MD, Charles S. White, MD,Pamela K. Woodard, MD, Jonathon A. Leipsic, MD
ABSTRACT
Ma
The intent of CAD-RADS - Coronary Artery Disease Reporting and Data System is to create a standardized method to
communicate findings of coronary CT angiography (coronary CTA) in order to facilitate decision-making regarding further
patient management. The suggested CAD-RADS classification is applied on a per-patient basis and represents the
highest-grade coronary artery lesion documented by coronary CTA. It ranges from CAD-RADS 0 (Zero) for the complete
absence of stenosis and plaque to CAD-RADS 5 for the presence of at least one totally occluded coronary artery and
should always be interpreted in conjunction with the impression found in the report. Specific recommendations are
provided for further management of patients with stable or acute chest pain based on the CAD-RADS classification. The
main goal of CAD-RADS is to standardize reporting of coronary CTA results and to facilitate communication of test results
to referring physicians along with suggestions for subsequent patient management. In addition, CAD-RADS will provide a
framework of standardization that may benefit education, research, peer-review and quality assurance with the potential
to ultimately result in improved quality of care. (J Am Coll Cardiol Img 2016;9:1099–113) © 2016 by the American
College of Cardiology Foundation.
CAD-RADSTM
Coronary Artery Disease – Reporting and
Data System:
Collaboration of SCCT, ACR, ACC & NASCI
Ricardo C. Cury, MD (Chair)Miami Cardiac and Vascular Institute8900 N Kendall Drive, Miami FL 33176(786) [email protected]
nuscript received April 7, 2016; revised manuscript received April 29, 201
Suhny Abbara, MDDepartment of Radiology5323 Harry Hines BlvdDallas, TX [email protected]
Stephan Achenbach, MDFriedrich-Alexander-Universität, Department ofCardiology, Ulmenweg 18, 90154 Erlangen,[email protected]
6, accepted May 26, 2016.
ABBR EV I A T I ON S
AND ACRONYMS
CAD-RADS = Coronary Artery
Disease Reporting and Data
System
Coronary CTA = coronary CT
angiography
BI-RADS = Breast Imaging
Reporting and Data System
LI-RADS = Liver Imaging
Reporting and Data System
Lung-RADS = Lung CT
Screening Reporting and Data
System
PI-RADS = Prostate Imaging
Reporting and Data System
ICA = invasive coronary
angiography
CAD = coronary artery disease
ACS = acute coronary
syndrome
N = non-diagnostic
S = stent
G = graft
V = vulnerability
Cury et al. J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 9 , N O . 9 , 2 0 1 6
Coronary Artery Disease – Reporting and Data System S E P T E M B E R 2 0 1 6 : 1 0 9 9 – 1 1 3
1100
Arthur Agatston, MDBaptist Health Medical Grp1691 Michigan AvenueMiami, FL [email protected]
Daniel S. Berman, MDCedars-Sinai Med Center8700 Beverly BoulevardTaper Building, Rm 1258Los Angeles, CA [email protected]
Matthew J. Budoff, MD1124 W. Carson StreetTorrance, CA 90502(310) [email protected]
Karin E. Dill, MD5841 South Maryland AveMC2026Chicago, IL 60637(773) [email protected]
Jill E. Jacobs, MD
550 First AvenueNew York, NY [email protected]Christopher D. Maroules, MDDepartment of Radiology, 5323 HarryHines Blvd Dallas, TX [email protected]
Geoffrey D. Rubin, MD2400 Pratt Street, Room 8020 DCRI Box17969 Durham, NC [email protected]
Frank J. Rybicki, MD, PhDThe Ottawa Hospital General Campus501 Smyth RdOttawa, ON, CA K1H 8L6(613) [email protected]
U. Joseph Schoepf, MD25 Courtenay Dr.Charleston, SC 29425(843) [email protected]
Leslee J. Shaw, PhD1256 Briarcliff Rd. NERm 529 Atlanta, GA 30324(404) [email protected]
Arthur E. Stillman, MD1364 Clifton Road, NEAtlanta, GA 30322(404) [email protected]
Charles S. White, MD22 S. Greene St. Baltimore MD 21201University of MarylandPh: 410 [email protected]
Pamela K. Woodard, MDMallinckrodt Instit of Radiology. 510 SKingshighway BlvdSt. Louis, MO 63110(314) [email protected]
Jonathon A. Leipsic, MDDepartment of Radiology j St. Paul’s Hospital2nd Floor, Providence Building 1081Burrard Street Vancouver,BC V6Z 1Y6 Canada(604) [email protected]
Staff contacts:Norm Linsky, MA, MPASCCT Executive Director415 Church St NE, # 204Vienna VA 22180(202) [email protected]
Grace RonanACC Team Lead, Policy Publication2400 N Street NWWashington DC [email protected]
Mythreyi Bhargavan Chatfield, PhDACR EVP for Quality & Safety.1891 Preston White Drive, Reston,VA 20191 P:[email protected]
Michele WittlingNASCI Executive Director. 1891 PrestonWhite DriveReston, VA [email protected]
1. INTRODUCTION
Coronary CT Angiography (coronary CTA) has madesubstantial progress since the introduction of 64-sliceCT scanners approximately 10 years ago (1), both
ataSy
stem
forPat
ient
sPre
sent
ingWithSt
able
Ches
tPain
nary
Sten
osis
Interp
retation
Furthe
rCa
rdiacInve
stigation
Man
agem
ent
sis)
Doc
umen
tedab
senc
eof
CAD*
Non
eRea
ssuran
ce.C
onside
rno
n-athe
roscleroticcauses
ofch
estpa
in
isor
plaq
ueMinim
alno
n-ob
structiveCA
DNon
eCo
nsider
non-athe
roscleroticcauses
ofch
estpa
inCo
nsider
prev
entive
therap
yan
dris
kfactor
mod
ification
Mild
non-ob
structiveCA
DNon
eCo
nsider
non-athe
roscleroticcauses
ofch
estpa
inCo
nsider
prev
entive
therap
yan
dris
kfactor
mod
ification
,pa
rticularly
forpa
tien
tswithno
n-ob
structiveplaq
uein
multiple
segm
ents.
Mod
eratesten
osis
Consider
func
tion
alassessmen
tCo
nsider
symptom
-guide
dan
ti-ische
mic
andprev
entive
pharmacothe
rapy
aswellas
riskfactor
mod
ification
pergu
ideline-directed
care***
Other
trea
tmen
tsshou
ldbe
considered
pergu
ideline-directed
care***
3-ve
ssel
disease
Seve
resten
osis
A:Co
nsider
ICA****
orfunc
tion
alassessmen
tB:ICAis
reco
mmen
ded
Consider
symptom
-guide
dan
ti-ische
mic
andprev
entive
pharmacothe
rapy
aswellas
riskfactor
mod
ification
pergu
ideline-directed
care***
Other
trea
tmen
ts(in
clud
ingop
tion
sof
reva
scularization)
shou
ldbe
considered
pergu
ideline-directed
care***
Totalco
rona
ryoc
clusion
Consider
ICAan
d/or
viab
ility
assessmen
tCo
nsider
symptom
-guide
dan
ti-ische
mic
andprev
entive
pharmacothe
rapy
aswellas
riskfactorsmod
ification
pergu
ideline-directed
care***
Other
trea
tmen
ts(in
clud
ingop
tion
sof
reva
scularization)
shou
ldbe
considered
pergu
ideline-directed
care***
Obs
truc
tive
CAD
cann
otbe
exclud
edAdd
itiona
lor
alternativeev
alua
tion
may
bene
eded
onape
r-pa
tien
tba
sisfortheclinically
mostreleva
nt(usually
high
est-grad
e)sten
osis.A
llve
sselsgrea
terthan
1.5mm
indiam
eter
shou
ldbe
grad
edforsten
osisseve
rity.
CAD-R
ADSwill
notap
plyforsm
allerve
ssels
hanon
emod
ifier
ispresen
t,thesymbo
l“/”(slash)shou
ldfollo
wea
chmod
ifier
inthefollo
wingorde
r:First:
mod
ifier
N(non
-diagn
ostic).Se
cond
:mod
ifier
S(stent).
Third
:mod
ifier
G(graft).
Fourth:mod
ifier
V**CA
D-R
ADS1–Th
iscatego
ryshou
ldalso
includ
ethepresen
ceof
plaq
uewithpo
sitive
remod
elingan
dno
eviden
ceof
sten
osis.*
**Guide
line-directed
care
perACC
Stab
leIsch
emicHea
rtDisea
seGuide
lines
(Fihnet
al.
giog
raph
y.
J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 9 , N O . 9 , 2 0 1 6 Cury et al.S E P T E M B E R 2 0 1 6 : 1 0 9 9 – 1 1 3 Coronary Artery Disease – Reporting and Data System
1101
concerning imaging technology and clinical validation.In parallel, several professional societies have issuedguidelines, expert consensus documents, and Appro-priatenessCriteria for coronaryCTA (2–8). Tomaximizethe clinical impact of coronary CTA, imaging protocolsmust be optimized with respect to image quality,diagnostic accuracy, and radiation dose. Training andinterpretation standards are important. Finally, stan-dardized reporting is helpful to decrease variabilityamong practitioners and may provide further benefitby linking the final impression in the report with sug-gestions for further patient management.
Other fields in medical imaging (notably, breastimaging with BI-RADS) have introduced standardizedreporting linked with actionable information to guidenext steps in patient management (9). BI-RADSstandardized reporting of screening mammogramsallows clinicians to interpret the clinical relevance ofreported findings and to take action. Moreover,BI-RADS facilitates collection of data for registriesand databases, allowing better tracking of individualpatient outcomes with specific imaging findings.
Next to BI-RADS, standardized reporting has beenintroduced for several other fields. They include, forexample:
� LI-RADSTM (Liver Imaging Reporting and DataSystem) for standardization reporting in patientswith chronic liver disease (10).
� Lung-RADSTM (Lung CT Screening Reporting andData System) for standardization reporting of high-risk smokers undergoing CT lung screening (11).
� PI-RADSTM (Prostate Imaging Reporting and DataSystem) for multi-parametric MR imaging in thecontext of prostate cancer (12).
The purpose of this document is to describe astandardized reporting system for patients undergo-ing coronary CTA. The report system is named CAD-RADS (Coronary Artery Disease Reporting and DataSystem) and is applicable to coronary CTA in patientswith suspected or known coronary artery diseaseeither in the outpatient, inpatient or emergencydepartment setting. It includes suggestions regardingfurther patient management, which, obviously will
TABLE 1 SCCT Grading Scale for Stenosis Severity
Degree of Luminal Diameter Stenosis Terminology
0% No visible stenosis
1-24% Minimal stenosis
25-49% Mild stenosis
50-69% Moderate stenosis
70-99% Severe stenosis
100% Occluded TABLE
2CA
D-R
ADSRep
orting
andD
Deg
reeof
Max
imal
Coro
CAD-R
ADS0
0%
(Noplaq
ueor
sten
o
CAD-R
ADS1
1-24
%-Minim
alsten
oswithno
sten
osis**
CAD-R
ADS2
25-4
9%
-Mild
sten
osis
CAD-R
ADS3
50-6
9%
sten
osis
CAD-R
ADS4
A-70
-99%
sten
osis
or B-Le
ftmain>50
%or
obstructive($
70%)
CAD-R
ADS5
100%
(total
occlusion)
CAD-R
ADSN
Non
-diagn
osticstud
y
TheCA
D-R
ADSclassification
shou
ldbe
applied
(<1.5mm
indiam
eter).
MODIFIERS:
Ifmoret
(vulne
rability).*
CAD–co
rona
ryartery
disease.
JACC
2012)(25).****ICA–inva
sive
corona
ryan
TABLE
3CA
D-R
ADSRep
orting
andDat
aSy
stem
forPat
ient
sPre
sent
ingWithAcu
teCh
estPain,
Neg
ativeFirstTr
opon
in,Neg
ativeor
Non
-diagno
stic
Elec
troc
ardiogra
man
dLo
wto
Inte
rmed
iate
Risk
(TIM
IRiskSc
ore<
4)(E
mer
gen
cyDep
artm
entor
Hos
pital
Setting)
Deg
reeof
max
imal
coro
nary
sten
osis
Inte
rpretation
Man
agem
ent
CAD-R
ADS0
0%
ACS
*high
lyun
likely
Nofurthe
rev
alua
tion
ofACS
isrequ
ired.
Consider
othe
retiologies.
CAD-R
ADS1
1-24
%**
ACS
high
lyun
likely
Consider
evalua
tion
ofno
n-ACS
etiology
,ifno
rmal
trop
onin
andno
ECGch
ange
s.Co
nsider
referral
forou
tpatient
follo
w-upforprev
entive
therap
yan
dris
kfactor
mod
ification
.
CAD-R
ADS2
25-4
9%***
ACS
unlik
ely
Consider
evalua
tion
ofno
n-ACS
etiology
,ifno
rmal
trop
onin
andno
ECGch
ange
s.Co
nsider
referral
forou
tpatient
follo
w-upforprev
entive
therap
yan
dris
kfactor
mod
ification
.Ifclinical
susp
icionof
ACS
ishigh
orifhigh
-riskplaq
uefeatures
areno
ted,
consider
hosp
ital
admission
withcardiology
consultation
.
CAD-R
ADS3
50-6
9%
ACS
possible
Consider
hosp
ital
admission
withcardiology
consultation
,func
tion
altestingan
d/or
ICA****
forev
alua
tion
andman
agem
ent.
Recom
men
dation
foran
ti-ische
mic
andprev
entive
man
agem
entshou
ldbe
considered
aswella
sris
kfactor
mod
ification
.Other
trea
tmen
tsshou
ldbe
considered
ifpresen
ceof
hemod
ynam
ically
sign
ificant
lesion
.
CAD-R
ADS4
A-70
-99%
orB-Le
ftmain>50
%or
3-ve
ssel
obstructivedisease
ACS
likely
Consider
hosp
ital
admission
withcardiology
consultation
.Fu
rthe
rev
alua
tion
withICAan
dreva
scularizationas
approp
riate.
Recom
men
dation
foran
ti-ische
mic
andprev
entive
man
agem
entshou
ldbe
considered
aswellas
riskfactor
mod
ification
.
CAD-R
ADS5
100%
(Total
occlusion)
ACS
very
likely
Consider
expe
ditedICAon
atimelyba
sisan
dreva
scularizationifap
prop
riate
ifacuteoc
clusion*
****
Recom
men
dation
foran
ti-ische
mic
andprev
entive
man
agem
entshou
ldbe
considered
aswellas
riskfactor
mod
ification
s.
CAD-R
ADSN
Non
-diagn
osticstud
yACS
cann
otbe
exclud
edAdd
itiona
lor
alternativeev
alua
tion
forACS
isne
eded
TheCA
D-R
ADSclassification
shou
ldbe
appliedon
ape
r-pa
tien
tba
sisfortheclinically
mostreleva
nt(usually
high
est-grad
e)sten
osis.A
llve
sselsgrea
terthan
1.5mm
indiam
eter
shou
ldbe
grad
edforsten
osisseve
rity.
CAD-R
ADSwill
notap
plyforsm
allerve
ssels
(<1.5mm
indiam
eter).
MODIFIERS:
Ifmorethan
onemod
ifier
ispresen
t,thesymbo
l“/”(slash)shou
ldfollo
wea
chmod
ifier
inthefollo
wingorde
r:First:
mod
ifier
N(non
-diagn
ostic).Se
cond
:mod
ifier
S(stent).
Third
:mod
ifier
G(graft).
Fourth:mod
ifier
V(vulne
rability).*
ACS
–acuteco
rona
rysynd
rome.
**CA
D-R
ADS1–Th
iscatego
ryshou
ldalso
includ
ethepresen
ceof
plaq
uewithpo
sitive
remod
elingan
dno
eviden
ceof
sten
osis.*
**CA
D-R
ADS2-Mod
ifier
2/Vcanbe
used
toindicate
vulnerab
le/high
-riskplaq
ue.
****ICA–inva
sive
corona
ryan
giog
raph
y.*****U
nlessthetotalco
rona
ryoc
clusioncanbe
iden
tified
asch
ronic(throu
ghCT
andclinical
characteris
tics
orpa
tien
thistory).
Cury et al. J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 9 , N O . 9 , 2 0 1 6
Coronary Artery Disease – Reporting and Data System S E P T E M B E R 2 0 1 6 : 1 0 9 9 – 1 1 3
1102
always need to be seen in light of the full clinical in-formation available to the treating physician. For thespecific setting of coronary CTA in patients with acutechest pain presenting to the emergency department,certain management recommendations have beenreported previously (13,14).
The goal of CAD-RADS, through standardization ofreport terminology for coronary CTA, is to improvecommunication between interpreting and referringphysicians, facilitate research, and offer mechanismsto contribute to peer review and quality assurance,ultimately resulting in improvements to quality ofcare. Importantly, CAD-RADS does not substitute theimpression section provided by the reading physicianand should always be interpreted in conjunction withthe more individual and patient-specific informationfound in the report.
2. CLINICAL VALUE OF
CORONARY CT ANGIOGRAPHY
Several recent prospective trials have evaluated theclinical utility of coronary CTA and the relevance ofCT findings in the context of suspected stable coro-nary artery disease. They include the PROMISE (15)and SCOT-HEART (16) trials, which demonstrated thatcoronary CTA is clinically useful as an alternativeto (PROMISE) or in addition to functional testing(SCOT-HEART).
Four large randomized trials (CT-STAT, ACRIN-PA,ROMICAT II and CT-COMPARE) compared coronaryCTA to the current standard of care in patients withacute chest pain (17–20). Complemented by “realworld” implementation data (21,22), they consistentlydemonstrate the safety of a negative coronary CTA toidentify patients for discharge from the emergencydepartment.
There are some limitations to the currentlymentioned available studies (for example, their over-representation of low risk patients). Other situations,such as the use of coronary CTA in patients withknown coronary artery disease, have not been eval-uated in appropriate clinical trials. Hence, while fullytaking into account the available data, this documentis based on expert consensus. This includes the sug-gested categories for reporting but also the sugges-tions for further patient management, which needto be interpreted in the context of other clinical in-formation that is available in any given patient.
3. CAD-RADS REPORTING SYSTEM
3.1. CAD-RADS CATEGORIES. CAD-RADS categoriesdepend on stenosis severity. For the grading of
FIGURE 1 CAD-RADS 0
Normal left main, LAD, LCX and RCA without plaque or stenosis.
FIGURE 2 CAD-RADS 1
J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 9 , N O . 9 , 2 0 1 6 Cury et al.S E P T E M B E R 2 0 1 6 : 1 0 9 9 – 1 1 3 Coronary Artery Disease – Reporting and Data System
1103
stenosis severity, a classification system suggested bythe Society of Cardiovascular Computed Tomographyis used (see Table 1). Tables 2 and 3 list the categoriesof the CAD-RADS reporting system for stable chestpain (Table 2) and acute chest pain (Table 3). Theyrange from CAD-RADS 0 (absence of atherosclerosis)to CAD-RADS 5 (presence of at least one total occlu-sion) in both settings. Categories should reflect theclinically most relevant finding per patient. Figures 1through 9 provide examples of CAD-RADS categoriesand sub-categories. It is important to note that CAD-RADS classification is meant to be complementaryto the final impression of the report, particularlybecause the report will provide specific informationregarding the location and extent of coronary plaqueand stenosis.
CAD-RADS categories 4 and 5 require some furtherconsideration. For CAD-RADS 4, recommendationsmay vary depending onwhether the leftmain or severeobstructive three-vessel disease (>70%) is affected ornot. If a left main coronary artery stenosis greater than50% is suspected or if the examination demonstratesthree-vessel obstructive disease, then further evalua-tion with invasive angiography and possible revascu-larization is recommended. For this reason, CAD RADS
4 is sub-divided into A and B:
CA
CAMinimal calcified plaque in the proximal LAD with minimal luminal narrowing
(less than 25% diameter stenosis).
D RADS 4A – Single-vessel or two-vesselsdemonstrating severe stenosis (70-99%).
D RADS 4B – This indicates presence of leftmain stenosis greater than 50% or three-vessel
obstructive disease (>70%). Further evaluationwith ICA and possible revascularization isusually recommended.
The clinical relevance of CAD-RADS 5 (totalcoronary occlusion) varies widely depending onthe clinical context. It may be acute or chronic,
FIGURE 3 CAD-RADS 2
Predominantly calcified plaque in the proximal LAD with 25-49% diameter stenosis (left). Invasive coronary angiography confirming 25-49%
stenosis (right).
Cury et al. J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 9 , N O . 9 , 2 0 1 6
Coronary Artery Disease – Reporting and Data System S E P T E M B E R 2 0 1 6 : 1 0 9 9 – 1 1 3
1104
and, in the context of chronic occlusion, factorssuch as lesion length, calcification particularly atthe proximal cap, and degree of collateralizationmay be of relevance for management decisions(Figure 8).
3.2. PATIENTS WITH KNOWN CAD. Managementrecommendations with regard to patients with pre-viously known CAD deserve special consideration.The main clinical benefit of coronary CTA is derivedfrom its high sensitivity and negative predictivevalue. The positive predictive value of coronary CTAis lower, and especially intermediate lesions maybe overestimated regarding their relevance. Manypatients with previously known CAD will includelesions that fall into this category, so that coronary
FIGURE 4 CAD-RADS 3
Predominantly calcified plaque in the mid LCX with 50-69% diameter s
angiography.
CTA will need to be complemented by further tests.Additionally, coronary CTA has low accuracy fordiagnosis of in-stent restenosis, particularly in stentssmaller than 3.0 mm diameter. Thus, the use ofcoronary CTA in patients with previously knownCAD should be carefully considered. Managementdecisions derived from coronary CTA results dependon other clinical findings as well as the patient-specific previous history, and should be made onan individual basis.
3.3. MODIFIERS. CAD-RADS categories can be com-plemented by modifiers to indicate that a study is notfully evaluable or non-diagnostic (N) or to indicatethe presence of stents (S), grafts (G), and vulnerableplaque (V).
tenosis. Left image: Coronary CTA. Right image: Invasive coronary
FIGURE 5 CAD-RADS 4A
Focal non-calcified plaque in the mid LAD (yellow arrow) with 70-99% diameter stenosis (left). Invasive coronary angiography confirming
70-99% stenosis in the mid LAD (yellow arrow, right).
J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 9 , N O . 9 , 2 0 1 6 Cury et al.S E P T E M B E R 2 0 1 6 : 1 0 9 9 – 1 1 3 Coronary Artery Disease – Reporting and Data System
1105
I. Modifier N – Non-diagnostic study
“N” can be used as a modifier or as a CAD-RADScategory, depending on context. If the study isnot fully diagnostic (i.e. not all segments > 1.5 mmdiameter can be interpreted with confidence) and astenosis is present in a diagnostic segment, thehighest stenosis should be graded in addition tothe modifier N if CAD-RADS is greater than 3. For
FIGURE 6 CAD-RADS 4B
Three-vessel obstructive disease (>70% stenosis), including in 70-99%
LAD (middle) and 70-99% stenosis of the mid LCX (right).
example, a patient with moderate stenosis (50-69%)in one segment and one or more non-diagnosticremote segments should be graded as CAD-RADS
3/N (Figure 10) and not CAD-RADS N, since furtherevaluation is needed, possibly with functionalimaging, and patient recommendations for anti-ischemic and preventive management apply. How-ever, for a patient with no stenosis (zero), minimal
stenosis of the proximal RCA (left), 70-99% stenosis of the proximal
FIGURE 7 CAD-RADS 4B
Distal left main stenosis with circumferential calcified plaque resulting in > 50% stenosis (arrow). Upper left panel:
oblique longitudinal plane of the left main coronary artery. Lower left panel: cross-sectional slice of the distal left main
coronary artery. Figures on the right: Invasive coronary angiography confirming focal severe stenosis in the distal left main
coronary artery.
FIGURE 8 CAD-RA
Two examples of ca
the proximal RCA (
small focus of “orp
chronic total occlus
Cury et al. J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 9 , N O . 9 , 2 0 1 6
Coronary Artery Disease – Reporting and Data System S E P T E M B E R 2 0 1 6 : 1 0 9 9 – 1 1 3
1106
(1-24%), or no more than mild stenosis (25-49%)in interpretable segments, CAD-RADS N should beused since Coronary CTA cannot be used to guidepatient management and further evaluation toexclude obstructive coronary artery disease is stillneeded.
DS 5
ses coded as CAD-RADS 5. Left: Focal, non-calcified occlusion of
arrow). Right: Total occlusion of the proximal LCX (arrow). A
han” calcium along the distal LCX supports the diagnosis of
ion.
II. Modifier S - Presence of a stent
The modifier “S” indicates the presence of atleast one coronary stent anywhere in the coronarysystem. For example, if a patient has a patent stentin the proximal left anterior descending coronaryartery (LAD) with no significant in-stent restenosisor occlusion and demonstrates mild non-obstructivedisease (25-49%) in the left circumflex artery (LCX)and right coronary artery (RCA), the case would beclassified as: CAD-RADS 2/S. If a patient demon-strates significant in-stent restenosis of a stent inthe proximal LAD, then the case would be classifiedas: CAD-RADS 4A/S (Figure 11). Similarly, a non-stenotic stent in the LAD and a new severe steno-sis in the RCA would be classified as CAD-RADS
4A/S. Finally, if a stent were non-evaluable, thecase would be classified as CAD-RADS N/S if there isno other stenosis greater than 50% in the coronarytree. Note: CAD-RADS was created to guide man-agement recommendations, so it does not matterwhether it is the stent or a non-stented vessel thathas a severe stenosis. Rather, what matters is thatthe patient has a severe stenosis and needs furtherwork-up.
III. Modifier G ¼ Presence of coronary bypass
grafts:
The modifier “G” indicates the presence of at leastone coronary-artery bypass graft (Figure 12). A
FIGURE 9 CAD-RADS N
Motion artifacts obscuring the left main, LAD and LCX arteries, which
renders these segments non-diagnostic (left). Motion artifacts in the
mid RCA (right).
FIGURE 10 CAD-RADS 3/N
Motion artifact obscuring the mid RCA (left, arrow), which renders this segment non-
diagnostic. There is also stenosis of the mid LAD with 50-69% luminal narrowing (right,
arrow), qualifying this lesion as CAD RADS 3. Although the mid RCA segment is non-
diagnostic, the presence of suspected obstructive disease within the LAD should be coded
as CAD RADS 3/N. If the LAD lesion were mild (less than 50% diameter stenosis), and no
other plaques were identified, the patient would be coded as CAD RADS N.
J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 9 , N O . 9 , 2 0 1 6 Cury et al.S E P T E M B E R 2 0 1 6 : 1 0 9 9 – 1 1 3 Coronary Artery Disease – Reporting and Data System
1107
stenosis bypassed by a fully patent graft is notconsidered for the CAD-RADS classification. Forexample, if a patient has a graft to LAD, withabsence of significant stenoses in the graft, distalanastomosis and run-off vessel, and demonstratesnon-obstructive lesions (25-49%) in the LCX andRCA, in addition to the “expected” proximal LADsevere stenosis, then the case would be classifiedas: CAD-RADS 2/G. If a patient demonstrates totalocclusion of a saphenous vein graft (SVG) to theRCA, and a patent LIMA to LAD and SVG to LCX,then the case would be classified as: CAD-RADS 5/G.The interpretation is that a total occlusion is pre-sent and further investigation and/or managementmay be required.
IV. Modifier V ¼ Presence of “vulnerable” or
high-risk plaque features
Data from recent coronary CTA studies havedescribed vulnerable plaque characteristics that areindependently associated with future ACS. Theyinclude positive remodeling, low-attenuationplaque, spotty calcification, and the napkin-ring sign(23,24).
If a coronary plaque clearly demonstrates two ormore high-risk features by coronary CTA, themodifier “V” (vulnerability) should be added(Figures 13 and 14). High-risk features include: lowattenuation plaque (less than 30 Hounsfield Units),positive remodeling, spotty calcification, and the“napkin ring sign” (see Figure 13).
For example, CAD RADS 2/V should be used for apatient with diameter stenosis between 25-49% anddemonstrating plaque with two or more high-riskfeatures (large non-calcified plaque, positive remod-eling, spotty calcification, low HU values andnapkin ring sign) (Figure 14). The features shouldbe described, particularly in patients presentingto the emergency department with acute chestpain. There is not enough published data to guidethe management of such patients. However,clinical and laboratory correlation and close obser-vation is recommended. Consider hospital admissionin high-risk clinical settings. If the patient is dis-charged, short-term clinical follow-up within a weekis suggested in the outpatient setting with a cardiol-ogist or primary care physician.
Studies coded with CAD-RADS 3/V (the presenceof high risk plaque with 50-69% diameter stenosis,excluding left main lesions) should prompt consid-eration for more aggressive management thanstudies coded with CAD-RADS 3, particularly in pa-tients presenting to the emergency departmentwith acute chest pain. This includes consideration
of further testing with invasive coronary angiog-raphy instead of non-invasive functional testing.However, management decisions should ultimatelybe made on an individual basis taking intoconsideration all supporting clinical and laboratorydata.
FIGURE 11 CAD-RADS 4A/S
In-stent stenosis of the proximal LAD with significant luminal narrowing (70-99% stenosis). Grading of in-stent stenosis should follow the
grading of normal coronary arteries (0% stenosis, 1-24% stenosis, 25-49% stenosis, 50-69% stenosis, 70-99% stenosis, and >99% stenosis).
In this case, severe in-stent restenosis designates a CAD-RADS 4A lesion, which would be followed by the stent modifier “S.”
Cury et al. J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 9 , N O . 9 , 2 0 1 6
Coronary Artery Disease – Reporting and Data System S E P T E M B E R 2 0 1 6 : 1 0 9 9 – 1 1 3
1108
V. If more than one modifier is present, the symbol“/” (slash) should follow each modifier in thefollowing order:
i. First: modifier N (non-diagnostic)
ii. Second: modifier S (stent)
iii. Third: modifier G (graft)
FIGURE 12 MODIFIER G
Coronary CTA demonstrating a patent left internal mammary artery to th
second obtuse marginal branch. No stenoses or luminal narrowing throu
demonstrating patent LIMA graft to the LAD (right). When evaluating co
segments proximal to the graft anastamoses should not be evaluated for
artery segments distal to and including the anastomosis should be evalu
iv. Fourth: modifier V (vulnerability)
For example:
i. Non-interpretable coronary stent without evidenceof other obstructive coronary disease: Modifier S [
CAD-RADS N/S
e LAD and patent saphenous vein grafts to the ramus intermedius and
ghout the grafts (0% stenosis, left). Invasive coronary angiography
ronary CTA of patients with bypass grafts, the native coronary artery
purposes of CAD RADS coding. Only the grafts and the native coronary
ated for CAD RADS coding.
FIGURE 13 High-Risk Plaque Features on Coronary CTA
These include (A) Spotty calcium, defined as punctate calcium within a plaque; (B) “napkin
ring sign”, defined as central low attenuation plaque with a peripheral rim of higher CT
attenuation (arrows); (C) Positive remodeling, defined as the ratio of outer vessel diameter
at the site of plaque divided by the average outer diameter of the proximal and distal vessel
greater than 1.1, or Av/[(Ap þ Ad)/2] >1.1; and (D) Low attenuation plaque, defined as
non-calcified plaque with internal attenuation less than 30 HU. Please note that a combi-
nation of two or more high-risk features is necessary to designate the plaque as high-risk for
CAD-RADS.
J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 9 , N O . 9 , 2 0 1 6 Cury et al.S E P T E M B E R 2 0 1 6 : 1 0 9 9 – 1 1 3 Coronary Artery Disease – Reporting and Data System
1109
ii. Presence of stent and a new moderate stenosisshowing a plaque with high-risk features:Modifiers S and V [ CAD-RADS 3/S/V (Figure 15)
iii. Presence of stent, grafts and non-evaluablesegments due to metal artifacts: Modifiers S and
G [ CAD-RADS N/S/G
iv. Presence of patent LIMA to the LAD and expectedoccluded proximal LAD. Mild non-obstructivestenosis in the RCA and LCX. Modifier G [
CAD-RADS 2/G.
v. For a patient with severe stenosis (70-99%)in one segment and a non-diagnostic area inanother segment, the study should be graded asCAD-RADS 4/N.
3.4. PRESENCE OF OTHER CARDIAC OR EXTRA-
CARDIAC FINDINGS. Patients undergoing coronaryCTA may demonstrate other significant, potentiallysignificant or non-significant cardiac or extra-cardiacfindings. CAD-RADS is intended to focus solely onthe classification of coronary artery stenosis and fur-ther management. However, other cardiac and extra-cardiac findings of relevance should be reported incoronary CTA studies and should be mentioned in thereport text. Specific follow-up and recommendationsshould be included depending on the pathology.
Finally, Figure 16 provides a sample standardizedreporting template for coronary CTA incorporatingCAD-RADS coding.
4. DISCUSSION
The use of coronary CTA to assess patients with stablechest pain in the outpatient setting or acute chest painpresenting to the Emergency Department has beenvalidated in various clinical trials. Major guidelinesare incorporating the use of coronary CT angiographyas appropriate for assessing low to intermediate riskpatients presenting with chest pain. Decreasing thevariation in reporting is one aspect that will contributeto wider dissemination in clinical practice, minimizeerror and to ultimately improve patient outcome. Themain goal of the CAD-RADS classification system is topropose a reporting structure that provides consistentcategories for final assessment, along with sugges-tions for further management.
CAD-RADS is intended to be a “living document”that undergoes continued development to provide up-to-date, evidence based recommendations to achieveits goal of being a tool that imagers can use tocommunicate with clinicians and to convey concisefindings using unambiguous and standardized termi-nology. Next to its utilization in clinical reporting,CAD-RADS will allow reliable and reproducible data
FIGURE 14 CAD-RADS 2/V
Focal non-calcified plaque in the mid RCA with 25-49% diameter stenosis.
The plaque demonstrates two high risk features, low attenuation (<30 HU)
and positive remodeling, thus coding with the modifier “V.”
Cury et al. J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 9 , N O . 9 , 2 0 1 6
Coronary Artery Disease – Reporting and Data System S E P T E M B E R 2 0 1 6 : 1 0 9 9 – 1 1 3
1110
collection, storage and retrieval for future researchtrials and audits.
Similar to other larger registries, such as theNational Radiology Data Registry (NRDR) and Na-tional Cardiovascular Data Registry (NCDR), CAD-RADS can provide the framework for standardizecollection of coronary CTA reports across multiplesites for quality improvement and benchmarking.
FIGURE 15 CAD-RADS 3/S/V
Example demonstrating a patent stent in the proximal RCA (0% stenos
stenosis. In isolation, the proximal LAD lesion would be coded CAD RADS
RCA stent is present, this patient would be coded as CAD RADS 3/S/V.
Further, it can provide the framework for collectingoutcome data in each of several sub-categories ofCAD-RADS, such as:
1. Follow-up of disposition of patients with positivecoronary CTA results;
2. Rate of downstream testing;3. Correlation with ICA;4. Rate of revascularization (percutaneous coronary
intervention and coronary artery by-pass graftsurgery)
5. Major adverse cardiac events, including cardio-vascular death and myocardial infarct.
Therefore, it is strongly encouraged that everycoronary CTA examination includes the CAD-RADSclassification for a final assessment. Residency andFellowship trainees should be required to use theCAD-RADS terminology, assessment categories andmanagement recommendations.
Similar to BI-RADS, peer-reviewed radiology andcardiology journals may also find the CAD-RADS ter-minology useful for standardized classification ofcoronary CTA results, which in turn will furtherpromote the use of CAD-RADS nationally andinternationally.
Finally, standardization in reports and manage-ment recommendations will not only improve theclarity of communication and comprehension ofimaging results by all members of the clinical careteam, but also will improve communication betweenhumans and computer-based systems. This will allowthe development of decision support technologiesand serve as the basis for developing artificial intel-ligence algorithms.
is) with high-risk plaque in the proximal LAD resulting in 50-69%
3/V. However, since CAD RADS is coded on a per-patient basis, and a
FIGURE 16 Reporting Template
CLINICAL HISTORY:
COMPARISON:
TECHNIQUE: scanner type0.5
ACQUISITION: Prospective Retrospective
MEDICATIONS: 100mg of oral metoprolol was administered prior to scanning .4mg sublingual nitroglycerine was administered immediately prior to scanning
TECHNICAL QUALITY: excellent, with no artifacts good, with minor artifact but good diagnostic quality; acceptable, with moderate artifacts; poor/suboptimal, with severe artifacts
FINDINGS:
IMPRESSION:
Sample standardized reporting template for Coronary CTA incorporating CAD-RADS coding.
J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 9 , N O . 9 , 2 0 1 6 Cury et al.S E P T E M B E R 2 0 1 6 : 1 0 9 9 – 1 1 3 Coronary Artery Disease – Reporting and Data System
1111
5. CONCLUSION
In conclusion, CAD-RADS has been developed basedon scientific data, expert guidance from leadersin cardiac imaging and a multi-disciplinary effortinvolving radiology and cardiology societies (Societyof Cardiovascular Computed Tomography, AmericanCollege of Radiology, American College of Cardiologyand North American Society for Cardiac Imaging).
It is meant to be an evolving document that willundergo continuous updates as new data are ac-quired. The main goal of CAD-RADS is to createreport standardization terminology for coronary CTAresults, and to improve communication of resultsto referring physicians in a clear and consistentfashion with a final assessment and suggestionsfor further management. In addition, CAD-RADSwill provide a framework to standardize education,
Cury et al. J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 9 , N O . 9 , 2 0 1 6
Coronary Artery Disease – Reporting and Data System S E P T E M B E R 2 0 1 6 : 1 0 9 9 – 1 1 3
1112
research, peer-review, quality assurance and ulti-mately result in improvement to patient care. Finally,compiling imaging data in a standardized mannerwill allow to link imaging findings with specific treat-ments and to better assess the impact on patientoutcomes.
ADDRESS FOR CORRESPONDENCE: Ricardo C. Cury,M.D., FAHA, FSCCT, FACC, Miami Cardiac andVascular Institute, Baptist Hospital of Miami, 8900 N.Kendall Drive, Miami, FL 33176. T: 786-5962314.E-mail: [email protected].
RE F E RENCE S
1. Cury RC. President’s Page: Ten Years of Inno-vation in Cardiac CT. J Cardiovasc Comput Tomogr2014 Jul-Aug;8(4):338–9.
2. Leipsic J, Abbara S, Achenbach S, et al. SCCTguidelines for the interpretation and reporting ofcoronary CT angiography: A report of the Societyof Cardiovascular Computed Tomography Guide-lines Committee. J Cardiovasc Comput Tomogr2014 Sep-Oct;8(5):342–58.
3. Abbara S, Arbab-Zadeh A, Callister TQ, et al.SCCT guidelines for performance of coronarycomputed tomographic angiography: a report ofthe Society of Cardiovascular Computed Tomog-raphy Guidelines Committee. J Cardiovasc ComputTomogr 2009 May-Jun;3(3):190–204.
4. Halliburton SS, Abbara S, Chen MY, et al.,Society of Cardiovascular Computed Tomography.SCCT guidelines on radiation dose and dose-optimization strategies in cardiovascular CT.J Cardiovasc Comput Tomogr 2011 Jul-Aug;5(4):198–224.
5. Achenbach S, Delgado V, Hausleiter J,Schoenhagen P, Min JK, Leipsic JA. SCCT expertconsensus document on computed tomographyimaging before transcatheter aortic valve implan-tation (TAVI)/transcatheter aortic valve replace-ment (TAVR). J Cardiovasc Comput Tomogr 2012Nov-Dec;6(6):366–80.
6. Taylor AJ, Cerqueira M, Hodgson JM, et al.ACCF/SCCT/ACR/AHA/ASE/ ASNC/NASCI/SCAI/SCMR 2010 Appropriate Use Criteria for CardiacComputed Tomography. A Report of the AmericanCollege of Cardiology Foundation Appropriate UseCriteria Task Force, the Society of CardiovascularComputed Tomography, the American Collegeof Radiology, the American Heart Association,the American Society of Echocardiography, theAmerican Society of Nuclear Cardiology, the NorthAmerican Society for Cardiovascular Imaging,the Society for Cardiovascular Angiography andInterventions, and the Society for Cardiovascu-lar Magnetic Resonance. J Am Coll Cardiol 2010Nov 23;56(22):1864–94.
7. Patel MR, White RD, Abbara S, et al., AmericanCollege of Radiology; American College of Cardi-ology Foundation. 2013 ACCF/ACR/ASE/ASNC/SCCT/SCMR appropriate utilization of cardiovas-cular imaging in heart failure: a joint report of theAmerican College of Radiology AppropriatenessCriteria Committee and the American College ofCardiology Foundation Appropriate Use CriteriaTask Force. J Am Coll Cardiol 2013 May 28;61(21):2207–31.
8. Wolk MJ, Bailey SR, Doherty JU, et al. AmericanCollege of Cardiology Foundation Appropriate Use
Criteria Task Force. ACCF/AHA/ASE/ASNC/HFSA/HRS/SCAI/SCCT/SCMR/STS 2013 multimodalityappropriate use criteria for the detection and riskassessment of stable ischemic heart disease: areport of the American College of CardiologyFoundation Appropriate Use Criteria Task Force,American Heart Association, American Society ofEchocardiography, American Society of NuclearCardiology, Heart Failure Society of America,Heart Rhythm Society, Society for CardiovascularAngiography and Interventions, Society of Car-diovascular Computed Tomography, Society forCardiovascular Magnetic Resonance, and Societyof Thoracic Surgeons. J Am Coll Cardiol 2014 Feb4;63(4):380–406.
9. Sickles EA, D’Orsi CJ, Bassett LW, et al. ACRBI-RADS� Mammography. In: ACR BI-RADS�
Atlas, Breast Imaging Reporting and Data Sys-tem. Reston, VA: American College of Radiology;2013.
10. Mitchell DG, Bruix J, Sherman M, Sirlin CB.LI-RADS (Liver Imaging Reporting and DataSystem): Summary, discussion, and consensus ofthe LI-RADS Management Working Group andfuture directions. Hepatology 2015 Mar;61(3):1056–65.
11. Kazerooni EA, Armstrong MR, Amorosa JK,et al. ACR CT Accreditation Program and the LungCancer Screening Program Designation. Journal ofthe American College of Radiology : JACR 2015;12:38–42.
12. Prostate Cancer Localization Using Multi-parametric MR Imaging: Comparison of Pros-tate Imaging Reporting and Data System (PI-RADS) and Likert Scales. Radiology 2013;269:482–92.
13. Raff GL, Chinnaiyan KM, Cury RC, et al. SCCTguidelines on the use of coronary computedtomographic angiography for patients presentingwith acute chest pain to the emergency depart-ment: A Report of the Society of CardiovascularComputed Tomography Guidelines Committee.J Cardiovasc Comput Tomogr 2014 Jul-Aug;8(4):254–71.
14. Cury RC, Feuchtner GM, Batlle JC, et al. Triageof patients presenting with chest pain to theemergency department: implementation of coro-nary CT angiography in a large urban health caresystem. AJR American journal of roentgenology2013;200(1):57–65.
15. Douglas PS, Hoffmann U, Patel MR, Mark DB,Al-Khalidi HR, Cavanaugh B, Cole J, Dolor RJ,Fordyce CB, Huang M, Khan MA, Kosinski AS,Krucoff MW, Malhotra V, Picard MH, Udelson JE,Velazquez EJ, Yow E, Cooper LS, Lee KL, PROMISE
Investigators. Outcomes of anatomical versusfunctional testing for coronary artery disease.N Engl J Med 2015 Apr 2;372(14):1291–300.
16. SCOT-HEART investigators. CT coronaryangiography in patients with suspected anginadue to coronary heart disease (SCOT-HEART): anopen-label, parallel-group, multicentre trial.Lancet 2015 Mar 13. pii: S0140-6736(15)60291–4.
17. Goldstein JA, Chinnaiyan KM, Abidov A,Achenbach S, Berman DS, Hayes SW, Hoffmann U,Lesser JR, Mikati IA, O’Neil BJ, Shaw LJ, Shen MY,Valeti US, Raff GL. The CT-STAT (CoronaryComputed Tomographic Angiography for Sys-tematic Triage of Acute Chest Pain Patients toTreatment) trial. J Am Coll Cardiol 2011;58:1414–22.
18. Litt HI, Gatsonis C, Snyder B, Singh H,Miller CD, Entrikin DW, Leaming JM, Gavin LJ,Pacella CB, Hollander JE. CT angiography forsafe discharge of patients with possible acutecoronary syndromes. N Engl J Med 2012;366:1393–403.
19. Hoffmann U, Truong QA, Schoenfeld DA,Chou ET, Woodard PK, Nagurney JT, Pope JH,Hauser TH, White CS, Weiner SG, Kalanjian S,Mullins ME, Mikati I, Peacock WF, Zakroysky P,Hayden D, Goehler A, Lee H, Gazelle GS,Wiviott SD, Fleg JL, Udelson JE. Coronary CTangiography versus standard evaluation in acutechest pain. N Engl J Med 2012;367:299–308.
20. Hamilton-Craig C, Fifoot A, Hansen M,Pincus M, Chan J, Walters DL, Branch KR.Diagnostic performance and cost of CT angi-ography versus stress ECG–a randomized pro-spective study of suspected acute coronarysyndrome chest pain in the emergency depart-ment (CT-COMPARE). Int J Cardiol 2014 Dec20;177(3):867–73.
21. Cury RC, Feuchtner G, Battle J, Pena CS,Janowitz WR, Katzen BT, Ziffer JA. Triage ofPatients Presenting with Chest Pain to theEmergency Department: Implementation of Cor-onary CTA in a Large Urban Hospital HealthcareSystem. Am J Roentgenol 2013 Jan;200(1):57–65.
22. Poon M, Cortegiano M, Abramowicz AJ,et al. Associations between routine coronarycomputed tomographic angiography andreduced unnecessary hospital admissions, lengthof stay, recidivism rates, and invasive coronaryangiography in the emergency departmenttriage of chest pain. J Am Coll Cardiol 2013;62(6):543–52.
J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 9 , N O . 9 , 2 0 1 6 Cury et al.S E P T E M B E R 2 0 1 6 : 1 0 9 9 – 1 1 3 Coronary Artery Disease – Reporting and Data System
1113
23. Motoyama S, Sarai M, Harigaya H, et al.Computed tomographic angiography characteris-tics of atherosclerotic plaques subsequentlyresulting in acute coronary syndrome. J Am CollCardiol 2009;54:49–57.
24. Puchner SB, Liu T, Mayrhofer T, et al. High-riskplaque detected on coronary CT angiographypredicts acute coronary syndromes independent of
significant stenosis in acute chest pain: resultsfrom the ROMICAT-II trial. J Am Coll Cardiol 2014;64:684–92.
25. Fihn SD, Gardin JM, Abrams J, et al. 2012ACCF/AHA/ACP/AATS/PCNA/SCAI/STS Guidelinefor the diagnosis and management of patientswith stable ischemic heart disease: a report ofthe American College of Cardiology Foundation/
American Heart Association Task Force on Prac-tice Guidelines, and the American College ofPhysicians, American Association for ThoracicSurgery, Preventive Cardiovascular Nurses As-sociation, Society for Cardiovascular Angiog-raphy and Interventions, and Society of ThoracicSurgeons. J Am Coll Cardiol 2012;60(24):e44–164.
