Application of Quantum Detection-Driven Artificial Intelligence Algorithms for Single-Molecule cfDNA Characterization in the Early Diagnosis of Prostate Cancer

Overview

This research project aims to develop a novel blood testing method integrating cutting-edge quantum sensing and artificial intelligence technologies to achieve precise, non-invasive early diagnosis of prostate cancer. The research will employ quantum sensors to perform ultra-high-sensitivity measurements of circulating free DNA (cfDNA) in blood, thereby training a dedicated AI diagnostic model. The ultimate objective is to establish the diagnostic efficacy of this approach through clinical validation, providing clinicians with a novel diagnostic tool capable of significantly reducing unnecessary prostate biopsy procedures.

Interventions

• Diagnostic test Quantum Detection
  This cohort will utilize archived plasma samples from a historical patient population with confirmed diagnoses (prostate cancer vs. controls). The objective is model development. The intervention involves analyzing these stored samples using the quantum sensing platform to extract multi-modal cfDNA features (e.g., fragmentomics, methylation). This data is then used to train and optimize the initial AI diagnostic algorithm, establishing the core model before prospective validation.
• Diagnostic test Quantum Detection
  This cohort will prospectively enroll new patients with suspected prostate cancer from the same institution as testing cohort. The objective is initial model validation. The intervention entails collecting pre-biopsy blood samples from these participants. The cfDNA from these fresh samples is analyzed using the locked model from the training phase. The model's predictions are then compared against the gold-standard prostate biopsy results to assess initial diagnostic performance.
• Diagnostic test Quantum Detection
  This cohort will prospectively recruit patients from multiple independent clinical centers. The objective is to test the model's generalizability. The intervention involves standardized blood collection across all external sites, with samples sent to a central lab for blinded cfDNA analysis using the finalized, locked-down model.

Primary outcome measures

• Area under the receiver operating characteristic curve (AUC-ROC) for the predictive model in the general population for prostate cancer. [Time frame: Through primary completion which may take 12 months.]
• Sensitivity of the predictive model in detecting prostate cancer within the general population. [Time frame: Through primary completion which may take 12 months.]
• Specificity of the predictive model in detecting prostate cancer within the general population. [Time frame: Through primary completion which may take 12 months.]

Secondary outcome measures (3)

• Area under the ROC curve for the predictive model in identifying prostate cancer within the PSA grey zone cohort. [Time frame: Through primary completion which may take 12 months.]
• Sensitivity of the predictive model in identifying prostate cancer within the PSA grey zone cohort. [Time frame: Through primary completion which may take 12 months.]
• The specificity of the predictive model in identifying prostate cancer among individuals in the PSA grey zone. [Time frame: Through primary completion which may take 12 months.]

Eligibility criteria

Inclusion criteria

• Male, aged 18-80 years;
• PSA > 4 ng/ml;
• Patients meeting criteria for prostate biopsy:

• fPSA/PSA < 0.16 or PSA D > 0.15 or PSA V > 0.75; ② Positive digital rectal examination (DRE); ③ Imaging studies (ultrasound/MRI) showing suspicious lesions.

Exclusion criteria

• Patients diagnosed with any malignant tumour within the past five years;
• Patients who have undergone transurethral resection or enucleation of the prostate;
• Patients who have previously received treatment for prostate cancer, including but not limited to endocrine therapy, targeted therapy, or immunotherapy;
• Patients on long-term anticoagulant or antiplatelet therapy (anticoagulants discontinued for less than one week);
• Patients who have received any form of tumour treatment prior to enrolment blood sampling, including surgery, radiotherapy/chemotherapy, endocrine therapy, targeted therapy, or immunotherapy;
• Concurrent severe systemic diseases deemed by the investigator likely to interfere with trial treatment, evaluation, or compliance, including serious respiratory, circulatory, neurological, psychiatric, gastrointestinal, endocrine, immunological, or urological disorders;
• Organ transplant recipients or individuals with prior non-autologous (allogeneic) bone marrow or stem cell transplantation;
• Subjects who have undergone blood transfusion within one month prior to blood sampling;
• Patients currently participating in other clinical trials, or who have participated in other clinical trials within the past year;
• Patients deemed unsuitable for this clinical trial by the investigator;
• Patients meeting any of the above criteria shall not be eligible for inclusion as subjects.

Criteria are shown verbatim from the registry (in English). Final eligibility is always assessed by the study center.

Healthy volunteers: Yes

Study design

Observational model

Cohort

Study locations

China · 7 centers

• Cancer Hospital, Chinese Academy of Medical Sciences — Beijing
• The First Affiliated Hospital of Guangzhou Medical University — Guangzhou
• Jiangsu Provincial People's Hospital — Nanjing
• The First Affiliated Hospital of Soochow University — Suzhou
• Shanghai Changzheng Hospital — Shanghai
• West China Hospital, Sichuan University — Chengdu
• Ningbo No. 1 Hospital — Ningbo

Identifiers

NCT: NCT07238959 · CAPS

Primary sources (government registries)

• ClinicalTrials.gov (US NIH)

View this study on ClinicalTrials.gov ↗