1
Invasive Diagnostic
Ik-Kyung Jang, MD, PhD
Associate Professor of MedicineHarvard Medical School
Massachusetts General Hospital
Angioplasty Summit 2008
2
Stable vs Vulnerable Plaque
• Low lipid conc. • High lipid conc.• Thick fibrous cap • Thin fibrous cap• Low mΦ density • High mΦ density
Stable Plaque Vulnerable Plaque
3
Current Technology for VP Identification
Non invasive- CTA- MR
Invasive- Anatomic- Physiologic
4
Intravascular Modalities
IVUSAngioscopyIV MROCT
ThermographySpectroscopy
Anatomic Information
Biochemical Information
5
Intravascular Modalities
IVUSAngioscopyIV MR OCT
ThermographySpectroscopy
6
IVUS
• Echolucency
• Remodeling
• Distensibility
LesionProximal Ref
Diastole Systole
7
IVUS
• Echolucency
• Remodeling
• Distensibility
8
Prevalence of Echolucent Area on IVUS
Yamagishi et al. JACC 2000
66.7
83.3
4.421.1
ACS Non-ACS
(%)
9
Echolucent Area: Unstable vs stable angina
Schoenhagen et al. Circulation 2000
0%
10%
20%
30%
40%
50%
Echolucent Echodense Mixed Calcified
Unstable Stable
10
Echolucency
• Previous histological studies have demonstrated that the discrimination of lipid is inconsistent using greyscale images alone.– Palmer et al. Eur Heart J., 1999– Peters et al. J Am Soc
Echocardiogr., 1994– Peters et al. Circulation, 1994
11
IVUS vs OCT (n = 145)
100100.01.82 quadrant lipid
9393.922.71 quadrant lipid
8897.06.3Superficial lipid
9597.017.9Deep lipid
PPV (%)Specificity (%)Sensitivity (%)
12
IVUS
• Echolucency
• Remodeling
• Distensibility
Lesion
Proximal Reference
13
Vascular remodeling:Unstable vs stable angina
Schoenhagen et al. Circulation 2000
0%10%20%30%40%50%60%70%
PositiveRemodeling
Absence ofRemodeling
NegativeRemodeling
Unstable Stable
14
IVUS
• Echolucency
• Remodeling
• Distensibility
Diastole
Systole
15
Arterial DistensibilityResults
Lipid Rich Mixed Fibrous PNumber 7 6 16
EEM CSA (mm2) 16.1 ± 7.2 16.9 ± 4.6 17.1 ± 0.4 NSLumen CSA (mm2) 7.2 ± 1.4 8.5 ± 3.6 8.5 ± 0.4 NSPlaque Volume (mm2) 8.9 ± 5.2 8.4 ± 3.5 8.6 ± 2.9 NS
Distensibility 2.8 ± 1.8 1.7 ± 0.9 1.0 ± 0.8 0.004Index (mmHg-1)
MacNeill & Jang
16
IVUS
• Echolucency
• Remodeling (Plaque volume)
• Distensibility
LesionProximal Ref
Virtual Histology
Palpography
17
FIBROUS
FIBROLIPIDIC
CALCIUM
LIPID CORE
MEDIA
VH Legend
Virtual Histology™ IVUS
Classification Tree
M a xP < -2 5.05T yp e: C ol la ge n
# of RO Is : 5
M a xP > -2 5.05T yp e: C ol la ge n
# of RO Is : 2 6
Int < -28 .65# of RO Is : 3 1
Int < -14 .8T yp e: C ol la ge n
# of RO Is : 5
Int > -14 .8T yp e: C ol la ge n
# of RO Is : 5
F at M a xP < 21 .04 5# of RO Is : 1 0
F at M a xP > 21 .04 5T yp e: C ol la ge n
# of RO Is : 1 7
M B F < -6 5.0 9# of RO Is : 2 7
Int < -4.1 95T yp e: C al c iu m
# of RO Is : 6
Int > -4.1 95T yp e: C ol la ge n
# of RO Is : 5
M B F > -6 5.0 9# of RO Is : 1 1
F at M a xP < 30 .03# of RO Is : 3 8
M a xP < -1 6.09 5T yp e: C ol la ge n
# of RO Is : 6
F at M a xP < 34 .27 5T yp e: N ecr otic
# of RO Is : 6
M B F < -6 6.6 6T yp e: F ib ro Li pi di c
# of RO Is : 5
M B F > -6 6.6 5T yp e: C ol la ge n
# of RO Is : 5
M a xP < -1 2.14 5# of RO Is : 1 0
M a xP > -1 2.14 5T yp e: F ib ro -Li pi di c
# of RO Is : 5
F at M a xP > 34 .27 5# of RO Is : 1 5
M a xP < -9 .9 15# of RO Is : 2 1
F at M a xP < 35 .5T yp e: C ol la ge n
# of RO Is : 8
F at M a xP > 35 .5T yp e: C ol la ge n
# of RO Is : 9
M a xP > -9 .9 15# of RO Is : 1 7
M a xP > -1 6.09 5# of RO Is : 3 8
F at M a xP > 30 .03# of RO Is : 4 4
Int > -28 .65# of RO Is : 8 2
M B F < -5 5.6 95# of RO Is : 1 13
M i nP < -1 7 .9 15T yp e: C ol la ge n
# of RO Is : 5
M B F < -5 3.1 5T yp e: C al ciu m
# of RO Is : 8
M B F > -5 3.1 5T yp e: C al ciu m
# of RO Is : 2 0
M i nP > -1 7.9 15# of RO Is : 2 8
M B F > -5 5 .6 95# of RO Is : 3 3
1 46# of RO Is :
TR E ER O OT
18
THIS IS THE PROBLEM !
ID (IVUS Defined) TCFAs
MP Margolis, MD, PhD
19
What is VH IVUS: VH slice thickness
300 µm
Longitudinal axis
MP Margolis, MD, PhDMP Margolis, MD, PhD
20
Histology slice thickness
4 µm
300 µm
MP Margolis, MD, PhDMP Margolis, MD, PhD
21
The Effect of Slice Thickness on VH Plaque Composition
Histopathology; Calcified FA
4 µm 4 µm 4 µm
300 µm
VH IVUS, Calcified FA
4 µm
MP Margolis, MD, PhDMP Margolis, MD, PhD
22
Effect of Lesion Thickness on VH Plaque Composition
220 µm
4 µm4 µm 4 µm4 µm 4 µm
20 µm 20 µm120 µm 60 µm
Histopathology: FA on right
Histopathology: Calcified TCFA on left
Lesion Analysis by Histopathology: Calcified TCFA
Calcified TCFA FA
VH IVUS:Calcified TCFA
MP Margolis, MD, PhDMP Margolis, MD, PhD
23
What is the “Gold Standard” for In-Vivo “histology”? Histopathology or Pathologists?
Example 1
Pathologists 1(March 2006)Pathologists 1(March 2006) Pathologists 2 (Aug 2006)Pathologists 2 (Aug 2006) Pathologists 3 (Sept Pathologists 3 (Sept 2006)2006)
MP Margolis, MD, PhDMP Margolis, MD, PhD
24
19) CCF 05104 B2Single Single
LesionLesion
PIT FA FA
Pathologists 1(March 2006)Pathologists 1(March 2006) Pathologists 2 (Aug 2006)Pathologists 2 (Aug 2006) Pathologists 3 (Sept Pathologists 3 (Sept 2006)2006)
2D ALAS
FA
MP Margolis, MD, PhDMP Margolis, MD, PhD
25
7) CCF 05065 B2Single Single
LesionLesion
Pathologists 1(March 2006)Pathologists 1(March 2006) Pathologists 2 (Aug 2006)Pathologists 2 (Aug 2006) Pathologists 3 (Sept 2006)Pathologists 3 (Sept 2006)
FC Ca FA FA
2D ALAS
Ca FA
MP Margolis, MD, PhDMP Margolis, MD, PhD
26
SOFT
HARD
27
Sensitivity and Specificity: in vitro
Sensitivity = 88%Specificity = 89%⇒
Schaar JA, et al. Circulation 2003
Elastogrampositive
Elastogramnegative
Histologypositive 20 3 23
Histologynegative 4 27 31
24 30 54
28
Smooth muscle cells
heavymediumminornone
Stra
in(%
)Strain and tissue components
p < 0.006p < 0.0001
3.0
2.5
2.0
1.5
1.0
.5
0.0
Macrophages
Stra
in(%
)1312326N =
3,0
2,5
2,0
1,5
1,0
,5
0,0
heavymediumminornone
29
Invasive Imaging Modalities
IVUSAngioscopyIV MROCT
ThermographySpectroscopy
30
The frequency of yellow plaque
0
10
20
30
40
50
60
70
80
AMI RMI OMI UA SAMizuno et al. Lancet 1991
(%)
31
Incidence of ACS: Angioscopic Finding
Uchida et al. Am Heart J 1995
68.4
7.63.3
0
10
20
30
40
50
60
70
Yellow Glistening(9/13)
Yellow (2/26)
White (4/118)
(%)
32
Invasive Imaging Modalities
IVUSAngioscopyIV MROCT
ThermographySpectroscopy
33
Top-spin® intracoronary MR catheter
34
100µ
100µ
100µ
CD68Movat’s
Vulnerable plaque
Thick cap fibroatheroma
Stable plaque
35
Correlation of lipid fraction determined by ic MR and histology
Fibrous Necrotic coreFoam cells
36
37
Invasive Imaging Modalities
IVUSAngioscopyIV MROCT
ThermographySpectroscopy
38
In vivo thermal heterogeneity within human atherosclerotic coronary arteries
Stefanadis et al. Circulation 1999
39
Risk of Adverse Cardiac Events
Stefanadis et al. JACC. 2001
40Stefanadis et al. J Mol Cell Cardiol. 2000
Clinical Presentation and the Temperature Difference: hsCRP
41
Atorvastatin and Plaque Temperature
Stefanadis et al. Eur Heart J
StatinsNo statin
Tem
pera
ture
diff
eren
ce
2.5
2.0
1.5
1.0
.5
0.0
-.5
P<0,001
42
Toutouzas et al reported correlation between temperature and expansive remodeling and MMP-9 concentration.
Verheye et al showed that temperature heterogeneity was reduced after change from high to low-cholesterol diet in rabbits.
43
Invasive Imaging Modalities
IVUSAngioscopyOCTIV MR
ThermographySpectroscopy
44
Near-IR Spectra of Human Aortic Samples
PlaqueRepeat plaque
Cholesterol
Collagen
Repeatnormal
Normal
Wavelength
Abs
orpt
ion
45
889593Negative predictive value
866890Positive predictive value
899393Specificity
847790Sensitivity
Inflam-mation
Thincap
Lipid pool
All values in %.
Characterization of plaque histology by NIR spectroscopy (ex vivo, no blood, no motion)
Moreno P et al. Circulation 2002
46
InfraReDx Spectroscopy System
• Three components: console, PBR, catheter (3.2 Fr, monorail, 0.014” compatible)
• Automatically scans artery
• Spectra processed by algorithm and displayed to user as a chemical image of lipid rich plaque probability (“Chemogram”)
47
Comparison of Chemogram with Histology
48
SPECTACL: No Lipid Signal at Culprit
NIR End
NIR Start
Culprit
• 55y, M
• BMI 32.4 kg/m2
• Prog Angina
•+DM, HTN, Chol
Culprit
Lahey Clinic
Lahey Clinic
Culprit
49
SPECTACL: Lipid Signal at Culprit
Beth Israel Deaconess Medical Center
Beth Israel Deaconess Medical Center
NIR End
NIRStart
Culprit
Culprit
Culprit• 71y, M
• BMI 28.3 kg/m2
• MI (<72 hrs)
• HTN
50
SPECTACL: Other Areas of Interest
William Beaumont Hospital
William Beaumont Hospital
NIR Start
NIR End
Culprit
Culprit
• 59y, F
• Atypical CP
• HTN, Chol, +FH
51
Multiple views of lipid-rich plaque probability
3-D Mesh
52
Intermediate Stenoses Caused by Lipid-rich vsFibrotic Plaques: Detection by NIR Spectroscopy
53
“Lipid Burden Index”
• Measure of overall Plaque Burden– Potentially useful as measure of risk or of pharmacologic
treatment efficacy
• Fraction of Chemogram image pixels above probability of 0.6– Scaled from 0 to 1000
• 0.85 AUC vs. fibroatheroma presence (0.79 – 0.91)
4
143
54
169 172 178 184 202 213 272 295 298 312 322
184
126 127 129 138 138 143 145 145 150 156 157
133 160 162
72 73 74 78 80 81 84 86 87 87 9677 102 106
58 58 59 62 63 64 65 66 67 68 6960 71 71
38 39 40 42 43 43 45 46 47 48 5041 53 54
23 26 27 29 34 34 35 35 36 36 3627 37 38
10 11 11 15 16 17 19 19 20 21 2313 23 23
3 3 4 4 5 5 6 7 7 8 84 9 10
0 0 0 0 0 0 1 1 1 2 20 3 3
LRP Burden Index Mosaic
55
Invasive Imaging Modalities
IVUSAngioscopyIV MROCT
ThermographySpectroscopy
56
Ex Vivo Study Results
.96NPV.97SPEC
.88PPV.87SENS
.95NPV1.0SPEC1.0PPV.95SENS
.97NPV.94SPEC
.81PPV.92SENS
Fibrous
Calcific
Lipid pool
Interobserver k = 0.88, Intraobserver k = 0.91
Yabushita, .. Jang, Bouma, Tearney. Circulation 2002
57
0
50
100
150
200
250
300
350
400
450
500
0 100 200 300 400 500OCT measurement ( )탆
y = 1.02x+3.8r = 0.89p < 0.0001
Correlation between OCT and histology
58
Linear NSD vs. CD68
100% (60 -100%)SPEC100% (70-100%)SENS
CD68 % area > 10 % -NSD cutoff 6.2%
Tearney, .. Jang, Bouma. Circulation 2003
59
Stable vs Vulnerable Plaque
• Low lipid conc. • High lipid conc.• Thick fibrous cap • Thin fibrous cap• Low m Φ density • High mΦ density
Stable Plaque Vulnerable Plaque
60
Plaque Characterization
• Goal: Determine plaque characteristics in various presentations of CAD
• Methods– Patients with CAD; N=57– OCT imaging culprit/remote lesions– OCT: 3.0 F catheter 8 cc saline purge
• Analysis– Clinical presentation:
• AMI (20)• ACS (20)• SAP (17)
– Two OCT readers consensus– Cap thickness– Macrophage density (~NSD)
Culprit Lesion
Remote Lesion
61
Cap Thickness
4753.8
102.6
0
20
40
60
80
100
120
140
160M
in. C
ap T
hick
ness
AMIACSSAP
P < 0.02
62
Prevalence of TCFA
Jang et al. Circulation 2005
45%
18%
65%
0%
10%
20%
30%
40%
50%
60%
70%
80%TC
FA P
erce
nt
AMIACSSAP
P < 0.01
63
Macrophage Density for Acute and Stable Clinical Syndromes
• Cap macrophage density is higher in acute clinical syndromes in both culprit and remote sites
*P < 0.001*
64
Limitations of OCT
1. Need to create blood free zone
2. No scanning capability
3. Shallow penetration depth
4. No functional information
65
Intravascular Diagnostics for VP
Modality Resolution Penetration Cap Lipid Inflam Use
IVUS 100 um good + ++ - +++
Angioscopy - poor + ++ - +
IV MR 160 um poor + ++ ++ +
OCT 10 um poor +++ +++ ++ +
Thermography - poor - - +++ ++
Spectroscopy - poor + ++ ++ +
Modality Resolution Penetration Cap Lipid Inflam Use
IVUS 100 um good + ++ - +++
Angioscopy - poor + ++ - +
IV MR 160 um poor + ++ ++ +
OCT 10 um poor +++ +++ ++ +
Thermography - poor - - +++ ++
Spectroscopy - poor + ++ ++ +
Modified from MacNeill and Jang. ATVB 2003
66
Ideal Invasive Diagnostics for VP
Combination of imaging and physiologic test
- OCT + thermography- OCT + spectroscopy- IV MR + thermography
67
Invasive Diagnostics for VP
Questions1. Can one justify the invasive
diagnostic tests?- invasive- high cost- pt acceptability as a screening tool
2. Which patients?3. When to perform the tests?4. When to treat the lesions??
68
Optimal Medical PreventionOptimal Medical Prevention
Cannon et al. NEJM 2004; 350: 1495–1504.
26.3% vs.
22.4%at 2y
PROVEPROVE--IT TrialIT Trial
69
Invasive Diagnostics for VP
Inherent limitations
1. Local information (systemic disease)
2. Superficial information
3. Difficulty in sampling the same site
4. Gold standard (??) – validation problem
5. Only when a local therapy is viable!!
70
Acknowledgements
Cardiology DivisionOC RaffelS. ChiaH. YabushitaM. Takano
M. Kawasaki A. Low D. DeJosephI. McNulty J. Healy
Wellman Laboratories of Photomedicine
B.E. Bouma S.H. Yun G.J. Tearney
Dept. of PathologyS. Houser
Dept. of Radiology
T. Brady U. HoffmannS. Abbara S. Achenbach