In vivo label-free structural and biochemical imaging of coronary arteries using an integrated ultrasound and multispectral fluorescence lifetime catheter system

Abstract

Existing clinical intravascular imaging modalities are not capable of accurate detection of critical plaque pathophysiology in the coronary arteries. This study reports the first intravascular catheter combining intravascular ultrasound (IVUS) with multispectral fluorescence lifetime imaging (FLIm) that enables label-free simultaneous assessment of morphological and biochemical features of coronary vessels in vivo. A 3.7 Fr catheter with a fiber-optic channel was constructed based on a 40 MHz clinical IVUS catheter. The ability to safely acquire co-registered FLIm-IVUS data in vivo using Dextran40 solution flushing was demonstrated in swine coronary arteries. FLIm parameters from the arterial wall were consistent with the emission of fluorophores present in healthy arterial wall (collagen, elastin). Additionally, structural and biochemical features from atherosclerotic lesions were acquired in ex vivo human coronary samples and corroborated with histological findings. Current results show that FLIm parameters linked to the amount of structural proteins (e.g. collagen, elastin) and lipids (e.g. foam cells, extracellular lipids) in the first 200 μm of the intima provide important biochemical information that can supplement IVUS data for a comprehensive assessment of plaques pathophysiology. The unique FLIm-IVUS system evaluated here has the potential to provide a comprehensive insight into atherosclerotic lesion formation, diagnostics and response to therapy.

Figure 1

System overview. The system, able to display IVUS images in real time using the clinical interface, is composed of a 3.7 Fr IVUS/FLIm catheter, an IVUS motor drive unit (MDU) modified to accommodate an optical channel, connected to a wavelength selection module (WSM). The WSM enables both coupling of the pulsed excitation light and spectral decomposition of the collected fluorescence signal. The signal from each band is sent to the photomultiplier tube (PMT) using different lengths of optical delay lines that enables measurement over four wavebands using a single detector. When acquiring bimodal data, co-registration is insured by sampling both FLIm and IVUS signal with a single digitizer.

Figure 2

Ability to acquire co-registered FLIm/IVUS data in vivo in coronary arteries was evaluated in swine. Intensity en face image of a section of left circumflex artery acquired in vivo in pig (a). A distance map derived from IVUS (b) enables distance correction of the intensity image (c). For each of the four channels, a lifetime image is derived (e). Channel 2 presents the most fluorescence due to presence of collagen and elastin and shows very uniform background as expected from a healthy pig (f). The FLIm data can be combined with IVUS to provide bimodal 3D renderings (g) as well as sections of the vessel (d) (30° orientation, corresponds to the dashed line of panel c).

Figure 3

The ability of FLIm to resolve targets using fluorescence lifetime and spectral contrast was evaluated in vivo in swine coronary arteries. In vivo images of stented section of swine circumflex artery. Fluorescent markers M1 & M2 are painted onto the stent to provide fluorescence contrast (a), present broad emission spectra (b). M1 emission presents a short (~3 ns) lifetime and can be easily discriminated from healthy vessel background (~5.3 ns) in channels 1 and 2. M2 has weaker emission in channels 1 and 2 but is readily identified in channel 4 (c). The ability to resolve fluorescent features based on both spectral and temporal parameters makes FLIm a powerful technique to discriminate areas of interest. Combining FLIm and IVUS images enables 3D rendering (d).

Figure 4

FLIm biochemical information supplements IVUS for the assessment of atherosclerotic lesion pathophysiology. Spectral ratio weighted lifetime images map the measured lifetime (color), as well as relative strength of the signal (brightness) for channels 1 (Collagen), 2 (Elastin, some collagen and lipids) and 3 (lipids, ceroid and lipofuscine) of the instrument. Section 1 presents a fibrocalcified lesion. Areas of superficial calcification (3 to 9 o’clock) present less fluorescence from collagen (channel 1) than the rest of the vessel. Section 2 presents diffuse intimal thickening (DIT, upper quadrant) as well as a fibrotic lesion with a deeper necrotic core (lower quadrant), easily identifiable in the IVUS cross section. The increased presence of collagen in the fibrotic area corresponds to high fluorescence intensity in channel 1. Infiltration of punctate foam cells (FC) in the lower right quadrant (see CD68) corresponds to increased lifetime in channel 2 and some low level fluorescence in channel 3. Section 3 presents a thin cap fibroatheroma (TCFA) in the lower right quadrant. The fluorescence signature of this area is characterized by an absence of channel 1 signal (no collagen) and long lifetime in channels 2 and 3 with a maximum emission in channel 3 (large amounts of lipids/ceroid/lipofuscine). Other locations in the section present a signature expected from DIT/fibrotic lesions (collagen/elastin). Calcification from 6–9 o’clock is readily detected by IVUS. Section 4 presents DIT (top, left) and a fibrotic lesion (bottom, right). The fibrotic area can be differentiated from DIT by its higher collagen content (increased channel 1 intensity, longer channel 2 lifetime). The absence of infiltrated FC (see CD68) in this fibrotic area is consistent with the much lower lifetime observed here with respect to the fibrotic region in section 1. Proteoglycans, identified by the alcian blue color in the 6 to 8 o’clock region of the Movat’s section may explain the increase in lifetime observed at this location in the channel 1 image. En face images of lifetime (e) and intensity ratio (f) used to compute ratio weighted lifetime images shown in (a).