This research constitutes a multi-centric, case-control designed investigation aimed at developing and implementing a blinded validation of a machine learning-powered, multi-cancer early detection model. This is to be achieved through the prospective collection of blood specimens from newly diagnosed cancer patients and individuals devoid of a confirmed cancer diagnosis
Cancerous tissues, their adjacent non-cancerous tissues, along with white blood cells (WBCs) and normal tissue samples will be utilized to identify potential methylation candidate markers and investigate variations in methylation patterns among patients diagnosed with distinct cancer types. Building upon previous research and current study, a comprehensive methylation signature panel tailored specifically to cancer patients will be established. We will prospectively collect blood samples from newly diagnosed cancer patients and non-cancer individuals to analyze and identify specific cancer signals via the detection of cfDNA methylation patterns. Following a rigorous and comprehensive research framework, a machine learning-driven model will be developed and validated through blinded testing in an independent cohort. The study aims to enroll approximately 2,650 cancer patients, with a focus on including early-stage cases to enhance the model's sensitivity in detecting cancers with favorable prognoses. Furthermore, around 2,400 control subjects, matched with cancer patients by age and gender and screened negative for cancer through routine tests, will participate as healthy or benign-condition volunteers in model development. Lastly, samples from an additional 300 patients with other tumors will be gathered to conduct interference testing, ensuring the robustness of the model's performance.
Study Type
OBSERVATIONAL
Enrollment
5,350
Fudan University
Shanghai, Shnaghai, China
RECRUITINGThe AUC, sensitivity, specificity and tissue origin accuracy of the multi-cancer early detection model in detecting cancer or non-cancer
Time frame: 12 months
The performance of the multi-cancer early detection model in early stage cancer and precancerous lesion cases
Time frame: 12 months
The performance of the multi-cancer early detection model in different subgroups (such as age, gender, cancer pathological classification, and clinical stage)
Time frame: 12 months
This platform is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare professional.