gms | German Medical Science

Deutscher Kongress für Orthopädie und Unfallchirurgie (DKOU 2022)

25. - 28.10.2022, Berlin

Tumor microenvironment of osteosarcomas: Can prognosis and response to immunological therapies be predicted by using AI?

Meeting Abstract

  • presenting/speaker Changwu Wu - Universität Leipzig, Institut für Anatomie, Leipzig, Germany
  • Siming Gong - Universität Leipzig, Institut für Anatomie, Leipzig, Germany
  • Sonja Kallendrusch - Health and Medical University, Potsdam, Germany
  • Georg Osterhoff - Klinik für Orthopädie, Unfallchirurgie und Plast. Chirurgi, Universitätsklinikum Leipzig AöR, Leipzig, Germany
  • Nikolas Schopow - Universitätsklinikum Leipzig AöR, Klinik für Orthopädie, Unfallchirurgie und Plast. Chirurgie, Leipzig, Germany

Deutscher Kongress für Orthopädie und Unfallchirurgie (DKOU 2022). Berlin, 25.-28.10.2022. Düsseldorf: German Medical Science GMS Publishing House; 2022. DocAB78-1150

doi: 10.3205/22dkou631, urn:nbn:de:0183-22dkou6315

Veröffentlicht: 25. Oktober 2022

© 2022 Wu et al.
Dieser Artikel ist ein Open-Access-Artikel und steht unter den Lizenzbedingungen der Creative Commons Attribution 4.0 License (Namensnennung). Lizenz-Angaben siehe http://creativecommons.org/licenses/by/4.0/.


Gliederung

Text

Objectives: The tumor microenvironment (TME) plays an important role in the development and progression of osteosarcoma. Immune checkpoint inhibitor (ICI) therapy could significantly improve the prognosis for patients with osteosarcoma. The aim of this study was to establish a TME-based prognostic index (TMEindex) for osteosarcoma which could then be used to predict prognosis and response rates to ICI therapy through machine learning algorithms.

Methods: Based on osteosarcoma samples from the TARGET database, the ESTIMATE algorithm was used to generate the ImmuneScore and StromalScore (which positively correlate with the presence of immune cell/stromal infiltration in tumor tissue). Through the combination of differentially expressed genes, the TMEindex could be formed by various regression analyses. The prognostic role of the TMEindex was validated in three independent datasets. The molecular and immunological properties of TMEindex and the impact on immunotherapy were then comprehensively investigated using various artificial intelligence algorithms.

Results and conclusion: The basis of the TMEindex is the gene expression of MYC, P4HA1, RAMP1 and TAC4. The TMEindex is an independent prognostic factor in osteosarcoma and correlates with worse overall survival, relapse-free survival, and metastasis-free survival of patients. This was confirmed in the three independent cohorts. These results support the TMEindex as a valid model for risk stratification.

A high TMEindex is associated with multiple pathways related to carcinogenesis, such as the MYC pathway. A low TMEindex is associated with immune-related pathways, such as the inflammatory response. Patients with a higher TMEindex had a more immune-cold TME and higher invasiveness. Therefore, patients with a lower TME index are more suitable candidates for treatment with ICI to compensate for immunosuppression and boost existing anti-tumor immunity. The TME index correlates with the sensitivity of 29 oncological drugs. In conclusion, the TME index is a promising biomarker to predict the prognosis of patients with osteosarcoma and their response to ICI therapy, and to discriminate between molecular and immunological characteristics.