Study of Predicting Lymph Node Metastasis of High-risk Prostate Cancer by Artificial Intelligence Multi-omics Analysis

Anhui Medical University2,000 enrolled

Overview

The pathological-omics and imaging-omics in this study are combined to construct an artificial intelligence (AI) model that can predict whether high-risk prostate cancer patients may have lymph node metastasis. The model determines whether the patient has lymph node metastasis based on the MRI results and the pathological section image information of the case combined with clinical data before radical resection of the prostate. This study is a multicenter, prospective clinical study to verify the model's ability to predict whether high-risk prostate cancer patients may have lymph node metastasis.

This is a multicenter, prospective clinical study designed to validate the radiopathology artificial intelligence (AI) model. The study will recruit patients with prostate cancer from the First Affiliated Hospital of Anhui Medical University, Nanjing Drum Tower Hospital, Cancer Hospital of Chinese Academy of Medical Sciences, Hospital General University Gregorii Maran and the First Affiliated Hospital of Bengbu Medical University, with Gleason score ≥8 or prostate specific antigen (PSA)≥20ng/ml. In addition, MRI examinations are required before prostate biopsy, and pathological sections are scanned after radical prostatectomy. Experienced radiologists and pathologists manually outline the tumor region of interest (ROI) on the image. The outlined MRI information and pathological section scan information are input into the model to obtain the probability of lymph node metastasis in the patient. Whether lymph node metastasis occurs is determined by pelvic lymph node dissection specimens. By comparing the probability of lymph node metastasis predicted by the model with the actual situation, the researchers calculate the predicted sensitivity, specificity, positive predictive value, negative predictive value, and overall diagnostic accuracy. This study verifies the high accuracy of the radiopathology AI model in predicting lymph node metastasis in patients with high-risk prostate cancer, and provides a basis for the precise treatment of high-risk prostate cancer patients.

Study Type

OBSERVATIONAL

Enrollment

2,000

Conditions

Prostate Cancer

Eligibility

Sex: MALEMin age: 50 Years

Medical Language ↔ Plain English

Inclusion Criteria: 1. Age ≥ 50 years 2. Patients must have histologically or cytologically confirmed prostate adenocarcinoma 3. PSA ≥ 20ng/ml or Gleason ≥ 8 4. Eastern Cooperative Oncology Group (ECOG) performance status (PS) score of 0-2 5. Life expectancy ≥ 6 months 6. Normal bone marrow function: absolute neutrophil count ≥ 1.5×109/L; platelets ≥ 75×109/L; hemoglobin ≥ 90g/L; white blood cell count ≥ 3.0×109/L 7. Normal liver function: alanine aminotransferase (ALT) or aspartate aminotransferase (AST) ≤ 2.5 times the upper limit of normal (ULN); for patients with liver metastasis, ALT/AST can be ≤ 5 times ULN 8. Total bilirubin ≤ 1.5 times ULN or total bilirubin \> 1.5 times ULN and direct bilirubin ≤ ULN; 9. Normal coagulation function: International Normalized Ratio（INR） ≤ 1.5, partial thromboplastin time (APTT) ≤ 1.5 times ULN, prothrombin time (PT) \< ULN + 4 seconds 10. Normal heart function: left ventricular ejection fraction (LVEF) ≥ 50%; corrected QT interval male \< 450ms, female \< 470ms, serum potassium ≥ 3.5mmol/L 11. Normal blood pressure: systolic blood pressure \< 140mmHg, diastolic blood pressure \< 90mmHg; patients with stable blood pressure assessment after appropriate clinical treatment can be enrolled 12. Normal renal function: serum creatinine ≤ 1.5 times ULN, and creatinine clearance ≥ 50 mL/min 13. Prospective subjects can understand and are willing to sign the informed consent form 14. Able to comply with the study visit schedule and other protocol requirements Exclusion Criteria: 1. Patients with contraindications to MRI examination, such as metal implants in the body, claustrophobia, etc. 2. Patients with any missing baseline clinical and pathological information 3. Patients with a clear history of neurological and psychiatric disorders, such as dementia, epilepsy, or seizures 4. In the judgment of the investigator, there are serious concomitant diseases that endanger the safety of the subjects or affect the subjects' completion of this study (such as severe diabetes, thyroid disease, and mental illness, etc.), or factors that affect the safety of the patients or affect the patients' provision of informed consent (including laboratory abnormalities), or any psychological, family, sociological or geographical conditions that affect the study plan and follow-up plan 5. The investigator believes that it is not suitable to participate in this clinical trial for any reason 6. Unable to provide informed consent

Outcomes

Primary Outcomes

The area under curve (AUC) of Receiver Operating Characteristic (ROC) curves of the radiopathology artificial intelligence model

The area under curve (AUC) of Receiver Operating Characteristic (ROC) curves of the radiomics and pathomics artificial intelligence model for predicting lymph node metastasis in high-risk prostate cancer patients

Time frame: baseline

Secondary Outcomes

The specificity of the radiopathology artificial intelligence model

The specificity of the radiomics and pathomics artificial intelligence model for predicting lymph node metastasis in high-risk prostate cancer patients.