Clinical data age, gender, and so on were obtained by reviewing the medical records. After adhering to these criteria, the study data set of 66 patients 40 males and 26 females, age 10—87 years, mean years from April to December was finalized. Table 1 lists the characteristics of patients in this study. We identified the primary tumors as the locations of lesions for the large lesions that involve more than one site.
MR images were all acquired using a Discovery MR 3. Other sequences such as T1-weighted fat-saturated images and general T1-weighted and T2-weighted images without fat saturation were excluded for analysis due to the retrospective data imperfection. All sequence parameters are described in Table 2. Prior to radiomic analysis, all images were transferred into itk-SNAP software version 3.
All lesions were segmented manually by a radiologist with more than 7 years of MRI diagnostic experience. Another radiologist with more than 10 years of MRI diagnostic experience was asked for confirming the accuracy of segmentation. A revise was made by two radiologists in case of disagreement on specific image segmentation after discussion.
Delineating the ROI, including the entire tumor, peritumoral edema, and periosteal reactions, was performed on one image of each sequence. The chosen slice or image demonstrated the maximum anteroposterior diameter of the lesion. Imaging features were extracted using Analysis Kit software A. Some features are reported in Table 3.
The principle of feature selection by the LASSO method is to restrain some feature coefficients to zero by adjusting the parameter. Then, the area under the receiver operating characteristic curve AUC could be achieved versus log by using tenfold cross-validation.
The LASSO method has an advantage of analyzing a large number of radiomic features with sparser samples [ 20 ]. The study flowchart is presented in Figure 1. A total of initial features in total were extracted from original MR data. We mainly focused on the analysis of the features extracted by using the LASSO method because the number of features extracted by the other two methods was relatively large, thereby increasing the difficulty for further analysis.
The selection result is shown in Table 4 b. In this paper, we assessed the ability of our newly established radiomic model based on using multiparametric MR data to help differentiate OS from EWS of the pelvis. Our radiomic model yielded favorable results and constituted a new technique for the discrimination of OS and EWS. In brief, we believe that the methodology developed in this work may serve as a reliable additional tool for differentiation OS from EWS.
Preoperative discrimination between OS and EWS of the pelvis is difficult for clinical practice, although it could be performed by using clinical data, such as age, gender, and other accessory examinations, such as medical images. OS is the most common primary malignant bone tumor, which originates from primitive bone-forming mesenchymal cells. It has a bimodal age distribution, which is during adolescence and older adulthood, respectively.
EWS of bones represents the third most common primary bone malignancy, which exceeded in prevalence only by OS and chondrosarcoma. The incidence rate peaks within the second decade of life. In terms of tumor location, EWS is commonly seen at diaphysis of long bones and the pelvis, though the most common site is still the extremities, which is quite similar as OS [ 22 ]. Both OS and EWS demonstrate aggressive features at radiography, reflecting their high-grade nature of malignancy.
Many common imaging characteristics are available, such as bone destruction with a moth-eaten to a permeative pattern, a variable amount of mineralized osteoid, various types of periosteal reaction, and adjacent soft tissue mass [ 23 , 24 ].
Among imaging modalities, MRI is routinely used. Marrow replacement and cortical destruction could be more clearly seen on MRI with heterogeneous signal intensity. It provides excellent tissue contrast, which is much better than computed tomography CT and conventional radiography; however, it could not clearly discriminate between OS and EWS due to overlapping imaging characteristics and tumor locations.
For instance, most of small cell OS a subtype of OS cases own the radiographic features including lytic bone destruction, a soft tissue mass, and periosteal reaction, which are also commonly seen in EWS cases. Furthermore, both their cells are small and have round, hyperchromatic nuclei; they might be mistaken from each other even by histologic analysis [ 25 ].
Some advanced sequences to make differential diagnosis have been investigated, such as diffusion-weighted imaging DWI [ 4 ] and intravoxel incoherent motion IVIM [ 26 ]. However, the ability of discrimination is not satisfactory.
Radiomics is a new methodology that uses advanced imaging features for differentiation, staging, monitoring tumors, and even detecting tumor genetics. Compared with traditional radiological diagnosis methods, radiomics could provide ample information and improve diagnostic reproducibility. Therefore, we decided to evaluate the performance of differentiation ability using radiomic analysis. The value of radiomics has been under investigation for the past decade.
Previous studies mainly focused on the establishment of a radiomics nomogram for staging malignant tumors [ 7 ], prediction of lymph node metastasis [ 8 , 9 ], tumor prognosis [ 10 ], prediction of treatment response [ 11 ], or even prediction of the cancer phenotype [ 12 , 13 ]. To the best of our knowledge, this study is the first that attempts to use radiomics to differentiate these two osseous malignant tumors.
The application we used is a mature platform, and the radiomic features it provides could describe the signal intensity, morphological characteristics, and texture information of the lesions, thereby possibly representing the information of the lesions comprehensively.
For example, in Table 4 , the strong associations of High Grey Level Run Emphasis extracted from the T2-FS data set could be explained from the fact that patients possess large and irregular high signal intensity on T2-weighted images in the inner portion of tumors, most likely representing necrotic areas.
The presence of these inner, long T2 signal irregular regions suggests that the tumor is rapidly increasing in size and might be more likely to be an OS. T2-FS could effectively depict the border of lesions and demonstrate the overall cellular density of lesions to a certain extent, whereas CET1 could describe the vascularity of lesions, reflect the degree of malignancy, and distinguish necrosis and solid tumors.
Furthermore, previous studies that used such sequences for radiomic analysis yielded favorable results [ 27 , 28 ]. Thus, we set up our model using these two sequences. Based on our results, features yielded from T2-FS perform better than those from CET1, with higher AUC and sensitivity, thereby possibly benefiting the patients who are allergic to contrast media.
Primary bone osteosarcoma in the pediatric age: state of the art. Lu, B. Luo, X. Osteogenic BMPs promote tumor growth of human osteosarcomas that harbor differentiation defects.
Machado, I. Virchows Arch. Mackintosh, C. Oncogene 31, — Mankin, H. The hazards of the biopsy, revisited. Members of the musculoskeletal tumor society. Bone Joint Surg. Meijer, D. Expression of aromatase and estrogen receptoralpha in chondrosarcoma, but no beneficial effectof inhibiting estrogen signaling both in vitro and in vivo. Sarcoma Res. Messerschmitt, P. Meyers, P. North Am. Mirabello, L. National osteosarcoma etiology study group.
Genetic variation at chromosome 8q24 in osteosarcoma cases and controls. Carcinogenesis 31, — Nakagawa, S. Nitric oxide synthases, cyclooxygenase-2, nitrotyrosine, and angiogenesis in chondrosarcoma and their relation to prognosis. Nathan, S. Elevated expression of Runx2 as a key parameter in the etiology of osteosarcoma. Nunez, A. CD syndecan-1 expression in bone-forming tumors. Papachristou, D.
Expression of integrin-linked kinase and its binding partners inchondrosarcoma: association with prognostic significance. Cancer 44, — Pedrini, E. Perbal, B. Prognostic value of CCN3 in osteosarcoma. Posthuma DeBoer, J. Molecular alterations as target for therapy in metastatic osteosarcoma: a review of literature. Metastasis 28, — Potikyan, G. Pradelli, E. Antagonism of chemokine receptor CXCR3 inhibits osteosarcoma metastasis to lungs. Robin, T. Rodriguez, N.
COX-2 expression correlates with survival in patients with osteosarcoma lung metastases. Rozeman, L. Dedifferentiated peripheral chondrosarcomas: regulation of EXT-downstream molecules and differentiation-related genes. Saini, V. Salinas-Souza, C. Glutathione S-transferase polymorphisms in osteosarcoma patients.
Genomics 20, — Sandhu, R. Serum fluoride and sialic acid levels in osteosarcoma. Trace Elem. Savitskaya, Y. Serum tumor markers in pediatric osteosarcoma: a summary review. Savola, S. Combined use of expression and CGH arrays pinpoints novel candidate genes in Ewing sarcoma family of tumors. Schrage, Y. COX-2 expression in chondrosarcoma: a role for celecoxibtreatment? Cancer 46, — Sharili, A. Expression of Snail2 in long bone osteosarcomas correlates with tumour malignancy.
Tumor Biol. Shen, T. Shimizu, T. Fibroblast growth factor-2 is an important factor that maintains cellular immaturity and contributes to aggressiveness of osteosarcoma. Singer, F. Using biochemical markers of bone turnover in clinical practice. Sun, X. Cancer Szuhai, K. Molecular pathology and its diagnosticuse in bone tumors.
The NFATc2 gene is involved in a novel cloned translocation in a Ewing sarcoma variant that couples its function in immunology to oncology.
Tan, P. Tirino, V. Human primary bone sarcomas contain CD cancerstem cells displaying high tumorigenicity in vivo. Vural, C. Wang, W. Biomarkers 15, — Wang, Y. Increased expression of insulin-like growth factor-1 receptor is correlated with tumor metastasis and prognosis in patients with osteosarcoma. Wang, Z. Functional and biological analysis of Bcl-xL expression in human osteosarcoma.
Bone 47, — Weber, K. Iowa Orthop. Wei, H. Expression of c-kit protein and mutational status of the c-kit gene in osteosarcoma and their clinicopathological significance. Won, K. Prognostic implication of immunohistochemical Runx2 expression in osteosarcoma. Tumori 95, — Woodgate, R. Xu, J. Expression of matrix metalloproteinase regulator, RECK, and its clinical significance in osteosarcoma.
C-erbB-2 expression and prognostic significance in osteosarcoma. Blood Cancer 51, — Yang, G. Arsenic trioxide up-regulates Fas expression in human osteosarcoma cells. Yang, J. Deletion of the WWOX gene and frequent loss of its protein expression in human osteosarcoma. Cancer Lett. Yang, Q. All-trans retinoic acid inhibits tumor growth of human osteosarcoma by activating Smad signaling-induced osteogenic differentiation.
Yao, Y. The expression of Crm1 is associated with prognosis in human osteosarcoma. Yasuda, T. Extraskeletal myxoid chondrosarcoma of the thigh with a t 9;17 translocation. Yu, L. Mitotic arrest defective protein 2 expression abnormality and its clinicopathologic significance in human osteosarcoma. APMIS , — Zambo, I. Nestin expression in high-grade osteosarcomas and its clinical significance.
Zenmyo, M. Gadd45b expression in chondrosarcoma: a pilotstudy for diagnostic and biological implications in histological grading. Zhang, Z. Serum and urinary metabonomic study of human osteosarcoma.
Proteome Res. Zhao, C. Ai Zheng 27, — Zuffa, E. P53 oncosuppressor influences selection of genomic imbalances in response to ionizing radiations in human osteosarcoma cell line SAOS The use, distribution or reproduction in other forums is permitted, provided the original author s or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice.
No use, distribution or reproduction is permitted which does not comply with these terms. Evola, robertoevola virgilio. Introduction Neoplastic bone disease includes metastatic lesions and primary bone neoplasms. Biomarkers There are several biomarkers and these can be specific or aspecific; diagnostic, prognostic, or therapeutic; serological, genetic, or histological. Serological Bone Formation and Resorption Biomarkers This group includes a series of serological markers which reflect the osteoblastic and osteoclastic activity of the bone.
Molecular Biomarkers In bone tumors, molecular diagnostics can increase the accuracy of the diagnosis and assist in subtyping bone tumors. Serum Markers Kaseta et al. Genetic Markers Perbal et al. This information focuses on the Ewing family of tumors in children and teens, but most of the content here including much of the treatment information applies to Ewing tumors in adults as well.
Osteosarcomas are the most common bone cancer in children and teens. They are described in Osteosarcoma. Most other types of bone cancers are usually found in adults and are uncommon in children. These include:. For more information on these cancers, see Bone Cancer in Adults. Many types of cancer that start in other organs of the body can spread to the bones.
These are sometimes referred to as metastatic bone cancers , but they are not true bone cancers. RESULTS A descriptive statistical analysis of patients' ages was performed after they were divided into 2 groups according to the type of tumor Table 1.
Rev Diag Tratamento. Sarcomas of bone. Cancer: principles and practice of oncology. Philadelphia: Lippincott; The role of surgery and resection margins in the treatment of Ewing's sarcoma. Clin Orthop. Outcome of radiation-related osteosarcoma after treating childhood and adolescent cancer: a study of 23 cases.
J Clin Oncol. Chemotherapy-induced tumor necrosis as a prognostic factor in localized Ewing's sarcoma of the extremities. Prognostic factors in nonmetastatic Ewing's sarcoma of bone treated with adjuvant chemotherapy: analysis of patients at the Istituto Ortopedico Rizzoli. Diagnostic imaging for therapy control of primary bone tumors.
Primary malignant bone tumors. Clinical aspects and therapy. Vienna Bone Tumor Registry. Widhe B, Widhe T. Initial symptoms and clinical features in osteosarcoma and Ewing sarcoma. J Bone Joint Surg Am. Sao Paulo: Lemar; Ewing's sarcoma family of tumors: Ewing's sarcoma of bone and soft tissue and the peripheral primitive neuroectodermal tumors.
Principles and practice of pediatric oncology. Metastatic lung tumors in children. Vopr Onkol. Link MP, Eilber F. Osteosarcoma and Ewing's sarcoma. Improved recovery chances in the past 25 years according to the experience of Balgrist Hospital. Schwinz Rundsch Med Prax. Ewing's sarcoma of bone: relation between clinical characteristics and staging. Oncol Rep. Lang S, Sulzbacher I.
0コメント