Image-Guided Surgery: Breast Cancer Detection Using Monoclonal Antibodies Conjugated to a Near-Infrared Fluorescent Dye in a Syngeneic Rat Model.

Abstract

Abstract Background: Incomplete tumor resections occur frequently in patients undergoing breast-conserving surgery. As the surgeon can only rely on palpation and visual inspection, real-time visualization of cancer cells during surgery is needed to increase the number of complete tumor resections. Near-infrared fluorescence (NIRF) imaging using an intra-operative camera system is a novel technique to assess the extent of disease during surgery. Advantages of NIRF light (700-900 nm) include high tissue penetration up to several centimeters deep and low autofluorescence providing sufficient signal-to-noise ratio. Currently, several NIRF dyes are available for chemical conjugation to molecules that can target tumors, e.g. antibodies. This study aimed to investigate whether monoclonal antibodies conjugated to a NIRF dye could detect primary breast carcinomas in a syngeneic rat model.Methods: The hormone-dependent syngeneic breast cancer rat model EMR86 and its derived cell line MCR86 were used. Tumors were induced by orthotopic implantation of fresh EMR86 tumor fragments of 1 mm3 into the flanks at four sites in female WAG/Rij rats (Harlan, the Netherlands). The mouse monoclonal antibody MG1 was used for tumor cell detection. MG1 (IgG2a) recognizes a membrane epitope of rat cancer cells of epithelial origin and is developed by our research group. MG1 was conjugated to the NIRF dye CW-800 (LI-COR, USA) and dye/protein ratio was determined using absorbance measurements (UltroSpec, Amersham, UK). Fluorescence imaging of cells, animals and organs was performed using IVIS Spectrum (Caliper, USA) and Odyssey (LI-COR, USA) equipment.Results: Conjugated antibodies MG1-CW800 bound specifically to cultured MCR86 cells. The signal intensity was linearly correlated with the number of cells and concentration of MG1-CW800 (R2=0.99, p<0.00001). Addition of unconjugated MG1 lowered the fluorescent signal in a competitive binding assay, showing that labeling did not influence binding properties. Intravenous injection of unconjugated MG1 (200 μg/rat) in 3 Wag/Rij rats and subsequent immunohistochemical analysis of tissue sections revealed a specific and exclusive binding of MG1 to the EMR86 tumors and no difference in staining intensity between 24 and 48hrs after injection. Intravenous injection of conjugated MG1-CW800 (200 μg/rat) demonstrated a clear fluorescent demarcation of EMR86 tumors and the surrounding tissue after spectral unmixing (paired t-test 8.17, p<0.00001). Mean fluorescence did not differ between 24 and 48hrs after injection (t=1.85, p=0.11). However, a two-fold increase of the dye/protein ratio led to a two-fold increase of the in vivo fluorescence (t=-6.40, p=0.0001).Conclusion: Intravenous injection of monoclonal antibodies MG1 conjugated to the NIRF dye CW800 enabled detection of primary breast carcinomas in the syngeneic hormone-dependent EMR86 rat model. This study demonstrates proof-of-concept for future labeling of clinically relevant antibodies (e.g. HER2/neu, EGF, VEGF). Translation of this technique to breast cancer patients can lead to intra-operative identification of optimal resection margins in order to increase the comple tumor resection rate. Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 5006.