Vaccinia virus GLV-1h153 is a novel agent for detection and effective local control of positive surgical margins for breast cancer

Abstract

Introduction: Surgery is currently the definitive treatment for early-stage breast cancer. However, the rate of positive surgical margins remains unacceptably high. The human sodium iodide symporter (hNIS) is a naturally occurring protein in human thyroid tissue, which enables cells to concentrate radionuclides. The hNIS has been exploited to image and treat thyroid cancer. We therefore investigated the potential of a novel oncolytic vaccinia virus GLV1h-153 engineered to express the hNIS gene for identifying positive surgical margins after tumor resection via positron emission tomography (PET). Furthermore, we studied its role as an adjuvant therapeutic agent in achieving local control of remaining tumors in an orthotopic breast cancer model.

Methods: GLV-1h153, a replication-competent vaccinia virus, was tested against breast cancer cell lines at various multiplicities of infection (MOIs). Cytotoxicity and viral replication were determined. Mammary fat pad tumors were generated in athymic nude mice. To determine the utility of GLV-1h153 in identifying positive surgical margins, 90% of the mammary fat pad tumors were surgically resected and subsequently injected with GLV-1h153 or phosphate buffered saline (PBS) in the surgical wound. Serial Focus 120 microPET images were obtained six hours post-tail vein injection of approximately 600 μCi of 124I-iodide.

Results: Viral infectivity, measured by green fluorescent protein (GFP) expression, was time- and concentration-dependent. All cell lines showed less than 10% of cell survival five days after treatment at an MOI of 5. GLV-1h153 replicated efficiently in all cell lines with a peak titer of 27 million viral plaque forming units (PFU) ( <10,000-fold increase from the initial viral dose ) by Day 4. Administration of GLV-1h153 into the surgical wound allowed positive surgical margins to be identified via PET scanning. In vivo, mean volume of infected surgically resected residual tumors four weeks after treatment was 14 mm3 versus 168 mm3 in untreated controls (P < 0.05).

Conclusions: This is the first study to our knowledge to demonstrate a novel vaccinia virus carrying hNIS as an imaging tool in identifying positive surgical margins of breast cancers in an orthotopic murine model. Moreover, our results suggest that GLV-1h153 is a promising therapeutic agent in achieving local control for positive surgical margins in resected breast tumors.

Figure 1

Cell lines MDA-MB-468, MDA-MB-231 and HCC38 show sensitivity to infection with GLV-1h153. A. Viral infection was concentration-dependent as shown by increasing number of GFP-expressing cells. B. Viral infection demonstrated to be time-dependent. GFP expression increased early after infection as shown from Day 1 to Day 3, but then decreased when cell death occurred (by Day 6). Illustrated were images from infection at an MOI of 1.

Figure 2

GLV-1h153 kills and replicated efficiently in breast cancer cell lines in vitro. A. GLV-1h153 kills breast cancer cells in a dose-dependent fashion. Cytotoxicity data of cell lines MDA-MB-231, MDA-MB-468 and HCC38 showed less than 10% viable cells after five days of infection at an MOI of 5. Multiplicity of infection (MOI) = plaque-forming units/cell. Experiment was performed in triplicates. B. GLV-1h153 replicated efficiently in all breast cancer cell lines. All cell lines supported viral replication as assessed by the viral plaque assay. Results demonstrated that in cell line MDA-MB-468, GLV-1h153 reached the highest titer of 2.7 × 107 PFU after 96 hours of infection, representing over a 10,000-fold increase in copy numbers from the initial viral dose. The experiment was performed in triplicates.

Figure 3

Surgical margins of breast cancer after resection can be detected via mCherry fluorescence imaging in vivo. A. Photographs of an animal pre- and post-surgical resection of the mammary fat pad tumor shown in upper panel. Tumor is outlined by dotted line. B. Bottom panel shows a photograph after surgical resection and infection with a closed wound. To the right is shown a mCherry Maestro fluorescence image to confirm that a positive margin was present prior to PET imaging.

Figure 4

GLV-1h153 can identify positive surgical margins of breast cancer after resection via hNIS-mediated radio-uptake. A. microPET images shown six hours after radiotracer injection of all three infected animals after surgical resection. B. Three-dimensional (3D) view of PET images of one infected animal at the surgical margin.

Figure 5

GLV-1h153 effectively prevents progression of residual breast tumor at surgical margins in vivo. Residual tumors at surgical margins of control mice grew from an average size of 25 mm³ to 169 mm³ (over a four-fold increase in tumor volume) compared to the average tumor size of treated mice which regressed from 20 mm³ to 14 mm³ in infected surgical margins within four weeks of treatment after resection (P < 0.05). Three out of six of the treated mice demonstrated complete regression of residual tumor.

Figure 6

GPF-expressing malignant cells are detected in positive surgical margins with fluorescent microscopy. Uninfected control tumors (treated with PBS) express mCherry only and no GFP signal (top panel). Bottom panel represents an infected remaining tumor margin four weeks after treatment with GLV-1h153. The tumor expresses mCherry as well as GPF signal.

Figure 7

H&E slides of untreated and treated positive surgical margins. Positive-margin controls (left) show malignant cells with large nuclei and irregular chromatin. Treated positive margins H&E slides (right) represent tissue margins four weeks after treatment with GLV-1h153. Large areas of fibrosis and necrosis can be seen without any viable cancer cells.