Fluorescently labeled therapeutic antibodies for detection of microscopic melanoma

Objectives/hypothesis: Detection of microscopic disease during surgical resection of melanoma remains a significant challenge. To assess real-time optical imaging for visualization of microscopic cancer, we evaluated three US Food and Drug Administration (FDA)-approved therapeutic monoclonal antibodies.

Study design: Prospective, basic science.

Methods: Melanoma cell lines (A375 and SKMEL5) were xenografted into the ears of immunodeficient mice. Bevacizumab, panitumumab, tocilizumab, or a nonspecific immunoglobin G (IgG) were covalently linked to a near-infrared (NIR) fluorescent probe (IRDye800CW) and systemically injected. Primary tumors were imaged and then resected under fluorescent guidance using the SPY (Novadaq, Toronto, Ontario, Canada), an NIR imaging system used in plastic and reconstructive surgeries to evaluate perfusion. Mice were also imaged with the Pearl Impulse small animal imager (LI-COR Biosciences, Lincoln, NE), an NIR imaging system designed for use with IRDye800CW. Postresection, small tissue fragments were fluorescently imaged and the presence of tumor subsequently confirmed by correlation with histology.

Results: All fluorescently labeled therapeutic monoclonal antibodies could adequately delineate tumor from normal tissue based on tumor-to-background ratios (TBR) compared to IgG-IRDye800CW. On serial imaging, panitumumab achieved the highest TBRs with both SPY and Pearl (3.8 and 6.6, respectively). When used to guide resections, the antibody-dye conjugates generated TBRs in the range of 1.3 to 2.2 (average, 1.6) using the SPY and 1.9 to 6.3 (average, 2.7) using the Pearl. There was no significant difference among the antibodies with either imaging modality or cell line (one-way analysis of variance).

Conclusions: Our data suggest that FDA-approved antibodies may be suitable targeting agents for the intraoperative fluorescent detection of melanoma.