Experimental animal model and surgical procedures

Our local animal ethics committee approved the animal study protocol, which followed the regulations for animal experiments of the Chinese PLA General Hospital. A total of 24 Sprague-Dawley male rats (age, 12 weeks; mean weight, 390 g) were housed in a temperature- and humidity-controlled environment with free access to food and water until the experiments. All surgical procedures and imaging studies were performed with continuous isoflurane inhalation anesthesia using a small animal anesthesia machine. During the anesthesia periods, the rats were placed on a heating pad maintained at 32°C and were administered lactated Ringer’s solution (10 mL/kg, intraperitoneal injection) at 1- to 2-hour intervals.

The in vivo experiments in the present study consisted of two parts (Fig 2). The first part involved imaging using the IVIS spectrum system. With the rat in the supine position, the femoral vessel and accompanying lymphatics were exposed via a longitudinal incision from the medial thigh to the groin.12 ICG or ICG-β-CD solution (1.25 mg/mL; 0.2 mL) were subcutaneously injected into the hind footpad using a 32-gauge needle. After injection, the rats were observed every 5 minutes for 60 minutes. The second part involved imaging using the clinical DIGI-MIH-001 system. The hair was removed from the lower abdomen and hindlimb area of the rats, and the imaging was performed without a skin incision. The rats were injected as described. To assess the clearance of ICG and the ICG-β-CD complex from the injection site, the right ankle was selected as the region of interest to measure the signal background ratio (SBR). NIR imaging was performed every 5 minutes for the first 60 minutes, every 15 minutes for the next 3 hours, and every 30 minutes for the final 8 hours.

The study flowchart. ICG, Indocyanine green; ICG-β-CD, indocyanine green–β-cyclodextrin.

All data are presented as the mean ± standard deviation. All statistical analyses were performed using GraphPad Prism, version 6 (GraphPad Software, Inc, La Jolla, CA). A P value < .05 was considered statistically significant. For two groups with repeated observations over time, two-way repeated measures analysis of variance with Sidak’s multiple comparisons test was initially performed. A potential nonlinear association between the time and SBR was explored using curve fitting with a third-order polynomial equation (cubic) and evaluated using the extra-sum-of-squares F test. Mixed models with the interaction of time and a grouping factor were used to compare the change in SBR over time between the two groups, which were implemented in the statistical software packages R (R Foundation for Statistical Computing, Vienna, Austria; available at: http://www.R-project.org) and EmpowerStats (X&Y Solution, Inc, Boston, MA; available at: http://www.empowerstats.com).