FFR : Clinical studies

Transcription

FFRCT: Clinical studies
Bjarne Nørgaard
Department Cardiology B
Aarhus University Hospital Skejby,
Denmark
Disclosures: Research grants:
Edwards and Siemens
Coronary CTA:
● High
diagnostic sensitivity
for detecting CAD, and the
best non-invasive test for
ruling-out CAD
BUT!!
Cheng V et al, JACC Imag. 2008
Meijboom W et al, JACC 2008
Schuijj JD et al JACC 2006
Coronary CTA:
● High diagnostic sensitivity for
detecting CAD, and the best noninvasive test for ruling-out CAD
BUT!!
• Poor correlation to stenosis
severity as assessed with
ICA
Coronary CTA:
● High diagnostic sensitivity for
detecting CAD, and the best noninvasive test for ruling-out CAD
BUT!!
• Poor correlation to stenosis severity
as assessed with ICA
• The hemodynamic
significance of a stenosis
cannot be assessed
“FFRCT: Clinical studies”
PET-CT
eeeeeeeeeeeeeeeeeeeeeee
FFRCT
eeeeeeeeeeeemmmmmmm
mmmmmmmmmmmmmm
mmmmmmmmmmmmmm
mmmmmmmmmmmmmm
mmmmmmmmmmmmmm
mmmmmmmmmmeeeeeee
eeeeeeeeeeeeeeee
Image source
Gaemperli O, EHJ 2011
PET-CT
CT-perfusion
FFRCT
eeeeeeeeeeeemmmmmmm
mmmmmmmmmmmmmm
mmmmmmmmmmmmmm
mmmmmmmmmmmmmm
mmmmmmmmmmmmmm
mmmmmmmmmmeeeeeee
eeeeeeeeeeeeeeee
Image source
PET-CT
CT-perfusion
FFRCT FFR
CT
eeeeeeeeeeeemmmmmmm
mmmmmmmmmmmmmm
mmmmmmmmmmmmmm
mmmmmmmmmmmmmm
mmmmmmmmmmmmmm
mmmmmmmmmmeeeeeee
eeeeeeeeeeeeeeee
Image source
• DISCOVER-FLOW
– Completed 2011
– N=104 patients
• DeFACTO
– Completed 2012
– N=252 patients
• NXT
– Completed 2013
– N~254 patients
“FFRCT: Clinical studies”, DISCOVER-FLOW trial
● 4 centers
● Patients with suspected or known CAD who underwent
cCTA, ICA and FFR
Koo B-K, JACC 2011
● 4 centers
● Patients with suspected or known CAD who underwent cCTA,
ICA and FFR
● Significant CAD by cCTA: stenosis >50% (>2 mm vessel)
● Ischemia def: FFRCT / FFR ≤0.80
● FFRCT and cCTA core lab reads, FFR local reads
Koo B-K, JACC 2011
● 4 centers
● Patients with suspected or known CAD who underwent cCTA,
ICA and FFR
● FFRCT and cCTA core lab reads, FFR local reads
● 104 patients, 159 vessels
● Mean age, 63 y, 20% with known disease (AMI / PCI),
56% with FFR ≤0.80
Koo B-K, JACC 2011
Koo B-K, JACC 2011
P=0.001
P=0.0001
Koo B-K, JACC 2011
“FFRCT: Clinical studies”, DeFACTO trial
● 17 centers
● Clinically non-emergent indicated ICA after cCTA <60 days
● ICA stenosis 30%-90% (>2 mm vessel)
%
Min J et al, JAMA 2012
● 17 centers
%
● 17 centers
●%FFRCT and cCTA core lab reads, FFR local reads
● 17 centers
● Ischemia def: cCTA stenosis >50%, FFRCT / FFR ≤0.80
●%FFRCT and cCTA core lab reads, FFR local reads
● Primary study end-point: Per-patient diagnostic accuracy of
FFRCT for the diagnosis of ischemia with a lower limit of 95%
CI >70%
● 252 patients, 406 vessels
Patient Characteristics
Age (years) [mean + SD]
63 ± 9
Male gender
71%
Prior MI
6%
Prior PCI
6%
Hypertension
71%
Diabetes mellitus
22%
Agatston score [mean + SD]
FFR ≤0.80
382 ± 401
37%
Primary end point
100
90
90
90
80
70
FFRCT <0.80
cCTA >50%
84
84
84
73
73
67
64
64
60
54
54
50
72
61
42
42
40
30
20
10
0
Accuracy
95% CI
FFRCT
cCTA
67-78
58-70
Sensitivity
84-95
77-90
Specificity
46-83
34-51
PPV
NPV
60-74
53-67
74-90
61-81
%
AUC , 0.68 (0.62-0.74)
vs 0.81 (0.75-0.86)
p<0.0001
“FFRCT: Clinical studies”, DeFACTO vs NXT trial design
Gaur S et al, JCCT 2013
“FFRCT: Clinical studies”, NXT trial, Study
objectives
• To determine the diagnostic performance of non-invasive
FFRCT using FFR as the reference standard
• To compare the diagnostic performance of FFRCT vs.
anatomic testing (coronary CTA or invasive coronary
angiography)
Abbara S et al. JCCT 2009
“FFRCT: Clinical studies”, NXT trial, Patient
recruitment
Inclusion Criteria:
•
•
Underwent >64-row CT and ICA scheduled
< 60 days between CT and ICA
Exclusion Criteria:
•
•
•
•
Prior CABG or PCI
Suspected ACS
Recent MI within 30 days of CT
Contraindication to nitrates, beta blockade or adenosine
ICA = Invasive coronary angiography; CABG = coronary artery bypass surgery; ACS = acute coronary syndrome;
MI = myocardial infarction; PCI = percutaneous coronary intervention
“FFRCT: Clinical studies”, NXT trial,
Endpoints
Primary Endpoint:
– Per-patient diagnostic performance as assessed by the area under
the receiver operating characteristic curve (AUC) of FFRCT vs.
coronary CTA for the diagnosis of ischemia in patients with stenosis
severity 30%-90% (vessel diameter >2 mm)
Secondary Endpoint:
– Diagnostic performance (accuracy, sensitivity, specificity, PPV and
NPV) of FFRCT, coronary CTA, and invasive coronary angiography
“FFRCT: Clinical studies”, NXT trial, Study
procedures
• Blinded core laboratories for FFR, FFRCT, and ICA
• CT: -Acquisition protocols according to societal guidelines1
-Image quality independently evaluated using a predefined scoring
system2
-Site-read stenosis severity >50%3
• ICA: -Core-lab read stenosis severity >50%
•
–
–
–
FFR: At maximum hyperemia during ICA
Adenosine 140 – 180 micg/kg/min IV
Positive: ≤0.804
≤ 0.804
1Abbara
S et al. JCCT 2009
B et al. JCCT 2013
3Raff GL et al. JCCT 2009
4Tonino PA et al. NEJM 2009
2Nørgaard
“FFRCT: Clinical studies”, NXT trial, Patient
Screened Cohort
(n=357)
enrollment
•
•
Study enrollment 9/2012 – 8/2013
10 sites in Europe, Asia, and Australia
• Image artifacts (n=44, 13%)
• Failed inclusion criteria (n=27)
Excluded
• FFR quality (n=22): Invalid FFR
(n=8), FFR in vessel <2 mm
(n=11), missing measure
location (n=3)
• Other (n=10)
Secondary endpoints
• Patients n=254
• Vessels n=484
Excluded from Primary Endpoint
• No cCTA stenosis 30%-90%
(n=3)
Primary Endpoint
• Patients n=251
“FFRCT: Clinical studies”, NXT trial, Patients
and CT characteristics
CT Characteristics
Patient Characteristics
Age (years) [mean + SD]
64 ± 10
Male gender
64%
Prior MI
Diabetes mellitus
Hypertension
Pre-test Likelihood of CAD
2%
23%
69%
58%
FFR ≤0.80
FFR assigned a value of
0.50 in 16 occluded vessels
32%
*Available for 214 patients
3%
– Nitrates
99.6%
– Beta Blockers
78%
– Heart Rate (bpm)
63
Range
– Prospective
mean dose (mSv)
– Retrospective
mean dose (mSv)
37-110
54%
3
46%
14
– Calcium score*
Mean
302
Range
0 – 3599
>300
33%
“FFRCT: Clinical studies”, NXT trial, Per-vessel
Pearson´s CC
= 0.82,
P<0.0001
FFR - FFRCT correlation
“FFRCT: Clinical studies”, NXT trial, Discrimination
Per-vessel
Per-Patient
(n=251)
Sensitivity
Per-patient
Per-Vessel
(n=484)
1 - Specificity
N=252
*Area under the receiver operating characteristics curve
of ischemia
“FFRCT: Clinical studies”, NXT trial, Per-patient
diagnostic performance
100
90
80
81
79
FFRCT < 0.80
65
70
%
86
92
60
50
40
30
20
10
0
Accuracy
N=254
95%
CI
Specificity
PPV
Sensitivity
NPV
100
*
90
80
70
%
60
94 91
*
81
79
86
92 93 92
FFRCT < 0.80
* p<0.0001
65
53
51
50
34
40
46
30
20
10
0
Accuracy
N=254
95%
CI
Specificity
PPV
CT > 50%
ICA > 50%
*
64
40
Sensitivity
NPV
100
*
90
80
70
%
60
94 91
*
81
79
86
92 93 92
FFRCT < 0.80
* p<0.0001
65
53
51
50
34
40
46
30
20
10
0
Accuracy
N=254
95%
CI
Specificity
PPV
CT > 50%
ICA > 50%
*
64
40
Sensitivity
NPV
100
*
90
80
70
%
60
94 91
*
81
79
86
92 93 92
FFRCT < 0.80
* p<0.0001
65
51
46
FFRCT correctly reclassified
68% of
40
40
34
30 CT false positive to true negatives
50
20
10
0
Accuracy
N=254
95%
CI
Specificity
PPV
CT > 50%
ICA > 50%
*
64
53
Sensitivity
NPV
FFRCT: Clinical studies”, NXT trial, Per-patient
LAD stenosis 70-90%
FFRCT 0.93
FFR 0.94
FFRCT Model
“FFRCT: Clinical studies”, NXT trial, Per-vessel
100
*
86
90
80
70
65
86
71
60
%
*
60
83 84 84
66
92 94 95
FFRCT < 0.80
* p<0.0001
61
33
40
30
20
10
0
Accuracy
N=484
95% C
Specificity
PPV
CT > 50%
ICA > 50%
*
50
40
Sensitivity
NPV
“FFRCT: Clinical studies”, NXT trial, Intermediate
100
93
*
80
90
80
*
79
50
91 92
85
cCTA >50%
FFRCT <0.80
*
70
60
lesions (30%-70%)
%
63
*p<0.05
51
40
32
37
30
20
10
0
Acc
n= 235 (93%)
Spec
PPV
Sens
NPV
“FFRCT: Clinical studies”, NXT trial, Coronary
calcification
100
90
87
85
80
83
85
82
75
FFRCT <0.80
86
77
73
70
60
Agatston score:
mean, 302
range, 0 - 3599
50
40
30
20
10
0
Acc
Ag:
N (214):
Sens
Spec
Acc
Sens
Spec
”0”
”1-300”
27
117
Acc
Sens
Spec
”>300”
70
“FFRCT: Clinical studies”, NXT trial, Coronary
calcification
100
90
*
*
80
*
*
FFRCT <0.80
cCTA >50%
*
*
70
*p<0.05
60
Agatston score:
mean, 302
range, 0 - 3599
50
40
30
20
10
0
Acc
Ag:
N (214):
Sens
Spec
Acc
Sens
Spec
Acc
Sens
”0”
”1-300”
”>300”
27
117
70
Spec
FFRCT: Clinical studies”, DISCOVER-FLOW, DeFACTO, and NXT trials
• FFRCT has high diagnostic accuracy and
discrimination for the diagnosis of ischemia
• When compared to anatomic interpretation by
coronary CTA or invasive angiography, FFRCT
leads to a marked increase in diagnostic
accuracy, specificity, and PPV
• When compared to anatomic interpretation by
coronary CTA or invasive angiography, FFRCT leads
to a marked increase in diagnostic accuracy,
specificity, and PPV
• FFRCT is performed from standard acquired CT
datasets without the need for additional
imaging, radiation or medication
“FFRCT: Clinical studies”, FFRct vs other non-invasive testing modalities
Desai RR, AJR 2013
Per patient
pooled sens / spec:
89% (95% CI, 84-93) /
85% (77-91)
Per-territory
pooled sens / spec:
88% (84-91) /
89% (87-90)
Desai RR, AJR 2013
● Single-center studies
Desai RR, AJR 2013
● # patients: 28 - 120
Desai RR, AJR 2013
● # patients: 28 - 120
● Local FFR reads
Desai RR, AJR 2013
● # patients: 28 - 120
● Local FFR reads
● The rate of actual FFR measurements
was <70% in 10 of 12 studies (ICA stenosis
>90% => FFR, 0.50, and in normal vessels
=> FFR, 0.95)
Desai RR, AJR 2013
● # patients: 28 - 120
● Local FFR reads
● The rate of actual FFR measurements
was <70% in 10 of 12 studies (ICA stenosis
>90% => FFR, 0.50, and in normal vessels
=> FFR, 0.95)
Desai RR, AJR 2013
22%-69% of FFR values were directly measured
Desai RR, AJR 2013
22%-69% of FFR values were directly measured
All FFR values were directly measured
Desai RR, AJR 2013
Diagnostic performance of Coronary diagnostic tests for
Functional (FFR ≤ 0.80) disease
Diagnostic performance of Coronary diagnostic tests for
Functional (FFR ≤ 0.80) disease
“FFRCT: Clinical studies”, Building
the Body of Evidence
● 609 patients
● 1051 vessels,
● FFR directly measured in 1035 vessels (98%)
DISCOVER-FLOW
NXT
DeFACTO
Cath lab
Population
Dx Accuracy
2012
30-90%
Stenosis
Dx Accuracy
2013
2014
2015
…
59
• Use of FFRCT to select patients for ICA and PCI may result in 30% lower
costs and 12% fewer events at one year compared to the most common
strategy of ICA
60
“FFRCT: Clinical studies”, Building
the Body of Evidence
● 609 patients
● 1051 vessels,
● FFR directly measured in 1035 vessels (98%)
DISCOVER-FLOW
NXT
2012
Randomized
Controlled
Trials
Analysis of existing
databases
DeFACTO
Cath lab
Population
Dx Accuracy
PLATFORM
30-90%
Stenosis
Dx Accuracy
2013
Stable
Angina
Outcomes and
resources
2014
ACS
(- enzyme)
Outcomes and
resources
2015
…
61
Thank you for the
attention
FFRCT false-negatives and FFR ”grey
zone”
• 24 vessels had false negative results of FFRCT when
compared to FFR.
• FRR 0.75-0.80 (”grey zone”), n=17
FFRCT false-negatives and FFR ”grey
zone”
• 24 vessels had false negative results of FFRCT when compared
to FFR.
• FRR 0.75-0.80 (”grey zone”), n=17
• Assuming, patients with false-negative FFRCT values in the grey
zone are in fact true negatives, the NPV for FFRCT increased to
98%.
• Thus, in vessels with FFRCT >0.80 the risk of having an FFR
<0.75 is only 2%.
Limitations
• No control over CT image acquisition protocol at clinical sites
• Did not interrogate every vessel with invasive FFR
• Did not solely enroll patients with intermediate stenosis
• Did not test whether FFRCT-based revascularization reduces
ischemia
Conclusions
•
FFRCT demonstrated improved accuracy over CT for diagnosis of patients and vessels with ischemia
– FFRCT diagnostic accuracy 73% (95% CI 67-78%)
• Pre-specified primary endpoint >70% lower bound of 95% CI
– Increased discriminatory power
•
FFRCT superior to CT for intermediate stenoses
•
FFRCT computed without additional radiation or imaging
•
First large-scale demonstration of patient-specific computational models to calculate physiologic
pressure and velocity fields from CT images
•
Proof of feasibility of FFRCT for diagnosis of lesion-specific ischemia

FFR : Clinical studies

Transcription

Similar documents

LEGO Mindstorms™ for informed metrics in virtual worlds

LTR 13 Abridged

Cranial Nerves Atlas

U libre de Bruxelle - International Communication Association

Ica news - Wilson Elser

Volcano Trail: Introduction

Multilinear Independent Components Analysis

toournation