Pedro R' Moreno, MD, FACC

1
Detection of Vulnerable Atherosclerotic Plaques
By Near Infrared Spectroscopy
Pedro R. Moreno, MD, FACC University of
Kentucky Lexington, Kentucky
2

• Biological tissues have unique absorbance in the
  NIR wavelength range
• NIR light has enough penetration that may obtain
  spectra through blood

3
Tissue Evaluation by Near-IR Spectroscopy
Absorbance peaks are caused by Combinations of
fundamental bonds (C-H, CC, CO) Electron
transitions in the heaviest atoms
4
Advantages of Near-IR Spectroscopy For Vulnerable
Plaque Research

• Analysis under 1 second
• Simultaneous, multi-component, non- destructive
  analysis
• Chemical, biological and molecular information
• Automated predictions using computer algorithms
• Detection limits can be very low (from picograms
  to planets)
• Cost per analysis is minimal (no reagents used)

Dempsey RJ Lodder RA. Applied Spectroscopy
19965018A-34A
5
Near-IR Spectroscopy to Identify Vulnerable
Plaques
Studies 1998-2001

• Human aortic plaques in-vitro
• Human coronary plaques in-vitro
• Rabbit aortic plaques in-vivo
• Coronary pig safety study
• Human coronary plaques in-vivo

6
Plaque Composition by Near-IR Spectroscopy
198 Human Aortic Plaques
Near-IR System

• Hypothesis
• NIRS will identify vulnerable
• plaques (1)
• Methods
• Spectrometer InfraAlyzer 500 (2)
• H E and Trichrome staining
• Identification Algorithm Model (3,4)
• 50 - Training set (histology)
• 50 - Validation set

Blinded prediction
1. Lipid pool (gt30), thin cap (lt 65 mm), and
macrophages 3. Chemometric software (Math 3.0,
Matlab 5.1,Speakease IV Eta) 2. (BranLuebbe,
Elmsford, N.Y.) 4. Regression by
principal component analysis
7
Near Infrared in Human Aortic Plaques
(n198)
Lipid
Thin Cap
Macrophages
Fibrotic
Thick Cap
No Macrophages
Moreno PR, et al. Circulation 2002105923-927
8
Correlation of Blinded Near-IR Spectroscopy Result
s with Histologic Findings
99 Aortic samples
HISTOLOGY
LIPID POOL
THIN CAP
MACROPHAGES

-

-

-

35
4
13
6
37
6
NEAR-INFRARED
SPECTROSCOPY
-
4
56
4
76
7
49
Moreno PR, et al. Circulation 2002105923-927
9
Correlation of Blinded Near-IR Spectroscopy Result
s with Histologic Findings
99 Aortic samples
Lipid Pool Thin Cap Macrophages

• Sensitivity () 90 77 84
• Specificity () 93 93 89
  
• PPV () 90 68 86
• NPV () 93 95 88

PPVPositive Predictive Value NPVNegative
Predictive Value
10
Near-IR Spectroscopy to Identify Vulnerable
Plaques
Studies 1998-2001

• Human aortic plaques in-vitro
• Human coronary plaques in-vitro
• Rabbit aortic plaques in-vivo
• Coronary pig safety study
• Human coronary plaques in-vivo

11
Coronary Composition by Near-IR Spectroscopy
147 Human Coronary Sections
Spectrometer

• Hypothesis
• Lipid pool in coronary plaques
• Methods
• Spectrometer Foss/NIRSystems
• H E and Trichrome staining
• Identification Algorithm Model
• Training Set (76 sections)
• Validation set (70 sections)

Blinded prediction
Moreno PR, et al. JACC 200137356A
12
Coronary Near-Infrared Spectra
Normal Artery
Fibrotic Plaque
Lipid-Rich, Calcified
Thick Cap Atheroma
Thin Cap Atheroma
Moreno PR, et al. JACC 200137356A
13
Coronary Plaque Lipid Pool Detection by Near-IR
Spectroscopy
Validation set (70 sections)

• Sensitivity () 95
• Specificity () 96
• PPV () 91
• NPV () 98

HISTOLOGY

-

21
2
NEAR-INFRARED
SPECTROSCOPY
-
1
46
Moreno PR, et al. JACC 200137356A
14
Near-IR Spectroscopy to Identify Vulnerable
Plaques
Studies 1998-2001

• Human aortic plaques in-vitro
• Human coronary plaques in-vitro
• Rabbit aortic plaques in-vivo
• Coronary pig safety study
• Human coronary plaques in-vivo

15
Identification of Lipid-rich Aortic
Atherosclerotic Plaques in Living Rabbits With a
Near-IR Spectroscopy Catheter
Normal
Normal

• Hypothesis
• Normal vs. atherosclerotic plaques
• Lipid-rich versus lipid-poor plaques
• Model
• Atherosclerotic Rabbit Model
• Pulsed 1 cholesterol x 8 months
• Normal rabbits (controls)
• Near-IR Spectroscopy
• Laser-driven catheter system
• Histology
• H E and Trichrome staining
• Computerized planimetry (Zedex software)

Atherosclerotic
Moreno PR, et al. JACC 2001 373A1039-21
16
In-vivo Detection of Groups of Lipid Plaques With
a Near-IR Spectroscopy Catheter
HISTOLOGY
True ()
False ()

-

19
0
NEAR-INFRARED
SPECTROSCOPY
-
5
6
True (-)
False (-)
Presence of lipid Sensitivity
79 Specificity 100 Lipid area gt0.75 mm2
Sensitivity 75 Specificity 78
Moreno PR, et al. JACC 2001373A
17
Near-IR Spectroscopy to Identify Vulnerable
Plaques
Studies 1998-2001

• Human aortic plaques in-vitro
• Human coronary plaques in-vitro
• Rabbit aortic plaques in-vivo
• Coronary pig safety study
• Human coronary plaques in-vivo

18
Percutaneous Coronary Near-IR Spectroscopy In
vivo A Safety Study

• Six Normal Swines
• Percutaneous, over-the wire NIR coronary
  catheterization performed in 2/3 coronary
  arteries
• Results
• Successful coronary catheterization in all cases
• Excellent angiographic and histologic results
  with not a single case of dissection, thrombosis
  or perforation.

3 French NIR catheter
Moreno PR and Fallon JT. University of Kentucky
and Mount Sinai School of Medicine, August, 2001
19
Near-IR Spectroscopy in Humans
3 French Catheter

• Phase I - Safety
• Stable angina / PTCA-Stent
• Reference normal segment
• Prospective Study
• Angioplasty/stenting
• Scan 3 major arteries
• Follow/up 12 months
• Correlation spectra/events

20
Trial of Detection Treatment of Vulnerable
Plaque
Randomize
Patients with angina Cath Lab PTCA/ Stenting
Patients with TCFA
1 year follow/up
Rx of VP
Placebo
Near-IR
UA AMI SCD
Patients w/out TCFA
TCFAThin-cap fibroatheroma UAunstable angina
- AMI acute myocardial infarction - SCDsudden
cardiac death
21
Near-IR Spectroscopy Vulnerable Plaques
Conclusions

• NIR spectroscopy can identify features of plaque
  vulnerability in vitro and lipid-rich plaques in
  vivo, through blood.
• A catheter-based system has been tested in-vivo
  with excellent performance in both swine and
  human coronary arteries.
• Additional clinical data are needed to
  definitively apply this technique to risk
  stratify human atherosclerotic lesions in the
  cath lab.