Non-invasive MRI biomarkers for assessment of neoadjuvant chemotherapy response in osteosarcoma: Current techniques and clinical perspectives

Background: Early prediction of chemotherapy response in osteosarcoma is critical for risk stratification, as the reference standard of histologic necrosis is only available after surgery. Conventional size-based imaging criteria are insufficient, prompting interest in advanced MRI biomarkers. To summarize and critically appraise the evidence on advanced MRI techniques for early prediction of response to neoadjuvant chemotherapy in high-grade osteosarcoma. Methods: A narrative review of studies evaluating diffusion MRI, perfusion MRI (DCE-MRI), oxygenation-sensitive MRI (BOLD/R2*), radiomics, radiogenomics, and complexity analyses during or before neoadjuvant chemotherapy. Associations with histologic response and survival outcomes were synthesized qualitatively. Results: Responders consistently demonstrated increased diffusivity (ADC/IVIM-D) and reduced perfusion during treatment. Among modalities, DCE-MRI relative wash-in rate (rWIR) showed the strongest and most reproducible association with histologic response and, in some cohorts, event-free survival. Radiomics and multiparametric models improved discrimination but lacked external validation. Performance depended on imaging timing, protocol standardization, and segmentation strategy. Conclusion: Advanced MRI provides valuable early, noninvasive indicators of chemotherapy response in osteosarcoma, particularly diffusion and perfusion metrics. Currently, these techniques support risk stratification and monitoring rather than treatment modification. Prospective multicenter validation and standardized protocols are required for clinical adoption.