Pathological Micro-Niches and Multimodal Imaging Macro-Niches for Predicting Clinically Significant Prostate Cancer and Metastasis: A Multicenter Study
Key facts
- Who runs it
- Guangxi Medical University
- Where
- China1 site in 1 countryChoose a country to see its sites
- Trial code
- Status
- Active, not recruiting
- People taking part
- 2,000 people, planned
- Started
- Jan 2014
- Expected to end
- Oct 2029
- Type of study
- Observational, 4 groups
- Record updated
- 13 November 2026
Who can take part
- Age
- 18 years and older
- Sex
- Men only
- Healthy volunteers
- Cannot take part
The registry criteria, word for word
Show all 10 criteriaShow less
Who can take part
- Age 18 years or older
- Evaluated, diagnosed, or followed for suspected or confirmed prostate cancer at a participating center
- A unique patient-level study ID can be established and the temporal order of key examinations, pathology sampling, and treatment can be confirmed
- At least one qualified imaging, pathology, clinical, or follow-up dataset required by a substudy is available
- Data use conforms to ethical approval, informed consent, or the approved retrospective consent-waiver scope
Who cannot take part
- Reliable de-duplication, or the before/after order of the prediction time point and endpoint, cannot be confirmed
- Core imaging, pathology, or outcome data are damaged and cannot pass quality control
- Required data use exceeds the original authorization and ethical approval has not been obtained
- Use is prohibited by laws or regulations, collaboration agreements, or an ethical decision
- For a specific substudy, the primary reference standard is missing or clear label leakage cannot be corrected
This trial has no sites in United States.
It has sites in one country. Choose a country to see its sites.
About the trial
What the trial wants to find out
This is an investigator-initiated, multicenter, observational ambispective cohort study, predominantly retrospective, with an optional prospective non-interventional silent validation. The study tests a cross-scale hypothesis that the pathological micro-niche of the primary prostate tumor (tumor epithelium, invasive front, immune-rich or immune-poor regions, ECM/collagen remodeling, neurovascular neighborhoods, and peritumoral field effect) is associated at the patient level with the imaging macro-niche captured by multiparametric MRI (diffusion, perfusion, heterogeneity, local invasion) and PSMA PET/CT (receptor expression, whole-body lesion distribution, tumor burden). A pathology teacher model provides patient- or lesion-set-level niche representations only during model development; at deployment and validation, only MRI, PSMA PET/CT, and clinical information available before the prediction time point are used, with no voxel-level pathology-to-imaging registration. Under one master protocol with three pre-specified substudies, the system predicts: (A) patient- and lesion-level clinically significant prostate cancer (csPCa, ISUP Grade Group 2 or higher) from pre-biopsy MRI; (B) pathological pelvic lymph node metastasis (pN1) after radical prostatectomy with pelvic lymph node dissection, primarily from MRI; and (C) concurrent distant metastasis (M1) and subsequent distant metastasis in baseline M0 patients, primarily from PSMA PET/CT. Models are evaluated for discrimination, calibration, clinical net benefit, uncertainty, abstention, and stability across centers, scanners, tracers, and disease spectra, using locked multicenter external validation and prospective silent validation. Model outputs are never shown to clinicians or patients and do not alter care.
The 4 groups in the trial
- Substudy A: csPCa Prediction Cohort
Patients who completed MRI before biopsy or surgery; patient- and lesion-level clinically significant prostate cancer (Grade Group 2 or higher) is predicted from MRI and pre-prediction clinical variables, with tiered pathology reference standards.
- Substudy B: pN1 Prediction Cohort
Patients undergoing radical prostatectomy combined with pelvic lymph node dissection; pathological pN1 is predicted preoperatively, primarily from MRI with PSMA PET/CT as an optional incremental input.
- Substudy C1: Concurrent M1 Prediction Cohort
Patients within the +/-90-day pre-treatment baseline staging window; concurrent distant metastasis (M1) is predicted primarily from PSMA PET/CT and confirmed by a blinded multidisciplinary composite reference.
- Substudy C2: Subsequent Metastasis Prediction Cohort
Patients confirmed M0 at baseline who are followed from a pre-specified landmark for the first distant metastasis, with treatment as time-dependent and death as a competing event.
Data and sources
- Registry record
- Listed on ClinicalTrials.gov.How to verify a trial in the registry
- Official title
- A Multicenter Study of Pathological Micro-Niches and Multimodal Imaging Macro-Niches for Predicting Clinically Significant Prostate Cancer and Metastasis: Patient-Level Cross-Scale Knowledge Transfer From Histology to MRI and PSMA PET/CT
- Record updated
- 13 November 2026