Mucins (MUC1–MUC24) are a family of glycoproteins involved in cell signaling and barrier protection. They have been implicated in the progression of numerous malignancies including gastric, pancreatic, ovarian, breast, and lung cancer. Mucins have also been extensively studied with respect to colorectal cancer. They have been found to have diverse expression profiles amongst the normal colon, benign hyperplastic polyps, pre-malignant polyps, and colon cancers. Those expressed in the normal colon include MUC2, MUC3, MUC4, MUC11, MUC12, MUC13, MUC15 (at low levels), and MUC21. Whereas MUC5, MUC6, MUC16, and MUC20 are absent from the normal colon and are expressed in colorectal cancers. MUC1, MUC2, MUC4, MUC5AC, and MUC6 are currently the most widely covered in the literature regarding their role in the progression from normal colonic tissue to cancer.

Poor visualization of polyps can limit colorectal cancer screening. Fluorescent antibodies to mucin5AC (MUC5AC), a glycoprotein upregulated in adenomas and colorectal cancer, could improve screening colonoscopy polyp detection rate. Adenomatous polyposis coli flox mice with a Cdx2-Cre transgene (CPC-APC) develop colonic polyps that contain both dysplastic and malignant tissue. Mice received MUC5AC-IR800 or IRdye800 as a control IV and were sacrificed after 48 h for near-infrared imaging of their colons. A polyp-to-background ratio (PBR) was calculated for each polyp by dividing the mean fluorescence intensity of the polyp by the mean fluorescence intensity of the background tissue. The mean 25 μg PBR was 1.70 (±0.56); the mean 50 μg PBR was 2.64 (±0.97); the mean 100 μg PBR was 3.32 (±1.33); and the mean 150 μg PBR was 3.38 (±0.87). The mean PBR of the dye-only control was 2.22 (±1.02), significantly less than the 150 μg arm (p-value 0.008). The present study demonstrates the ability of fluorescent anti-MUC5AC antibodies to specifically target and label colonic polyps containing high-grade dysplasia and intramucosal adenocarcinoma in CPC-APC mice. This technology can potentially improve the detection rate and decrease the miss rate of advanced colonic neoplasia and early cancer at colonoscopy.

Introduction: Gastric cancer has a global incidence of ~1 million cases per year. Currently, complete surgical resection is the only opportunity for cure for patients. Bright labeling of tumor margins with tumor specific markers could improve outcomes. This report demonstrates the advantages of targeting gastric cancer with humanized anti-CEA and TAG-72 antibodies in mouse models. Methods: A cell line of human gastric cancer, MKN45, and a patient-derived tumor, KG8, were used to create xenograft models. To establish orthotopic models for each, ~1mm tumor fragments were harvested from subcutaneous models of these lines and sutured to the lesser curvature of the stomach. Humanized versions of TAG-72 (huCC44) and CEA (M5A) were conjugated to the near infrared dye IRDye800CW. IgG was conjugated to IRDye800CW and used as a control. For CEA-IR800 and IgG-IR800, mice received 50 µg IV and were imaged after 72 hours. For Tag72-IR800, mice received 75 µg IV and were imaged after 96 hours. Fluorescence imaging was performed on the mice using the LI-COR Pearl Imaging System. A tumor to background ratio (TBR) was calculated by dividing mean fluorescence intensity of the tumor versus normal gastric tissue. Results: Humanized anti-CEA and TAG-72 brightly labeled multiple mouse models of human gastric cancer (Table 1). The TBR for TAG-72 labeling of KG8 was 2.777. CEA had a superior TBR of 3.528 compared to 0.606 for the control (IgG) in KG8 models. In MKN models labelled with CEA, an excellent average TBR of 7.407 were seen compared to 0.668 for the control. Conclusion: Humanized anti-CEA and TAG-72 antibodies conjugated to near-infrared fluorescent dyes provide specific labeling of gastric cancers in orthotopic models. These tumor-specific fluorescent antibodies are promising potential clinical tools for visualization of gastric cancer margins at the time of surgical resection and improved surgical outcomes. Orthotopic Model Antibody Tumor to Background Ratio KG8 anti-TAG-72 2.777 KG8 anti-CEA 3.528 KG8 Control 0.606 MKN anti-CEA 7.407 MKN Control 0.668 Citation Format: Kristin E. Cox, Michael A. Turner, Siamak Amirfakhri, Thinzar M. Lwin, Pradipta Ghosh, Marygorret Obonyo, Kaitlyn M. Kelly, Robert M. Hoffman, Paul J. Yazaki, Michael Bouvet. Fluorescence labeling of human gastric cancer using novel tumor specific near infrared labeled antibodies [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 2380.