Evaluation of Lung Quality by Near-Infrared Fluorescent Imaging during Ex Vivo Lung Perfusion

Abstract

Improved real-time assessment of donor lung quality during Ex Vivo Lung Perfusion (EVLP) could enhance safe utilization of lungs for transplantation. Near-infrared (NIR) fluorescent imaging with indocyanine green (ICG) has been used in various clinical intraoperative applications to evaluate tissue perfusion. We investigated the feasibility of NIR imaging during EVLP and aimed to detect and quantify lung perfusion and permeability. Pig lungs with extended 24-hour cold ischemia (n=4) and human lungs rejected from clinical transplantation (n=2) were assessed with NIR imaging during EVLP. Serial NIR imaging was performed with SPY Elite imaging system after ICG (0.3, 0.6 or 1.0 mg) administration through pulmonary artery (Fig A). Tissue perfusion was quantified using SPYQ analysis software, and tissue ICG accumulation was correlated with lung damage. Strong NIR signal was detected immediately after ICG administration with 0.6 mg (pig lungs) or 1.0 mg (human lungs) doses resulting in optimal fluorescence for tissue perfusion evaluation. In contrast to the transient kinetics and fast liver metabolism during clinical in vivo ICG administration, a single ICG dose during EVLP resulted in stable NIR signal for over 10 hours. Perfusion quantification of pig lungs with extended cold preservation revealed peripheral areas with compromised tissue perfusion that gradually recovered over time with recruitment of additional pulmonary vascular beds (Fig B). ICG accumulation was detected in sites with damage and increased vascular permeability in pig and human lungs (Fig C and D). NIR imaging with a clinically-approved intravascularly-delivered ICG fluorescent probe can dynamically quantify tissue perfusion and detect damage-induced vascular permeability. This novel and promising imaging modality could be used during EVLP to assess lung quality, or to determine effects of therapeutics that target tissue perfusion and vascular permeability.