Optimal Standard Treatment Selection for Solid Tumor Patients by Biologically-informed Multi-agent System (SINGULARITY)

NING LI

Status and phase: Not yet enrolling; Phase 4

Conditions: Advanced Solid Tumors

Treatments: Drug: Biologically-informed multi-agent system (Quasar) including targeted drugs Osimertinib, chemotherapy pemetrexed, immunotherapy pembrolizumab et al. approved by China CDE.

Study type: Interventional

Funder types: Other

Identifiers

NCT06824792

SINGULARITY-001

Details and patient eligibility

About

This study is an exploratory cohort study conducted under real-world conditions, aiming to evaluate the feasibility of an artificial intelligence (AI)-guided standard treatment selection model for advanced solid tumors, as well as its superiority compared to clinician-selected treatment plans. A multi-agent system based on multimodal AI models will rank the priority of standard treatment options based on the personalized information of the patients, including including demographics, clinical information, and multi-omics data. The final treatment plan will be jointly selected by the patient and the clinician from the AI-recommended options, thereby delivering a personalized treatment.

Full description

This study is an exploratory cohort study conducted under real-world conditions, aiming to evaluate the feasibility of an artificial intelligence (AI)-guided standard treatment selection model for advanced solid tumors, as well as its superiority compared to clinician-selected treatment plans. The study will prospectively collect patient data of multiple dimensions, including demographics, clinical information (pathological classification, tumor staging, imaging findings, previous treatment regimens and their effectiveness, performance status scores), and multi-omics data (DNA gene panel testing, whole-exome sequencing, transcriptome sequencing, etc.). A multi-agent system based on multimodal AI models will rank the priority of standard treatment options based on the personalized information of the patients. The final treatment plan will be jointly selected by the patient and the clinician from the AI-recommended options, thereby delivering a personalized treatment.

Enrollment: 3,000 estimated patients

Sex: All

Ages: 18+ years old

Volunteers: No Healthy Volunteers

Inclusion criteria

• Voluntarily participate in the clinical study, fully understand and be informed about the study, sign the informed consent form, and be willing and able to comply with and complete all trial procedures.
• Aged ≥18 years, no gender restrictions.
• Patients with advanced or metastatic malignant tumors confirmed by histology or cytology.
• Able to provide tumor tissue and peripheral blood samples for multi-omics testing, or able to provide qualified whole-exome sequencing and transcriptomics data.

Exclusion criteria

• As assessed by the investigator, no standard treatment is available, or the patient is unsuitable for guideline-recommended anti-tumor therapies.
• Other conditions deemed unsuitable for participation in this study by the investigator.

Trial design

Primary purpose: Treatment

Allocation: N/A

Interventional model: Single Group Assignment

Masking: None (Open label)

---

3,000 participants in 1 patient group

Quasar

Experimental group

Description:

This arm involves the prospective collection of individual patient data, including demographic information, clinical details (such as pathological classification, tumor staging, imaging findings, prior treatments and their efficacy, and performance status scores), and multi-omics data (DNA gene panel testing, whole-exome sequencing, and transcriptome sequencing). An artificial intelligence model (namely, Quasar) integrates this multidimensional information to prioritize standard treatment options and identify the optimal personalized treatment plan for each patient. Based on the AI-recommended treatment list, the final treatment plan is jointly selected by the patient and the physician. If treatment adjustments are required due to tumor progression, intolerance, or other reasons, the AI model will generate a new optimal treatment plan based on updated patient characteristics. This iterative process continues until the patient withdraws from the study.

Treatment:

Drug: Biologically-informed multi-agent system (Quasar) including targeted drugs Osimertinib, chemotherapy pemetrexed, immunotherapy pembrolizumab et al. approved by China CDE.