JACC: CARDIOVASCULAR IMAGING V O L . 15, NO.

1, 2022
ª 2022 THE AUTHORS. PUBLISHED BY ELSEVIER ON BEHALF OF THE AMERICAN

COLLEGE OF CARDIOLOGY FOUNDATION. THIS IS AN OPEN ACCESS ARTICLE UNDER

THE CC BY LICENSE (http://creativecommons.org/licenses/by/4.0/).

ORIGINAL RESEARCH

Dynamic Myocardial Perfusion CT for the

Detection of Hemodynamically
Significant Coronary Artery Disease
Fay M.A. Nous, MD,a,b Tobias Geisler, MD, cMariusz B.P. Kruk, MD, P D,HHatem
d Alkadhi, MD, e

KakuyaKitagawa,MD,fRozemarijnVliegenthart,MD,PHD,gMichaelaM.Hell,MD,hJörgHausleiter,MD,i
PatriciaK.Nguyen,MD,j,k,lRicardoP.J.Budde,MD,PHD,a,b KonstantinNikolaou,MD,MBA,m
CezaryKepka,MD,PHD,dRobertManka,MD,nHajimeSakuma,MD,oSachinB.Malik,MD,p,qAdriaanCoenen,MD,a,b
FelixZijlstra,MD,PHD ,bErnstKlotz,DIPLPHYS,rPi mvanderHarst,MD,PHD,s ChristophArtzner,MD,c
AdmirDedic,MD,P,HD,bFrancescaPugliese,MD,PHD,tuFabianBamberg,MD,PHD,v,*KoenNieman,MD,PHDa,b,w,*

ABSTRACT

OBJECTIVES In this international, multicenter study, using third-generation dual-source computed tomography (CT), we
investigated the diagnostic performance of dynamic stress CT myocardial perfusion imaging (CT-MPI) in addition to cor-
onary CT angiography (CTA) compared to invasive coronary angiography (ICA) and invasive fractional flow reserve (FFR).

BACKGROUND CT-MPIcombinedwithcoronaryCTAintegratescoronaryarteryanatomywithinduciblemyocardialischemia,
showing promising results for the diagnosis of hemodynamically significant coronary artery disease in single-center studies.

METHODS At 9 centers in Europe, Japan, and the United States, 132 patients scheduled for ICA were enrolled; 114
patients successfully completed coronary CTA, adenosine-stress dynamic CT-MPI, and ICA. Invasive FFR was performed in
vessels with 25% to 90% stenosis. Data were analyzed by independent core laboratories. For the primary analysis, for
each coronary artery the presence of hemodynamically significant obstruction was interpreted by coronary CTA with CT-
MPI compared to coronary CTA alone, using an FFR of #0.80 and angiographic severity as reference. Territorial absolute
myocardial blood flow (MBF) and relative MBF were compared using C-statistics.

RESULTS ICA and FFR identified hemodynamically significant stenoses in 74 of 289 coronary vessels (26%). Coronary
CTA with $50% stenosis demonstrated a per-vessel sensitivity, specificity, and accuracy for the detection of hemody-
namically significant stenosis of 96% (95% CI: 91%-100%), 72% (95% CI: 66%-78%), and 78% (95% CI: 73%-83%),
respectively. Coronary CTA with CT-MPI showed a lower sensitivity (84%; 95% CI: 75%-92%) but higher specificity
(89%; 95% CI: 85%-93%) and accuracy (88%; 95% CI: 84%-92%). The areas under the receiver-operating characteristic
curve of absolute MBF and relative MBF were 0.79 (95% CI: 0.71-0.86) and 0.82 (95% CI: 0.74-0.88), respectively. The
median dose-length product of CT-MPI and coronary CTA were 313 mGy$cm and 138 mGy$cm, respectively.

CONCLUSIONS Dynamic CT-MPI offers incremental diagnostic value over coronary CTA alone for the identification of
hemodynamically significant coronary artery disease. Generalized results from this multicenter study encourage broader
consideration of dynamic CT-MPI in clinical practice. (Dynamic Stress Perfusion CT for Detection of Inducible Myocardial
Ischemia [SPECIFIC]; NCT02810795) (J Am Coll Cardiol Img 2022;15:75–87) © 2022 The Authors. Published by Elsevier
on behalf of the American College of Cardiology Foundation. This is an open access article under the CC BY license
(http://cre ativecommons.org/l icenses/by/4.0/ ).

From the aDepartment of Radiology and Nuclear Medicine, Erasmus University Medical Center, University Medical Center
Rotterdam, Rotterdam, the Netherlands; bDepartment of Cardiology, Erasmus University Medical Center, University Medical
Center Rotterdam, Rotterdam, the Netherlands; cDepartment of Cardiology, University of Tuebingen, Tuebingen, Germany;
d Coronary Disease and Structural Heart Diseases Department, Institute of Cardiology, Warsaw, Poland; eInstitute of Diagnostic
and Interventional Radiology, University Hospital Zurich, University of Zurich, Zurich, Switzerland; fDepartment of Advanced
Diagnostic Imaging, Mie University Graduate School of Medicine, Tsu, Japan; gDepartment of Radiology, University Medical

ISSN 1936-878X https: //doi.org/10.1016/j.jcm g.2021.07.021

 Nous et al JACC: CARDIOVASCULAR IMAGING, VOL. 15, NO. 1, 2 0 2 2

Diagnostic Performance of Dynamic Perfusion CT JANUARY 2022:75–87

Coronarycomputed
ABBREVIATIONS METHODS
AND ACRONYMS

STUDY DESIGN.The SPECIFIC (Dynamic Stress

AUC = area under the receiver-
operating characteristic curve
Perfusion CT for Detection of Inducible Myocardial
CAD = coronary artery disease onary artery disease (CAD) in various Ischemia) study is an international, multicenter,
CT = computed tomography clinical settings. However, coronary CTA prospective, observational cohort study designed to
tends to overestimate angiographic investigate the diagnostic accuracy of dynamic CT-
CTA = computed tomography
angiography severity, and it cannot measure functional MPI (NCT02810795). Study participants were
CT-MPI = computed significance (1,2). Hence, clinical manage- recruited at 9 hospitals in Europe, Japan, and the
tomography myocardial
perfusion imaging ment decisions often require further func- United States. The study protocol was compliant with
tional testing (3). New techniques such as the Declaration of Helsinki and received approval
FFR = fractional flow reserve
CT-derived fractional flow reserve (CT- from the research ethics committee at each institu-
ICA = invasive coronary
angiography FFR) and CT myocardial perfusion imaging tion. All participants provided written informed
MBF = myocardial blood ow
(CT-MPI) consent.
Dynamic may address
CT-MPI this limitation
can quantify (4,5).
myocardial
fl
blood flow (MBF) during pharmacologic hy- STUDY POPULATION. Symptomatic patients aged
NPV = negative predictive
value peremia from the myocardial enhancement $ 21 years with suspected stable CAD and scheduled
PPV = positive predictive value patterns after injection of contrast medium for ICA were eligible for the study. Study exclusion
QCA = quantitative coronary (6). Absolute measures of MBF offer poten- criteria were as follows: 1) hemodynamically unstable
angiography
condition; 2) prior myocardial infarction; 3) coronary
tial advantages to quantify the ischemia
bypass surgery; 4) percutaneous coronary interven-
severity and identify balanced ischemia. Despite
tion for myocardial infarction; 5) significant other
favorable diagnostic performance in single-center
cardiovascular diseases affecting CT-MPI perfor-
studies (5,7,8), CT-MPI has not yet found wide-
mance (eg, heart failure, severe valvular regurgita-
spread clinical use because of the relative
tion); 6) estimated glomerular filtration rate
complexity of the test and radiation exposure. The
latest-generation CT scanners offer higher spatial of < 60mL/kg/min;7)bodymassindexof>35kg/m2;
and temporal resolution with wider detector arrays 8) atrial fibrillation or other significant arrhythmias
and lower radiation doses, providing a more effec- ( >6 ectopic beats/min); 9) allergy to iodinated
tive imaging approach (9,10). This prospective inter- contrast medium; 10) pregnancy; and 11) contraindi-
cations to adenosine. Patients were excluded from
national multicenter study aims to evaluate the
the analysis if CT-MPI, coronary CTA, or ICA was not
diagnostic performance of dynamic CT-MPI in addi-
performed.
third-generation
tion to coronary CTA by using
IMAGINGPROTOCOL.
dual-source CT compared to invasive coronary angi- Patientpreparation.Patients
ography (ICA) and invasive FFR as the reference underwent a noncontrast scan, followed by CT-MPI,
standard.

Center Groningen, University of Groningen, Groningen, the Netherlands; hDepartment of Cardiology, Faculty of Medicine, Frie-
drich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany; iDepartment of Cardiology, Ludwig-Maximilians University,
Munich, Germany; jVeterans Affairs Palo Alto Healthcare System, Cardiology Section, Palo Alto, California, USA; kStanford Uni-
versity, Division of Cardiovascular Medicine, Stanford, California, USA; lStanford Cardiovascular Institute, Stanford, California,
USA; mDepartment of Radiology, University Hospital of Tübingen, Tübingen, Germany; nDepartment of Cardiology, University
Heart Center and Institute of Diagnostic and Interventional Radiology, University Hospital Zurich, University of Zurich, Zurich,
Switzerland; oDepartment of Radiology, Mie University Graduate School of Medicine, Tsu, Japan; pVeterans Affairs Palo Alto
Healthcare System, Thoracic and Cardiovascular Imaging Section, Palo Alto, California, USA; qStanford University, Division of
CardiovascularImaging(Affil iated),Stanford,California,USA;rS iemensHealthineers,Forcheim,Germany;sDepartmentofCar-
diology, University Medical Center Groningen, University of Groningen, Groningen, the Netherlands; tCentre for Advanced Car-
diovascular Imaging, William Harvey Research Institute, Barts National Institute for Health Research Biomedical Research
Centre, Queen Mary University of London, London, United Kingdom; uBarts Heart Centre, St Bartholomew’s Hospital, Barts
Health Na- tionalHealthServiceTrust,WestSmith field,London,UnitedKingdom;vDepartmentofRadiology,MedicalCenter-
Universityof Freiburg, Faculty of Medicine, University of Freiburg, Freiburg, Germany; and the wStanford University School of
Medicine and Cardiovascular Institute, Stanford, California, USA. *Dr Bamberg and Dr Nieman contributed equally to this work.
Harvey Hecht, MD, served as Guest Editor for this paper. The authors attest they are in compliance with human studies
committees and animal welfare regulations of the authors’ institutions and Food and Drug Administration guidelines, including
patient consent where appropriate. For more information, visit the Author Center.

Manuscript received March 1, 2021; revised manuscript received July 14, 2021, accepted July 21, 2021.
JACC: CARDIOVASCULAR IMAGING, VOL. 15, NO. 1, 2022 Nous et al
JANUARY 2022:75–87 Diagnostic Performance of Dynamic Perfusion CT

F I G U R E 1 Study Protocol and Analysis

A Non-contrast scan CT-MPI CCTA

≥3 min 5 min

Adenosine - Nitroglycerin
infusion - Beta-blockers
(140 ug/kg/min) (if needed)

B HU HU C
AIF
ml/100 ml/
min 284.1

TAC

Time

Max Slope Tissue Attenuation Curve (TAC) 0.0

MBF = Max. Atrial Input Function (AIF)

(A) Dynamic stress CT-MPI and coronary CTA study protocol. (B) CT-MPI postprocessing: AIF curve and TAC to calculate MBF. (C) CT-MPI
analysis: volumes of interest (circles) placed on a color-coded polar map. AIF
¼ atrial input functional; CT-MPI ¼ computed tomography
myocardial perfusion imaging; CTA ¼ computed tomography angiography; CCTA ¼ coronary computed tomography angiography;
HU ¼ Hounsfield units; MBF ¼ myocardial blood flow; TAC ¼ time-attenuation curve.

and coronary CTA on a third-generation dual-source CT time, 66-ms temporal resolution, and tube voltage of
scanner (SOMATOM Force, Siemens Healthineers) 70-80 kV using the automated exposure control (300
(Figure 1A). Patients were asked to refrain from mAs/rotation at 80 kV as reference). The 3.0-mm-
caffeine-containing beverages for 12 hours and nico- thick slices were reconstructed with 2.0-mm overlap.
tine for 3 hours before the examination. Sublingual CT- MPI data were evaluated at an independent core
lab- oratory (Centre of Advanced Cardiovascular
nitroglycerin was given before coronary CTA, as well as
intravenous beta-blockers if the heart rate was >75 Imaging, Barts Cardiovascular Biomedical Research
beats/min. Center, London, United Kingdom). Image quality was
assessed using a 4-point Likert scale. CT-MPI images
Dynamic stress CT-MPI.Hyperemia was
induced by with poor image quality were excluded from the
intravenous adenosine (140 mg/kg/min) over $3 mi-analysis.
nutes. The standard contrast injection protocol was a Coronary CTA.Coronary CTA scan was acquired
45-mL contrast bolus at 5.5 mL/s (iopromide, Bayer) 5 minutes after CT-MPI using prospective electrocar-
(370 mg/mL), followed by 40 mL saline, with minor diogram-triggered axial or high-pitch spiral scans.
modification at 2 sites because of availability. The CT- Tube current and voltage were (semi)automatically
MPI scan started 4 seconds after contrast injection, selected based on body size. Scan timing was deter-
using alternating table positions (shuttle mode) for mined with a 10-mL contrast test bolus plus 40 mL
complete myocardial coverage. The data set consisted saline or using bolus tracking. For coronary CTA, the
of 10-15 CT data samples over 30 seconds. The cardiac contrast volume was 65 (IQR: 55-75) mL, injected at
rhythm was continuously monitored, and the blood 5.0 (IQR: 4.9-5.4) mL/s with a 40-mL saline bolus
pressure was measured at regular intervals. The CT- chaser. Images were reconstructed with a medium-
MPI scan parameters were as follows: 2 smooth kernel, 0.6-mm slice thickness, and 0.4-mm
96 0.6-
mm collimation resulting in a 105-mm z-axis increment. For 34 patients, adequate-quality coro-
nary CTA was clinically performed within 4 months of
coverage by shuttle mode, 250-ms gantry rotation
Nous et al JACC: CARDIOVASCULAR IMAGING, VOL. 15, NO. 1, 2 0 2 2

Diagnostic Performance of Dynamic Perfusion CT JANUARY 2022:75–87

study enrollment. coronary

In these CTA
patients,
was notthe
performed.
research compromised. The most severely affected coronary
The coronary CTA data were transferred to a branch determined per-territory disease classifica-
coronary CTA core labora- tory (University of tion. To calculate MBF per-vessel territory, a region of
Tubingen, Tubingen, Germany). Coronary CTA images interest (corresponding to $0.5 cm3 of sub-
assessed as poor quality (6 vessels in 4 patients) endocardial myocardium) was sampled onto the MBF
were not excluded but were classified as positive for polar maps for each vessel territory, either in the
obstructive CAD. area of suspected ischemia or centrally within
CT-MPI data postprocessing.CT-MPI source images territories without suspected ischemia. The reference
were processed using commercial software (Syngo.CT MBF was defined as the 75th percentile of the
Myocardial Perfusion, Siemens Healthineers). A mo- automatically generated global endocardial MBF,
tion correction algorithm was applied to align the which represents a robust measure of normal MBF in
serial samples. The left ventricular myocardium was a specific patient and a specific examination that is
isolated using a method of blood pool removal based relatively unaf-
fected
ischemia
by or
territorial
artifacts (15). The
on attenuation value thresholds. The arterial input relative MBF was calculated per vessel territory as the
function was derived from attenuation values absolute MBF divided by the reference MBF.
measured in the descending aorta sampled in both ICA and FFR. ICA was performed following local stan-
dynamic image stacks. Time-attenuation curves were dards with a median of 2 days (IQR: 1-6 days) after CT-
created for each myocardial volumetric element MPI and 3 days (IQR: 1-23 days) after coronary CTA. By
(voxel) within the left ventricle volume of interest.
protocol, intermediate coronary lesions with visual
Dedicated parametric deconvolution based on a 2-
diameter stenoses of 25% to 90% were interrogated by
compartment model of intra- and extravascular space
FFR, if considered technically feasible and safe by the
was applied to fit the time-attenuation curves and
operator. An FFR pressure wire was positioned distal
compute MBF (11). MBF was calculated as the ratio
between the maximum slope of the fit curve and the to the stenosis of interest, after which hyperemia was
peak arterial input function (Figure 1B). The data were induced by intravenous adenosine at 140 mg/kg/min.
then processed using prototype software (Cardiac ICA images and FFR data were evaluated by an ICA core
laboratory (Erasmus Medical Center, Rotterdam, the
Functional Analysis Prototype, Siemens Healthi- Netherlands) and an FFR core laboratory (University
neers) to automatically segment the left ventricle
the
Medical Center Groningen, Groningen,
based on a heart model (12) and generate 17-segment
Netherlands) for independent reading, blinded to the
polar maps representing the MBF distribution within
CT findings. Quantitative coronary angiography (QCA)
the subendocardial layer of the left ventricular
software (Caas, Pie Medical Imaging) was used to
myocardium (Figure 1C) (13).
measure the angiographic stenosis severity in all cor-
Integration of coronary CTA and CT-MPI data.A onary segments with a diameter of >1.5 mm. Hemo-
comprehensive coronary CTA and CT-MPI core labo- dynamically significant CAD was defined as an FFR
ratory reading was performed by Christoph Artzner of #0.80, or angiographic stenosis severity of >90% if
(coronary CTA core laboratory), Francesca Pugliese
FFR could not be performed. The absence of hemody-
(CT-MPI core laboratory), and Koen Nieman (principal
namically significant disease was defined as an FFR of
investigator) to visually match the coronary anatomy
>0.80, or angiographic stenosis of <25% if FFR was not
with the subtended myocardial territories and assess
performed. Numerous studies have demonstrated that
myocardial hypoperfusion per coronary branch. The
visual interpretation overestimates tight stenoses and
readers were blinded to the ICA and FFR results. First,
underestimates mild stenoses when compared to QCA
coronary stenoses were classified per vessel following
(16). In addition, a threshold of 70% stenosis by QCA
Society of Cardiovascular Computed To- mography
has shown a 98% specificity for the presence of FFR-
criteria (14). Second, CT-MPI maps were used for side-
positive CAD (17). Therefore, very severe stenosis
by-side comparison to the coronary CTA images. The
(>90%) or the absence of stenosis (<25%) interpreted
patient’s coronary anatomy on coronary CTA was
by the clinical operators at the time of the catheteri-
used to assign myocardial perfusion defects to
zation required QCA confirmation by the ICA core
specific coronary vessels. Based on the interpreta-
laboratory of at least >70% stenosis or <40% stenosis,
tion of available coronary CTA and CT-MPI images,
respectively. Vessels with intermediate stenosis and
the presence of hemodynamically significant CAD was
no FFR were excluded from the analysis.
determined per vessel territory. If coronary CTA and
CT-MPI findings were discordant, then myocar- STATISTICAL ANALYSIS. Continuous variables are
presented as mean
dial perfusion overruled coronary CTA stenosis categorical variables areSD
given as frequencies
or median (IQR) and and
severity, unless CT-MPI image quality was
Nous et al JACC: CARDIOVASCULAR IMAGING, VOL. 15, NO. 1, 2 0 2 2

Diagnostic Performance of Dynamic Perfusion CT JANUARY 2022:75–87

narrow CIs (<0.15). Statistical analyses were per-

TABLE 1 Patient Demographics
formed using SPSS version 25 (IBM Corp) and R (R
Age, y
64 8 Core Team 2019, version 3.6.2, DTComPair package).
Men
75 (66) Med-Calc version 19.5.3 (MedCalc Software) was used
Body mass index, kg/m2
26 4
Risk factors to compare the AUC. A P value of <0.05 was consid-
ered statistically significant.
Current or previous smoker 61 (54)
RESULTS
Diabetes mellitusa 22 (19)
Dyslipidemiaa 83 (73)
Hypertensiona 81 (71)
Family history of coronary artery diseaseb 59 (52) STUDY POPULATION.Between July 2016 and
Previous percutaneous coronary intervention 5(4) September 2019, 132 patients were enrolled, of whom
Symptoms 123 completed all examinations (Figure 2). No severe
Typical angina 39 (34) cardiac events or study-related complications were
Atypical angina 35 (31) encountered during CT-MPI and invasive FFR. Image
Nonanginal symptom 40 (35) quality of CT-MPI was adequate for analysis in 114
ICA and FFR patients (93%). The mean age was 64
Patients with coronary lesion causing ischemia, %c 54/111
8 years, 66%
were men, 39 (34%) patients had typical angina
Single-vessel disease, % 37 (33)
2-vessel disease, % 14 (13) symptoms, and 5 (5%) patients had previously un-
3-vessel disease,% 3(3) stable
dergone CAD
stenting
(Table for
1). After
Number of vessels evaluated 289 exclusion of 53 vessels with indeterminate hemody-
Vessels with stenosis on QCA of $50% Vessels 84 (29) namically significance of disease by ICA/FFR, 289
with stenosis on QCA of $70% Vessels with 29 (10) coronary territories in 111 patients were available for
coronary lesion causing ischemiac 74 (26)
the primary analysis. Functionally significant steno-
Right coronary artery 18 (6)
sis was present in 74 vessels (26%) and 54 patients
Left main/left anterior descending coronary artery 41 (14)
(49%) based on an FFR of #0.80 (n
Left circumflex artery 15 (5) ¼ 20). Of¼these,
54) or37
very se-
(33%)
CT-MPI vere angiographic stenosis (n
had single-vessel, 14 (13%) had 2-vessel, and 3 (3%)
Heart rate baseline, beats/min 66 (60-75)
had 3-vessel disease of hemodynamic significance.
Heart rate during adenosine stress, beats/min 83 (70-93)
Functionally significant stenosis was absent in 215
Image quality
(74%) vessels and 57 (51%) patients based on an FFR
Excellent 66 (58)
Good 39 (34)
of >0.80 (n
Moderate 9(8) ¼ 141). The median dose-length products ofstenosis
¼ 74) or absent angiographic CT-MPI
(n
Dose-length product, mGy$cm 313 (237-448) and coronary CTA were 313 mGy$cm (IQR: 237-448)
Coronary CTA and 138 mGy$cm (IQR: 76-280), respectively.
Beta-blocker administered 37 (32) DIAGNOSTIC PERFORMANCE OF CORONARY CTA
Image quality Coronary CTA showed coronary calcium
Excellent 60 (53)
Good 40 (35) AND CT-MPI.
Moderate 10 (9) in 95 (83%) patients, stenoses of $50% in 131 vessels in
Poor 4(4) 78 patients, and stenoses of $70% in 45 vessels in 37
Dose-length product, mGy$cm 138 (76-280) patients. CT-MPI showed 85 ischemic territories in 60
patients (Figure 3). Coronary CTA with stenosis
Values are mean SD, n (%), or median (IQR). aBased on medication use. bFamily history of
coronary artery disease having first- or second-degree relatives with premature coronary artery of $50% demonstrated a per-vessel sensitivity, speci-
disease (age: 55 y). cFunctionally significant coronary lesion defined as FFR of #0.80 or visual ficity, PPV, NPV, and accuracy for the detection of
diameter narrowing of $90% combined with a QCA of $70%.
hemodynamically significant stenosis of 96% (95% CI:
CTA ¼ computedtomographyangiography;CT-MPI ¼ computedtomographymyocardial
invasive coronary angiography;
perfusion imaging; FFR ¼ fractional flow reserve; ICA ¼ 91%-100%), 72% (95% CI: 66%-78%), 54% (95% CI:
quantitativecoronaryangiography.
QCA ¼ 46%-63%), 98% (95% CI: 96%-100%), and 78% (95% CI:
73%-83%), respectively (Table 2, Central Illustration).
Stenosis of $70% on coronary CTA demonstrated
the Kruskal-Wallis test and Mann-Whitney U test.
higher specificity (94% vs 72%) but lower sensitivity
Differences in image quality of dynamic CT-MPI be-
(45% vs 96%) for the detection of hemodynamically
tween experienced (>50 scans) and inexperienced
centers (#15 scans) were tested using the chi-square significant stenosis. Coronary CTA with CT-MPI
test. Based on a predicted rate of 1.5 stenosed ves- sels demonstrated a per-vessel sensitivity, specificity,
per patient and a 50% functionally significance rate by PPV, NPV, and accuracy for the detection of hemody-
FFR, we determined that 120 cases would result in a namically significant stenosis of 84% (95% CI: 75%-
sensitivity and specificity with acceptably 92%), 89% (95% CI: 85%-93%), 73% (95% CI: 63%-83%),
JACC: CARDIOVASCULAR IMAGING, VOL. 15, NO. 1, 2022 Nous et al
JANUARY 2022:75–87 Diagnostic Performance of Dynamic Perfusion CT

F I G U R E 3 Case Examples

A B C D

LCX

LAD
Dg
LAD
LAD

E F G H

J K L M

LCX

LAD
LAD LAD

Case 1: (A) Discrete narrowing in the LAD on CT (coronary CTA, arrows) and (E, F) an apical defect by perfusion imaging (CT-MPI, arrows) with (J) concordant ICA and an
FFR of 0.76. The color bar in A displays the myocardial blood flow range from normal (red) to low (green and blue). (B) The same patient had a second stenosis in the
LCX, with (E, F) a posterolateral perfusion defect (arrowheads), concordant with (K) ICA and FFR of 0.74. Case 2: (C) Diffuse, partially calcified narrowing and focal
dilatation in the LAD on coronary CTA and a (G) CT-MPI perfusion defect in the distal septum and apex, confirmed by (L) ICA and an FFR of 0.56. Case 3: (D) Coronary
CTA shows severely calcified plaque of uncertain angiographic stenosis severity in the LAD and a predominantly noncalcified severe stenosis in a large Dg. (H) There is
a distinct anterolateral perfusion defect subtended by the Dg (arrow) but normal blood flow in the LAD territory. (M) ICA confirms the severe Dg stenosis (FFR: 0.68)
and functionally nonsignificant, moderate mid-LAD stenosis (FFR: 0.83). CT
¼ computed tomography; CT-MPI ¼ computed tomography myocardial perfusion imaging;
CTA ¼ computed tomography angiography; Dg ¼ diagonal branch; FFR ¼ fractional flow reserve; ICA ¼ invasive coronary angiography; LAD ¼ left anterior descending
coronary artery; LCX ¼ left circumflex coronary artery.

94% (95% CI: 91%-97%), and 88% (95% CI: 84%-92%), than coronary CTA with stenosis of $70% (94%; P <
respectively. Coronary CTA with CT-MPI demon- 0.05). However, the sensitivity of coronary CTA with
strated a higher specificity than coronary CTA CT-MPI was higher than that of coronary CTA for
stenosis of $50% (89% vs 72%; P < 0.001) but lower stenosis of $70% (84% vs 45%; P < 0.001) but lower
specificity
JACC: CARDIOVASCULAR IMAGING, VOL. 15, NO. 1, 2022 Nous et al 83
JANUARY 2022:75–87 Diagnostic Performance of Dynamic Perfusion CT

CENTRAL ILLUSTRATION Diagnostic Accuracy of Computed Tomography Angiography

and Dynamic Perfusion Computed Tomography for Hemodynamically Significant Coronary
Artery Disease

Coronary CTA Dynamic CT-MPI Invasive Angiography & FFR

LCX
LCX

RCA

LAD

LAD

FFR = 0.76 FFR = 0.74

96%
Sensitivity 45%
84%
72%
Specificity 94%
89%
78%
Accuracy 82%
88%
CTA Stenosis ≥50% CTA Stenosis ≥70% CTA + Dynamic CT-MPI

Nous, F.M.A. et al. J Am Coll Cardiol Img. 2022;15(1):75–87.

Coronary computed tomography angiography(CTA) and invasive angiography demonstrating moderate stenosis (arrow) in the left anterior
descending coronary artery and severe stenosis (arrowhead) in the left circumflex coronary artery. Dynamic stress computed tomography
myocardial perfusion imaging demonstrated corresponding perfusion defects (yellow-blue) in the apex and lateral wall, indicating inducible
ischemia, as confirmed by fractional flow reserve. The bar graph below summarizes the diagnostic performance of CTA with a coronary
stenosis threshold of 50% and 70% to CTA combined with perfusion imaging.

alone. This study included centers with a range of prior QUANTITATIVE MBF ANALYSIS. Dynamic CT-MPI
CT-MPI experience but comparable technical and calculation of absolute MBF can be helpful in
performance, providing encouragement for broader multivessel disease with balanced ischemia or
clinical implementation. Invasive FFR and MPI are both microvascular disease (8,15,20). However, a challenge
functional tests, but each is based on different for dynamic CT-MPI is cardiac motion and myocardial
physiologic principles. Because of these mechanistic displacement during the long breath-hold (21).
differences, even a perfect perfusion test could not be Consequently, reported MBF cutoff values that
expected to exactly match the pressure drop over an signify hemodynamic significance vary substantially,
epicardial stenosis in every single patient. Coronary CTA from 75-164 mL/min/100 mL among studies (5,7-9).
at a low stenosis threshold is very sensitive but not very Therefore, several studies showed that MBF values
specific. Therefore, it is virtually unavoid- able that the normalized to remote myocardium outperform abso-
addition of CT-MPI, or other functional tests that
lute MBF values (20,22). However, more recent
improve specificity and overall accuracy, will
studies contradicted these findings (15,23), and also,
underestimate a number of lesions with an FFR of
in the present study, we observe no significant dif-
#0.80.
ference between absolute (AUC: 0.79) and relative
JACC: CARDIOVASCULAR IMAGING, VOL. 15, NO. 1, 2022 Nous et al 85
JANUARY 2022:75–87 Diagnostic Performance of Dynamic Perfusion CT

F I G U R E 5 Correlation Between MBF and Invasive FFR

A
1.20
250.00
1.00
200.00
MBF
ml/min)

.80
Absolute

150.00
(ml/100

100.00 .60

50.00 .40

.00 .00
.00 .20 .40 .60 .80 1.00 .00 .20 .40 .60 .80 1.00
Invasive Fractional Flow Reserve Invasive Fractional Flow Reserve

B ** p = 0.007 ** p = 0.001
300.00 1.20

1.00
(ml/100 ml/min)
Absolute MBF

Relative MBF

200.00 .80
*
.60 *
*
100.00 .40
5

5
.7

.7

.20
≤0

≤0

*p < 0.001 *p < 0.001

0

0
.8

.8

.00 .00
-0

-0
76

76
1

1
.9

.9
0.

0.
5

5
.8

.8
≥0

≥0
-0

-0
81

81
0.

0.
0

Invasive Fractional Flow Reserve

.9

Invasive Fractional Flow Reserve

.9
-0

-0
86

86
0.

0.

(A) Scatterplots comparing absolute (left) and relative (right) MBF with FFR with a correlation of 0.51 and 0.53, respectively. Horizontal and
vertical lines are placed at the cutoff values of absolute MBF, relative MBF, and FFR. (B) Boxplots show median values (IQR) of absolute (left)
and relative MBF (right). *P value from Kruskal-Wallis test. **P value from Mann-Whitney U test. FFR ¼ fractional flow reserve;
MBF ¼ myocardial blood flow.

stenoses, whereas MPI reflects the combination of CONCLUSIONS

epicardial and microvascular disease. By study
design, the CAD prevalence was relatively high in our Dynamic CT-MPI offers incremental diagnostic value
cohort, and
ulations
extrapolation
with lower
of disease
the results
probabilities
to pop- over coronary CTA alone for the identification of he-
should be done with care. Furthermore, we did not modynamically significant CAD. Generalized results
adjust for potential correlation between multiple from this multicenter study encourage broader
vessels in the same subject, which might have consideration of dynamic CT-MPI in clinical practice.
underestimated the SEs of our analyses.
ACKNOWLEDGMENTS The authors thank all partici-
Additionally, absolute and rela- tive MBF thresholds
pants for their dedication to this research. We also
indicating myocardial ischemia vary among studies
thank Jochen von Spiczak, MD, and Malgorzata
and may be affected by the type of CT scanner and
postprocessing software. Thus, future investigation
TABLE 3 Image Quality by Dynamic Computed Tomography
is warranted to confirm our re- sults on different CT Myocardial Perfusion Imaging Experience
scanners and postprocessing software. Similar to
prior studies, and for a range of Excellent Good Moderate Poor

Experienced centers (>50 scans) 30 (45) 29 (44) 7 (11) 0(0)

including reasons
practical discrepancies between visual and QCA Inexperienced centers (#15 scans) 37 (62) 15 (25) 6 (10) 2(3)
stenosis severity, FFR was not per- formed in all
vessels with an intermediate stenosis severity (30). Values are n (%). Differences in image quality were tested using the chi-square test
(P ¼ 0.072).
86 Nous et al JACC: CARDIOVASCULAR IMAGING, VOL. 15, NO. 1, 2 0 2 2

Diagnostic Performance of Dynamic Perfusion CT JANUARY 2022:75–87

Polacin, MD, both from the Institute of Diagnostic and

Interventional Radiology, University Hospital Zurich, ADDRESS FOR CORRESPONDENCE: Dr Koen
University of Zurich, Switzerland. Nieman,
FUNDING SUPPORT AND AUTHOR DISCLOSURES Stanford University School of Medicine and Cardio-
vascular
Stanford,Institute, 300 Pasteur
California Drive, Room
94305, USA. H2157,
E-mail:
[email protected].
This study was supported by unrestricted grants from Siemens
Healthineers and Bayer Healthcare. Dr Nguyen’s research is sup-
ported by the National Institutes of Health (R01HL134830-01). Koen PERSPECTIVES
Nieman’s research is supported by the National Heart, Lung, and
Blood Institute of the National Institutes of Health (R01HL141712;
R01HL146754). Dr Geisler has received research grants from Med- COMPETENCY IN MEDICAL KNOWLEDGE: Dy-
tronic and Edwards Lifesciences. Dr Kitagawa has received an namic CT-MPI with coronary CTA provided incremen-
endowed chair position supported by Siemens Healthineers. Dr
tal diagnostic value over coronary CTA alone in
Vliegenthart has received an institutional research grant from
Siemens Healthineers. Dr Hausleiter has received receiving speaker patients with suspected CAD. This encourages broader
honoraria and research support from Abbott Vascular and Edwards considerations of dynamic CT-MPI in the clinical
Lifesciences; and has served as a consultant for Edwards Lifesciences. evaluation of patients with suspected CAD.
Dr Pugliese has received research support from Siemens Healthineers.
Dr Budde has received institutional research support to the Erasmus
MC from Siemens Healthineers. Dr Nikolauo has received research TRANSLATIONAL OUTLOOK: The potential role
grants
and Bayer
from
Healthcare;
Siemens Healthineers,
and has served GE
as Healthcare,
a consultant for Siemens of coronary CTA with dynamic CT-MPI to provide
Healthineers; and Bayer Healthcare. Dr Sakuma has received interventional cardiologists and cardiac surgeons
departmental research grants from FUJIFILM Toyama Chemical Co,
with an anatomy and functional noninvasive roadmap
Ltd, and Guerbet Japan KK. Dr Klotz is a retired employee of and
serves as a consultant for Siemens Healthineers. Dr Bamberg has
for the decision making of myocardial revasculariza-
received research grants from Siemens Healthineers and Bayer tion strategy needs to be investigated. Additionally,
Healthcare; and has served as a consultant for Siemens Healthineers, studies on different CT scanners and postprocessing
Bayer Healthcare, and Bracco. Dr Nieman has received unrestricted
software are needed to better understand how ab-
institutional research support from Siemens Healthineers and
HeartFlow Inc; has served as a consultant for Siemens Medical
solute and relative MBF can be used in clinical
Systems USA; and holds equity in Lumen Therapeutics. All other practice.
authors have reported that they have no relationships relevant to the
contents of this paper to disclose.

REFERENCES
1. Douglas PS, Hoffmann U, Patel MR, et al. Out-
comes of anatomical versus functional testing for dynamic CT-based myocardial perfusion imaging. generation dual-source CT. 2020;21:58–67.
Korean J Radiol.
coronary artery disease. N Engl J Med. 2015;372: Radiology. 2011;260:689–698.
11. Mahnken AH, Klotz E, Pietsch H, et al. Quan-
1291–1300. 6.Rossi A, Merkus D, Klotz E, Mollet N, de titative whole heart stress perfusion CT imaging as
2. Litt HI, Gatsonis C, Snyder B, et al. CT angiog- Feyter PJ, Krestin GP. Stress myocardial perfusion: noninvasive assessment of hemodynamics in cor-
raphy for safe discharge of patients with possible onary artery stenosis: preliminary animal experi-
imaging with multidetector CT. Radiology.
ence. Invest Radiol. 2010;45:298–305.
2014;270:25–46.
acute coronary syndromes. N Engl J Med. 7. Rossi A, Dharampal A, Wragg A, et al. Diagnostic 12. Ebersberger U, Marcus RP, Schoepf UJ, et al.
2012;366:1393–1403. performance of hyperaemic myocardial blood flow Dynamic CT myocardial perfusion imaging: per-
3. Fihn SD, Blankenship JC, Alexander KP, et al. index obtained by dynamic computed tomogra- formance of 3D semi-automated evaluation soft-
phy: does it predict functionally significant coro- ware. Eur Radiol. 2014;24:191–199.
2014 ACC/AHA/AATS/PCNA/SCAI/STS focused
nary lesions? Eur Heart J Cardiovasc Imaging. 13. Cerqueira MD, Weissman NJ, Dilsizian V, et al.
update of the guideline for the diagnosis and
2014;15:85–94.
management of patients with stable ischemic 8. Wang Y, Qin L, Shi X, et al. Adenosine-stress
heart disease: a report of the American College of
Cardiology/American Heart Association Task Force Standardized myocardial segmentation and
on Practice Guidelines, and the American Associ- dynamic myocardial perfusion imaging with nomenclature for tomographic imaging of the
ation for Thoracic Surgery, Preventive Cardiovas- second-generation dual-source CT: comparison heart. A statement for healthcare professionals

for
cular Nurses Association, Society with conventional catheter coronary angiography from the Cardiac Imaging Committee of
Cardiovascular Angiography and
Interventions, and SPECT nuclear myocardial perfusion imaging. the Council on Clinical Cardiology of the Amer-
and Society of Thoracic Surgeons. J Am Coll Car- AJR Am J Roentgenol. 2012;198:521–529. ican Heart Association. Circulation. 2002;105:
diol. 2014;64:1929–1949. 9. Pontone G, Baggiano A, Andreini D, et al. Dy- 539–542.
4.George RT, Mehra VC, namic stress computed tomography perfusion 14. Leipsic J, Abbara S, Achenbach S, et al. SCCT
al. Chen MY, et with a whole-heart coverage scanner in addition to guidelines for the interpretation and reporting of
Myocardial CT perfusion imaging and SPECT for coronary computed tomography angiography and coronary CT angiography: a report of the Society
the diagnosis of coronary artery disease: a head- fractional flow reserve computed tomography of Cardiovascular Computed Tomography Guide-
to-head comparison from the CORE320 Multi- derived. J Am Coll Cardiol Img. 2019;12:2460– lines Committee. J Cardiovasc Comput Tomogr.
center Diagnostic Performance Study. Radiology. 2471. 2014;8:342–358.
2014;272(2):407–416. 15. Rossi A, Wragg A, Klotz E, et al. Dynamic
10.Takafuji M, Kitagawa K,
5. Bamberg F, Becker A, Schwarz F, et al. Detec- computed tomography myocardial perfusion im-
tion of hemodynamically significant coronary ar- Ishida M, et al. aging: comparison of clinical analysis methods for
Myocardial coverage and radiation dose in dy-
tery stenosis: incremental diagnostic value of namic myocardial perfusion imaging using third-