Early Detection and Diagnosis
Real-time intraoperative detection of melanoma lymph node metastases using recombinant vaccinia virus GLV-1h68 in an immunocompetent animal model†
Version of Record online: 30 SEP 2008
Copyright © 2008 Wiley-Liss, Inc.
International Journal of Cancer
Volume 124, Issue 4, pages 911–918, 15 February 2009
How to Cite
Kelly, K. J., Brader, P., Woo, Y., Li, S., Chen, N., Yu, Y. A., Szalay, A. A. and Fong, Y. (2009), Real-time intraoperative detection of melanoma lymph node metastases using recombinant vaccinia virus GLV-1h68 in an immunocompetent animal model. Int. J. Cancer, 124: 911–918. doi: 10.1002/ijc.24037
Conflict of Interest: A financial conflict of interest does exist for co-authors affiliated with Genelux Corporation. Specifically, Dr. Yong A. Yu and Dr. Aladar Szalay are shareholders of and are employed by Genelux Corporation. Dr. Nanhai Chen is employed by Genelux Corporation.
- Issue online: 11 DEC 2008
- Version of Record online: 30 SEP 2008
- Accepted manuscript online: 30 SEP 2008 12:00AM EST
- Manuscript Accepted: 12 SEP 2008
- Manuscript Received: 19 JUN 2008
- National Institute of Health. Grant Number: R01 CA 76416
- Flight Attendant Medical Research Institute (FAMRI)
- malignant melanoma;
- sentinel lymph node biopsy;
- oncolytic virus
There is a clinical need for improved intraoperative detection of lymph node metastases from malignant melanoma (MM). We aimed to investigate the use of recombinant vaccinia virus GLV-1h68, expressing green fluorescent protein (GFP), for real-time intraoperative detection of melanoma lymph node metastases in an immunocompetent animal model. Mice bearing foot pad tumors received intratumoral injections of GLV-1h68, and 48 hr later were evaluated for popliteal lymph node metastasis using noninvasive bioluminescence imaging and fluorescence imaging. Histologic analysis of lymph nodes was performed to determine sensitivity and specificity of virus-mediated detection. Intratumoral injection of GLV-1h68 into primary foot pad melanoma tumors resulted in viral transmission to popliteal lymph nodes, infection of lymphatic metastases, and transgene expression that was reliably and easily detected. Histologic confirmation demonstrated favorable operating characteristics of this assay (sensitivity 80%, specificity 100%, positive predictive value [PPV] 100%, negative predictive value [NPV] 91%). Detection of marker gene expression by GLV-1h68 allowed the detection of lymphatic metastases in an immunocompetent animal model of MM. This assay is rapid, sensitive, specific and easy to perform and interpret. As a candidate gene therapy virusfor killing cancer, GLV-1h68 may also have significant concomitant diagnostic utility in the staging of cancer patients. © 2008 Wiley-Liss, Inc.
Malignant melanoma is the most lethal form of skin cancer, accounting for 75% of skin cancer deaths in 2007.1 At present, ∼60,000 new cases of melanoma are diagnosed yearly. Early diagnosis and treatment are paramount. The 5-year survival rate for patients with localized, early stage disease is 99%, but this drops to 65 and 15% for patients with regional and distant stage disease, respectively.
The most important predictor of survival for melanoma patients is the tumor status of the regional lymph node basin.2, 3 Sentinel lymph node biopsy (SLNB) is the standard procedure for determining regional lymph node status for patients with intermediate-thickness melanoma. SLNB is a procedure whereby dye or radioactive tracer is injected into the site of cancer. As the dye travels along the lymphatics, the first regional lymph node that shows dye or radioactive tracer uptake is biopsied. This test that maps lymphatic trafficking is used to identify the node with the highest likelihood of cancer for biopsy and histologic documentation of presence or absence of nodal metastases. Detection of subclinical lymph node metastases identifies those patients in need of complete lymphadenectomy, the only therapeutic option for local control.
SLNB, however, has limitations. Intraoperative analysis of sentinel lymph nodes (SLNs) by frozen section lacks sensitivity and frequently fails to identify nodal disease.4–6 The majority of patients must undergo lymphadenectomy as a second procedure after detection of nodal disease by permanent section, immunohistochemistry and/or molecular analysis.
Oncolytic vaccinia viruses are replication-competent, attenuated viruses that selectively infect and kill cancer cells, sparing normal cells. The therapeutic efficacy of these viruses has been demonstrated in a wide variety of human malignancies in vitro and in experimental models.7–13 Vaccinia viruses are well-suited to act as oncolytic agents as they have a natural tropism for targeting tumors when administered in vivo. In addition, the large genome of vaccinia virus can accommodate the insertion of multiple transgenes without loss of stability or diminished replicative ability. GLV-1h68 is an oncolytic vaccinia virus genetically modified by insertion of various marker genes, one of which replaces the gene encoding viral thymidine kinase (TK). As a result of these modifications, this virus can only productively infect and replicate in rapidly dividing, transformed cells with upregulated cellular TK. GLV-1h68 carries the marker genes encoding GFP, Ruc and β-gal. Expression of these transgenes by cancer cells infected with the virus can be detected by fluorescence and bioluminescence imaging modalities.
This study investigates the potential of GLV-1h68 to track and detect lymph node metastases after direct injection into primary melanoma tumors, and aims to enable real-time intraoperative detection of lymph node metastases by fluorescence imaging for GFP expression in an immunocompetent animal model.
Material and methods
The human metastatic melanoma cell lines SK-Mel 2, SK-Mel 119 and SK-Mel 197 were a generous gift from Dr. Alan Houghton at Memorial Sloan-Kettering Cancer Center (MSKCC). They were grown in Roswell Park Memorial Institute (RPMI) medium supplemented with 10% fetal calf serum (FCS), 100 U/mL penicillin and 100 mg/mL streptomycin. The syngeneic malignant melanoma cell line B16-F10 was obtained from the American Type Culture Collection (ATCC, Rockville, MD) and grown in Dulbecco's modified Eagle's medium-high glucose (DME-HG) supplemented with 10% FCS, 100 U/mL penicillin, 100 mg/mL streptomycin and 1.5 g/L sodium bicarbonate. All cells were maintained in a humidified incubator at 37°C supplied with 5% CO2.
GLV-1h68 is a replication-competent, recombinant vaccinia virus derived from the vaccinia virus LIVP strain (Lister strain from the Institute of Viral Preparations, Moscow) whose construction was previously described.13 GLV-1h68 carries 3 gene cassettes inserted into the viral genome: a Renilla luciferase-GFP (ruc-gfp) fusion cassette in the F14.5L locus, a reverse inserted human transferrin receptor and ß-galactosidase cassette in the J2R locus (which encodes TK), and a ß-glucuronidase cassette in the A56R locus (encoding hemagglutinin).
Cells were plated at 2 × 104 per well in 12-well plates in 1 mL of medium per well. After 6 hr of incubation, GLV-1h68 in 100 μL Dulbecco's modified Eagle's medium (DMEM) with 2% FCS per well was added at multiplicities of infection (MOIs) of 10, 1.0 and 0.1 for B16-F10 cells, and 1.0, 0.1 and 0.01 for human cell lines. All conditions were performed in triplicate. At time intervals of 6, 12, 24 and 48 hr postinfection, cells were examined with a fluorescence inverted microscope (Nikon Eclipse TE300, Nikon Japan) for GFP expression and photographs taken.
5-bromo-4-chloro-3-indolyl-B-D-galactopyranoside (X-gal) cytochemistry
Cells were plated at 2 × 104 per well in 24-well plates with 1 mL of medium per well. After incubation for 6 hr, cells were infected with GLV-1h68 at an MOI of 5. At varying intervals, cells were fixed with 1% glutaraldehyde for 5 min. Cells were stained with 1 mg/mL X-gal in an iron solution of 5 mmol/L K4Fe(CN)6, 5 mmol/L K3Fe(CN)6 and 2 mmol/L MgCL2 for 4 hr. Cells were then washed with phosphate-buffered saline (PBS) and examined by light microscopy.
Cells were plated at 1 × 104 per well in 96-well plates with 100 μL of medium per well. After incubation for 6 hr, human cell lines were infected with GLV-1h68 at MOI 1.0 and B16-F10 at MOI 10.0 in 50 μL of medium per well. At varying intervals, cells were lysed and ß-galactosidase activity measured with an Enhanced ß-Galactosidase assay kit (Gene Therapy Systems, San Diego, CA) using a spectrophotometer (EL321e, Bio-Tek Instruments, Winooski, VT) at 570 nm. All samples were measured in triplicate.
Cells were plated at 2 × 104 per well in 96-well plates in 100 μL of medium per well. After incubation for 6 hr, cells were infected with GLV-1h68 at an MOI of 1.0 (human cell lines) or 10.0 (syngeneic cell line) for varying intervals. Coelenterazine (0.25 μg; Biotium, Hayward, CA) diluted in 50 μL of PBS was added per well for 10 min and luciferase activity was detected with a cooled high-sensitivity charge coupled device (CCD) camera (Xenogen IVIS 100 Caliper, Hopkinton, MA). Emitted photons were measured for 30 sec from ventral views. Images were analyzed with Living Image 2.5 software (Caliper, Hopkinton, MA). All samples were analyzed in triplicate.
Cells were plated at 2 × 104 cells per well in 12-well plates with 1 mL of medium. After incubation for 6 hr, cells were infected with GLV-1h68 at MOIs of 0, 0.01, 0.1 and 1.0 for the human melanoma cell lines and 0, 1, 10, 25, 50 and 100 for the syngeneic cell line. Vehicle-treated cells served as controls. Viral cytotoxicity was tested at daily intervals. Medium was removed daily and cells were lysed with 1.35% Triton-X solution to release intracellular lactate dehydrogenase (LDH). Standard LDH assay was performed using the Cytotox 96 nonradioactive cytotoxicity assay (Promega, Madison, WI) and a microplate reader (EL321e, Bio-Tek Instruments). Results are expressed as the percentage of surviving cells relative to control wells.
Establishment of an immunocompetent animal model of melanoma lymphatic metastasis
Four-to-6-week old homozygous C57BL/6J-TyrC-2J/J albino mice (carrying a spontaneous mutation in the tyrosinase gene) were purchased from Jackson Laboratories (Bar Harbor, ME). All animals were provided with food and water ad libitum. Animals received humane care in accordance with the “Guide for the care and use of laboratory animals (NIH)” and the animal protocol was approved by the animal care and use committee (IACUC) at MSKCC. Anesthesia was induced by a mixture of Isoflurane (2 L/min) and oxygen (4 L/min) in an induction chamber and was maintained by a nasal cone. B16-F10 cells (2 × 105 in 50 μL PBS) were injected to the right hind footpad. Gentle pressure was applied to the right hind limb just above the foot during injection to prevent extravasation of the cell suspension out of the footpad and up the leg.
Lymph node imaging
Detection of fluorescence
Three weeks after tumor implantation, 20 animals with tumor and 5 control animals (previously injected with PBS only) received intratumoral injections of 1 × 107 pfu of GLV-1h68 in 50 μL of PBS. Forty-eight hours after injection of virus, animals were sacrificed with CO2 inhalation. The medial aspect of the right hind limb was surgically exposed from the site of the footpad tumor to the right groin. Animals were imaged using the Leica MZFL3 Stereomicroscope (Leica Microsystems, Wetzlar, Germany) in both bright-field and fluorescent modes. The Retiga EX digital CCD camera (Qimaging, Burnaby, Canada) was used for image capture.
Detection of bioluminescence
Just prior to sacrifice for fluorescence imaging, animals were anesthetized with a mixture of Isoflurane (2 L/min) and oxygen (4 L/min) in an induction chamber. Anesthesia was maintained by a nasal cone. Animals underwent retro-orbital injection of coelenterazine (5 μL of 0.5 μg/μL solution diluted in 95 μL PBS). Anesthetized animals were then placed in a dark box for photon counting (IVIS 100 Caliper, Hopkinton, MA). Emitted photons were measured for 30 sec. Images were analyzed with Living Image 2.5 software (Caliper, Hopkinton, MA). Animals harboring tumor treated with intratumoral injection of PBS only (n = 3) and animals without tumor treated with GLV-1h68 (n = 3) served as controls.
Histologic analysis of lymph nodes
Immediately after imaging, lymph nodes were harvested and frozen in Tissue-Tek optimal cutting temperature (OCT) Compound (Sakura Finetek USA, Torrance, CA). Tissues were cut into 5-μm-thick sections on glass slides. Cryosections were fixed and stained with hematoxylin and eosin (H&E) and X-gal (1 mg/mL) in an iron solution of 5 mmol/L K4Fe(CN)6, 5 mmol/L K3Fe(CN)6, and 2 mmol/L MgCL2 to identify virally mediated lacz expression. Cryosections from each sample were also fixed and left unstained for light microscopic examination for the presence of black-pigmented melanoma cells.14
GLV-1h68 infection allows successful transgene expression in B16-F10 melanoma cells and in a panel of human melanoma cell lines
In all cell lines, GFP expression increased with increasing MOI. At all MOIs tested, the percentage of GFP-expressing cells increased gradually over time (Fig. 1). Very few cells expressed GFP at 6 hr after infection, but by 12 hr GFP expression was visible in all cell lines. Of the human melanoma cell lines, SK-Mel 119 demonstrated the greatest GFP expression, followed by SK-Mel 2 and SK-Mel 197. B16-F10 cells showed very few GFP-expressing cells at MOI 1.0, but significant expression at all time points at MOI 10. In all cell lines tested, abundant GFP expression was present by 48 hr after virus infection.
Luciferase activity paralleled GFP and β-galactosidase expression, gradually increasing over time. SK-Mel 119 cells demonstrated the greatest luciferase expression with 3.6 × 107 ± 5 × 106 photons/s/cm2/sr at 48 hr after infection with MOI 1.0. B16-F10 cells displayed 2.4 × 107 ± 9.2 × 106 photons/s/cm2/sr at 48 hr after infection with MOI 10.0 (Fig. 2). Quantified luciferase activity correlated well with visual findings by GFP microscopy. The genes encoding Ruc (ruc) and GFP (gfp) are present in the same gene fusion cassette in the GLV-1h68 virus genome and so are expected to display similar levels of expression.
X-gal cytochemistry and β-galactosidase assay
X-gal staining demonstrated results similar to GFP fluorescence microscopy. For all MOIs tested, the percentage of blue cells increased gradually over time, with very few blue cells at 6 hr, more at 12 hr and an abundance of cells by 48 hr in all cell lines. In addition to this qualitative data, lacz expression was assessed quantitatively with a β-galactosidase activity assay. At 12 hr, low levels of β-galactosidase were measured (0.030–0.080 mU). Increasing levels were measured at 24 and 48 hr. At 48 hr after infection with MOI 1.0, SK-Mel 2 demonstrated the greatest amount of β-galactosidase (2.03 ± 0.15 mU), followed by SK-Mel 119 (1.69 ± 0.24 mU), and SK-Mel 197 (0.39 ± 0.09 mU). At 48 hr after infection with MOI 10.0, B16-F10 cells demonstrated 1.64 ± 0.15 mU β-galactosidase activity. As expected, determination of β-galactosidase enzyme activity correlated with qualitative visual findings by X-gal cytochemistry (Fig. 3).
GLV-1h68 is cytotoxic to human and murine melanoma cells
All human melanoma cell lines demonstrated significant sensitivity to GLV-1h68, with >90% cells killed in SK-Mel 119 and SK-Mel 197 cell lines and 85% in SK-Mel 2 by day 7 at MOI 1.0. The murine melanoma cell line was significantly more resistant, requiring MOI 50 to achieve a successful cytotoxic effect. In all cell lines, the cytotoxic effect of the virus was dose dependent (Fig. 4).
GLV-1h68-guided detection of lymph node metastases
We found that GLV-1h68-mediated GFP expression in primary melanoma tumors and lymph node metastases can be detected with fluorescence imaging in vivo. All 20 tumor-bearing mice demonstrated a strong fluorescent signal at the primary tumor site, whereas control mice (those with PBS-treated tumors (n = 6) and those with no tumors that were virus treated; n = 6) had no signals. Four of 20 mice (20%) demonstrated a fluorescent signal at the right popliteal lymph node. Figure 5 illustrates these findings in animals with microscopic and gross nodal disease. No signals were detected over the thorax or other areas in any of the animals.
We found that GLV-1h68-mediated luciferase expression in primary melanoma tumors and popliteal lymph nodes can be detected with bioluminescence imaging in vivo. Again, all 20 tumor-bearing, virus-treated animals demonstrated luminescence at the primary tumor site, whereas control mice (those with PBS-treated tumors and those with no tumors but virus-treated; n = 6) had no luminescent signals. The same 4 of 20 mice (20%) demonstrated bioluminescence at the right popliteal area, indicating the presence of nodal metastasis (Fig. 6a).
Histologic analysis of lymph nodes
Histologic examination confirmed the presence of metastatic melanoma in popliteal lymph nodes in 5 of 20 (25%) animals. Lymph node metastases were easily detected by light microscopic examination for the presence of black melanin pigmentation. Three of 5 animals had grossly detectable invasion of lymph nodes with nodal tissue nearly replaced by tumor. Two of the 5 had microscopic metastasis.
Four of these 5 animals (80%) demonstrated measurable luminescent and fluorescent lymph node signals (2 with gross and 2 with microscopic nodal metastases). Areas of GFP expression localized to tumor foci in lymph nodes (Fig. 7). Absence of tumor and GFP in negative lymph nodes (as determined by fluorescent imaging) was similarly confirmed by microscopic analysis. Finally, X-gal cytochemistry detected GLV-1h68-mediated lacz expression in the popliteal lymph nodes of the same 4 animals with positive findings by bioluminescent and fluorescent imaging, further confirming successful transgene expression by the virus in areas of metastatic disease. The sensitivity, specificity, positive predictive value [PPV], and negative predictive value [NPV] of virally mediated fluorescent detection of nodal metastasis was 80, 100, 100, and 93%, respectively (Table I).
Many oncolytic viruses have been designed for specific, targeted killing of cancer. The genetic alterations of many of these viruses include insertion of transgenes, such as immune-stimulating genes for enhanced therapeutic efficacy or marker genes for monitoring of therapy. Vaccinia virus is particularly well suited for oncolytic therapy because it is known to be safe in humans as it has been given to millions of people with acceptable toxicity in the form of the smallpox vaccine, it has a large genome that is able to incorporate multiple transgenes, and it is highly immunogenic so that infected cancer cells will incite an immune response, further contributing to the anticancer effect.15, 16 GLV-1h68 is an attenuated oncolytic vaccinia virus carrying 3 marker genes that has been shown to be highly cancer specific.13 We have demonstrated for the first time that GLV-1h68 successfully tracks lymph node metastases from malignant melanoma and allows for rapid and simple detection of metastatic disease in an immunocompetent animal model.
Multiple studies have shown that for patients with intermediate-thickness cutaneous melanoma, regional lymph node status is the most important predictor of recurrence and survival.17–20 The current standard technique of injecting dye or radiotracer for SLNB successfully identifies the sentinel node, but provides no information about the status of that node until resection and processing. It has been shown that intraoperative frozen section of SLNs of patients with cutaneous melanoma is only 38% sensitive and is therefore not justified for routine use.4–6 Imprint cytology has been investigated in attempts to improve sensitivity of intraoperative detection of sentinel node disease. Results have shown, however, that imprint cytology is equivalent to but not better than frozen section analysis, with reported sensitivities of 38%21 and 33%22 for detection of melanoma nodal metastases. Without a reliable assay for intraoperative assessment, melanoma patients with positive SLNs must undergo lymph node dissection as a second procedure. Even permanent section histologic analysis has been shown to lack sensitivity relative to newer molecular detection techniques such as immunohistochemistry (IHC) and reverse-transcriptase polymerase chain reaction (RT-PCR).23
GLV-1h68 was able to track lymph node metastases 48 hr after injection into the primary tumor in an immunocompetent animal model of metastatic melanoma. Vaccinia virus-mediated GFP expression offers the possibility of immediate identification of regional lymph nodes harboring metastatic disease in the operative setting by the presence or absence of easily identifiable green fluorescence. In our study, the transgenes ruc and lacz were used to confirm that green fluorescence detected in lymph nodes was due to GFP expression as opposed to autofluorescence.
We found that GLV-1h68-mediated, real-time fluorescence detection of nodal disease was 80% sensitive and 100% specific for identification of sentinel node disease. One animal with macroscopic disease of the popliteal lymph node did not display GLV-1h68 transgene expression. Of note, the fluorescent and bioluminescent signals from the primary footpad tumor in this animal were particularly weak when compared with those of other animals in the study, indicating that there may have been a technical problem with injection of virus in this animal.
It is also possible that GLV-1h68 may allow for intraoperative detection of nonsentinel node disease. Our model is limited for assessing this utility as the regional nodal basin draining the mouse footpad consists of a solitary popliteal node. The next lymph node in the drainage system is the inguinal node, which is significantly proximal. In our study, no animals had inguinal node metastases. An alternative model to test the ability of the virus to detect nonsentinel node disease would be a mammary fat pad or truncal tumor with multiple lymph nodes in a regional drainage basin. Eisenberg et al.24 demonstrated that GFP-expressing herpes simplex virus is able to identify nonsentinel node disease in a xenograft model of breast cancer metastatic to the axillary lymph nodes.
In addition to offering increased sensitivity of intraoperative detection, GLV-1h68 may confer an additional therapeutic advantage. A recombinant vaccinia virus, rV-B7.1, expressing a T-cell costimulatory molecule was shown to be cytotoxic to human melanoma cells and to produce biologic effects with an acceptable safety profile in a phase I clinical trial.10, 25, 26 GLV-1h68 has been shown to be an effective oncolytic agent against human breast cancer, anaplastic thyroid cancer, and malignant pleural mesothelioma cell lines in vitro and in vivo.11, 13 We have shown that GLV-1h68 was cytotoxic to malignant melanoma cell lines in vitro and in vivo (unpublished data).
Fluorescence detection of lymphatic metastases with oncolytic vaccinia virus is a novel, sensitive and specific technique for identification of metastatic melanoma. Advantages of this method compared with standard techniques include real-time intraoperative detection and the potential for a simultaneous therapeutic effect. In addition, this assay is both easy to perform and interpret, and it is cost-effective. This technique may benefit an enormous population of melanoma patients at least by allowing SLNB and lymph node dissection to be performed in a single procedure, and potentially by offering enhanced staging accuracy and a therapeutic effect.
This research was supported by grants R01 CA 76416 (Y.F.) from the National Institute of Health, the Flight Attendant Medical Research Institute (FAMRI), and by grant support to the Research and Development Division of Genelux Corporation (A.A.S.). The authors thank the editor, Meryl Greenblatt, for her outstanding assistance in this work.
- 6A case of penile melanoma illustrating the low sensitivity of frozen sections in the assessment of sentinel lymph nodes. Can J Urol 2003; 10: 1947–9., , , .