AI-powered multi-omics model to predict breast cancer distant and organ-specific metastasis

A Multi-omics Breast Cancer Distant Metastasis Prediction Model Empowered by Artificial Intelligence and Its Clinical Translation Research

This project will test whether combining tumor images, protein data, and gene sequencing with artificial intelligence can predict if and where breast cancer will spread for women who have stored metastatic tissue samples at the hospital.

Quick facts

• Study type: Observational
• Enrollment: 2000 (estimated)
• Sex: Female
• Sponsor: Second Affiliated Hospital, School of Medicine, Zhejiang University (Academic / other)
• Locations: 1 site (Hangzhou, Zhejiang)
• Trial ID: NCT07252986 on ClinicalTrials.gov

What this trial studies

This observational project will build a precision prediction system by integrating digital pathology images, immunohistochemistry, proteomics, and gene sequencing with artificial intelligence. Researchers will use clinical data and tissue specimens from female breast cancer patients treated at the lead hospital or participating centers between 2000 and 2028, focusing on cases with biopsy or surgical samples from distant metastatic sites. The model aims to predict overall distant metastasis and organ-specific spread to bone, lung, liver, and brain while analyzing molecular patterns to clarify underlying mechanisms. There are no experimental treatments; the work is retrospective and prospective observational data analysis to develop and validate predictive algorithms.

Who should consider this trial

Good fit: Women with pathologically confirmed breast cancer who have a biopsy or surgical specimen from a distant metastatic site stored at the Second Affiliated Hospital, Zhejiang University (or participating centers) from 2000 to 2028 are ideal candidates.

Not a fit: Patients without stored metastatic tissue specimens at the study sites, with incomplete clinical records, or with a concurrent metastatic cancer from another primary are unlikely to benefit from this project.

Why it matters

Potential benefit: If successful, the model could help clinicians monitor high-risk patients more closely and tailor surveillance or treatment to prevent or detect metastasis earlier.

How similar studies have performed: Previous AI and genomics efforts have shown promise for predicting breast cancer outcomes, but comprehensive multi-omics models that combine digital pathology, proteomics, and sequencing for organ-specific metastasis prediction remain relatively novel.

Eligibility criteria

Inclusion Criteria:

1. Female patients pathologically diagnosed with breast cancer at our hospital or participating centers between January 1, 2000, and December 30, 2028.
2. Must meet one of the following: 1) Initially diagnosed with breast cancer at our hospital, has developed distant metastasis (liver, bone, lung, or brain), and has a biopsy or surgical specimen from the metastatic site stored at our hospital, with complete clinical information available. 2) Primary tumor specimen is not at our hospital, but a biopsy or surgical specimen from a distant metastatic site (liver, bone, lung, brain) is stored at our hospital.

Exclusion Criteria:

1. Concomitant metastatic malignancy from another primary cancer.
2. Incomplete clinical information.

Where this trial is running

Hangzhou, Zhejiang

• The Second Affiliated Hospital, Zhejiang University School of Medicine — Hangzhou, Zhejiang, China (Recruiting)

Study contacts

• Study coordinator: Fengbo Huang, MD

Conditions Breast; Breast Cancer; Pathology