We would also like to thank the National Cancer Institute for generously donating the 14.GD2a antibody.
Survival outcomes for patients with osteosarcoma have remained stagnant over the past 30 years. Targeting of ganglioside GD2, a glycosphingolipid on the cell surface of some tumors, with immunotherapy has resulted in improved outcomes for patients with neuroblastoma. In the current study, the expression pattern of GD2 was examined in osteosarcoma.
Immunohistochemistry was performed on osteosarcoma samples from patients at the time of initial biopsy, definitive surgery, and disease recurrence. The intensity and location of staining were scored. Cell-based enzyme-linked immunoadsorbent assay was performed on osteosarcoma cell lines to quantitate the level of GD2 expression.
Forty-four osteosarcoma samples were evaluated by immunohistochemistry, including 8 samples from the initial biopsy, 28 samples from the definitive surgery, and 8 samples from the time of disease recurrence. GD2 was expressed on all 44 osteosarcoma samples. Osteosarcoma tissue obtained at the time of disease recurrence demonstrated a higher intensity of staining compared with samples obtained at initial biopsy and definitive surgery (P = .016). The majority of osteosarcoma cell lines expressed GD2 at higher levels than the neuroblastoma cell line BE(2)-C.
Osteosarcoma is the most common primary bone tumor in childhood. Despite multiple clinical trials, cure rates for patients with osteosarcoma have not significantly improved over the past 30 years, and survival for patients with metastatic or recurrent disease remains dismal.[1, 2] Novel approaches to the treatment of patients with osteosarcoma are needed to improve outcomes.
Monoclonal antibodies targeted against cell surface antigens specific to tumor cells have been proven to be effective in patients with breast cancer, lymphoma, and neuroblastoma.[3-5] Disialoganglioside GD2 is a glycosphingolipid that is expressed on the cell surface of limited normal adult tissue: the central nervous system, peripheral nerves, skin melanocytes, and mesenchymal stromal cells.[6-8] GD2 is also expressed on tumor cells and has been shown to be uniformly expressed on the surface of neuroblastomas and many melanomas.[6, 9, 10] Varying expression of GD2 has been shown in tumors in patients with lung cancer, central nervous system tumors, and sarcomas.[11, 12] Due to the relatively isolated expression of GD2 on malignant cells, it is an attractive target for antibody-mediated therapy and anti-GD2 antibodies have been shown to improve survival for patients with high-risk neuroblastoma.[5, 13]
To the best of our knowledge, the expression of GD2 on the cell surface of osteosarcomas has not been fully evaluated, although prior reports suggest it is expressed. In the current study, we explored the usefulness of GD2 as a potential target for antibody-mediated therapy in patients with osteosarcoma.
MATERIALS AND METHODS
Patient Samples and Cell Culture
Human osteosarcoma tumor tissue was obtained from patients at the time of initial biopsy, definitive surgery, or at disease recurrence. All tissue microarray samples were fixed in formalin and embedded in paraffin. The majority of samples were decalcified and the tissue microarray was constructed as previously described.[14, 15] All tumor samples used for enzyme-linked immunoadsorbent assay (ELISA) were obtained from patients with osteosarcoma who were treated at Memorial Sloan-Kettering Cancer Center in New York City. Tumor samples obtained at the time of definitive surgery all had ≤ 50% necrosis to be able to accurately assess GD2 expression across specimens. All specimen collection and analysis were performed in accordance with an Institutional Review Board-approved protocol and all patients or their guardians provided written informed consent. Primary cell cultures were generated by standard collagenase disaggregation of tissues. All isolated cells were cultured in Eagle minimum essential medium supplemented with 20% fetal calf serum and antibiotics in a 5% carbon dioxide humidified atmosphere at 37°C. All short-term cultures were established within 20 passages. The neuroblastoma cell line BE(2)-C and the fibrosarcoma cell line HT1080 were obtained from the American Type Culture Collection (Manassas, Va) and were maintained in the suggested media and additives.
Slides that contained the paraffin-embedded microarray and control tissues were baked and subsequently deparaffinized. Endogenous peroxidase activity was quenched using 0.3% hydrogen peroxide in methanol. Antigenic proteins were unmasked using universal antigen retrieval solution (IHC World, Woodstock, Md). The tissue was blocked with 10% normal goat serum in 5% bovine serum albumin (BSA) in Tris-buffered saline and stained with 100 ug/mL 14.G2a antibody diluted in 5% BSA overnight (14.G2a antibody was a generous gift from Dr. Karen Muszynski at the National Cancer Institute). Commercially available paraffin-embedded melanoma tissue known to uniformly express GD2 (BioChain Institute, Hayward, Calif) was used as positive control, and the purified immunoglobulin (Ig) G2a diluted in the same diluent as above was used instead of the primary antibody as the negative control. Slides were placed in a humidified chamber during incubation with primary antibody. Detection of the antibody-binding reaction was performed with biotinylated secondary antibody coupled with streptavidin-horseradish peroxidase. Avidin-biotinylated enzyme complex (Vectastain ABC System; Vector Laboratories, Burlingame, Calif) was used according to the manufacturer's instructions. The tissue was treated with 3,3′-diaminobenzidine (BioFx Laboratories, Owings Mills, Md) to identify sites of additional antibody binding and counterstained with hematoxylin. The tissue was then dehydrated with alcohol, permeated with xylene, and mounted with Permount organic mounting solution (Fisher Scientific Inc, Pittsburgh, Pa). Stained slides were viewed using a Nikon Inverted Microscope ECLIPSE TE200 (Nikon Instruments Inc, Tokyo, Japan) attached to a CCD (Diagnostic Instruments, Sterling Heights, Mich).
The intensity and location of tissue staining were assessed by a comparison with the positive and negative controls. Staining was considered positive (+++) only if the location and intensity of staining was consistent with that of the control melanoma tissue with 67% to 100% of cells staining positive. Tissue was considered negative if there was a definitive absence of staining. Slides were considered to demonstrate sporadic staining (+) if 1% to 33% of cells stained positive or intermediate staining (++) if 34% to 66% of cells stained positive. Tissue was assessed and graded by 2 independent observers who were blinded to the timing at which the specimens were obtained (initial biopsy, definitive surgery, or disease recurrence). Specimens with discordant assessments were rereviewed by the 2 observers and a consensus was reached.
Twenty-four short-term cultures (OS71, OS191, OS194, OS223, OS229, OS231, OS233, OS238, OS242, OS248, OS268, OS269, OS290, OS291, OS297, OS300, OS301, OS303, OS307, OS308, OS314, OS319, OS322, and OS323) described previously were used for ELISA. A total of 10,000 cells were plated in 96-well plates and incubated overnight (37 °, 5% carbon dioxide) in serum-free media (Eagle minimum essential medium, 1% Pen-Strep [Life Technologies, Grand Island, NY], and 1% nonessential amino acids). Cells were fixed with 10% buffered formalin and incubated at room temperature. Plates were washed and incubated with blocking solution (2% BSA in phosphate-buffered saline). After blocking, plates were washed and incubated with primary antibody (100 ng of 14G2a antibody [Santa Cruz Biotechnology, Santa Cruz, Calif]) while shaking slowly (50 revolutions per minute on an I2400 New Brunswick incubator shaker). After additional washes, secondary antibody was added (1:1000 dilution of horseradish peroxidase-conjugated antimouse IgG2a in phosphate-buffered saline; Santa Cruz Biotechnology) and incubated. Plates were then washed and 100 μL of TMB substrate was added to each well. After the substrate turned blue, 100 μL of stop solution was added (1M phosphoric acid). Plates were read at 450 nanometers on a Benchmark Plus microplate reader (BioRad, Hercules, Calif).
Standard curve preparation
Purified human GD2 protein (American Research Products, Belmont, Mass) was plated in a 96-well plate at concentrations from 12.5 ng/well to 1600 ng/well. Plates were incubated overnight. The ELISA procedure was identical to that used for the osteosarcoma cells with the exception that no fixing step was performed.
Results of the immunohistochemistry assays are reported as frequencies of negative staining (-), sporadic staining (+: 1%-33% of cells stained positive), intermediate staining (++: 34%-66% of cells stained positive), and positive staining (+++: 67%-100% of cell stained positive) sorted by the time at which the specimen was obtained (initial biopsy, definitive surgery, and disease recurrence). Fisher's exact test was used to determine the difference between intermediate and positive staining (++ and +++) versus negative and sporadic staining (- and +) in the recurrence samples versus the biopsy and definitive surgery samples. A P value < .05 was considered to be statistically significant.
Forty-four human osteosarcoma tissue samples from 41 individual patients were available to evaluate GD2 expression. Approximately 18% of samples were obtained from the initial biopsy, 64% were obtained at the time of definitive surgery, and 18% were obtained at the time of disease recurrence (Table 1). One recurrent specimen was obtained from a metastatic bone lesion. All other specimens obtained at the time of disease recurrence were resected lung metastases.
Table 1. Immunohistochemistry Analysis of GD2 Protein Expression in Osteosarcoma Samples
Samples were scored as follows: - indicates negative staining; +, sporadic staining; ++, intermediate staining; and +++, positive staining. A P value of .016 represents the statistical difference between intermediate and positive staining (++ and +++) versus negative and sporadic staining (- and +) in the recurrence samples versus the biopsy and definitive surgery samples.
GD2 Distribution in Osteosarcoma Specimens
A tissue microarray analysis with 44 samples was used. The samples were stained with 14G2a, a murine monoclonal anti-GD2 antibody, and the distribution and intensity of GD2 staining was qualitatively graded compared with positive and negative control samples (Fig. 1). There was 96% concordance between the independent observers.
The results of the tissue microarray demonstrated the presence of GD2 in 100% of the patient samples (Table 1). No significant difference was noted in GD2 staining between samples obtained from the initial biopsy and samples obtained from the definitive surgery. However, the percentage of cells staining positive for GD2 was significantly greater in tissue obtained at the time of disease recurrence compared with tissue obtained at initial biopsy or during the definitive surgery. One hundred percent of samples obtained at the time of disease recurrence displayed intermediate or positive staining compared with 53% of samples obtained at the initial biopsy or definitive surgery (P = .016).
To evaluate GD2 expression quantitatively, a cell-based ELISA was performed on 24 osteosarcoma short-term cultures (Fig. 2). Neuroblastoma and fibrosarcoma are malignancies known to express GD2. Thus, the neuroblastoma cell line BE(2)-C and the fibrosarcoma cell line HT1080 were used as positive controls for this experiment. Nearly all of the osteosarcoma cell lines demonstrated levels of GD2 expression that were comparable to the neuroblastoma and fibrosarcoma cell lines. Approximately 83% and 46% of osteosarcoma cell lines expressed GD2 at higher levels than BE(2)-C and HT1080, respectively. Approximately 41% of osteosarcoma cell lines expressed GD2 at a level that was twice that of BE(2)-C and 12.5% of osteosarcoma cell lines demonstrated expression that was 5 times that of BE(2)-C.
Over the past decade, the identification and targeting of proteins specific to cancer cells has become increasingly important to improving outcomes for patients with cancer. In recent years, targeted therapy such as imatinib, trastuzumab, and CH14.18 has improved survival for patients with acute lymphoblastic leukemia, breast cancer, and neuroblastoma, respectively.[4, 5, 17] Similar to this study, a prior study has suggested that GD2 is expressed in osteosarcoma. In contrast to the prior study, the current study used an antibody that recognizes the same epitope as the therapeutic antibody CH14.18 that will soon potentially be licensed by the US Food and Drug Administration. Our findings that GD2 is highly expressed on the surface of osteosarcoma cells provide promising data supporting the development of clinical trials assessing the efficacy of anti-GD2 therapy in patients with osteosarcoma.
The results of the current study demonstrated that specimens of recurrent osteosarcoma demonstrated higher expression of GD2 compared with samples obtained at initial biopsy and definitive surgery. This finding suggests that osteosarcoma cells with high expression of GD2 may be more resistant to chemotherapy and provides additional support for targeting the antigen in clinical trials. To validate this finding, prospective studies assessing changes in GD2 expression between primary and recurrent specimens from individual patients will need to be conducted.
Numerous preclinical and clinical trials have assessed targeting GD2 in patients with cancer, focusing on 2 murine monoclonal antibodies, 3F8 and 14G2a, and a human-mouse chimeric antibody, CH14.18.[9, 10, 13, 18-28] The majority of these studies have focused on patients with neuroblastoma and melanoma, due the high surface expression in these malignancies. A few patients with osteosarcoma have been included in early phase 1 studies of GD2 antibodies. Varying responses were noted in these patients, and 1 patient with multiple bone metastases who was treated with 14G2a plus interleukin-2 (IL-2) had a complete response (Table 2).[5, 9, 10, 13, 18, 19, 21-29] The addition of IL-2 and granulocyte-macrophage-colony-stimulating factor to anti-GD2 therapy demonstrated increased antibody-dependent, cell-mediated cytotoxicity.[30, 31] A recent study by Yu et al demonstrated that Ch14.18, an anti-GD2 antibody, improved survival for patients with high-risk neuroblastoma when administered with granulocyte-macrophage-colony-stimulating factor and IL-2. Patients with no evidence of residual disease appeared to derive the greatest benefit from the antibody therapy, and patients received the treatment after intensive induction chemotherapy, autologous stem cell transplant, surgery, and radiotherapy. These findings may have implications for the design of protocols incorporating CH14.18 in the treatment of patients with other malignancies that express GD2 on the cell surface.
Table 2. Prominent Clinical Trials Using Anti-GD2 Antibody
Current anti-GD2 therapy induces cytotoxicity by stimulating the host immune system to clear tumor cells bound by antibody.[30, 31] It is difficult to assess the cytotoxicity of anti-GD2 therapy in vitro and osteosarcoma xenograft models are frequently in an immunosuppressed background. Thus, although it is feasible to demonstrate the antibody binds osteosarcoma cells, it is difficult to clearly assess tumor response and cytotoxicity. One potential approach will be to assess the effectiveness of anti-GD2 antibodies with cytokines in canine models of osteosarcoma because the dogs have fully functional immune systems. These studies should address tumor response, time to disease progression, and overall survival in dogs with osteosarcoma who are treated with anti-GD2 antibody therapy. In addition, it is unclear whether the GD2 antigen remains on the cell surface of osteosarcoma cells after treatment with anti-GD2 antibody, similar to neuroblastoma.[29, 32] Canine studies should assess the persistence of surface GD2 antigen after antibody treatment and could assess the usefulness of GD2 expression as a predictive biomarker.
The poor survival of patients with metastatic and recurrent osteosarcoma, despite decades of clinical trials, highlights the need for novel anticancer agents. The current study finding that GD2 is highly expressed on osteosarcoma cells when paired with recent data demonstrating the effectiveness of anti-GD2 therapy support the development of clinical trials in patients with metastatic and recurrent osteosarcoma.
Supported by the Foster Foundation, Swim Across America, and the Paul Calabresi Career Development Award for Clinical Oncology (to Dr. Roth) No. K12 CA-132783-04 from the National Cancer Institute.