Multimodal Large Model-Driven Risk and Prognosis Assessment for Brain Metastases in Lung Cancer

Not yet recruiting

• Conditions: AI (Artificial Intelligence); NSCLC Brain Metastasis

• Registration Number: NCT07107035
• Lead Sponsor: Fudan University

Brief Summary

The goal of this nationwide, multicenter observational study is to develop and externally validate multimodal large models that can (1) predict the risk of brain metastases and (2) estimate long-term prognosis in patients with non-small cell lung cancer (NSCLC).

The main questions it aims to answer are:

* Can a multimodal large model that fuses imaging, pathology, genomic, and clinical data accurately identify NSCLC patients at high risk of developing brain metastases?

* Can a multimodal large model reliably forecast intracranial progression-free survival, progression-free survival, and overall survival across diverse real-world treatment settings? (ie, patients receiving distinct treatment regimens, in different treatment lines and with or without intracranial local therapies).

Because this is an observational study, there are no investigational treatments; instead, researchers will compare outcomes among patients who receive standard-of-care therapies (surgery, radiotherapy, systemic therapy) to determine how well the model's predictions align with observed events.

Participants will:

* Allow use of their routinely collected clinical information, imaging (chest CT, brain MRI), pathology slides, and molecular test results for model training and validation

* Undergo standard-of-care follow-ups

* Complete optional quality-of-life questionnaires during scheduled visits

Detailed Description

Not available

Study Timeline

• Status Verified: 2025-07
• Study First Submitted: 2025-07-30
• Study First Submitted QC: 2025-07-30
• Last Update Submitted: 2025-07-30
• Study First Posted: 2025-08-06
• Last Update Posted: 2025-08-06
• Study Start: 2025-09-01
• Primary Completion: 2027-08
• Study Completion: 2030-07-01

Recruitment & Eligibility

• Status: NOT_YET_RECRUITING
• Sex: All
• Target Recruitment: 20000

Inclusion Criteria

• Age≥18 years old;
• KPS score≥70;
• Pathologically confirmed lung cancer;
• Receiving guideline-concordant standard-of-care therapy, defined as: radical surgical resection for early- to mid-stage non-small cell lung cancer (NSCLC); stereotactic radiotherapy for early-stage NSCLC deemed medically inoperable; radical chemoradiotherapy for locally advanced NSCLC; or systemic therapy for advanced-stage NSCLC.
• Complete systemic imaging before treatment initiation, including contrast-enhanced brain MRI and contrast-enhanced chest CT;
• Informed consent of the patient.

Exclusion Criteria

• Multiple primary or metastatic tumors (except early skin cancer, cervical carcinoma in situ that has been treated radically, with no recurrence or progression for more than 5 years);
• Uncontrolled epilepsy, central nervous system disease, or history of mental disorders, judged by the researcher to potentially interfere with the signing of the informed consent form or affect patient compliance;
• Physical examination findings, clinical laboratory abnormalities, or other uncontrolled medical conditions identified by the investigator as potentially interfering with study results interpretation or increasing the patient's risk of treatment complications
• Pregnant or lactating women.

Study & Design

• Study Type: OBSERVATIONAL
• Study Design: Not specified

Primary Outcome Measures

Name	Time	Method
The predictive performance of the multimodal large models	2 years	Including Sensitivity (SEN), Specificity (SPE), Matthews Correlation Coefficient (MCC), Area Under the Receiver Operating Characteristic Curve (AUC) with its corresponding 95 % confidence interval (CI)

Secondary Outcome Measures

Name	Time	Method
Progression-Free Survival (PFS)	2 years
Intracranial Progression-Free Survival (iPFS)	2 years
Overall Survival (OS)	2 years