Abstract Purpose: Sensitivity of preoperative chemotherapy, evaluated by a histological analysis using resected tumor tissue, is the most important prognostic factor for osteosarcoma (OS). Methotrexate, doxorubicin, cisplatin (MAP regimen), and ifosfamide (IFO) are recognized as key drugs for OS, although it is still debatable whether addition of IFO for all OS patient is beneficial or not. The aim of this study is to identify possible genomic markers for the prediction of chemosensitivity of preoperative chemotherapy for pediatric OS patients using CGH and NGS technologies. Methods: Pre-therapeutic biopsy tumor samples of 30 pediatric conventional OS patients treated in our institute who had homogeneous clinicopathological and therapeutic background were analyzed. The patients were divided into three groups according to the histological response to chemotherapy (A: MAP-sensitive, B: IFO-sensitive and C: resistant to both). Twenty-two tumor samples were used as a learning set for searching candidate markers. Additionally, 8 tumor samples were used for independent analysis as a validation set. Genomic DNAs were prepared from frozen biopsy samples and used for CGH analysis with Agilent 44k DNA microarray. Target-sequencing by using the Ion AmpliSeq Comprehensive Cancer Panel containing 409 cancer-related genes were also conducted for 10 samples. Results: Survival analysis of this cohort assured that chemosensitivity was significantly correlated to the patient outcome. Differential aberration analyses of the array CGH data resulted in selection of 8 putative genomic markers for classifying into A, B and C groups with differential chemosensitivity. Scoring system to predict chemosensitivity of each individual was developed with these genomic markers and subsequent validation analysis indicated that the scoring system was remarkably consistent to the patient properties. Target re-sequencing of 409 cancer-related genes by using Ion Proton sequencer is currently ongoing to identify additional genomic changes in the tumors of highly chemoresistant group. Conclusion: Our results indicated that copy number changes in OS biopsy samples could be good predictors of response to preoperative chemotherapy. These signatures as well as global mutation information may be useful for the future genomics-based personalized chemotherapy for OS patients. Citation Format: Shintaro Iwata, Tsukasa Yonemoto, Hagime Kageyama, Sana Yokoi, Hiroki Nagase, Akira Nakagawara, Takeshi Ishii, Miki Ohira. Characterization of genomic alterations in pediatric osteosarcoma with differential chemosensitivities: Construction of genome-based prediction system using preoperative biopsy samples. [abstract]. In: Proceedings of the AACR Special Conference on Pediatric Cancer at the Crossroads: Translating Discovery into Improved Outcomes; Nov 3-6, 2013; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2013;74(20 Suppl):Abstract nr A49.