Detection of Lipid Core Coronary Plaques in Autopsy Specimens With a Novel Catheter-Based Near-Infrared Spectroscopy System
Craig M. Gardner, PhD*,*,
Huwei Tan, PhD*,
Edward L. Hull, PhD*,
Jennifer B. Lisauskas, MS*,
Stephen T. Sum, PhD*,
Thomas M. Meese, BS*,
Chunsheng Jiang, PhD*,
Sean P. Madden, PhD*,
Jay D. Caplan, BS, MBA*,
Allen P. Burke, MD ,
Renu Virmani, MD ,
James Goldstein, MD ,
James E. Muller, MD*
* InfraReDx, Inc., Burlington, Massachusetts
Division of Cardiovascular Pathology, Armed Forces Institute of Pathology, Washington, DC
CVPath Institute, Gaithersburg, Maryland
Division of Cardiology, William Beaumont Hospital, Royal Oak, Michigan
View larger version (117K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1 Apparatus for Fixing an Arterial Sample, Gross and Histologic Views, and Close-Up of NIRS Catheter
(A) A coronary artery segment mounted in a fixture to ensure registration of intraluminal near-infrared spectroscopy (NIRS) signals with histology. The vertical rods permit precise sectioning at 2-mm intervals. (B and C) Gross and microscopic cross-sections. Histomorphometry outlines show the lumen (red), lipid core (blue), and a small lipid pool (yellow). The angle subtended by the lipid core is shown in green. Radii used to measure cap and core thickness are in black. (D) The imaging tip of the NIRS catheter lying on a U.S. penny.
|
|
View larger version (25K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2 Flow Chart of Inclusion Criteria
Flow diagram describing the selection of data that met prospectively specified histologic and spectral criteria for inclusion in the blinded validation. NIRS = near-infrared spectroscopy.
|
|
View larger version (50K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3 A NIRS Scan Correlates Well to Histologic Findings in Coronary Artery From an 85-Year-Old Male With a History of MI
(A) Chemogram image indicating artery wall lipid content (x axis = pullback in millimeters; y axis = rotation in degrees). Each pixel is marked with red for low probability and yellow for high probability of lipid core plaque of interest (LCP). The lipid core burden index (top right) indicates amount of lipid in scanned artery on a 0 to 1,000 scale. (B) Summary (block chemogram) of LCP presence at 2-mm intervals in 4 probability categories. (C) Map of histologic classifications (yellow = LCP; light orange = small or thick-capped fibroatheroma; dark orange = intimal xanthoma and pathologic intimal thickening; red = all other types). (D) Movat cross-sections from locations along the artery (dotted lines). Black bars denote 1 mm. Image interpretation: The chemogram shows prominent lipid core signal at 2 to 16 mm, occupying 180°. The block chemogram shows that the strongest LCP signals extend 5 to 11 mm. The NIRS signals at 18 and 42 mm correctly indicate absence of LCP. MI = myocardial infarction; other abbreviations as in Figure 1.
|
|
View larger version (15K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5 ROC Curves Showing Agreement Between NIRS Signals (as Indicated by Block Chemogram) and Histology
(A) Prespecified end point analysis to discriminate 2-mm blocks containing LCP from blocks containing all other tissue types in blocks with lumen diameters  3 mm. (B) Exploratory analyses conducted in blocks with lumen diameters 2.5 mm. (B1) Discrimination of blocks containing noncalcified LCPs versus blocks without lipid. (B2) Discrimination of LCPs identified by requiring block chemograms that span 3 or more consecutive 2-mm blocks from all other tissue types. AUC = area under the curve; ROC = receiver-operator characteristic; other abbreviations as in Figures 1 and 3.
|
|
View larger version (24K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6 Sources of False Positive and False Negative
The graph displays medians (horizontal lines) and interquartile ranges (boxes) of block chemogram readings for different histologic classifications. (A) The LCP-negative class expanded into histologic subtypes. Higher y-axis values for LCP-negative result in false-positive readings. Fibroatheroma that were too small or thick-capped to be defined as LCP (FA) were the major source of false-positive readings. (B) The LCP-positive class expanded into histologic subtypes. Lower y-axis values for LCP-positive result in false-negative readings. The LCPs with calcified cores (Early FA-CC and Late FA-CC) were the main source of false-negative readings. AIT = adaptive intimal thickening; BF = bland fibrous plaque; CF = calcified fibrous plaque; CN = calcified nodule; -diff = spatially diffuse lipid; FA = fibroatheroma without calcified core; FA-CC = fibroatheroma with calcified necrotic core; IX = intimal xanthoma; NML = normal tissue; PIT = pathologic intimal thickening; -pool = spatially localized lipid pools; TCFA =thin-capped fibroatheroma; other abbreviations as in Figure 3.
|
|
View larger version (14K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7 LCBI Values for Normalized Lipid Core Volume per Artery Segment
Medians (horizontal lines) and interquartile ranges (boxes) of lipid core burden index (LCBI) values for no, low, intermediate, and high ranges of normalized lipid core volume per artery segment based on histologic findings.
|
|
View larger version (13K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8 ROC Analysis of Agreement Between LCBI and Histologic Classification
(A) Prospective end point evaluated in all scanned artery segments of adequate quality for detecting presence of lipid core by analysis of NIRS LCBI. (B) Retrospective analysis conducted in extreme scanned artery segments with either 0 or >0.2 mm3 of lipid core volume per millimeter of artery length. Abbreviations as in Figures 1, 5, and 7.
|
|
|
