Integrating peritumor and tumor CT radiomics features in predicting local control after SBRT in patients with pulmonary oligometastases

Yao Lu ^1,2, Yiqi Wang ¹, Yuxi Ding ³

¹Department of Radiation Oncology, The First Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, Zhejiang, 310003, P.R. China.
²Department of Health Technology and Informatics, The Hong Kong Polytechnic University, Hung Hom, Hong Kong, Kowloon, China.
³School of Basic Medical Sciences and Forensic Medicine, Hangzhou Medical College, Hangzhou, Zhejiang, 310053, P.R. China.
⁴Cancer Center, Zhejiang University, Hangzhou, Zhejiang, 310014, P.R. China.
⁵Department of Radiology, The First Affiliated Hospital, Zhejiang University School of Medicine, No. 79 Qingchun Road, Hangzhou, Zhejiang, 310003, P.R. China.
⁶Department of Pathology, The First Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, Zhejiang, 310003, P.R. China.
⁷Department of Radiation Oncology, The First Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, Zhejiang, 310003, P.R. China. yansenxiang@zju.edu.cn.
⁸Cancer Center, Zhejiang University, Hangzhou, Zhejiang, 310014, P.R. China. yansenxiang@zju.edu.cn.
⁹Department of Health Technology and Informatics, The Hong Kong Polytechnic University, Hung Hom, Hong Kong, Kowloon, China. gary-ge.ren@polyu.edu.hk.
¹⁰Department of Clinical Oncology, The University of Hong Kong-Shenzhen Hospital, Shenzhen, P.R. China. gary-ge.ren@polyu.edu.hk.
¹¹Department of Radiation Oncology, The First Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, Zhejiang, 310003, P.R. China. zju_zhaofeng@zju.edu.cn.
¹²Cancer Center, Zhejiang University, Hangzhou, Zhejiang, 310014, P.R. China. zju_zhaofeng@zju.edu.cn.

⁰Department of Radiation Oncology, The First Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, Zhejiang, 310003, P.R. China.

Radiation Oncology (london, England) +

|

August 15, 2025

View abstract on PubMed

Summary

A new radiomics model accurately predicts local control after stereotactic body radiotherapy (SBRT) for pulmonary oligometastases. Integrating tumor and surrounding features with clinical data improves prediction accuracy for optimizing SBRT strategies.

Area Of Science

Radiomics and Medical Imaging
Oncology
Radiation Therapy

Background

Accurate local control prediction is vital for optimizing stereotactic body radiotherapy (SBRT) in patients with pulmonary oligometastases.
Radiomics offers potential for non-invasive prediction of treatment outcomes.

Purpose Of The Study

To develop and validate a predictive radiomics model integrating tumor-intrinsic, peritumoral features, and clinical factors for enhanced local control prediction in pulmonary oligometastases treated with SBRT.
To assess the added value of peritumoral features and clinical data in improving prediction accuracy.

Main Methods

Analysis of 223 tumors from 146 patients, split into training (n=165) and validation (n=58) sets.
Extraction of radiomic features from gross tumor volume (GTV) and peritumoral regions (pGTV) in CT images.
Development of prediction models using Multilayer Perceptron (MLP) with SHAP analysis, incorporating GTV, pGTV, and clinical factors.

Main Results

Model-G (GTV features) achieved a validation AUC of 0.806.
Model-P (pGTV features) achieved a validation AUC of 0.708.
Model-GPC (GTV, pGTV, and clinical features) demonstrated the highest validation AUC of 0.902, indicating robust integration of data for accurate prediction.

Conclusions

The developed Model-GPC accurately predicts post-SBRT local control in pulmonary oligometastases.
Incorporating peritumoral features and clinical data significantly enhances prediction accuracy.
This model provides valuable insights for optimizing SBRT strategies in this patient population.

Keywords:

Local control prediction Peritumoral features Pulmonary oligometastases Radiomics Stereotactic body radiotherapy (SBRT)