Near infrared optical lymphography for cancer diagnostics

Abstract

A new molecular imaging modality has been developed to detect and locate positive axillary and sentinel lymph nodes non-invasively in breast cancer patients undergoing lymphoscintigraphy. The modality is based on fluorescent photon detection to locate the presence of indocyanine green (ICG) in the lymph subsequent to peritumoral injection of ICG into the breast. The imaging system consists of a gain-modulated intensified charge-coupled device (ICCD) camera, which captures low-intensity, near-infrared, and frequency-modulated photons. A four-fold ‘optical lymphography’ study was conducted to (1) examine fluorescence depth penetration and ICCD system accuracy at clinically relevant depths, (2) compare image quality of the ICCD system vs. conventional gamma imaging, (3) measure ICG pharmacokinetics in vivo, and (4) develop a clinical protocol while examining pre-clinical factors such as the outcome of combining ICG with sulfur colloids used in lymphoscintigraphy. The frequency-domain ICCD system was found to precisely detect modulation amplitude, IAC, and phase, θ, at depths up to 9 cm and with IAC accuracy less than 20% and θ less than 2º using an 80-mW laser incident on phantoms having ranging tissue optical properties. Significant differences in the mean depth of penetration owing to 0.62-ns lifetime and 100-MHz frequency increases were detected. An in vivo optical vs. nuclear image quality comparison demonstrated statistically similar (α=0.05) target-to-background ratios for optical (1.4+/-0.3) and nuclear (1.5+/-0.2). Alternatively, resulting image signal-to-noise ratios (SNR) from the ICCD system were greater than that achieved with a conventional gamma camera (pvalue<<0.01). Analysis of SNR versus contrast showed greater sensitivity of optical over nuclear imaging for subcutaneous tumors. In vivo and rapid detection of ICG in the blood-stream of nude mice was accomplished with a home-built avalanche photodiode dynamic fluorescence measurement system. Intensity data upon i.v. injection were regressed with a pharmacokinetic model describing the partitioning of ICG from the blood to the surrounding tissues. ICG blood-clearance was detected approximately 15 min after injection. Lastly, a human subject protocol was written, practiced, and federally approved for the application of optical lymphography. Furthermore, ICG was unaffected when mixed with sulfur colloids thus supporting the feasibility for combining fluorescence imaging with lymphoscintigraphy in breast cancer patients.