Comparative Quantification of Arterial Lipid by Intravascular Photoacoustic-Ultrasound Imaging and Near-Infrared Spectroscopy-Intravascular Ultrasound

Abstract

Intravascular photoacoustic-ultrasound (IVPA-US) imaging and near-infrared spectroscopy-intravascular ultrasound (NIRS-IVUS) are two hybrid modalities that detect arterial lipid, with comparison necessary to understand the relative advantages of each. We performed in vivo and ex vivo IVPA-US imaging of the iliac arteries of Ossabaw swine with metabolic syndrome (MetS) and lean swine to investigate sensitivity for early-stage atherosclerosis. We repeated imaging ex vivo with NIRS-IVUS for comparison to IVPA-US and histology. Both modalities showed significantly greater lipid in MetS vs. lean swine, but only IVPA-US localized the lipid as perivascular. To investigate late-stage atherosclerosis, we performed ex vivo IVPA-US imaging of a human coronary artery with comparison to NIRS-IVUS and histology. Two advanced fibroatheromas were identified, with agreement between IVPA-measured lipid area and NIRS-derived lipid content. As confirmed histologically, IVPA-US has sensitivity to detect lipid content similar to NIRS-IVUS and provides additional depth resolution, enabling quantification and localization of lipid cores within plaques.

Keywords: Atherosclerosis; Human; Intravascular imaging; Lipid core plaque; Near-infrared spectroscopy; Perivascular adipose tissue; Photoacoustic imaging; Swine.

Fig. 1

IVPA-US imaging probe and in vivo imaging procedure. a Schematic of the IVPA-US collinear catheter design, in which the acoustic and optical paths overlap, after a series of reflections off the rod mirror and optical fiber surfaces. b Schematic of the in vivo procedure, in which the IVPA-US catheter was advanced 80 mm into the iliac artery of swine. c Angiogram with contrast to visualize vascular anatomy. d Angiogram without contrast to confirm IVPA-US catheter placement at 80 mm distal to the introducer sheath

Fig. 2

Representative imaging and histology from a MetS Ossabaw swine iliac artery showing no intimal lipid deposition and only early-stage neointimal thickening. a In vivo IVPA-US still frame, in which no lipid deposition (absence of red) and mild neointimal hyperplasia (three-layer appearance, grayscale IVUS channel) was observed. b In vivo single-modality grayscale IVUS still frame, depicting the three-layer appearance. c Ex vivo IVPA-US still frame, in which no lipid deposition and the three-layer appearance were observed. d Ex vivo NIRS-IVUS still frame, in which no lipid deposition (absence of yellow circumferentially) and the three-layer appearance (grayscale IVUS channel) were observed. e Verhoeff–Van Gieson stained histological section, showing typical artery morphology with a small area of neointimal thickening (NEO) on the luminal side of the internal elastic lamina (IEL). M media, A adventitia. Horizontal and vertical axis tracings are 1 mm apart

Fig. 3

Representative in vivo IVPA-US imaging from a MetS Ossabaw swine iliac artery showing a lipid signal by both modalities. a Unmerged IVUS channel still frame showing a three-layer appearance of the iliac artery. b Unmerged IVPA channel still frame showing circumferential distribution and depth of lipid signal (red). c Merged composite showing depth-resolved lipid signal localized to the perivascular region. d NIRS-IVUS still frame showing grayscale IVUS channel and lipid signal (yellow) circumferentially, but not the precise radial depth into the arterial wall and perivascular regions. e Verhoeff–Van Gieson stained histological section showing abundant perivascular adipose tissue (PVAT, inset), noted by the presence of round and empty adipocytes. I intima, M media, A adventitia. Horizontal and vertical axis tracings are 1 mm apart

Fig. 4

Average cross-sectional lipid area in iliac arteries as determined by IVPA-US imaging. The average cross-sectional lipid area, as calculated by photoacoustic signal intensity per frame above a pre-determined threshold, is significantly greater in the iliac arteries of swine with MetS as compared to lean swine, both by in vivo (a) and ex vivo (b) imaging; *** p < 0.0001; n refers to number of cross-sectional frames analyzed

Fig. 5

Correlation between modalities in a fresh human coronary artery. a Ex vivo angiogram of the artery, in which the sheath (I), side branches (II, V), and lesions locations (III, IV) are identified. Location III refers to Lesion 1 and location IV to Lesion 2. b, c Corresponding cross-sectional still frames are shown from NIRS-IVUS (b) in which lipid is shown with yellow and from IVPA-US (c) in which lipid is represented by red. Notable side branches are indicated by arrowheads and calcification by arrows. d Movat’s pentachrome stained histological sections of lipid core lesions (*) with calcification (arrows). Horizontal and vertical axis tracings are 1 mm apart