Abstract
1022
Objectives: The introduction of contrast agents for fluorescence-guided surgery (FGS) has equipped surgeons with real-time imaging to complement standard surgical navigation techniques1,2. The addition of a radioactive label to the FGS agent adds further utility by providing quantitative measurements of drug distribution2. Although clinical radiotracers are ideal models for dual labeling based on their proven diagnostic accuracy, maintaining target specificity following dye conjugation is challenging. Accordingly, we developed a multimodality chelator (MMC) that acted as a “radioactive linker”, and performed biorthogonal dye conjugation to the neuroendocrine tumor (NET) imaging agent, 68Ga-DOTA-TOC. This strategy produced a fluorescent 68Ga-DOTA-TOC analog with excellent retention of receptor binding properties in vitro3. Here, we performed multiscale evaluation of the tumor-targeting properties of the dual-labeled fluorescent analog and examined the use of dual labeling for cross-validation.
Methods: Tyr3-octreotide (TOC) was conjugated to the MMC on solid-phase and IRDye 800CW in solution-phase in a site-specific manner to produce MMC(IR800)-TOC. The dual conjugate was radiolabeled with 67Ga and 68Ga using cation exchange chromatography. Mice implanted with somatostatin receptor subtype-2 (SSTR2) overexpressing HCT116 cells (HCT116-SSTR2) were used for in vivo and ex vivo near-infrared fluorescence (NIRF) imaging, and biodistribution analysis (n=5). In vivo specificity was evaluated by NIRF imaging of mice with non-SSTR2-overexpressing HCT116 tumors (HCT116-WT) and the non-targeted MMC analog, Ga-MMC(IR800), in HCT116-SSTR2 mice (n=5). Histological and microscopic analyses were performed on tumor and non-tumor frozen sections.
Results: 67Ga/68Ga-MMC(800)-TOC was synthesized with crude radiochemical yields of 82.0±4.1% and 79.9±8.1%, respectively. In vivo NIRF imaging at 3 h showed some tumor delineation, alongside prominent kidney signal and background fluorescence in the thoracic and abdominal walls. These results were confirmed by ex vivo NIRF imaging, and tumor/background ratios (TBRs) ranging from 1.4±0.4 (lung) to 4.6±0.5 (muscle) were determined in key organs. 68Ga-biodistribution data were in agreement with NIRF results and showed a 3.5±0.9 %IA/g (injected activity/gram of tissue) in the tumor and high renal excretion (45.6±3.8 %IA/g). Delayed NIRF imaging (24 and 48 h) showed excellent tumor delineation and time-dependent contrast enhancement, suggesting increased agent uptake, elimination from background tissues, or a combination of the two. Maximal contrast in key sites of NET formation was obtained at 48 h, with TBRs of 4.3±1.3 (pancreas), 7.6±2.0 (lung), and 17.2±3.5 (small intestine). The corresponding 67Ga-biodistribution showed tracer retention in the tumor (6.7±0.9 %IA/g) and significant elimination from background tissues, yielding TBRs of 34.2±13.9 (muscle) and 81.0±35.3 (blood). Specificity studies showed clear Ga-MMC(IR800)-TOC signal in HCT116-SSTR2 tumors but not in HCT116-WT tumors. Ex vivo NIRF imaging supported the in vivo results and subsequent image analysis showed a higher average radiant efficiency emitted from the HCT116-SSTR2 xenograft compared to the WT tumors and non-targeted analog (P < 0.01). Meso- and microscopic analyses of frozen sections further confirmed the in vivo and ex vivo imaging results, and showed specific Ga-MMC(IR800)-TOC uptake only in HCT116-SSTR2 tumors.
Conclusions: These results show that 68Ga-DOTA-TOC can be converted into a fluorescent analog with SSTR2-specificity in vivo. The high TBRs in tissues where NETs typically originate suggest excellent intraoperative utility for Ga-MMC(IR800)-TOC. Importantly, the use of dual labeling provided definitive measurements of drug distribution for cross-validation of NIRF imaging data.