Background
VS38c is a monoclonal antibody that recognises an intracellular antigen identified as the rough endoplasmic reticulum (rER) cytoskeleton-linking membrane protein P63 (CLIMP 63) which has also been termed P63 and cytoskeletal-associated protein 4 [
1]. CLIMP-63 is a stable, abundant rER protein that plays a role in protein transport. The VS38c monoclonal antibody strongly stains immunoglobulin-producing plasma cells and facilitates the immunohistochemical diagnosis of myeloma [
2,
3]. VS38c staining of tumour cells in neuroendocrine carcinoma, melanoma and osteosarcoma has also been reported [
4‐
7]. A common feature of all tumour cells that stain with VS38c is that they ultrastructurally contain numerous cytoplasmic ribosomes and abundant rER.
Regular protein production is a functional characteristic of viable cells, and loss or inactivation of ribosomes and rER has been noted in necrotic tumour cells [
8,
9]. Several types of cell death involve changes in ER [
10,
11], and ER stress has been shown to play a role in the necrosis of osteosarcoma cells [
12‐
14]. Osteosarcomas are usually treated pre-operatively by a course of neoadjuvant chemotherapy with assessment of the response evaluated by determining histologically the extent of necrosis in the resected tumour specimen [
15‐
19]; this percentage measure of tumour necrosis provides important prognostic information and gives a guide to as to whether adjuvant chemotherapy is required. Osteosarcoma cells can exhibit nuclear abnormalities post-chemotherapy and it is not certain whether such cells are viable or apoptotic in this context [
20‐
26]. There are no formal morphological criteria that can be used to determine whether an osteosarcoma cell is viable or not and it can be difficult to assess accurately the extent of necrosis in an osteosarcoma specimen.
In this study, we have sought to determine the diagnostic and prognostic utility of VS38c in the assessment of osteosarcoma and other primary bone tumours. We have examined VS38c staining in a wide range of benign and malignant bone tumours, including osteosarcoma subtypes, in order to assess the sensitivity and specificity of this marker for the diagnosis of osteosarcoma. In addition, as ER changes have been noted in necrotic tumour cells, and active protein production is a feature of cell viability, we have examined VS38c staining in osteosarcomas that have received neoadjuvant chemotherapy in order to determine if expression of this marker can be used to identify viable tumour cells and to facilitate the evaluation of percentage tumour necrosis in osteosarcoma specimens.
Discussion
Osteosarcoma is the most common primary malignant tumour of bone. Histological diagnosis of osteosarcoma is based on the identification of cytologically malignant tumour cells that form osteoid/bone. These matrix protein-producing malignant cells ultrastructurally contain numerous ribosomes and rER. In this study, we have shown that tumour cells in all high-grade and low-grade osteosarcomas are stained by the monoclonal antibody VS38c which recognises the CLIMP-63 protein on rER. Although absence of VS38c expression effectively excludes the diagnosis of osteosarcoma, the diagnostic specificity of this antibody is limited because, as noted previously [
5], VS38c also stains other bone tumours. Absence of staining with VS38c was noted in cultures of actinomycin D/cisplatin-treated osteosarcoma cells and in necrotic areas of (post-neoadjuvant therapy) osteosarcoma specimens, a finding which may be useful in evaluating percentage tumour necrosis in osteosarcoma.
In this study we characterised the pattern of VS38c staining in primary benign and malignant bone tumours. In addition to osteosarcoma and other benign bone-forming tumours, VS38c strongly stained other high-grade primary malignant bone tumours, including undifferentiated spindle-cell/pleomorphic sarcoma and Grade II/III chondrosarcoma. VS38c staining was also noted in most Ewing sarcomas, although the number of stained cells was highly variable. Atypical cartilaginous tumour/chondrosarcoma grade 1 and chordoma were either negative or showed little staining with VS38c. In a number of giant-cell rich lesions of bone, including giant-cell tumour, non-ossifying fibroma, chondroblastoma and aneurysmal bone cyst, VS38c strongly stained mononuclear cells but not osteoclastic giant cells. VS38c staining of such a wide range of bone tumours clearly limits its diagnostic specificity but, given its strong and consistent staining of osteosarcoma, it may be useful as a highly sensitive if not specific positive marker of this tumour.
Although a number of factors, (e.g. tumour site, size and histological subtype) influence disease survival in osteosarcoma, the extent of chemotherapy-induced necrosis has been shown to correlate most strongly with prognosis. Multi-drug neoadjuvant chemotherapy for high-grade osteosarcoma is now routinely employed and systematic pathological evaluation of the resection specimen of a treated osteosarcoma has been shown to yield important prognostic information about the chemotherapeutic response and to provide a guide as to whether adjuvant chemotherapy is required [
16‐
19]. In cases where chemotherapy-induced necrosis is 90% or more, the 5 year disease free survival is more than 80% [
27,
29‐
36]. Loss of ribosomes and rER is seen in necrotic tumour cells [
10,
11]. We noted that, in osteosarcoma specimens that had received neoadjuvant therapy, VS38c staining was strongly positive in areas of residual viable tumour but largely absent in morphologically necrotic areas of tumour; this pattern of staining was reflected in osteosarcoma cell cultures, where VS38c strongly stained viable but not necrotic tumour cells.
Qualitative assessment of the percentage area of tumour necrosis in pathological specimens of osteosarcoma is believed to correlate well with quantitative morphometric analysis [
15‐
19]. However, there are no studies, to our knowledge, which have proved that morphological assessment of tumour necrosis in osteosarcoma is accurate. Most histopathologists rely on the identification of cytoplasmic and nuclear alterations indicative of cell death, such as pyknosis, karyorrhexis and karyolysis, to determine tumour necrosis in chemotherapy-treated osteosarcomas. Necrotic areas of tumour often contain ghost cells which show a variable degree of loss of nuclear and cytoplasmic detail. However, irreversible cell death or necrosis is difficult to distinguish morphologically from reversible degenerative change or apoptosis affecting individual cells or small groups of cells. Preoperative neo-adjuvant chemotherapy of osteosarcomas not only results in necrosis, but also changes in the tumour stroma which maybe partly calcified, contain hypermineralised bone or areas of reparative fibrosis, neovascularisation and inflammation [
22‐
24,
26]. A major challenge to the surgical pathologist is the interpretation of atypical cells containing vacuolated cytoplasm and hyperchromatic nuclei with smudged or clumped chromatin found in some post-treatment osteosarcoma specimens. These atypical cells are often found in areas where there is not only necrosis but also fibrosis and it can be difficult to determine morphologically whether such cells are apoptotic/necrotic or viable. We found that most of these cells were VS38c negative, indicating that they did not contain rER. However, a few of these cells were VS38c positive. Cells with pyknotic nuclei are not uncommonly seen in areas of tumour necrosis in post-chemotherapy osteosarcoma specimens. Pyknosis is a nuclear feature seen in both apoptotic and necrotic cells and we noted that some tumour cells with pyknotic nuclei were VS38c positive; this indicates that such cells contain rER and that they are thus potentially capable of protein production, a feature characteristic of viable tumour cells.
In conclusion, our findings indicate that VS38c is a highly sensitive but not specific diagnostic marker for osteosarcoma. VS38c staining was seen in other primary benign and malignant bone tumours but it was not prominent in some low-grade tumours such as atypical cartilaginous tumour. VS38c staining outlined areas of residual viable tumour in post-chemotherapy osteosarcoma specimens and aided assessment of percentage tumor necrosis. However, although there was little or no VS38c staining in most areas of tumour necrosis in osteosarcoma specimens, there was variable staining of atypical cells with clumped chromatin and vacuolated cytoplasm and cells with pyknotic nuclei. The status of these VS38c positive cells in these necrotic areas is uncertain and more studies are needed to determine whether the morphological assessment of percentage tumour necrosis truly reflects the extent of viable and necrotic tumour in osteosarcoma specimens.
Authors’ contributions
EH collected and analysed the data and contributed to writing the manuscript. ZO and YU collected data. AG, AH, and UO interpreted the data and contributed to the writing of the manuscript. NA conceived the experiments, analysed the data and wrote the manuscript. All authors read and approved the final manuscript.