Predicting resistance to neoadjuvant chemotherapy in osteosarcoma using machine learning with clinical data and T2-weighted MRI radiomics
Clinical Snapshot
PICO Framework
| P — Population | Patients with osteosarcoma receiving neoadjuvant chemotherapy (n=115 internal cohort, n=49 external validation) |
| I — Intervention | Machine learning model using baseline T2-weighted MRI radiomics features and clinical parameters |
| C — Comparator | Histological response assessment as reference standard |
| O — Outcomes | Prediction of chemoresistance with sensitivity, specificity, and AUROC as diagnostic accuracy measures |
Bottom Line
This diagnostic accuracy study demonstrates that combining T2-weighted MRI radiomics features with clinical parameters can predict chemoresistance in osteosarcoma with good accuracy (AUROC 0.84 internally, 0.77 externally validated). The support vector machine model achieved 83.3% sensitivity and 72.7% specificity in the internal test set, with external validation confirming generalizability despite reduced specificity. While promising for personalized treatment planning, the approach requires standardized imaging protocols and specialized analysis software. The positive likelihood ratio of 3.06 provides moderate diagnostic value for identifying chemoresistant cases. Clinical implementation would need workflow integration and validation in larger, multicenter cohorts. This represents a significant advance toward precision medicine in osteosarcoma management, potentially enabling early identification of patients who may benefit from alternative treatment strategies.
Key Findings
P Value: Not reported
Effect Size: AUROC 0.84 (internal test), 0.77 (external validation)
Primary Outcome: Prediction of chemoresistant osteosarcoma using combined radiomics and clinical model
Nnt Or Sensitivity: Sensitivity 83.3% (internal), 88.5% (external); Specificity 72.7% (internal), 47.8% (external)
Confidence Interval: Not reported for sensitivity/specificity
Clinical Application
Requires standardized T2-weighted MRI protocols and radiomics analysis software; manual segmentation currently limits routine implementation Relevant to Australian sarcoma centers; MRI is standard imaging under Medicare; could inform PBS-funded treatment decisions and referral to specialized sarcoma units Patients with high-grade osteosarcoma being considered for neoadjuvant chemotherapy
Abstract
OBJECTIVES: Identifying patients at risk of chemoresistant osteosarcoma enables risk-adapted management. This study aimed to predict chemoresistant osteosarcoma using baseline clinical and magnetic resonance (MRI)-derived radiomics features, with histological response as the reference standard and external validation included. MATERIALS AND METHODS: This retrospective single-center study included 115 patients with osteosarcoma from an institutional registry as the internal cohort, divided into training and test sets, and 49 patients from another institution as an external validation cohort. Tumor and peritumoral regions were manually segmented on baseline fat-suppressed T2-weighted MRI. Radiomics features were extracted using PyRadiomics, followed by two feature selection methods to identify potential predictors. Six machine learning models with varying feature combinations were trained to classify histologic chemoresistance in the internal training set. Model performance was assessed in the internal test set, and the best model was externally validated. RESULTS: The support vector machine model combining eight tumor radiomics features and four clinical-imaging parameters (presence of tumor necrosis > 50% on contrast-enhanced MRI, age, body mass index, and presence of metastasis at presentation) demonstrated the best performance. In the internal test set, it achieved a sensitivity of 83.3%, a specificity of 72.7%, an area under the receiver operating characteristic curve (AUROC) of 0.84, and a positive likelihood ratio of 3.06. External validation yielded a sensitivity of 88.5%, a specificity of 47.8%, and an AUROC of 0.77. CONCLUSION: A model combining tumor radiomics and clinical parameters at diagnosis showed strong performance in predicting chemoresistant osteosarcoma, with results confirmed by external validation. This approach may support personalized treatment strategies in high-grade osteosarcoma. RELEVANCE STATEMENT: The validated model may support early, individualized osteosarcoma management. KEY POINTS: Baseline T2-weighted MRI radiomics and clinical data can predict chemoresistant osteosarcoma. Tumor radiomics combined with clinical features achieved strong predictive accuracy. The support vector machine model reached an AUROC of 0.84 for the internal testing and of 0.77 for the external validation. The validated model may support early, individualized osteosarcoma management.
References
- 1.Inkeaw, P., Papangkorn, Pruksakorn, D., Dumnoensun, Angkurawaranon, Salita, Morakote, W., Boonsri, P., Phettom, Rattanathon, & Kanthawang, T. (2026). Predicting resistance to neoadjuvant chemotherapy in osteosarcoma using machine learning with clinical data and T2-weighted MRI radiomics. European Radiology Experimental. https://doi.org/10.1186/s40644-019-0283-8
This content is for educational purposes for healthcare professionals only and does not constitute clinical advice. Clinical decisions should be based on individual patient assessment, current guidelines, and appropriate specialist consultation. Editorial Standards · Privacy Policy · Terms of Service