A proposed radiological model for preoperative chemotherapy response prediction in patients with skeletal Ewing sarcoma

Hisaki Aiba ^1,2, Paolo Spinnato ³, Ayano Aso ^4,5

¹Clinica Ortopedica e Traumatologica III a Prevalente Indirizzo Oncologico, IRCCS Istituto Ortopedico Rizzoli, Via Pupilli 1, 40136, Bologna, Italy. hisakiaiba@yahoo.co.jp.
²Department of Orthopedic Surgery, Nagoya City University, Nagoya, Japan. hisakiaiba@yahoo.co.jp.
³Diagnostic and Interventional Radiology, IRCCS Istituto Ortopedico Rizzoli, Bologna, Italy.
⁴Clinica Ortopedica e Traumatologica III a Prevalente Indirizzo Oncologico, IRCCS Istituto Ortopedico Rizzoli, Via Pupilli 1, 40136, Bologna, Italy.
⁵Department of Medical and Surgical Sciences, University of Bologna, Bologna, Italy.
⁶Department of Pathology, IRCCS Istituto Ortopedico Rizzoli, Bologna, Italy.
⁷Department of Information Sciences, Tokyo University of Science, Noda, Japan.
⁸Department of Orthopedic Surgery, Nagoya City University, Nagoya, Japan.
⁹Department of Biomedical and Neuromotor Sciences (DIBINEM), Alma Mater Studiorum University of Bologna, Bologna, Italy.
¹⁰Osteoncology, Bone and Soft Tissue Sarcomas, and Innovative Therapies, IRCCS Istituto Ortopedico Rizzoli, Bologna, Italy.

⁰Clinica Ortopedica e Traumatologica III a Prevalente Indirizzo Oncologico, IRCCS Istituto Ortopedico Rizzoli, Via Pupilli 1, 40136, Bologna, Italy. hisakiaiba@yahoo.co.jp.

Skeletal Radiology +

|

October 11, 2025

View abstract on PubMed

Summary

A new imaging model predicts Ewing sarcoma chemotherapy response. This helps identify patients needing adjusted treatment for better surgical outcomes and effective strategies.

Area Of Science

Oncology
Radiology
Medical Imaging

Background

Ewing sarcoma is a rare bone cancer requiring effective preoperative chemotherapy.
Histological response to chemotherapy is crucial for treatment planning and surgical outcomes.
Accurate prediction of treatment response can optimize patient management.

Purpose Of The Study

To develop and validate a predictive model using imaging data to estimate histological response to preoperative chemotherapy in Ewing sarcoma patients.
To identify key radiological parameters that correlate with treatment response.

Main Methods

Analysis of 133 patients with Enneking stage IIB/IIIB Ewing sarcoma treated between 2003-2020.
Radiological parameters assessed via MRI, including necrotic grade and volume change, before and after chemotherapy.
Least Absolute Shrinkage and Selection Operator (LASSO) regression used for parameter selection and model development.

Main Results

Key predictors identified: volume change, radiological necrotic grade, extraskeletal component regression, and gadolinium-enhancement disappearance.
The predictive model demonstrated good discrimination: AUC of 0.89 (training) and 0.77 (test).

Conclusions

The developed imaging-based model can accurately monitor preoperative chemotherapy efficacy in Ewing sarcoma.
Early identification of poor responders facilitates surgical margin planning and tailored treatment strategies.
This model aids in optimizing treatment for improved patient outcomes.

Keywords:

Ewing sarcoma Histological response Predictive model Preoperative chemotherapy