Study on Radiogenomics Features Associated With Radiochemotherapy Sensitivity in Gliomas

The MRI data were collected from patients with gliomas before surgery, 2 weeks before initiating radiochemotherapy, 1 month after completing the radiotherapy (for lower-grade gliomas, LGG), or 4 and 10 months after completing the radiochemotherapy (for high-grade gliomas, HGG). Radiochemotherapy sensitivity labels were constructed based on the MRI images obtained before and after radiochemotherapy, following the RANO criteria. Radiomics features were extracted from preoperative MRI images and combined with transcriptomic information obtained from tumor tissue sequencing. This process allowed the construction of a radiogenomics model capable of predicting the response of gliomas to radiochemotherapy.

In this prospective cohort study, we will recruit patients with gliomas who have undergone craniotomy and received postoperative radiotherapy or radiochemotherapy (in cases of LGG and HGG, respectively). MRI images of the same sequences will be collected at corresponding time points, and transcriptomic sequencing will be performed on tumor tissue obtained during surgery. The established model will be applied to predict radiochemotherapy sensitivity and compared with the 'true' radiochemotherapy sensitivity labels, which are constructed based on the RANO criteria, to evaluate the predictive performance of the model.

Study Overview

• Status: Recruiting
• Conditions: Glioma
• Intervention / Treatment: Diagnostic test: Assess the response glioma to radiochemotherapy using radiogenomics-based AI model

Detailed Description

This trial aims to recruit 100 cases of LGG and 100 cases of HGG based on statistical calculations. MRI data, including T1-weighted, T2-weighted, T1 contrast-enhanced, and T2-Fluid Attenuated Inversion Recovery (FLAIR) sequences, will be collected before surgery, 2 weeks before initiating radiochemotherapy, 1 month after completing the radiotherapy (LGG), or 4 and 10 months after completing the radiochemotherapy (HGG).

The collected MRI images before and after radiochemotherapy will be used to assess changes in tumor volume. The RANO criteria will be employed to determine the tumor's sensitivity to radiochemotherapy: a complete response and partial response will be classified as sensitive, while stable disease and disease progression will be considered insensitive.

Radiomics features will be extracted using the open-source 'PyRadiomics' python package after performing image preprocessing and segmentation. Transcriptomic data will be obtained by conducting RNA sequencing analysis on tumor samples collected during surgery. Selected radiogenomic features will be incorporated into a pre-constructed machine learning model to predict the sensitivity of gliomas to radiochemotherapy. The model's performance will be evaluated using metrics such as classification accuracy (ACC), area under the receiver operating characteristic curve (AUC), positive predictive value (PPV), and negative predictive value (NPV).

• Study Type: Interventional
• Enrollment (Estimated): 200
• Phase: Not Applicable

Contacts and Locations

• Study Contact: Name: Tao Jiang, MD and PhD; Phone Number: +86 10 67021832; Email: taojiang1964@163.com
• Study Contact Backup: Name: Yinyan Wang, MD and PhD; Phone Number: +86 13581698953; Email: tiantanyinyan@126.com

Study Locations

• China
  • Beijing
    • Beijing, Beijing, China, 100071
      • Recruiting
      • Beijing Tiantan Hospital
      • Contact: Yinyan Wang, MD and PhD; Phone Number: +86 13581698953; Email: tiantanyinyan@126.com

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Adult; Older Adult
• Accepts Healthy Volunteers: No

Description

Inclusion Criteria:

• Patients aged 18 or older
• Histologically confirmed glioma
• No history of other brain tumors or previous cranial surgeries
• No history of preoperative radiotherapy or chemotherapy
• Available preoperative, pre-radiotherapy(postoperatively), and post-radiotherapy magnetic resonance imaging (MRI) data

Exclusion Criteria:

• Those who do not meet any of the inclusion criteria

Study Plan

How is the study designed?

Design Details

• Primary Purpose: Diagnostic
• Allocation: N/A
• Interventional Model: Single Group Assignment
• Masking: None (Open Label)

Number of Arms

1

Arms and Interventions

Participant Group / Arm	Intervention / Treatment
Other: Evaluate the response of patients with glioma to radiochemotherapy; The response of patients with glioma to radiochemotherapy will be assessed by the RANO criteria and the established radiogenomics-based artificial intellegent model.	Diagnostic test: Assess the response glioma to radiochemotherapy using radiogenomics-based AI model; Predict the radiochemotherapy sensitivity of patients with glioma using an established radiogenomics-based artificial intellegent mode

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Sensitivity of the AI model in predicting radiochemotherapy respone	Sensitivity = TP/(TP+FN)	1 month after radiotherapy (LGG); 4 and 10 months after radiochemotherapy (HGG)
Specificity of the AI model in predicting radiochemotherapy respone	Specificity = TN/(TN+FP)	1 month after radiotherapy (LGG); 4 and 10 months after radiochemotherapy (HGG)
Area under the Receiver Operating Characteristic curve (AUC)	AUC measures the entire two-dimensional area underneath the entire ROC curve	1 month after radiotherapy (LGG); 4 and 10 months after radiochemotherapy (HGG)

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Accuracy of the AI model in predicting radiochemotherapy respone	Accuracy of radiotherapy sensitivity prediction AI model = (TP+TN)/ (TP+TN +FP+FN)	1 month after radiotherapy (LGG); 4 and 10 months after radiochemotherapy (HGG)
Positive predictive value (PPV) of the AI model in predicting radiochemotherapy respone	PPV of radiotherapy sensitivity prediction AI model = [TP/(TP+FP)]*100	1 month after radiotherapy (LGG); 4 and 10 months after radiochemotherapy (HGG)
Negative predictive value (NPV) of the AI model in predicting radiochemotherapy respone	NPV of radiotherapy sensitivity prediction AI model = [TN/(FN+TN)]*100	1 month after radiotherapy (LGG); 4 and 10 months after radiochemotherapy (HGG)

Collaborators and Investigators

• Sponsor: Beijing Tiantan Hospital
• Investigators: Principal Investigator: Yinyan Wang, MD and PhD, Beijing Tiantan Hospital

Study record dates

Study Major Dates

• Study Start (Actual): January 23, 2024
• Primary Completion (Estimated): November 30, 2024
• Study Completion (Estimated): December 31, 2024

Study Registration Dates

• First Submitted: June 6, 2024
• First Submitted That Met QC Criteria: June 6, 2024
• First Posted (Actual): June 12, 2024

Study Record Updates

• Last Update Posted (Actual): June 12, 2024
• Last Update Submitted That Met QC Criteria: June 6, 2024
• Last Verified: June 1, 2024

More Information

Terms related to this study

• Additional Relevant MeSH Terms: Neoplasms by Histologic Type; Neoplasms; Neoplasms, Glandular and Epithelial; Neoplasms, Neuroepithelial; Neuroectodermal Tumors; Neoplasms, Germ Cell and Embryonal; Neoplasms, Nerve Tissue; Glioma
• Other Study ID Numbers: 82072786

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No